H'EPA
             United States
             Environmental Protection
             Agency
             Environmental Research
             Laboratory
             Athens GA 30613-7799
EPA/600/3-88/038
August 1989
             Research and Development
Terrestrial Ecosystem
Exposure Assessment
Model (TEEAM)

-------
                                              EPA/600/3-88/038
                                              August 1989
                TERRESTRIAL ECOSYSTEM
          EXPOSURE ASSESSMENT MODEL (TEEAM)
                         by

J.D. Dean, K.A.  Voos, R.W.  Schanz, and B.P.  Popenuck
             Woodward-Clyde Consultants
             500 12th Street, Suite 100
               Oakland, CA  94607-4014
               Contract No.  68-03-6304
                   Project Officer
                    Lee A. Mulkey
                  Assessment Branch
          Environmental Research Laboratory
               Athens, GA  30613-7799
          ENVIRONMENTAL RESEARCH LABORATORY
         OFFICE OF RESEARCH AND DEVELOPMENT
        U.S.  ENVIRONMENTAL PROTECTION AGENCY
               ATHENS, GA  30613-7799

-------
                                  DISCLAIMER

      The information in this document has been funded wholly or in part by
the United States Environmental  Protection Agency under Contract No.  68-03-
6304 with Woodward-Clyde Consultants.   It has been subject to the Agency's
peer and administrative review,  and it has been approved for publication as
an EPA document.  Mention of trade names or commercial products does  not
constitute endorsement or recommendation for use by the U.S. Environmental
Protection Agency.

-------
                                   FOREWORD


      As environmental controls become more costly to implement and the
penalties of judgment errors become more severe, environmental quality
management requires more efficient analytical tools based on greater
knowledge of the environmental phenomena to be managed.  As part of this
Laboratory's research on the occurrence, movement, transformation, impact,
and control of environmental contaminants, the Assessment Branch develops
management or engineering tools for exposure and risk evaluations of
pesticides and other toxic substances.

      In this work, a simulation model for toxic chemical exposures to
terrestrial wildlife was developed as part of the Ecological Risk Assess-
ment Research Program.  The initial focus of the Terrestrial Ecosystem
Exposure Assessment Model was on pesticide threats to small- and medium-
sized birds in agricultural settings.  Using TEEAM, the environmental
analyst can compute the probability of wildlife exposure in evaluating
the registration or regulation of specific pesticides.

                                     Rosemarie C. Russo, Ph.D.
                                     Director
                                     Environmental Research Laboratory
                                     Athens, Georgia
                                     iii

-------
                             ABSTRACT

      The Terrestrial Ecosystem Exposure Assessment Model  is a computer
code that simulates toxic organic chemical exposures to wildlife.   The
approach was to build a code generally applicable to a diverse range of
terrestrial ecosystems that could be parameterized to represent various
ecosystem types.  The initial focus, however, was on pesticide exposure
to small and medium-size birds in agricultural settings.  Using TEEAM,
the environmental analyst can compute the probability of wildlife  ex-
posure in evaluating the registration or regulation of pesticides.

      The model, which consists of seven computational modules, simulates
the environmental concentrations of pesticides in air, ephemeral surface
ponds, soil, soil water and soil gas, plant roots and aboveground  plant
biomass, and animals in the terrestrial food chain.  These media serve as
vectors for end-point species exposure to pesticides.  The model computes
both toxicant loadings to, and whole body concentrations in, the end-
point species.  To compute the probability of wildlife exposures to
these environmental concentrations, the model is equipped with a Monte
Carlo pre- and post-processing capability.

      The model documentation contains a discussion of model theory, code
installation and execution, parameter guidance and programmer's-level
model description.  Also described is ATEEAM, a simplified analytical
version of the food chain portions of the model.

      This report was submitted in partial fulfillment of Contract No.
68-03-6304 by Woodward-Clyde Consultants under the sponsorship of the
U.S. Environmental Protection Agency.  This report covers a period from
November 1986 to November 1987, and work was completed as of August 1988.
                                   iv

-------
                                  CONTENTS

                                                                        Page

Disclaimer	   ii
Foreword	iii
Tables    	   ix
Figures   	xii
Acknowledgments 	  xiv

Section

  1.0  Introduction 	    1

       1.1  Background	    1
       1.2  Objectives and Scope  	    1
            1.2.1   Long-Term Ecological Changes 	    2
            1.2.2  Short-Term Toxicant-Induced Ecusystem Changes  ...    2
            1.2.3  Typical MOdel Applications 	    4
            1.2.4  Priority Ecosystems  	    5
            1.2.5  Outputs of Interest  	    5
            1.2.6  Chemical Release Scenarios 	    6
       1.3  Report Organization   	    6

  2.0  Model Overview   	    7

       2.1  Previous Work	    7
       2.2  Model  Features  	    9
            2.2.1   Proceses Simulated 	    9
            2.2.2  Model Spatial Features 	   13
            2.2.3  Model Temporal Features  	   15
            2.2.4  Software Features  	   17
       2.3  Recommendations for Further Work	   18
            2.3.1   Additions to Existing Code	   18
            2.3.2  Parameter Estimation and Model  Verification  ....   19

  3.0  TEEAM Modules and Processes  	   20
       3.1  Toxicant Application/Deposition Submodel  	   20
            3.1.1   Introduction	   20
            3.1.2  Module Development 	   25
       3.2  Terrestrial Fate and Transport Module (TFAT)  	   40
            3.2.1   The Basic Fate and Transport Model (PRZM)	   40
            3.2.2  Enhancements to PRZM	   41
            3.3.3  Mathematical Description of the Terrestrial
                   Fate and Transport Processes	   43
       3.3  Plant Growth Module (PLTGRN)  	   72
            3.3.1   Introduction   	   72
            3.3.2  Development of TEEAM Plant Growth Module as
                   Adapted from EPIC	   73

-------
                                                                      Page
          3.3.3  EPIC Modifications for Inclusion in TEEAM 	     76
     3.4  Plant Contamination Transport Module (PLTRNS)   	     78
          3.4.1  Introduction	     78
          3.4.2  Background	     79
          3.4.3  Development of Module 	     79
     3.5  Terrestrial Animal Exposure Mpdule 	     84
          3.5.1  Introduction	     84
          3.5.2  Module Development  	     85
     3.6  The Monte Carlo Module (MC)	     95
          3.6.1  Description of Monte Carlo Parameter Distributions.     97
          3.6.2  Uncertainty in Correlated Variables 	     99
          3.6.3  Generation of Random Numbers  	    102
4.0  Model Installation and Execution  	    103

     4.1  IBM-PC Compatible Environment Requirements 	    103
          4.1.1  Hardware	    103
          4.1.2  Software	    103
     4.2  Loading Executable Codes and Test Data Files	    104
          4.2.1  Executing Test Data Inputs	    105
          4.2.2  Verifying Test Data Outputs	    105
     4.3  General Procedures for TEEAM Execution 	    105
     4.4  Machine and Compiler Dependencies  	    106
5.0  Input Sequence Development  	    108
     5.1  Overview of TEEAM Input Data	    108
     5.2  Description of Input Files for TEEAM Modules 	    109
          5.2.1  Execution Supervisor  	    109
          5.2.2  Input Data for the FSCBG Module	    Ill
          5.2.3  Input Data for the Spray Grid Definition Module .  .    112
          5.2.4  Input Data for the Terrestrial Fate and Transport
                 Module	    113
          5.2.5  Plant Growth and Translocation Location Modules .  .    115
          5.2.6  Input Data for the Terrestrial Animal Exposure
                 Module	    116
          5.2.7  Input Data for the Monte Carlo Module	    117
6.0  Parameter Estimation  	    179
     6.1  Introduction	    179
     6.2  FSCBG Parameters   	    179
          6.2.1  Aerial Spray Application  	    180
          6.2.2  Ground Spray Application  	    187
     6.3  GRDDEP Parameters  	    190
     6.4  TFAT Parameters	    191
          6.4.1  Original PRZM Parameters	    191
          6.4.2  Infiltration and Ponding	    238
          6.4.3  Volatilization and Pond Chemistry	    241
          6.4.4  Granular Formulations 	    242
          6.4.5  Soil Surface Temperature Regression Coefficients  .    245
     6.5  PLTGRN Parameters  	    245
     6.6  PLTRNS Module	    246
          6.6.1  RW—Root Reflection Coefficient	    246

                                   vi

-------
                                                                       Page

           6.6.2  LAMBDA—Degradation Rate of the  Contaminant
                  Within the Plant	    247
           6.6.3  KOW—Octanol-Water Partition Coefficient   	    248
           6.6.4  KP—Partition Coefficient	    249
           6.6.5  RHONA—Ratio of Dry Weight to Wet  Weight	    249
      6.7  APUM Module	    249
           6.7.1  Species Abundance 	    249
           6.7.2  Animal Movement	    251
           6.7.3  Feeding, Uptake, and Depuration  	    253
      6.8  Sensitivity of TEEAM to Input Parameters  	    263
           6.8.1  Sensitivity Analysis Approach 	    264
           6.8.2  Sensitivity Analysis Results  	    264

 7.0  Model Output	    269

      7.1  Introduction	    269
      7.2  INPREA	    269
      7.3  FSCBG	    269
      7.4  TFAT	    269
      7.5  APUM	    276
      7.6  MCARLO	    282
 8.0  Example Application 	    287

      8.1  General Problem Setting  	    287
      8.2  FSCBG and GRDDEF Inputs	    287
      8.3  TFAT Inputs	    288
      8.4  PLTGRN and PLTRNS Inputs	    288
      8.5  APUM Inputs	    288
      8.6  TEEAM Simulation Results 	    289

 9.0  Model Architecture  	    294

      9.1  Code Architecture	    294
           9.1.1  Pesticide Application/Deposition  	    297
           9.1.2  Terrestrial Fate and Transport	    302
           9.1.3  Plant Growth	    304
           9.1.4  Plant Contaminant Transport 	    305
           9.1.5  Terrestrial Animal Exposure 	    305
           9.1.6  Monte Carlo Simulation  	    306
      9.2  Intermodule Communication  	    308
      9.3  Coding Conventions	•	    308
      9.4  File Utilization	    308

10.0  Simple Models for Predicting Toxicant Accumulation
      in Terrestrial Wildlife:  ATTEAM 	    324
      10.1   Model Description 	    324
            10.1.1  Model 1.  Single Toxicant Application with
                    First-Order Soil Decay  	    324
            10.1.2  Model 2.  Continuous Toxicant  Application
                    with First-Order Decay  	    327
            10.1.3  Model 3.  Steady-State Concentrations under
                    Continuous Deposition 	    327
      10.2   Solution of Model Equations 	    327


                                    vii

-------
                                                                      Page

     10.3  Model  Acquisition,  Installation,  and Execution 	     330
           10.3.1  Hardware   	     331
           10.3.2  Software   	     331
           10.3.3  Installation and Execution Instructions   ....     331
     10.4  Model  Input Sequence Development    	     332
           10.4.1  Input Parameter Description  	     332
           10.4.2  Batch Input Sequence 	     336
           10.4.3  Interactive Input  	     337
     10.5  Parameter Estimation 	     341
           10.5.1  Mass Estimates	     344
           10.5.2  Rate Constants	     345
           10.5.3  Initial  Pesticide Concentrations 	     347
           10.5.4  Estimation  of Parameters  for Monte Carlo
                   Analysis	     347
     10.6  Output Files	     350
     10.7  Example Problem	     352

11.0  References	     356

Appendix.  ATTEAM Example Output using Default Parameter Values .  .     366
                                  vm

-------
                                LIST OF TABLES
Number                                                                   Page
5-1   Input formats for theexecution supervision module (EXESUP) ....  120
5-2a  Input formats for FSCB6 module 	  124
5-2b  Explanation of FSCBG options (ISW values)  	  137
5-3   Input formats for the spray grid definition module 	  140
5-4   Input formats for the TFAT meteorology file	143
5-5   Input formats for the terrestrial fate and transport module (TFAT)  144
5-6   Variable designations for TFAT time series file	157
5-7   Input formats for the plant growth (PLTGRN) and plant
      translocation (PLTRNS) modules   	  159
5-8   Input formats for the terrestrial animal exposure module (APUM)  .  164
5-9    Input formats for the Monte Carlo module (MC)	171
5-10   PNAME labels used to identify Monte Carlo input variables ....  174
5-11   SNAME labels used to identify Monte Carlo output variables  ...  176
6-1    FSCBG parameter specifications for four meteorological regimes  .  181
6-2    Aircraft-specific model parameters  	  183
6-3    Typical FSCBG spray drop size distribution parameters 	  184
6-4    Spray drop size range, approximate recovery rate, and
       recommended use of various spray nozzle types  	  185
6-5    Drop size distribution of aerosols and sprays, cumulative
       percent by volume   	186
6-6    Actual daytime hours for latitudes 24° to 50°  north of equator  .  195
6-7    Indications of the general magnitude of the soil/erodi-
       bility factor, K	197
6-8    Values of the erosion equation's topographic factor, LS, for
       specified combinations of slope length and steepness  	  198
6-9    Values of support-practice factor, P  	  199
6-10   Generalized values of the cover and management factor,C, in
       the 37 states east of the Rocky Mountains	200
6-11   Interception storage for major crops  	  203
6-12   Agronomic data for major agricultural  crops in the United States   204
                                     IX

-------
Number                                                                   Page
 6-13  Runoff curve numbers for hydrologic soil-cover complexes ....   205
 6-14  Method for converting crop yields to residue   	   206
 6-15  Residue remaining from tillage operations  	   206
 6-16  Reduction in runoff curve numbers caused by conservation
       tillage and residue management 	   207
 6-17  Values for estimating WFMAX in exponential  foliar model  ....   208
 6-18  Pesticide soil application methods and distribution  	   213
 6-19  Degradation rate constants of selected pesticides on foliage   .   215
 6-20  Physical  characteristics of selected pesticides for use in
       development of partition coefficients (using water solu-
       bility) and reported degradation rate constants in soil
       root zone	   217
 6-21  Octanol water distribution coefficients and soil  degrada-
       tion rate constants for selected chemicals  	   222
 6-22  Coefficients for linear regression equations for prediction
       of soil water contents at specific matric potentials 	   227
 6-23  Hydrologic properties by soil texture  	   230
 6-24  Descriptive statistics and distribution model for field
       capacity (percent by volume) 	   231
 6-25  Descriptive statistics and distribution model for wilting
       point (percent by volume)	   232
 6-26  Mean bulk density for five soil textural classifications ....   234
 6-27  Descriptive statistics for bulk density  	   235
 6-28  Descriptive statistics and distribution model for organic
       matter	   236
 6-29  Representative saturated hydraulic conductivity ranges
       for sedimentary materials  	   238
 6-30  Values of Green-Ampt parameters for SCS hydrologic soil groups .   239
 6-31  Descriptive statistics for saturated hydraulic conductivity  . .   240
 6-32  Estimated values of Henry's constant for selected pesticides . .   243
 6-33  Measured rate constants for release of pesticides from granules.   244
 6-34  Example regressions of surface soil temperature on air
       temperature	   246
 6-35  Plant growth parameters, typical ranges  	   247
 6-36  Plant growth parameters, crop specific values  	   248
 6-37  Examples of microphytic feeders and of carnivores that act as
       secondary and tertiary consumers within or on top of the soil  .   251
 6-38  Populations of earthworms in different habitats  	   252

-------
Number                                                                  Page
 6-39  Use of various habitats by mallard ducks in Nebraska  	   253
 6-40  Food intake rates and assimilation efficiencies for various
       common soil-dwelling organisms	   255
 6-41  Proportion by volume of plant and animal foods in the
       esophagi of mallards collected during breeding seasons of
       1974-80 in south central  North Dakota   	   258
 6-42  Depuration rates for various pesticides in various animal
       species found in the literature   	   260
 6-43  Input parameters used in  sensitivity analyses and their
       assumed distribution properties 	   265
 6-44  Significant parameters controlling pesticide dosage
       to the target species   	   267
 6-45  Significant parameters controlling pesticide concentrations
       in the target species   	   267
 9-1    List of subroutines by module and description of their
       functions   	   310
 9-2    Common block names, topics, and include file names  	   319
 9-3    Parameter statements, parameter definitions, and include
       file names	   320
10-1    Default values for ATTEAM model control parameters  	   334
10-2    Defalut values for system descriptive parameters  	   335
10-3    Formats for batch input file	   337
                                     XI

-------
                               LIST OF FIGURES
Number
       Ecosystem components and intercompartmental  transfer processes
       (soil/water/atmosphere/plant system)  simulated by TEEAM  .  .  .
Page

  10
2.1

2.2   Ecosystem components and intercompartmental  transfer processes
      (faunal system) simulated by TEEAM   	     11
2.3   Spatial structure of terrestrial  exposure model   	     14
2.4   Schematic of an ecosystem food chain   	     16
3.1   Schematic diagram showing geometry used in constructing the
      distance xr and effective source height H' for the case when
      the settling velocity in the jth size category at height
      H(Vj;H) is less than u   	     27
3.2   Schematic plan view showing the line source geometry with
      respect to a calculation point at R(e,6,z) for a wind
      direction e	     29
3.3   PRZM, release 1, model components  	     41
3.4   Schematic representation of a TFAT module soil layer	     45
3.5   SCS trapezoidal runoff hydrograph  	     59
3.6   Chemical transport and fate in ponded water	     61
3.7   Leaf area index as function of above ground biomass minus yield     74
3.8   Cover as function of leaf area index	     78
3.9   Schematic of plant contaminant transport module (PLTRNS  ....     80
5.1   Sample execution supervisor input data file (TMRUN.DAT)  ....    110
6.1   Pan evaporation correction factors   	    193
6.2   Diagram for estimating soil evaporation loss   	    194
6.3   Diagram for estimating SCS soil hydrologic groups  	    209
6.4   1/3-Bar soil moisture by volume	    228
6.5   15-Bar soil moisture by volume	    229
6.6   Mineral bulk density   	    233
6.7   Estimation of drainage rate AD versus number of compartments . .    237
6.8   Correlation matrix used in the sensitivity analyses  	    266
7.1   Example of a portion of the TFAT input echo	    270
7.2   Example of the PLT6RN/PLTRNS input echo	    271

-------
Number                                                                   Page
 7.3   Example of a portion of the APUM input echo	    272
 7.4   Example of a portion of the 6RDDEF input echo	    273
 7.5a  Example of the FSCBG input echo	    274
 7.5b  Example of the FSCBG output	    275
 7.6   Example of habitat hydrologic status output  	    277
 7.7   Examples of the habitat pesticide mass status output   	    278
 7.8   Example of the habitat pesticide concentration output  	    279
 7.9   Example of the habitat time source output    	    280
 7.10  APUM dosage breakdown output file	    281
 7.11  APUM time series output file   	    282
 7.12  Monte Carlo cummary output file  	    283
 7.13  Monte Carlo parameters output file   	    286
 8.1   TEEAM output for the example application   	    290
 9.1   TEEAM main program structure   	    296
 9.2   Batch input module (BATENT) structure  	    297
 9.3   Executive supervisor (EXESUP) structure  	    298
 9.4   TEEAMAIN program structure   	    299
 9.5   Module FSCBG structure   	    301
 9.6   Module TFAT structure	    303
 9.7   Module PLTGRN structure  	    304
 9.8   Module PLTRNS structure  	    305
 9.9   Module APUM structure	    306
 9.10  Module MCARLO structure  	    307
 9.11  FSCBG linkage to TEEAM habitats  	    309
10.1   Schematic of ATEEAM model  structure  	  '325
10.2   Screen 1 ATEEAM model title screen   	    338
10.3   Screens 2, 3, and 4 from ATEEAM model	    339
10.4   Screens 5 and 6 from ATEEAM model	    340
10.5   Screens 7 and 8 from ATEEAM model running in interactive mode  .    342
10.6   Screens 9 and 10 from running ATEEAM model  in interactive mode .    343
10.7   Example of MCARLO.OUT output file	    350
10.8   Example of TLEV1.0UT output file when model is run in Monte
       Carlo mode   	    351
                                    xm

-------
Number                                                                   Page

10.9   Example of TLEVL.OUT output file when model  is run in deter-
       ministic mode   	352

10.10  Example of histogram printer plot found in user-specified
       output file	353

10.11  Example of CDF printer plot found in user-specified output file .   354
                               ACKNOWLEDGMENTS

      The authors would like to acknowledge the support of the U.S.  Environ-
mental Protection Agency for this work.  Dr. Craig McFarlane and Dr. John
Emlen of EPA's Environmental Research Laboratory, Corvallis OR, are  thanked
for their information, advice and assistance regarding uptake and transloca-
tion of xenobiotic chemicals by vascular plants and ecological modeling,
respectively.  Mr. Lee Mulkey and Dr. David Brown of EPA's Environemntal
Research Laboratory, Athens GA, are appreciated for their confidence in, and
patience with, the authors and for their leadership role in defining project
objectives.  The staff of EPA's Ecological Effects Branch, Office of Pesticide
Programs, are thanked for helpful comments.
      Portions of the code and documentation were written originally by staff
of H.E. Cramer, Inc., of Salt Lake City UT and Aqua Terra Consultants of
Mountain View CA.  The authors also acknowledge the assistance of word pro-
cessors, editors, and graphic artists of Woodward-Clyde Consultants.
                                     xiv

-------
                                 SECTION 1

                                INTRODUCTION
1.1  BACKGROUND

    The U.S. Environmental Protection Agency is continually faced with
issues concerning the regulation, restriction, or banning of chemical
substances.  Decisions concerning these chemicals are based upon the
concept that humans or animals may be subjected to concentrations of
chemicals through various routes of exposure.  In the case of humans,
chemicals are normally subject to restrictions upon their use if the sum of
the loadings from these various routes exceeds an acceptable intake level
(ADI, RFD).  Until now, the Agency has had no computer-based methodology
for calculating exposures to terrestrial animals to compare to existing
toxicological data.  Data on wildlife contamination levels and reports of
wildlife kills from pesticide use indicate that the impact of these
exposures is substantial (RSPB 1965; DeWeese et al. 1986).

    Most pesticides are registered for use in terrestrial environments.
Other xenobiotics are released to the atmosphere and are subsequently
deposited on soil and plant surfaces, where they are subject to cycling in
terrestrial ecosystems, providing exposures to plants and animals.  Little
is known concerning the magnitude of exposures among the various exposure
pathways.  Generation of this information requires simultaneous simulation
of environmental concentrations in food items and the intake of
contaminants via the food chain and other exposure pathways to the species
of interest.

    In order to establish risk, probabilities of various levels of exposure
must be made utilizing the model.  This requires that variations in
naturally occurring processes which serve to cycle xenobiotics through the
terrestrial ecosystem and the uncertainty in model parameters be
considered, as well as the impact of management strategies.  All of these
factors have direct implications for risk analysis and risk management.

1.2  OBJECTIVES AND SCOPE

    The objective of this model development effort was to build a
simulation model which can be used by the Agency to evaluate the magnitude
of exposure from organic toxicants to plants and animals in terrestrial
                                     1

-------
ecosystems.   The simulation of events occurring in a terrestrial  ecosystem
which affect chemical  exposure to plants and  animals is a complex problem
because the interactions which occur in such  systems are complex.
Ecosystems are characterized by an intricate  web of feedback mechanisms
which keep the ecosystem in balance.  Many varied processes may affect this
balance.

    It was not possible, in this model development effort, to account for
all of these processes or their effects.  Therefore, a number of
limitations to the scope of the development effort were necessary.  These
are discussed in the following sections under the following categories:

       • Long-Term Ecological Changes
       • Short-Term Toxicant-Induced Ecosystem Changes
       • Typical Model Applications
       • Priority Ecosystems
       • Outputs of Interest
       • Chemical Release Scenarios

1.2.1  Long-Term Ecological Changes

    Some ecological processes operate over long periods of time (e.g.,
forest succession) and may be due to internal changes (for instance, in
nutrient balances) or external changes (e.g., climate).  It was beyond the
scope of this modeling effort to consider these types of changes.
Furthermore, it is probably inappropriate to consider them within the
context of this work.   The history of organic chemical production extends
back several decades.   The useful life of most chemicals encompasses
relatively short periods of time due to continual advances in technology.
In addition, the persistence of organic chemicals in the environment is
relatively short compared to ecosystem changes of the type mentioned above.

1.2.2  Short-Term Toxicant-Induced Ecosystem Changes

    Of greater pertinence to this work are ecosystem perturbations which
might be caused by toxicant insult.  There is little evidence to
demonstrate that toxicants produce effects of such significance to cause
terrestrial ecosystems to shift to new equilibrium points.  This
supposition is supported by the recent work such as that of Biederbeck et
al. (1987) and Baker et al. (1986).  However, toxicant insult may have
important short-term effects which may dramatically affect exposure to
species of interest.  For instance, the toxicant might temporarily reduce
the population of the primary food item of the species of  interest.  This
species, then, might


      •  Leave,  seeking a  new  source of this food  item

                                     2

-------
       • Switch to an alternative food source, thereby decreasing exposure

       • Switch to an alternative food source, increasing exposure

       • Prey on the contaminated, reduced population, resulting in the
         species'  own sickness or death

    In general, the toxicant effects which might affect exposures can be
broken down into two broad categories:

       • Avoidance/attraction

         Loss of habitat
    Avoidance/attraction behavior can be direct or secondary.  Direct
effects would be repulsion or attraction to the chemical itself.  A
secondary response would be, for instance, a repulsion or attraction to
prey which had been weakened by exposure to the toxicant.  There is
research currently underway to investigate the avoidance/attraction
behavior exhibited by various species in response to a variety of chemicals
and various formulations of those chemicals.  However, at this time, this
information cannot be generalized among species, chemicals, or formulations
so that avoidance/attractance mechanisms cannot be modeled realistically.
However, this is definitely an area in which further development of the
model is warranted as understanding of the processes advances.

    Loss of habitat can be subdivided into several effects.  Loss of
nesting habitat could occur due to spraying of an herbicide, directly
destroying cover or other characteristics of the nesting site.  Loss of
feeding habitat could result due to mortality or other loss of preferred
food items, such as failure of reproduction in lower trophic level fauna or
failure of fruit production in plants.  Although the impact of these
effects on wildlife exposures to chemicals is probably substantial, the
cause and effect relationships that govern the relationships between loss
of habitat and behavioral responses of specific species are not
generalizable.  For many species, habitat requirements have not even to
date been well documented.

     Therefore,  it  is  felt  that  simulation of  complex sequences  of
 behavioral responses  resulting  from  toxicant  insult is  not  possible at this
 time.   An  important  assumption  which derives  from this  conclusion  is:

        • Toxicants do not  affect the state of the ecosystem.  These
         characteristics which  describe  the  "state" include:

         - population levels of animals  (birth, death rates)

-------
         - assemblage of species in the plant community

         - behavior of animals (migration, feeding, etc.)

    Collectively, these assumptions require that the ecosystem is at a
point of equilibrium which will  be user defined at the outset of the
simulation and will not change over the simulation period.  In essence,
then, TEEAM is an exposure model in the setting of a terrestrial ecosystem,
and not an ecosystem model which can account for the effects of toxicants
on the ecological state of the system.

1.2.3  Typical Model Applications

    The model  is expected to be used by the Environmental Protection Agency
as a regulatory tool.  The utility of the model is to aid in the
formulation of strategies for mitigating damage to "terrestrial ecological
systems" based upon model simulation output.  This requires that the model
be capable of simulating exposure to terrestrial plant or animal species of
interest and that the impact of regulatory decisions on exposures can be
estimated by the model.  Regulatory strategies for pesticides might
include:

       • Banning

       • Restriction of application rates

       • Restriction on the timing of applications

       • Restriction of application in or within certain distances
         of sensitive ecosystems

       • Restriction on certain types of chemical formulations

        •  Restriction on  the cumulative quantity of a chemical which could
          be applied  in the same ecosystem over a given time period

The  model has been  constructed with the capability to address these
strategies.   Furthermore, the model is designed to be applied to simulate  a
generalized ecosystem.   For instance, the model might be applied to
evaluate  exposure  to sage grouse  in wheat field ecosystems  in the northern
Great  Plains.   It  would  not likely be applied  to a specific field in a
specific  township,  range, etc.  Therefore,  the need for  the capability to
simulate  on a high  level of spatial detail  is  deemed unnecessary.

     Obviously,  since the model  is to  be utilized for risk management,
probabilities of events  must  be generated by the model as well  as the

                                    4

-------
events themselves.  This requires a framework by which these probabilities
can be generated.  Therefore, the typical  model application may involve the
use of the Monte Carlo simulation capability.

1.2.4  Priority Ecosystems

    There are a number of ecosystems which are of potential interest to the
Agency.  Certainly ecosystems in which the regulation of chemicals may have
a large economic impact are of concern.  Other less economically important
ecosystems which might contain endangered or threatened species may also be
of concern.  It seemed unnecessary to prespecify ecosystems in which the
Agency has an interest for the purpose of the model development effort, and
this was not done.  To some extent, this may have aided in the development
of a more generalized approach. However, at this point in time the model
focuses on avians in agricultural ecosystems with pesticides as the
toxicant of greatest concern.  It is possible, through judicious selection
of parameters, that forest ecosystems may be simulated.  However,
simulation of ecosystems such as wetlands on tidal marshes is clearly
beyond the scope of the current model.

1.2.5  Outputs of Interest

    The most important question concerning model output is that of relevant
toxicological endpoints.  These endpoints might include:

       • Death of species of interest

       • Some measure of health of the population or an individual or
         population of the species (e.g.,  reproductive failure, egg shell
         thinning)
                           •
       • Behavioral modifications (e.g. nonphysical effects)

    Since most of the toxicological information available on the effects of
pesticides or xenobiotics on animals is in the form of a lethal dose (e.g.,
LD50, LD10), tne output of interest is most likely the dosage (through time
or cumulative) to the species of interest.  Because deciding upon
mitigation strategies may involve looking at major pathways of exposure, it
was also thought to be desirable to be able to view the total dosage in
terms of its components (e.g., inhalation, ingestion, absorption) or to
further categorize into concentrations and food/air/water intake rates.
This implies that tissue burden in plants or animals in lower levels of the
trophic systems, in the soil, water, and air, are also important outputs.

    Other effects may ultimately be simulated based upon concentrations in
plants and animals (e.g., avoidance/attraction behavior).  At some point,
it may be desirable to have the capability to simulate and/or output

-------
concentrations in various organs of the plants or animals in order to
determine certain effects of the toxicant's presence.   However,  this has
not been implemented in the current version of model.

1.2.6  Chemical Release Scenarios

    The terrestrial ecosystem exposure model was formulated for use by both
the Office of Pesticide Programs (OPP) and the Office  of Toxic Substances
(OTS).  Therefore, there was a desire to address pesticides as well as
other xenobiotics.  Once the chemical is placed into the ecosystem, it is
postulated that the processes involved in cycling of organic toxicants are
no different.  Therefore, the key difference in terms  of modeling
pesticides versus xenobiotics is the application/deposition process.
Pesticides are released to terrestrial environments for the most part in
ground or aerial spraying events, either to bare soil  and/or to plant
canopies.  Application may also take the form of soil  injection or
application in irrigation waters (chemigation).  Irrigation may take on
various forms (spraying, flooding, drip, etc.).  Toxics other than
pesticides, with the exception of spills, are almost entirely deposited
from atmospheric sources.  Deposition may occur in the form of gases,
dissolved in precipitation or adsorbed to particulate matter.  Since the
initial focus of this modeling effort is on pesticides, the modeling of
aerial or ground spray deposition events is addressed in this report rather
than toxicant deposition modeling from atmospheric sources.  Chemigation is
not addressed by the model.
1.3  REPORT ORGANIZATION

    This report is divided into 10 major sections inclusive of this
introduction.  Sections 2 and 3 provide documentation of the content of the
model.  Section 2 is an overview and Section 3 describes model theory,
algorithms, and solutions to equations in detail.

    Sections 4 through 8 form a model user's guide.  Section 4 contains
information on model installation and execution; Section 5, input sequence
development; Section 6, parameter estimation Guidance; Section 7, a
description of model output; and Section 8, some example model
applications.

    Section 9 constitutes the model programmer's guide.  It contains
information on model structure, coding conventions, intermodule
communication and file manipulation.

    Section 10 describes a simplified analytical food chain exposure model
(A-TEEAM).  The discussion includes theory, solution of equations, user's
guidance, and example problems.

-------
                                 SECTION 2

                               MODEL OVERVIEW
    The purpose of this section is to give an overview of the model  in its
current state of development.  This level of detail should suffice for
those readers who are not interested in the more mathematical descriptions
contained in Section 3.  In evaluating the appropriateness of this model
development effort it is important to keep in mind the objectives and scope
discussed in Section 1.  It is anticipated that the model will continue to
evolve.  Therefore, the model has been built in a modular style.  In
addition, "hooks" (that is, interfaces) have been left in places to
facilitate the addition of pieces of software at a later date.  Because of
this, model capabilities may seem incomplete to some reviewers, or an
analyst of the code may notice that certain calculations are performed and
not subsequently used.  It is hoped that the majority of these occurences
have, in fact, been planned, and will not interfere with the application of
the current version of the code.

    This section begins with a brief review of some previous terrestrial
ecosystem-type exposure assessment modeling to indicate the state-of-the-
art.  It then overviews the features and limitations of the current model
and concludes with recommendations for continued development.

2.1  PREVIOUS WORK

    Historically, exposure modeling in support of risk assessment for
organic toxicants has concentrated on aquatic exposure resulting either
from transport from the terrestrial system (e.g., due to runoff) or direct
discharge (e.g., toxics in wastewater effluents).  Notable examples of
modeling procedures which provide this capability include HSPF  (Johanson et
al. 1980), TOXIWASP (Ambrose et al. 1983), EXAMS (Burns et al.  1982), and
SWRRB-EXAMSII (Offutt, personal communication).  The same general approach
for toxics was adopted by OTS with models developed for generalized
multimedia analyses.  UTM-TOX (Patterson et al. 1984) and TOXSCREEN
(Hetrich and McDowell-Boyer 1984) simulate the transport and fate of
toxicants from their sources through the air and their deposition on the
land or through forested watersheds and idealized urban environments,
respectively.  All the above referenced models treat the terrestrial
environment as a source or a sink.  Spatial distribution within the

-------
terrestrial environment and concentration profiles for ecologically
important compartments (e.g., plants and other wildlife food sources)  are
largely ignored in these models.  However, with some modifications, such
models could feasibly have been used as a framework on which to develop a
terrestrial ecosystem exposure assessment model.


    Some limited work has been done in the past several years to address
the problem of estimating chemical exposures in terrestrial ecosystems.  A
program called FEAST (Kiekebusch et al. 1981) was developed to evaluate
exposures to humans by tracing toxicant movement through soil, uptake by
plants and livestock, and food processing.  A similar model, 6EOTOX, was
recently developed which provides the same type of capability, although it
is more refined (McKone and Layton 1986).  GEOTOX is a compartmental,
multimedia model which investigates the fate and transport of organic
toxicants in air, soil, water, and human food products, and evaluates
health effects.  The model solves a set of differential equations (one for
each compartment) either for steady state or time varying applications.
The code also evaluates mass transfer coefficients for the advection and
diffusion processes which are assumed to move contaminants between
compartments.  Process descriptions are kept extremely simple, consistent
with the screening-level application intended for the model.

    Another model called PATHWAY (Kirchner and Whicker 1984) simulates the
transport of radionuclides in agroecosystems and ultimate exposure to
humans.  PATHWAY is a dynamic continuous simulation model and has been
validated with historical data on toxicant fallout and uptake into
vegetation and livestock.  As for FEAST and GEOTOX, the focus of this model
is human exposure as opposed to wildlife.

    In 1984, a modeling effort was undertaken to evaluate exposure of
pesticides to ducks and upland game birds in wheat field ecosystems.  The
model made use of PRZM (the Pesticide Root Zone Model) (Carsel et al. 1984)
to describe the fate and transport of pesticides in the soil and the FSC8G
model (Dumbauld, Bjorklund, and Saterlie 1980) to describe application/
deposition and drift during aerial spraying events.  Elementary soil
arthropod and plant uptake and translocation models were added to PRZM  in
order to simulate pesticide concentrations in the food items of these avian
species.  The approach is fully described in Dean et al. (1984), and
various components are described in Donigian and Dean  (1985).  This work
has been utilized as a building block for the current model development
effort.


     Recommendations  for  additional capabilities needed for simulation of
exposure  in  a generalized terrestrial ecosystem were made  as a result of
the  above mentioned  modeling  effort.  They included:

                                     8

-------
       • A more accurate description of partitioning of pesticide between
         foliage and soil during the spray events, accommodating different
         types of sprayers, leaf and plant geometry

       • A submodel to describe foliar absorption and translocation into
         the plant

       • Simulation of organisms on plant foliage and toxicant uptake by
         these organisms

       • The addition of types of soil organisms other than macroarthropods
         (for example, oligochaetes) and descriptions of organism growth
         and mortality dynamics in lieu of assuming a fixed population

       • Food chain dynamics and biomagnification (prey/predator
         relationships)

    Some of these recommendations have been incorporated as components of
the TEEAM code, described below.

2.2  MODEL FEATURES

    The utility of a piece of software can be judged by its ease of
application in the opinion of the user and its appropriateness to solving
the problem at hand.  The model has been conceived with both of these
factors in mind.  In the following sections, the processes simulated, the
model's spatial and temporal structure, and user-oriented features of the
software are discussed.

2.2.1  Processes Simulated

    In order to realistically simulate exposure to plants and animals in
terrestrial ecosystems, the model contains the following simulation
capabilities:

       • Toxicant application/deposition
       • Soil/atmosphere fate and transport
       • Plant uptake, fate, and translocation
       • Terrestrial food chain bioaccumulation and biomagnification
                                                                    •

     Figures  2.1  and 2.2  show a  schematic overview of the processes which
the  model  simulates.

     Figure 2.1  overviews the processes  involved  in the  soil/plant/
atmosphere system.  The  toxicant  is deposited in an application event, or
to either  plant  or  soil  surfaces.  Drift into non-target areas can be

-------
quantified by using  the FSCBG module.   FSCBG is  an  analytical model  which
simulates spray deposition from an  elevated line source.  Currently,
exposure to animals  (inhalation, dermal  contact,  etc.)  during spray  events
is not  simulated.  Deposition is allowed to occur to various user-specified
levels  within the canopy,  although  this  information is  not used  in the
current version of the  model.  The  application event may be either aerial
or ground spraying,  in  which case drift  to adjacent habitats may be
quantified.  Ground  spraying events are  simulated by manipulating certain
parameters of the aerial  spray model,  as described  in Section 6.   Foliar,
soil surface, and soil  subsurface applications are  also addressed.
Application/deposition  due to toxicant fallout and  chemigation are not
simulated at this time.   Toxicants  may be applied in granular form from
which they may slowly be  released into the soil  or  be consumed directly by
terrestrial animals.  Application in the form of treated seeds may also be
simulated.
  Drift to
  nontarget
  habitats
     •*—
  ( during
   spray
   events )
                t
Diffusion
Atmosphere
                       Volatilization
                     Deposition
                           Plant Surface
                           • Degradation
                        Fauna
                                 Exudation
           Plant
            ( roots, above ground biomass )
                • Translocation
                • Metabolism
                       Ephemeral   Surface  Pond
   Runoff
   Erosion
                                  Infiltration
                                                         Other Fauna / Predators
                                                    Uptake
           Upper  Soil
            Degradation
            Solid / Liquid / Gas •
            Equilibria
                Fauna
   Leaching
Lower Soil
* Degradation
Equilibria

Fauna


          Figure 2.1   Ecosystem components and intercompartmental transfer processes
                     (soil/water/atmosphere/plant  system) simulated by TEEAM.
                                        10

-------
    Atmosphere
     Plant   Surface

         Ingestion	^
Fauna
• metabolism
                  Ingestion
                  Ingestion
Ingestion
      Plant  (roots, above ground biomass )
      Ephemeral  Surface  Pond
                                       Ingestion
                                       Ingesllon
                                                                Inhalation
                               Other  Fauna/  Predators

                                    •  metabolism
                                          Ingestion
                                   Ingestion
   Upper  Soil

      (Water, Soil, Gas)
                            Fauna
                                  metabolism
                                          Absorption
                                  Ingestion
                                                          migration
  Lower  Soil

      (Water, Soil, Gas)
                            Fauna
                                . metabolism
                                           Absorption
                                                      Ingestion
          Figure 2.2  Ecosystem components and intercompartmental  transfer
                    processes (faunal  system) simulated by TEEAM.
    Terrestrial  fate and transport calculations  are made using  a  modified
version of  the PRZM code.  Toxicants deposited on plant surfaces  may be
washed off  during rainfall events and moved  to the soil surface,  returned
to the atmosphere via volatilization, or degraded on the plant  surface.
Foliar absorption and subsequent translocation throughout the plant are not
currently simulated.

    Toxicants  on the soil surface or in the  soil  may be present in
adsorbed, dissolved, or gaseous phases.  Phase concentrations are
determined  assuming equilibrium among phases  and  making use of  adsorption
partition coefficients and Henry's Law constants.  Toxicants  in the upper
soil may subsequently be leached into the  lower  soil, degraded, taken up by
plants, returned to the atmosphere via volatilization, or lost  to the
system through erosion by water and runoff.   Similar processes  operate in
the lower soil profile (below the root zone)  with the exception of plant
uptake.  Throughout the soil, transport occurs due to advection,

                                       11

-------
dispersion, and liquid and gas phase diffusion.  Following precipitation
events, water may be temporarily ponded at the soil surface.  Various
chemical fate processes may act upon the chemical within the pond,
including volatilization, degradation, and infiltration into the soil.

    An adaptation of the plant growth module from the EPIC model  is used
for simulation of plant growth.  Plant growth affects the  concentration of
pesticide within the plant and micrometeorological  processes which control
spray deposition and volatilization.  Plant uptake  is controlled  by the
rate of evapotranspiration, chemical adsorption in  soil,  and reflection at
the root surface.

    Within the plant, toxicants are partitioned between roots and
aboveground biomass and may be degraded while traveling through the
plant.  Advection of the chemical through the plant is modified via
partitioning into the nonaqueous phase.  Chemicals  which travel completely
through the plant are exuded back into the atmosphere.  Currently, this
exuded quantity exits the ecosystem and does not contribute to
concentrations within the plant canopy.   In the atmosphere within the
plant canopy, various degradation processes may serve to diminish the
chemical mass.  In this model, the chemical mass is removed from the system
only by diffusion into the atmosphere above the plant canopy height.

    The soil, water, air, and plants in the terrestrial system provide
vectors for uptake of pesticides by animals.  TEEAM computes the dosages
to, as well as concentration in, the biomass of the modeled species.  The
components of the fauna! system and associated transfer processes are shown
in Figure 2.2.  In general, animals may be present in ar on the soil and
depending upon their physiological capabilities and behavior, they may
migrate between various layers in the soil.  They may also migrate between
ecosystem habitats as described in the next section.  Soil fauna may feed
on organic detritus, microbes  (bacteria) in the lower trophic levels and on
earthworms, arthropods, crustaceans, molluscs, and the like in higher
trophic levels.  Animals may also consume living above ground plant
biomass.  They may also absorb chemicals from the soil system,
bioconcentrating them in their tissues.  Animal mortality and the return of
toxicants to the soil in this manner is not simulated.  Biomass levels of
each of the modeled species are assumed to be fixed for a given
simulation.  It is also the case that the uptake terms in the animal
exposure model do not have equivalent sink terms in the terrestrial
model.  Therefore, the effects of removal of pesticide by animals from the
terrestrial system are not accounted for.  This should not be a major
problem as the amount of chemical residing in the animal biomass is
probably small compared to the total amount of chemical in the system.
                                     12

-------
    Higher trophic level predators are also considered.  These would  be  the
species of interest to the exposure assessment and may ultimately  include
various higher level terrestrial herbivores, carnivores, and  avians.
However, as stated earlier, the model is focused at this time on avians.

    Mass balances as calculated within the terrestrial fate and transport
module, the plant translocation module, and the animal exposure module.
Currently, a global mass balance calculation is not performed.  However,
simulations using the TEEAM code match reasonably well with ATEEAM (See
Section 10) simulations, indicating that mass transfers between the
terrestrial and animal exposure modules are handled correctly.

    The above discussion describes the processes which occur at a single
point in the ecosystem.  However, due to heterogeneity of the area which
this species inhabits, the model must be capable of simulating subareas
(habitats) of various types.  The model's spatial features are described
below.

2.2.2  Model Spatial Features

    The spatial structure of the model is a collection of habitats, having
a one-dimensional vertical structure as depicted in Figure 2.3.  The
modeled ecosystem is broken into these habitats because there is some
uniqueness about them in the real ecosystem.  For instance, habitat 1 may
be the agricultural field to which the chemical is applied which may also
serve as a feeding habitat for the species of interest.  Habitat 2 might be
the nesting habitat of a species that feeds primarily in Habitat 1.
Habitat 3, on the other hand, might be an area which received drift during
the chemical application event, is visited infrequently by the species of
interest, and therefore is different, in terms of exposure potential, from
habitats 1 or 2.

    In general, habitats are differentiated because there are:

       • Distinct differences in plant or animal  species in various parts
         of the ecosystem

       • Distinct differences in behavior and/or feeding preferences when
         various animal species are in these various habitats

       • Distinct differences in pesticide or toxicant distribution among
         habitats

Within each habitat, conditions are assumed to be laterally homogeneous.
The effect of minor deviations from this assumption can be addressed through
the use of Monte Carlo simulation.   It is possible to transfer mass between
compartments to account for the effects of wind and animal  migration;

                                     13

-------
ATMOSPHERE
      SOIL
                 Figure 2.3  Spatial structure of terrestrial exposure model,
  however,  the  dependence of concentrations In a given compartment on the
  processes occurring in an adjacent compartment is assumed to be weak.  Each
  compartment has  a vertical structure, the discretization of which depends
  upon  the  dynamics of the processes being simulated.  Concentrations in the
  profile are considered to be strongly dependent upon processes which
  transfer  mass vertically (e.g., volatilization, deposition, percolation of
  water through the soil, etc.), hence the greater detail in the vertical
  dimension. As many habitats may be used as are necessary to describe the
  ecosystem, within the limits of practicality.  It is likely that a greater
  knowledge of  the ecosystem will result in greater compartmentalization.
  Each  habitat  specified, however, requires a corresponding application of the
  terrestrial fate model.  Since lateral transfers between habitats are
  assumed to be of relatively minor importance, the order of execution of
  computations  among habitats, is also of little importance.  Exact
  directional relationships between compartments are required by the

                                       14

-------
deposition model to simulate pesticide application events.  However, for the
operation of the remainder of the modules, exact spatial relationships need
not be known.

    Animal species are allowed to migrate among habitats.  Species in lower
trophic levels may be permanent habitat residents, depending upon the area
of the habitat selected.  As with the differentiation of habitats, as many
trophic levels and as many species within each trophic level may be used, as
is necessary, to represent the food chain.  A schematic of a food chain is
shown in Figure 2.4.  The food chain also includes plants, soil, and
water.  The species of primary interest is that in the highest trophic level
(1.1).

    The boxes in Figure 2.4 represent a single species of interest.  In
terms of model implementation, however, each box may represent an
individual, a group of individuals, or the entire species population within
the ecosystem.  Thus, each box may be subdivided as necessary to simulate
variations among species.  For instance, species (3,1) may be present
ubiquitously in the soil and relatively immobile.  If the ecosystem is
spatially segregated into habitats, several subgroups which represent the
portion of the community in each habitat would be required.  For a species
in a higher trophic level, (i.e., species 2,1) a group of several
individuals might be simulated. These individuals may or may not be
constrained to specific spatial habitats, depending upon their mobility.
The range of each species is denoted by specifying habitats that each
species is allowed to visit.  Thus, the model spatial structure and the
trophic structure are integrated.  Even though individuals or subpopulations
may have the same behavioral tendencies (on the average), a stochastic
modeling approach to animal movement and feeding incorporated  into the model
can be utilized to yield estimates of variability of dosage or body burden
of toxicant caused by random movement.

2.2.3  Model Temporal Features

    There are several important time-related features of the model  which
must be pointed out.  The model operates in a daily time step.   The length
of simulation is flexible and can be selected by the model  user.   For short-
lived chemicals which do not accumulate to any great extent in animal
tissues, the magnitude of single exposure events to the species of interest
is of greatest importance.  For less acutely toxic, more lipophilic
chemicals, the body burden over the life cycle of the species  of interest is
of greatest importance.  In these cases, the simulation could  be as short as
a single exposure period (days to weeks) or the species life-cycle (months
to years).  In other cases, the interest may be in carryover of pesticide
from generation-to-generation within the species and/or long-term effects on
population levels.  In these scenarios, simulation periods of  many years may


                                     15

-------
be of interest.  The focus of this model  is currently  short-term lethal
exposures; however, longer term simulations may  be  run.   It  is  recommended
that relatively short periods of time be  simulated  due to the  lack  of a
comprehensive representation of seasonal  effects on animals  (e.g.,  behavior,
metabolic rates).

    Aside from overall simulation length  and  time step there are other time
scale phenomena which are also important.  For  instance,  there  may  be
distinct seasonal or diurnal behavioral patterns in the species of
interest.  In addition, there may be distinct behavioral  differences which
are a function of life-cycle stage in the species of interest.   Animal
behavior (e.g., feeding rates, food preference,  movement) is functionally
described in the model and governed by  sets of  user input parameters.  For
instance, animal movement is simulated  using  a  single  lag Markov
(autoregressive) process.  The process  is parameterized by the  use  of a
transition matrix.  These such functions  themselves do not have seasonal or
life-stage effects built into them directly and, currently,  there is no way
of changing input values to account for seasonal effects.
           0.
           o
                                                      Denotes a
                                                      predatory
                                                      relationship
                                    Species

                   Figure 2.4 Schematic of an  ecosystem food chain.
                                     16

-------
2.2.4  Software Features

    Two significant features have been incorporated into the TEEAM model:
1) the ability to simulate multiple habitats, allowing for the simulation of
wind drift of an aerial spray applied to one habitat onto an adjacent
habitat as well as animal movement among habitats, and 2) a Monte Carlo
driver which permits the identification of input parameter uncertainty and
which allows the quantification of the uncertainty in  selected  output
parameters.

    The TEEAM execution software has been designed for the software user.
Extensive error checking and reporting has been incorporated into the
software.  Most errors are trapped within the software, and an  error message
displayed to the screen explaining what the error was  and where it
occurred.  If the error is fatal, the program is halted with all  files
closed permitting the user to examine the files containing input  data echo
for potentially inappropriate data values.  The TEEAM  software  also has a
'TRACE' option which displays which subroutine the software is  currently
executing.  Thus, if an error does occur, the user will be able to determine
the subroutine in which it occurred and the path to that subroutine.

    The TEEAM software has been designed to be compatible with  the IBM PC
compatible microcomputer environment.  IBM PC specific features include a
screen generator which displays the status of the simulation and  time series
output files which can be directly imported into popular spreadsheet
programs for graphing and analysis.  While the software has been  designed
with microcomputer applications in mind, relatively few changes should be
required for adaptation to alternate computer systems.  INCLUDE files are
used to define array sizes, file unit numbers, and other potentially
compiler/computer system-specific parameters.  Most,  if not all,  changes
necessary for adaption to alternate computer systems can be accomplished by
modifying these INCLUDE files without modifying the FORTRAN code  itself.

    TEEAM was formed by linking three preexisting computer codes, the
Pesticide Root Zone Model  (PRZM) (Carsel et al. 1984), and FSCBG  (Dumbauld
et al. 1980), and plant growth model in EPIC (Williams et al.  1988) to code
that was written specifically for this application.  As a result, different
programming conventions and styles are apparent in the FORTRAN  Code.  The
conventions and style of PRZM and the code written for this project are
similar.  The most significant contrast in programming conventions and
styles are evident between FSCBG and the remainder of  the code.  No attenpt
has been made to either restructure or make cosmetic changes to FSCBG.
Neither has any attempt been made to unify input and output styles or
formats.
                                     17

-------
    This nonuniformity has both advantages and disadvantages.  The
advantages are that for users familiar with the preexisting versions of the
individual codes, the similarities will be comforting.  It is also more
apparent, when one is inputting data or reading output, the module to which
the input or output belong.  Disadvantages include some redundancy,  at
times, in input, calculations and/or output, and attendant difficulties in
interpreting output or debugging/modifying code.

2.3  RECOMMENDATIONS FOR FURTHER WORK

    The recommendations for further work with the TEEAM code fall into two
categories:

       • Additions to existing code and
       • Parameter estimation and model verification

2.3.1  Additions to Existing Code

    Some additions to the code would augment its ability to realistically
simulate toxicant exposures to wildlife.  The first recommendation would be
to bring the simulation of the animal populations to the same level  as that
of the plant biomass model.  Thus, population models for the species of
interest and seasonal effects on animal growth, metabolism, and behavior
could be added.

    A second recommendation would be to enhance the animal exposure model to
simulate chemical fate and transport in various organs of plants and
animals.  For plants, this might include roots, stems, leaves, seeds, fruit,
etc. and for animals, distribution into brain, liver, etc.  Along with this
might be added other exposure routes.  For instance, foliar absorption and
translocation could be added for plants, and percutaneous absorption for
animals.  Although the capability exists within the model to simulate
inhalation exposure during spray events, this is not currently available.
The adaptations should be made to gain advantage of this capability.

    A third category of additions would be the simulation of the effects of
the toxicant on the ecosystem.  For  instance, the effects of growth
inhibition of herbicides on certain  plants or the effects on seed production
might be of concern.  Simulation of  avoidance/attraction effects of
pesticides on food ingestion by animals would significantly improve the
capability to accurately simulate exposure.


    Two additional extensions would  enhance the applicability of the
model.  One would be the addition of algorithms to  simulate the  loading of
atmospheric toxicants to the terrestrial system.  Currently only discrete
spray or application "events" can be simulated.  A  second would  be the

                                     18

-------
extension to animals other than avians.  This would chiefly be done by
giving appropriate parameter guidance for the generalized movement and
feeding algorithms already in the model.

2.3.2  Parameter Estimation and Model Verification

    The third category of recommendations is in the area of validating and
improving, as necessary, the accuracy of the model.  Two ways of doing this
are improving estimates of input parameters and verifying the model
algorithms.  This could be done by using a short term (weeks to months)
field trial for the current model, looking at acute toxic effects and
toxicant concentrations in various media over time.  System data would be
used to improve parameter estimates through calibration and identify
inappropriate process algorithms.  It would also be appropriate to
parameterize a number of preselected sets of ecosystems in order to
facilitate the application of the model for regulatory purposes.
                                    19

-------
                                  SECTION  3

                         TEEAM MODULES AND PROCESSES
    This section describes the TEEAM modules and details the processes which
are simulated by each.  Presently, there are six computational  modules:

       • Toxicant application/deposition
       • Soil/atmosphere terrestrial fate and transport
       • Plant growth
       • Plant contaminant transport
       • Terrestrial animal exposure
       • Monte Carlo simulation

3.1  TOXICANT APPLICATION/DEPOSITION SUBMODEL

3.1.1  Introduction

    A prerequisite to determining the fate of pesticides in the environment
and the resulting effects on the ecosystem is the ability to realistically
characterize the processes which bring them into the environment at the
point of application.  This requires knowledge of the methods of application
as well as the subsequent transport and deposition processes.  The
application of pesticides may result in exposure both during and after the
spray event.  During the event, high concentrations of sprayed material may
be inhaled by animal species.  Therefore, it is important to predict these
concentrations to determine overall exposure.  The exposures which occur
after the event result from deposition which has occurred to soil or
vegetation either within or out of the targeted spraying area.  Therefore,
both on- and off-site deposition rates are calculated and provided as  input
to the exposure models.

    A variety of application/deposition scenarios are possible.  These have
been discussed under "Chemical Releases" in Section 1.2.4.  In this version
of the model, spray application events are assumed to be of central
importance.  Pesticide spraying methods can be divided into two main
categories:  (1) aerial applications, and (2) ground-based spray application
methods.  The aerial spraying technique is widely used and has the greatest
potential for off-target drift of any of the common techniques.  Off-site

                                    20

-------
drift is a process of key interest to the Agency.  This discussion of the
aerial spraying model is followed by a discussion of techniques to simulate
ground-based spraying and user options for other types of applications.

3.1.1.1  Aerial Spraying--
    Aerial spraying of pesticides, herbicides, and fungicides is commonly
used to control insects, weeds, and plant or tree diseases in forests and
crop lands.  Aerial spraying models simulate the behavior of sprayed
material from the time it is released from the aircraft until it has been
deposited on the soil and/or vegetation.  These models should also be able
to simulate the concentration of the sprayed material during application;
the extent of coverage and the drift beyond targeted areas; and the
deposition of the material above, within, and below vegetative canopies.
The processes and most significant factors which affect the fate of aerially
sprayed material are discussed in the following paragraphs.

    Five groups of factors can be recognized which will contribute to the
transport and ultimate fate of chemicals applied aerially.  These are:

    1) Spray system characteristics
    2) Vegetation characteristics
    3) Spray depletion characteristics
    4) Target characteristics
    5) Meteorological factors

These factors are elaborated upon below.

    Spray system characteristics—Important characteristics of the spray
system include:

       • Aircraft weight, speed, wing span, aspect ratio, etc.
       • Application rate, swath width, release altitude
       • Source dimensions, spray boom and nozzle type
       • Physical and chemical properties of spray material:  drop-size
         distribution, density, and volatility
       • Emission location

    Observations of aerial spraying show that the drops emitted from
aircraft spray nozzles or tanks are quickly swept into the wakes of fixed
wing or rotary aircraft.  These drops are entrained into the aircraft-
generated vortex system which, in general, sinks below the aircraft, grows
in size, and decays in intensity as time elapses.  The position, size, and
shape of the vortex are controlled by the physical characteristics of the
aircraft.
                                    21

-------
    During the first minutes after the spray release,  the  wake  vortices
principally control the growth of the spray cloud.   Except for  the  lateral
translation of the vortex system by a crosswind,  the vortices also  control
the position in space of the spray cloud.

    After the first few minutes, when vortex circulations  have  decayed,
meteorological factors in conjunction with gravitational  settling govern the
transport, diffusion, and deposition of the spray.   If the aircraft altitude
is less than about 1.5 to 2.0 wingspans above the deposition surface and the
spray drops have sufficient settling velocities,  the effective  swath width
and deposition of the spray is largely governed by the descent, growth,  and
subsequent decay of the vortex wake system  generated by the aircraft.
Thus, for aircraft spraying at higher altitudes,  simple procedures  appear to
be sufficient in describing the vortex sink rate.

    Emission rates and the location and duration of emission must  obviously
be considered for accurate prediction of the fate of the sprayed material.

    Vegetation characteristics—The important vegetation characteristics to
be considered include:

       • Type of vegetation and dimensions, especially height
       • Foliage density
       • Spatial distribution

    The presence or lack of vegetative cover is an important factor to be
considered.  If a vegetative canopy (i.e., forest canopy or tall vegetation)
exists, its characteristics play an important role in determining  the wind,
thermal, and turbulence structures within the canopy and its effects on
evaporation and impaction of spray on the target area.

    If it is desired to determine the spatial distribution of deposited
spray material within and below a canopy, methods which account for the
canopy effects are needed.  A canopy penetration model can be used to
accomplish this.  The canopy penetration model should be able to calculate
the percentage of material which, after entering the top of the canopy, is
retained at various levels within the canopy and on the soil surface
underneath.  Vegetative factors that affect the transport and deposition
within and below the canopy include the spacing and type of foliage,  height
of canopy and coverage of the ground surface.

    Spray depletion characteristics—Three  factors are of major importance
 in determining  deposition:

        • Gravitational  settling
        • Evaporation
        • Impaction
                                    22

-------
    Gravitational settling acts as a depletion mechanism in drift
calculations, and is the major mechanism ensuring effective canopy
penetration.  Spray dispersion models should account for gravitational
settling in a manner that is consistent with conservation of mass and still
be able to permit turbulent dispersion and partial reflection at the ground
surface or canopy top of smaller drops.

    Evaporation can significantly alter the airborne aerosol drop-size
distribution as the spray cloud descends from the aircraft.  The net effect
of evaporation is to reduce the size of the drops in the cloud,  thereby
reducing gravitational settling velocities and deposition near the source
and thus increasing downwind drift from the spray cloud.

    The amount of spray reaching a given height within a canopy, following
canopy penetration, depends upon drop impaction losses at higher levels
within the canopy.  These impaction losses depend on the drop size and the
collection efficiency of the vegetation through which the drops are falling.

    Target characteristics—Important target characteristics to be
considered include:

       • Geographic location and dimensions of spray area
       • Topography and other surface features

    The physical dimensions of the target, the topography and surface
features in the target area are all important features that should be
considered in the model formulation and in the selection of model input
parameters.  Other data, such as geographic location and total size of the
spray block, must be known to specify spray line lengths and number of
swaths required.


    Meteorological factors—Important meteorological factors to be
considered include:

       • Vertical profiles of turbulence, wind speed and direction,
         temperature, and humidity
       • Insolation
       • Mixing depth

    Deposition of sprayed material depends on the specification of
meteorological conditions that exist during the spray release period (i.e.,
atmospheric stability).  Parameters which reflect these conditions include
vertical profiles of wind speed and temperature.  After vortex effects have
dissipated, wind speeds are required to calculate the time for drops of
various sizes to reach the canopy and the flux of drops at any distance from
                                  •23

-------
the flight line.   Wind speed profiles within the canopy may also be required
to calculate the  vertical trajectory of drops and the depletion by
deposition and impaction within the canopy.

    Temperature and relative humidity are needed if evaporation of the drops
is to be considered either above or below the top of the canopy.

    Estimates of  vertical and lateral turbulence above the canopy are needed
to calculate vertical and lateral dispersion of the spray cloud for drift
calculations.  In addition, knowledge of the depth of the surface mixing
layer is important in estimating spray drift levels at longer distances from
the flight line.

3.1.1.2  Ground-Based Spraying—
    Ground-based  applications of pesticides, herbicides, fumigants, and
fungicides are routinely used to control weeds, insect damage to plants, and
plant or tree diseases in forests and crop lands.  There is a wide spectrum
of application scenarios with ground-based application equipment.  Typical
application equipment includes disc sprayers, hydraulic sprayers, air
sprayers and mist blowers, and fog applicators or aerosol generators.  These
application devices may be carried by hand,  mounted on trucks, drawn by
tractor, or incorporated in mobile units such as self-propelled sprayers.
Application rates associated with spraying equipment are generally
classified as high, low, or ultra-low volume.

    Spraying recommendations vary with crop type, and may even vary for a
particular crop.   For example, on one crop,  two different systems may be
necessary.  Disease (fungus) control might require spraying pressures of 200
to 400 psi and an application rate of 75 to 150 gallons per acre
(gal acre  ), and weed control may require spraying pressures of 40 to 80
psi and an application rate of 10 to 20 gal  acre'1.  Due to such diversity,
development of a universally applicable ground spray simulation model, or
even a set of models, to encompass all ground spraying situations is
difficult.

    The five groups of factors identified and discussed  in the previous
section on aerial spraying also apply to ground-based spraying methods.
Obviously, the major difference between the  two is that  aircraft wake
effects are no longer of concern.  Basically, only the spray system
characteristics are different and are listed below as they apply to ground
spraying.

    Spray system  characteristics—Important  elements of  ground spray systems
include:

       • Ground sprayer, speed, and other source characteristics
       • Application rate, swath width, release height,  and coverage

                                     24

-------
       • Source dimensions,  spray boom and nozzle type
       • Physical  and chemical  properties of spray material:   drop-size
         distribution, density, and volatility
       • Emission  location

3.1.2  Module Development

3.1.2.1  Background--
    Previous discussions (Section 3.1.1.1 and Section 3.1.1.2) have
identified processes and five major groups of factors which must be
quantified in order to accurately describe the fate of sprayed material.
The need to simulate these processes formed a set of criteria by which
existing models were evaluated and selected.  A literature review was
conducted to identify both aerial spraying and ground spraying models which
could be used to simulate the processes described above.  The FSCBG model
(Dumbauld et al. 1980) was chosen to provide the capability to simulate both
aerial and ground  spray operations.

    The model is applicable for modeling spray emissions from fixed wing
aircraft and helicopters having high forward speeds.  The types of ground-
spraying systems which produce spray clouds that can be potentially modeled
as line sources include those methods in which the spray is delivered above
or near the top of the plant canopy, if one exists, and where the spray is
directed vertically, or nearly so, downward.  Sprays that are directed
upward or have significant velocities above the horizontal plane are less
likely to be successfully modeled using the elevated line-source model.

    Examples of ground-spraying equipment that can be modeled as line
sources include:

        •  Some  truck-mounted  and  tractor-drawn spraying  systems
        •  Self-propelled,  high-clearance  sprayers
        •  Mobile boom sprayers

     For those  situations  where the  line-source model cannot be used, either
 because sufficient  data  are  not  available to adequately characterize the
 source  or because the spraying method being modelled is not physically
 represented  by a  line source (e.g., an orchard air  sprayer),  simplistic
 ground  spray models  must  be  used.   This  is a result of  the lack of more
 mechanistic  models  found  in  the  literature.

     Even  though the  version  of FSCBG used has a canopy  penetration model,  it
 has  not been included in  this  simulation package.   Its  drawbacks include
 extensive  input requirements,  the need to determine parameters of the plant
 or tree foliage (which may be  known for  various types of trees but are not
 well  defined for the wide variety of vegetative canopies expected to be
 encountered),  and the execution  time required for the simulation of a large
 number  of  drops.
                                    25

-------
    The canopy penetration method selected for use in TEEAM is based on
theoretical  and experimental  work described by Uk and Courshee (1982) and
Bache and Uk (1975).  This method is based on an exponential decrease of the
deposition density of the spray material with depth from the exposed canopy
top.  This simplified canopy penetration model was developed for optional
use with either the aerial or ground spray models.

3.1.2.2  Description of the Spray Application/Deposition Model--
    The following components make up the spray/deposition portions of TEEAM:

       • An aerial spray deposition model (FSCBG)
       • A ground spray model (FSCBG minus aircraft wake effects,
         essentially, an elevated line-source model)
       • An alternative ground-based application model for cases in which
         the elevated line-source model is inappropriate
       • A simplified canopy penetration model

The theoretical bases of these components are described in the following
sections.
    Aerial spray model (FSCBG)--The FSCBG model as implemented for use with
the TEEAM model is described in this section.  The discussions of FSCBG are
taken primarily from Dumbauld et al. (1980).  The major processes involved
are simulation of aircraft wake effects, followed by the simulation of
growth and gravitational  settling of the spray cloud.   The  dosage  and
deposition calculations provided in FSCBG have been  incorporated along with
the drop evaporation algorithms.  These are described  in  the  following
subsections.

    Wake settling—The version of the FSCBG code  described  in this report
    contains  a simple model.  Prandtl and Tietjens (1934) express  the  sink
    rate w(m  s'1) of a vortex system as
                   8g W         ,
           u  =   	 *	 (lo'-3)                                    (3-1)
               TV  p. b  V
                  ~ A     -a
                   r\     u

    where

    Wa = weight of the aircraft (kg)
     g = acceleration due to gravity (9.8 m s  )
    p. = air  density (g cm  )
     b = aircraft wing span (m)
    Va = aircraft speed (m s  )

    Equation  (3-1) is strictly applicable to fixed-wing aircraft;  however,
    helicopter-produced vortices resemble fixed-wing vorticas at high

                                     26

-------
forward speeds  (Dumbauld et al.  1980,  p. 33) and thus  Equation (3-1)  is
applicable  to  high speed helicopters.

In the spray dispersion models,  the source parameters  correspond to the
spray cloud properties when the  cloud  has reached  approximate
atmospheric equilibrium.  The  vortex sink rate calculated from Equation
(3-1) and the  observations that  the vortex tube stops  descending at a
distance b/2 above the ground  surface  (or above the  canopy) are used  to
calculate source parameters for  the models.  Figure  3.1 is a schematic
diagram showing the geometrical  considerations used  in specifying the
distance downwind XR at which  cloud stabilization  occurs and the
effective release height Hj.   As shown in Figure 3,1,  the cloud descent
from the aircraft at height H  forms the angle tan'1  (w/u) with the
horizontal  when the gravitational  settling velocity  for the j   drop-
size category  at height H(Vj.H)  is less than w.   In  general, the FSCBG
program calculates the distance  downwind from the  flight path where the
cloud reaches  a height b/2 above the surface from  the  expression
        XR = QtJ
                                                       (3-2)
where
                 z
                 A
                                    WIND DIRECTION
  2«0
             H1
T"
£
v7
                     ^___             ^
                                	I_   \ u  '
     /////////
                                                    Source. Dumbuald, Rafferty and Bjorklund, 1977
             Figure 3.1. Schematic diagram showing geometry used in constructing the distance xr and
                    effective source height H' for the case when the settling velocity in the jth size
                    category at height H(Vj;H) is less than 
-------
t1.: = effective time for the cloud centroid to reach the height b/2
  J      [H - (b/2) - Hc]/« ; a, > Vj;H

         tj;b/2             ; « < V.;H                            (3-3)

VJ.H = gravitational settling velocity for the j   drop-size category at
       the aircraft flight altitude

       = time for the drop in the j   category to reach the height b/2
         above the ground or canopy, calculated from the evaporation
         model

As discussed in more detail in the description of the spray dispersion
models, the dispersion models are formulated under the assumption that
the cloud axis descends at an angle tan   (V,-/u) with the horizontal.
It is therefore assumed that the effective Declination of the cloud axis
downwind from an effective source height H- is given by the angle tan"
            where Vj-b/2 is the gravitational settling velocity
calculated by the program for the drop at £he height b/2.  Based on this
assumption, the effective release height Hj is

       H; - Hc + (b/2) + t: V..b/2                                (3-4)

All gravitational settling velocities are calculated in the FSCBG
program using a technique suggested by McDonald (1960).  The graph given
by McDonald expressing the relationship between the drag coefficient of
spheres and their Reynolds number has been fitted for Reynolds numbers
less than 2x10  and is used in the FSCBG program.

Spray dispersion model s--In this section the models used in the FSCBG
program to calculate dosage and deposition downwind from nearly
instantaneous elevated line sources oriented at an arbitrary angle with
respect to the mean wind direction are described.  The axis of the spray
cloud is assumed to be inclined from the horizontal plane by an angle
that is proportional to Vj/u, where V,- is the gravitational settling
velocity for the j   drop-size category and u is the mean cloud
transport speed.  When evaporation is negligible, this inclination angle
is invariant with distance from the line source.  For evaporating drops,
the angle changes with distance from the source because V- depends on
the drop size.  It is assumed that drops dispersed upwards by turbulence
are totally reflected downwards at the top of the surface mixing layer,
but the fraction of drops reflected at the ground is a variable input
parameter for each jth category.  The models use the Cartesian
coordinate system shown in Figure 3.2 for a line source of length L at a
height H1 and a calculation point at R(e, 6, z).
                                  28

-------
                                Source: Oumbuild. Rilferty ind Blorklund. 1977
              Figure 3.2. Schematic plan view showing the line source geometry with respect
                     to a calculation point at R (e, 6, z) for a wind direction 8.
    Dosage  model— Here dosage  is  the integral of the concentration  in
the atmosphere within the spray cloud over time.  Thus it represents  a
time-averaged  concentration which can be used (together with inhalation
rates) to model inhalation exposure to species of concern.  In the
current version of TEEAM, inhalation exposures during spray events  are
not computed.   The dosage (mass x time/unit volume)  above the canopy  or
soil surface  is the sum of contributions from the drops in the dis-
persing cloud  and from vapor produced by drop evaporation.
    The dosage  for the contribution from drops is given by the
expression
                     J               ,        1/2        Q  2
                     I  fj  { I    [YJ   (-^-}   {exp  (-ff-}
                    j=l     i=0
{erf (Y1/2  (^^ )) -

       1/2       D2
     kaAsine  u
                                                             - P}     (3-5)
       L)1/2

       1/2
                                      (  +   ))- erf  (T
1/2
                                                                  r))}]J
                                   29

-------
where


S = Qk/2TrL                                                         (3-6)
                                                                   (3~8)

p = [-2            2
     9
     2
            7
       1/2 k V
    7

       1/2
      2    [-   (D + 2) + (n -     ) cote]                      (3-11)
       M     U


    72l
                                                                 (3-13)
     1/2  k V
X - ITS- M  - (n-^) cote]                        (3-15)
     M     U

C = 2iHm - HJ - (V.,xv/u)                                         (3-16)


D = 21Hm + H^ + (V..xv/u)                                         (3-17)


a = 21/2o^ (x1  + xv - i sine)                                    (3-18)


b = 21/2o^ (x1  + xv)                                             (3-19)


n = (x1  + xv) cote + x'tane = (e/sine)+(6/cose) + xvcote         (3-20)

 i
x  = (e + 6 tane) cose = e cose + 6 sine
                                   30

-------
x  = virtual  distance                                            (3-21)
                fo
         ~ XR = o  ~ XR
j, = effective line length                                        (3-22)
      6 + e COte  ;6 -I- e COt9L
    The following parameters used in the preceding equations are based
on meteorological measurements or inferred from meteorological
measurements:
    al  =  standard deviation of the wind azimuth angle in
         radians
    a£  =  standard deviation of the wind elevation angle in
         radians
     k  =  constant relating oi and OF                              (3-23)
       	   I /  I              "      ^

    Hm  =  depth of the  surface mixing layer below a capping  inversion
     u  =  mean transport wind speed above the canopy
    9  =  angle between a  line perpendicular to the line
         source and  the mean wind direction (see Figure 3.2)
    AU  =  vertical wind-speed shear
    The  following parameters are source  inputs required  for use  in  the
model:

   Q  =  total source strength emitted along the  length
         L of the line source
   H  =  aircraft flight altitude above the ground
   V. =  gravitational settling velocity  for the  median
                             J.U
         drop by mass in the j   drop-size category
   f. =  fraction of the total source strength in the
         th
         j   drop-size category
   YJ =  reflection coefficient for the median drop by
                     th
         mass in the j   drop-size category
   a  =  standard deviation of the cloud  distribution at
         the distance XR
                                   31

-------
   L  = length of the line source


    As noted above, the vapor emitted from evaporating drops also
contributes to the total dosage above the canopy or soil surface.  In
the program, the evaporation model  determines the mass of vapor emitted
in Ah height units along the cloud  axis as it descends to the canopy.
The distance downwind from the line source where the vapor is emitted is
given by ut,  where t is the time (from the evaporation model)  when the
cloud axis passes through the midpoint of the height interval Ah.  The
source dimension of the vapor cloud emitted over the Ah height interval
is
   .„

under the assumption that the vapor is rectangularly distributed at the
point of emission.  The vapor dosage is then calculated from
Equation (3-5) with Vj set equal to zero, Y- set to equal  unity,
and Xn set equal to ut.  Finally, the total  dosage at the  calculation
point is determined in the program by summing the contributions from the
drops and vapor clouds.

    Deposition model— In TEEAM, FSCBG is used to compute the deposition
of chemicals to the top of canopy or bare soil surface if  no canopy
exists.  The deposition, expressed in units  of mass per unit area, at
the point (e, 6, 0) downwind from line sources at an angle 0 with the
mean wind direction is given by the expression

                                 2
                   1/91   r
       - exp [ - (F1/2 (1 -   ))
                                              4FJ/

       [erf (F1/2 (I-fp)) -erf (F1/2 (^ - fp))]

                   C exp J_ _ p
         1 /9  1    r   9
[exp [-(F1/* (± - fp))2]



                         (3-25)
                                      GBu1/2 exp
                                    JCTt1/2 exp (^r - P)
                                     32

-------

                                  1/O1    I/           1/11   I/
                            \~~c(c*-/£ fi   K. ^n   n~f(c*-/£ (*•   K ">ini
      ^j 	^372-lerf^E    ^b-2E^-erf^E    ^-2E^1}}


                                                                 (3-26)

    ,1/2   V.Bk?
G = -^T— (-^ + (n - -^]  cote]                               (3-27)
     °A      u

F - k¥ +fn - -!-)2                                            (3-28)
               0
      1/2  V Ck
J = ^- [-L- + (n - ^ cote]                               (3-29)
                                                                 <3-30'
     1/2  V Dk
            r- + (n - 55) cote]                               (3-31)
      M     U

                                                                 <3-32>
and the remaining parameters, except a and b, are identical to those
defined for the dosage model.  The definitions of a and b for the dosage
model given by Equations (3-18) and (3-19) are reversed in the case of
the deposition model, i.e.,

   a = 21/Z  a^(x' + xv)                                         (3-33)

   b = 21/2  oj^x1 + xv - Jtsino)                                 (3-34)

Spray droplet evaporation--The FSCBG computer program allows the user to
account for the effects of the evaporation of drops in the calculation
of dosage and deposition.  The first step in the calculation of
evaporation effects is the specification of the time-rate change of the
drop diameter for the initial j drop-size categories.  The program user
has the option of selecting an automated procedure for the theoretical
calculation of the time-rate change in drop diameter (D,-), or of
supplying values of the constants A, B, and C in the quadratic equation

     D. = A. + B.t + C.t2                                        (3-35)
      J    J    J     J

                                 33

-------
specifying the change of drop diameter with time t after release.
The theoretical calculation of the time-rate  change of  drop  diameter in
the FSCBG program is based on the expression


     °f,j = °i,j + f"                                         <3-36>

where

D-  .  =  final diameter of the median drop in the j   drop-size
  'J     category after the time increment At
D.  .  =  initial diameter of the median drop  in the  j   drop-size
  'J     category
Frossling (see Fuchs 1959, p. 44) defines the change  in diameter of
water drops due to evaporation by

   jJDj     (4xl08) Mfc D  pA(es-eJ  _                           (3-37)
    dt   ~      Mm D. pn (Pfl-eJfv
where            m  J  «• l A  *}

M   =  molecular weight of evaporating vapor  from the drop (g mol~ )
Mm  =  mean molecular weight of the resulting vapor-air mixture in the
       transfer path which is approximated by that of air (MA) in the
       FSCBG program (g moT1)
D   =  molecular diffusivity of the evaporating vapor in air at the drop
       temperature (cm  s~ )

D.  =  drop diameter (ym)
 J                      _3
p.  =  air density (g cm  )
                                        3
p   =  density of the liquid drop (g cm  )
 Xf
e   =  partial pressure of the evaporating vapor at the drop surface (mb)
em  =  partial pressure of the evaporating vapor at an infinite distance from
       the drop (mb)

P.  =  air pressure (mb)
7y  =  ventilation factor (dimensionless)

The air density in the model is calculated from the relationship

    P, = -T*  (iO"4)                                              (3-38)
        R Tv
where R* is the universal gas constant (8.31432 j mole   °K~1) and Tv is
given in terms of the mixing ratio r  .  According to Beard and
Pruppacher (1971),
                                    34

-------
   TV  •  V1+)/(l + r                                 (3-39)
where

   T.  =  air temperature ( K)
   r   =  0.62197 e /(P.-e }                                     (3-40)
    GO              CO x f\  CO '

The diffusivity DV of the vapor into air at the temperature of the
surface of the drop Tr depends on the liquid being considered.  For
water drops, the FSCBG program uses the expression (Pruppacher and
Rasmussen 1979)
               T  L94  P
   Dv = 0.211 [f ]     [^]                                      (3-41)
                o        A
where TQ is 273.16 °K, PQ is 1013.25 mb, and Tf is determined from the
relation
               L D  M0 (ec- e )
   Tr  =  TA -    v  V s   ~J                                  (3-42)
    r      H       k R  Tf
where

 L  =  latenUheat of vaporization (cal g~ )
    = 597.3 [] °                                               (3-43)
 a  =  0.107 +3.67 x 10"  Tr                                    (3-44)
 k  =  thermal conductivity (cal cm'1 s-1 °K"1)                  (3-45)
    =  kd[l - (1.17 - 1.02 kv/kd)(e
-------
The vapor pressure at e^is also obtained from Equation (3-49) when Tp is
replaced by TA in Equation (3-50) and 6 is replaced by (RH/100), where
RH is the relative humidity in percent.  To find the drop temperature,
vapor pressure at the drop surface, and diffusivity, Equations (3-41)
through (3-49) are solved by iteration in the FSCBG program.

The ventilation factor 7  for water (Pruppacher and Rasmussen 1979) is
given by

j  =  r 0.78 + 0.308 Sc1/3 Re1/2    1.4 < Sc1/3 Re1/2 < 51. 4i     (3-52)
 v     1.00 + 1.108 Sc1/3 Re1/2      0 < Sc1/3 Re1/2 <  1.4
where
Re  =  Reynolds number
Sc  =  Schmidt number
                                                                 (3-54)
        vA                           l  .1
MA  =  absolute viscosity of air (g cm   s  )

    _  (7.6342 x IP"2) ,  TA  >                                   ,,
    ~   [TA +296. 16 J  1296.16J                                   (

The time-rate change of drop diameter for non-water drops can also be
calculated.  In this case, temperature and vapor pressure of the  drop
are automatically calculated from the expression (Picot 1979)

A exp  (B - C/Tr) - e$ = (rJT) (TA - Tr)                         (3-56)

                         [PA - A exp (B - C/Tt)j

where

A   =  P0/760
                                                _o
C^  =  molal concentration of the liquid (mol cm  )

k   =  thermal conductivity of the gas mixture  surrounding the drops  at
       the drop surface temperature

B,C =  constants obtained from tables expressing variations of vapor
       pressure with temperature (see, for example, page D-140, Handbook
       of Chemistry, 58th Edition, published by Chemical Rubber Co.)
                                  36

-------
    Values of k, Dv, and L for non-water drops must also be specified.   The
    value of emfor non-water drops can probably be set equal  to zero.   When
    the parameters required by Equation (3-56) have been specified,  the
    program uses Newton's iteration procedures to determine the drop
    temperature and vapor pressure and the time-rate change of diameter,
    using the following expression (Fuchs 1959) for the ventilation
    coefficient:
    where

    Sh  =  Sherwood number

        =   2(1 + a Sc1/3 Re1/2)                                     (3-58)

    and a is set equal to 0.3 in the FSCBG program as suggested by Pi cot
    (1979).

Finally, for both water and non-water drops, the results obtained using the
evaporation model for above-canopy calculations are fitted with Equation (3-
35) by least squares over the time period required for the drop to evaporate
to a diameter of 5 micrometers or the settling velocity to reach 0.02 ms  ,
whichever is greater.

    Alternative pesticide application methods—As previously mentioned,
options are available for simulating ground based pesticide applications.
These options would be used to simulate spraying events where conditions
invalidate the use of FSCBG, granular applications and/or soil incorporation
methods.  The first involves a partitioning of deposited material between
plants and soil using an exponential model.  The second involves the user
specification of deposition rates to foliage and soil.  Even though these
algorithms are in reality a part of the terrestrial fate and transport
module (TFAT) they are presented here for clarity of organization.

    Exponential method—This method distributes the pesticide between the
    plant canopy and the soil surface based on the extent of areal
    vegetative coverage.  The fraction of applied pesticide intercepted by
    the foliage is given by:

         CF =  (1 - e~yW) R                                            (3-59)

    where y  is the canopy interception coefficient (m2 kg-1) W is the
    current value of the above ground biomass (kg m~2) and R is the
    application rate in grams of active ingredient per square centimeter
    (g of cnf^).  Cp is not a concentration on the leaf surface, but the

                                     37

-------
    deposition based on projected surface area.   The user may also specify
    the depth of penetration of the soil  applied chemical so that soil
    incorporation techniques can be simulated.

    Similarly, the fraction of applied pesticide intercepted by the soil
    surface is given by:

        SF = Re"yW                                                   (3-60)

    where Sp is the concentration on soil surfaces not covered by the
    foliage.

    This method does not account for any off-site drift or losses due to
    evaporation.  Thus, all of the pesticide mass that is applied is
    deposited on either the plant or soil surfaces.

    User-specified distribution—This method distributes the applied
    pesticide to the vegetative canopy and soil  surface, based upon user
    supplied percentages for each.  The equations used are:

       CF = PFR                                                      (3-61)

       Sp = PSR                                                      (3-62)

    where Cp and Sp (both in g cm~^) are the quantities of the applied
    pesticide apportioned to the foliage and soil surface, respectively,  Pp
    and PP are the specified fractions of material allocated to the foliage
    and soil and R is the application rate in g of cm  .  The user may  also
    specify the depth of penetration of the soil applied chemical so that
    soil incorporation techniques can be simulated.  Once the application
    rate to foliage Cp is known, this quantity can be distributed into  the
    canopy using one of the following techniques.

    Simplified canopy penetration model—The simplified canopy penetration
model can optionally be used in conjunction with either the aerial  (FSCBG)
or ground application submodels.  This model allows the user either to  input
the percent of spray material  retained on the canopy at different levels and
on the under-canopy soil or it will  compute the  vertical  profile  of
deposited spray material on the vegetive  canopy.   The exponential decay
Jethod is described in the following paragraphs  followed by a brief
discussion of the user specified method.

    Exponential decay canopy penetration  method—The exponential  decay
    method is based on work by Uktand Courshee  (1982) and Bache and Uk
    (1975).  The model describes an exponential  decrease of the deposition
    density of the spray with depth from  the exposed canopy top,  modified to
    account for conservation of the sprayed material.

                                    38

-------
  If  the  entire  vegetative  canopy  is  represented  by  a  series  of  discrete
  canopy  levels,  the  total  deposition of the  spray material on a given
  level is described  by:

     Di ' DoPi                                                      (3-63)

          hi           hi-l
     P1 = J0e-6h  dh - J0e '6hdh                                     (3-64)

  where
                                                      o
  0^  = deposition of  material  in canopy level  i  (g cm~^)
  h.j  = distance  downward from  the  top of the  canopy  to the bottom of
      canopy  level i  (cm)
  DQ  = initial deposition of material at the  top  of  the canopy (g cm~2)
  Pj  = percent of material  deposited  in the interval h^ to h.j_j  (canopy
      interval  i)
   B  = attenuation coefficient  (cm )

  The total amount of material deposited on the plant  canopy  is  given by

     DT -z  0.                                                     (3-65)

•  where
                                               p
  DT  = total deposited  material on canopy  (g  cm   )
   n  = number of  canopy levels

  The remaining spray material that is not intercepted by the plant canopy
  is  deposited on the  under-canopy soil, D_,  given by:

   Ds =  DQ -  DT                                                  (3-66)

 The  only external parameter which the user needs to supply is the
 attenuation coefficient, 6.   This parameter  depends on the foliage
 density  (foliage area per  unit volume), impaction efficiency, and the
 leaf area  index.

 Discrete distribution canopy  penetration  method—This method allows the
 user to  specify the  percent of material that will be  distributed to each
 canopy level,  including the under-canopy  soil surface.  The  submodel
 then calculates  the  deposition density based on  the top-of-canopy
 quantity (computed by the  aerial  spray or ground spray models)  and the
 percentage retained  at each level.

 Situations where the use of the discrete  distribution canopy penetration
 method would  be appropriate are:

                                  39

-------
    •  If the linear method is being used to model a sparse vegetative
       canopy, none of the  pesticide would be deposited in the soil below
       the canopy, which could realistically occur.  But if the discrete
       distribution canopy penetration method is used, the pesticide applied
       to the plant can be redistributed so that a specified fraction
       reaches the under-canopy soil.

    •  When an orchard, or other crop, is being sprayed from below canopy-
       top level, the exponential decay canopy penetration method cannot be
       used. If the approximate vertical distribution can be quantified, the
       discrete distribution canopy penetration method can be used.

3.2  TERRESTRIAL FATE AND TRANSPORT MODULE (TFAT)

    This section describes the dynamic terrestrial fate and transport module
(TFAT) which is used to simulate the vertical movement of pesticides in
terrestrial compartments. -These compartments are made up of the soil,
including the solid soil particles, the soil-air, and the soil-water, and
air overlying the soil column.  It is required to have the capability to
simulate each of the significant chemical transport components and ancillary
processes identified in Section 2.  An existing transport code was selected
and enhanced to serve this purpose, as described in below.

    The terrestrial fate and transport module was developed in two steps.
First, PRZM (Pesticide Root Zone Model, Carsel et al. 1984) was selected as
the basic model.  Dean et al. (1984) reviewed six pesticide fate and
transport models for applicability to a similar exposure assessment
problem.   They selected PRZM for that application primarily because it  used
acceptable theory and readily obtainable input, and had  good
documentation.   PRZM was similarly selected for inclusion in TEEAM  and
modifications and enhancements were added to enable it to produce the
pesticide concentrations required for assessment of wildlife exposure in
terrestrial ecosystems and to link it with the other TEEAM modules.  The
most important additions to the basic model include two  groups of
subroutines which simulate vapor phase transport and chemical  volatilization
from the soil surface and enhance the surface water hydraulics to  simulate
surface water ponding.

3.2.1  The Basic Fate and Transport Model (PRZM)

    The Pesticide Root Zone Model is a comprehensive, dynamic, compartmenta1
model for use in simulating chemical movement in the unsaturated soil
systems within and immediately below the plant root zone.  The model
components are depicted in Figure 3.3.

    PRZM has two major components:  hydrology and chemical transport, which
simulate runoff, erosion, plant uptake, leaching, decay, and foliar washoff
                                     40

-------
                   Figure 3.3  PRZM, Release 1, model  components.
of a pesticide after application to an agricultural field.  The hydrologic
component for calculating runoff and erosion is based on the Soil
Conservation Service curve number technique and the Universal Soil Loss
Equation.  Evapotranspiration is estimated from pan evaporation data or by
an empirical formula if input pan data are unavailable.  It is divided among
evaporation from crop interception, evaporation from soil and ponded water,
and transpiration by the crop.  Water percolation is simulated by the use of
generalized soil parameters including field capacity, wilting point, and
saturation water content.  Time-varying transport by both advection and
dispersion are represented in the program.  However, transport in only
dissolved phases is simulated in the original program.  Dissolved and
adsorbed concentrations in the soil are estimated by simultaneously
considering the processes of pesticide uptake by plants, surface runoff,
erosion, decay, advection, foliar loss, dispersion, and retardation.  The
transport equations are solved utilizing a backwards-difference implicit
finite-difference numerical approximation.

3.2.2  Enhancements to PRZM

    The basic PRZM model was enhanced to more adequately simulate chemical
fate and transport processes required for exposure assessment in terrestrial
ecosystems.  This required the addition of two processes which are used to
simulate:
                                     41

-------
       • Time-dependent infiltration and ponding

       • Volatilization and vapor phase transport

Time-Dependent Infiltration and Ponding--
    Soil surface hydraulics in PRZM are simple,  and the enhancement of these
surface water processes for the TFAT model  allow for a more realistic
simulation of surface water behavior and soil  moisture distribution in the
underlying soil during a precipitation event.   In the previous releases of
the PRZM model, incoming precipitation was  partitioned into runoff and
infiltration using the SCS curve number procedure.  However, no distinction
was made between water which immediately infiltrated into the subsurface and
that which initially ponded on the surface  and then infiltrated after some
period of time.  In the TFAT model, the distinction between water which
infiltrates immediately and that which ponds is made, and the evolution and
dissipation of the surface ponds is simulated  through time.

    In order to estimate chemical exposure  from surface ponds, TFAT
estimates the depth of the pond at a given  time, the length of time for
which water is ponded, and the concentration of pesticide in the ponded
water.  These parameters are a function of  meteorological
variables—especially precipitation and temperature; soil variables—soil
moisture, texture, permeability, etc.; and  pesticide chemistry and
application history.  Most of these variables  are input data for other
subroutines in PRZM.  For example, precipitation and the soil hydraulic
properties are required for the redistribution of moisture within the soil
profile.  The algorithm to simulate ponding was selected to impose minimal
additional requirements data and utilize existing inputs to the extent
possible.

    One major deficiency in the structure of PRZM for the simulation of
ponded water is the time step.  PRZM, and the  other TEEAM modules, operate
on a 24-hour time step, which is too large  to  capture the dynamic, ephemeral
nature of surface ponds.  Using a finer time step in the ponding algorithm
requires that each of the other hydrologic  variables—precipitation, runoff,
snowmelt, evapotranspiration--be disaggregated to the smaller time step and
that results be reaggregated for linkage with  the rest of the TEEAM model.

Volatilization--
    Potential  volatility of a chemical  is related to its inherent vapor
pressure, but  actual vaporization rates depend on environmental conditions
and all  other  factors that control behavior of the chemical at the
soil/air/water interface.  Several factors control volatilization from the
soil  surface,  including pesticide properties,  adsorption characteristics of
the soil, pesticide  concentration, soil water content, and  soil
temperature.   Sealing of the  soil surface by ponding effectively eliminates


                                     42

-------
volatilization.  As with the ponding algorithm, the volatilization algorithm
was designed to use a minimum of new input data and merge easily with the
existing transport equations.

3.2.3  Mathematical Description of the Terrestrial Fate and Transport
       Processes

    The mathematical description of the processes simulated by TFAT are
broken down in the following discussion into five categories:

       • Transport in Soil
       • Water Movement
       • Soil Erosion
       • Surface Water Ponding
       • Volatilization

    The first three categories were simulation options previously available
in PRZM Release 1.  Since the capability to simulate ponding is new, the
mathematical basis of the ponding algorithm is described in detail.  The
final process, volatilization, was not available in previous releases of
PRZM, and its theoretical basis is also described in detail.

Transport in Soil--
    The pesticide transport equations in soils were derived from the
conceptual, compartmentalized representation of the soil profile shown in
Figure 3.4.  From consideration of Figure 3.4, it is possible to write mass
balance equations for both the surface zone and the subsurface zones.
Addition of the vapor phase and ponded water compartments in TFAT require
the consideration of additional terms.  The surface zone expressions for
each of the dissolved, adsorbed, and vapor phases can be written as:

       AAX a(Cwe)                                                    (3-67)
       — ^   —  = JD * V JDW ~ JU ' JQR + JAPP + JFOF

       AAX 3(CsPs)                                                   (3-68)
       —        = -JDS - JER
        AAX
            at     ~  GD ~DG
 where
                     ,    ,                                            (3-69)
                   ~ J   ~J
     A  = cross-sectional  area of soil  column (cm )
     AX = depth dimension  of compartment (cm)
     Cw = dissolved concentration of pesticide (g cm)
     C$ = sorbed concentration of pesticide  (g g"^)
                                    43

-------
    Cg = gaseous concentration  of  pesticide  (g  cm"3)
    e  = volumetric water  content  of  soil  (cm3  cm"3)
    a  = volumetric air content of the soil  (cm  cm"  )
    p  = soil  bulk density (g cm"  )
    t  = time  (d)
    JQ = rate  of mass loss due  to  dispersion and  diffusion
         of dissolved phase (g  day  )
    Jv = rate  of mass loss due  to  advection  of  dissolved
         phase (g day"1)
    JQQ  = rate of mass loss due to dispersion  and  diffusion
           in  vapor phase  (g day  )
    JDW  = rate of mass loss due to degradation in  the  dissolved
           phase (g day"1)
    JQQ  = rate of mass loss due to degradation in  the  vapor
           phase (g day"1)
    Jy = rate  of mass loss by plant uptake of dissolved phase
         (g day"1)
    JQR  = rate of mass loss by removal in runoff (g  day)
    JApp = mass gain due to pesticide deposition  on the soil
           surface, including infiltration from surface ponds (a day  )
         = mass gain due to washoff from plants to  soil (g day  )
         = rate of mass loss due to degradation of  sorbed phase
           chemical (g day  )
    J£R  = rate of mass loss by removal on eroded sediments
           (g  day"1)

    Equations  for the subsurface zones are identical  to Equations (3-67),
(3-68), and (3-69) except  that  JQR, JFQF»  and ^ER are not Inc1uded.  JAPP
applies to subsurface zones only when pesticides  are incorporated into the
soil.  For subsurface layers below the root  zone, the term Jy is also not
utilized.

    Each term  in Equations (3-67)  through (3-69)  are now further defined.
Dispersion and diffusion in the dissolved phase are combined and are
described using Fick's law as

               AAX D  32C  e

        J° ' ' —*                                                  '3
 where

     Dw = diffusion-plus dispersion  coefficient for the dissolved
          phase, assumed constant  (cm2 day"  )
     Cw = dissolved concentration  of  pesticide  (g cm" j)
     e  = volumetric soil  water content (cm3 cm"3)
     x  = soil depth dimension  (cm)
                                  44

-------
(Surface L
Runoff)
(Surface Layer:
rros1on)




ayer: JOR
t JER
I
Diffusion
I.
SOLIDS
cs
DS
Adsorption/ 	
Desorption _. —
-kCs (JCs)


1
Leaching Diffusion (Surface Layer:
hv | Volatilization)
WATER
cw
e
•^-



— •* —
tr 1 1 \
~ W DW'
Diffusion
JD



GAS
C9
n
— »- Gas/L1qu1d
	 Equilibria
-kCg (JDQ)
^ *
Leaching Diffusion
^v JGD
Ju
•> Plant Uptake
	 f Animal Uptake






                  Cwe + Cgn
                   cw
             Figure 3.4  Schematic  representation of a TFAT module soil layer.
    In a similar manner,  dispersion and diffusion in the vapor  phase are
described by Pick's  law as
               AAX Dn  a2 (Cna)
        JGD
                                                                   (3-71)
                      ax
where
D  =
a  =
         Molecular diffusivity of the pesticide in the air filled
         pore space  (cm  day*  )
         vapor-phase concentration of pesticide (g cm"3)
         volumetric  air  content (cm  cm" )
    The dependence of the molecular diffusivity of the pesticide  in  air
filled pore space on the volumetric air content is described by the
Millington-Quirk expression  (Jury  et al. 1983a)
         Dg .
                                                                   (3-72)
where

    a
    n
     the air-filled porosity  (cm3 cm
     total  porosity (cm3 cm"3)
                                         ~3
                                    45

-------
    Da = molecular diffusivity of the chemical in air, assumed
         constant (cm2 day " )

    The mathematical theory underlying the diffusive and dispersive flux of
pesticide in the vapor phase within the soil and into the overlying air can
be found in the section describing volatilization.

    The advective term for the dissolved phase, Jv, describes the movement
of pesticide in the bulk flow field and is written as

             A AX V a(C e)
        j  = - "—                                            (3-73)
                  3X
 where
     V = velocity of Water movement (cm day  )
     Vapor-phase advection has not been included as a flux in the transport
 equation.

     Degradation of a pesticide in or on soil  may be due to such processes as
 hydrolysis, photolysis and microbial decay.  If these processes follow
 pseudo first-order kinetics, the rate coefficients may be combined into a
 single decay coefficient.  Assuming the same rate constants for the solid,
 gaseous and dissolved phases, the rate of change of chemical out of each
 phase due to decomposition may be written as:

        JDW = K$Cwe AAX                                               (3-74)

        JDS - Kscsps AAX                                              (3-75)

        JDG = KsCg a AAX                                              (3~76)

 where

     KS = lumped, first-order decay constant for solid and dissolved
          phases (day  )

     Plant uptake of pesticides is modeled by assuming that uptake of a
 pesticide by a plant is directly related to transpiration rate.  The uptake
 is given by:
        J.,  = f C  9 e AAX                                             (3-77)
         u       w
 where
     Ju = uptake rate of pesticide (g day'1)

                                     46

-------
    f  = the fraction of total water in the zone used for transpiration
         (day'1)
    e  = an uptake efficiency factor or reflectance coefficient
         (dimensionless)

    Erosion and runoff losses as well as inputs to the surface zone from
foliar washoff are considered at the soil surface.  The loss of pesticide
due to runoff is
       JQR -   - C  A                                                <3-78>
in which

    JQR  = pesticide loss due to runoff (g day  )
    Q    = the daily runoff (cm3 day'1)
    AW   = watershed area (cm2)

and the loss of pesticide from the entire modeled  area due to erosion is

       i   _ P  e rom  s                                             ,3
       JER -       ^                                                (J-

in which

    JER  = the pesticide loss due to erosion (g day"1)
    Xe   = the erosion sediment loss (tons day  )
    rom  = the enrichment ratio for organic matter (g g  )
    p    = a units conversion factor (g tons  )

    Pesticides can be applied to either bare soil  if pre-plant conditions
prevail or to a full or developing crop canopy if  post-plant treatments are
desired.  The pesticide application is an input mass rate which is
calculated by one of the application/deposition models discussed in Section
3.1 or input directly by the user.  It is partitioned between the plant
canopy and the soil surface, and the rate at which it reaches the soil
surface is designated
    Pesticides applied to the plant canopy can be transported to the soil
surface as a result of washoff by rainfall.  This term, Jp' 1S defined as:
       JFOF = ExPr M A                                               (3-80)

where

    Ex = foliar extraction coefficient (cm  ) (Smith and Carsel 1984)
    Pr = daily rainfall depth (cm day  )
    M  = mass of the pesticide on the plant surface,
         projected area basis (g cm  )

                                    47

-------
The foliar pesticide mass, M, is further subject to degradation and loss
through volatilization.  Its rate of change is given by
 where
            =  _K  MA  -  Jrni- + Apb A                                     (3-81)
         dt     r             "
     Kf = lumped  first-order  foliar degradation  constant  (day   )
     Ap = application  rate  to the  plant  (g  ba    day   )
     b   = a units conversion  factor (ha  cm"  )

     Adsorption and  desorption in  Equations  (3-67) through  (3-69)  are treated
 as instantaneous,  linear,  and reversible processes.   Using this  assumption,
 the sorbed phase concentration can be related to  the dissolved phase
 concentration by:

        Cs = Kd Cw                                                    (3-82)

 where

     K^ = partition  coefficient between  the  dissolved and
          solid phases (cm3 g-1)

     A  similar expression can be developed  to express the vapor phase
 concentration in terms of  the dissolved phase concentration as follows

        Cg = KHCW                                                     (3-83)

 where

     KH = Henry's constant, i.e.,  distribution-coefficient
          between liquid phase and vapor phase  (cm  cm" )

     Summing Equations (3-67), (3-68),  and  (3-69)  and utilizing (3-82) and
 (3-83), the following expression  results  for  the  mass balance of pesticide
 in the uppermost soil layer.
        3[Cw(9+Kdp$+aKH)]      32(Cwe)       32(CWKH)    aCweV
                                w     2      a       2
                  at            w   3x^      y     3x        3x px r  K >
                                          KgaKH) + fee
                                                          W         W
                                                                      (3-84)
                         AXA     AX
     Boundary and initial conditions are required to solve Equation 3-84.
 Initial conditions are input by the user as a total pesticide mass or
 concentration for each soil layer.  Boundary values must be provided for the


                                      48

-------
dissolved and vapor phase concentrations.  The condition at the upper
boundary for dissolved phase advection is a constant flux condition for each
time step.  If takes on a zero value if there is no mass flux during the
time step and a positive value if there is a mass input from plant washoff
or infiltration from surface ponds.  The condition at the lower boundary for
transport is a zero gradient

              1 . 0                                                  (3-85)
where C^ and C-j+j are dissolved concentrations in the ith and i+lst
compartments.

    For the vapor phase, the condition at the lower boundary is also a zero
gradient.  The condition at the upper boundary is disscussed under the
description of volatilization from soil and ponded water.

    The system of equations are written in finite difference form, using
central differences for diffusive terms and a backwards difference for the
advection terms.  The backwards difference formulation has the problem of
producing some numerical dispersion.  The equations are solved using a fully
implicit scheme which is unconditionally stable.  The Thomas algorithm is
used to solve for end-of-timestep dissolved phase concentrations in each
layer.

Water Movement —
    Because velocity and water content of the soil are not generally known
or measured as part of routine monitoring programs, it is necessary to
develop additional equations for these variables in order to solve the
transport equations.  In the general case, Darcy's law can be combined with
the continuity equation to yield the Richards equation (Richards 1931):

                                                                     <3-86>
    K(e) = hydraulic conductivity at various water contents (0)
    e    = soil water content

and

       V = -K(e) |£                                                  (3-87)

or, in simpler terms

       o y     d V                                                     / ^i or> \
       at = - w                                                     (3-88)
                                     49

-------
where
       V = soil  water velocity (cm day  )
Writing (3-88)  in a backwards  finite difference  form  yields

       AX (ot+1-0*) = (V.  - V.^)  At                                 (3-89)

or

       0t+1 AX = (V1 - V^JAt + 0*Ax                                (3-90)

    In these equations, t  and  t + 1 denote the beginning and end of time
step values, respectively, and i is the soil  layer index.  These equations
can be further simplified  by substituting the nomenclature SW for OAX so
that

       SWt+1 = SW* + (V. - V.^) At                                  (3-91)

where

    SW = soil water content (cm).

    The velocities shown in Equation (3-91) are a function of inputs to the
soil (precipitation, infiltration) and outflows from the soil
(evapotranspiration, runoff).

    Water balance equations are separately developed for (a) the surface
zone, (b) horizons comprising the active root zones, and (c) the remaining
lower horizons within the unsaturated zone.  The equations are:

    Surface Zone

       (SW)*+1 = (SW)* + INF - Ij - E1 - Uj                          (3-92)

    Root Zone

       (SW)*+1 = (SW)* + IM - U. - I.                              (3-93)

    Below Root Zone

       (SW)*/1 = (SW)* + 1.^ - I.                                   (3-94)
                                    50

-------
where

    (SW)^ = soil water in layer "1" on day "t" (cm)

    E..    = evaporation (cm day" ).
    U.    = transpiration (cm day" )
    I.    = percolation out of zone i (cm day" )
    INF   = infiltration into layer 1 (cm day" )

    Daily computation of soil moisture in the soil profile using the above
equations requires the additional calculations for infiltration,
evaporation, transpiration, and percolation.

    Infiltration is calculated as

       INF = P + SM - Q - ADP - E - U                                (3-95)

where

    P  = precipitation as rainfall, minus crop interception
         (cm day'1)
    SM = snowmelt (cm day  )
    Q  = runoff depth (cm day" )
    E  = evaporation (cm day'1)
    U  = evapotranspiration (cm day  )
    ADP= increase in depth of ponded surface water (cm day'1)

    The calculation of precipitation, snowmelt, and runoff on a daily time
step as calculated in TFAT are described below.  The disaggregation of these
values and the calculation of the change in the depth of ponding on a finer
time step is included in the section describing the simulation of ponded
surface water.

    Precipitation pan evaporation and/or air temperature are first read from
an input file.  Incoming precipitation is first partitioned between snow or
rain, depending upon temperature.  Air temperatures below 0.0°C produce snow
and may result in the accumulation of a snowpack.  Precipi-tation first
encounters the plant canopy and once the interception storage is depleted,
the remaining depth is available for the runoff or infiltration.

    The runoff calculation partitions the precipitation between infiltrating
water and surface runoff.  Infiltrating water may be ponded on the soil
surface for a period of time before it infiltrates (this  process  is
described in a subsequent section).   Runoff  is calculated by a modification
of the USDA Soil  Conservation Service curve  number approach  (Haith  and  Loehr
1979).  Snowmelt is estimated on days in  which a snow pack exists  and  above
freezing temperatures occur as

                                     51

-------
       SM = CMT                                                      (3-97)

where

    Cft = degree day snowmelt factor (cm °C   day" )
    T  = average daily temperature (°C)

    The precipitation and/or snowmelt are inputs to  the SCS runoff equation
used to compute runoff depth Q:

       n _ (P + SM - 0.2S)2                                          ,3_98)
       U    P + SM + 0.8S                                            (* ybj

where S, the watershed retention parameter, is estimated by

       S = 1000/RCN - 10                                             (3-99)

where RCN = SCS runoff curve number

    Curve numbers are a function of soil type, soil  drainage properties,
crop type, and management practice.  Typically, specific curve numbers for a
given rainfall event are determined by the sum of the rainfall totals for
the previous 5 days, known as the 5-day antecedent moisture condition.  In
TFAT, as in the original version of PRZM, the curve numbers are continuously
adjusted each day as a function of the soil water content in the upper soil
layers.  These algorithms were developed and reported by Haith and Loehr
(1979).

    The daily evaporative demand is divided among evaporation from canopy,
ponded surface water, soil evaporation, and crop transpiration.  Total
demand is first estimated and then extracted sequentially from crop canopy
storage, ponded surface water, and then from each layer until wilting point
is reached in each layer or until total demand is met.  Evaporation occurs
down to a user-specified depth.  The remaining demand, crop transpiration,
is met from the active root zone.  The root :one growth function is
activated at crop emergence and increases stepwise until maximum rooting
depth is achieved at crop maturity.

     Actual transpiration from a soil  layer is estimated as:

        U. = MIN  [(SW. - WP.) fd1, ETp - V U.]                     (3-100)

 where

     U^   = the actual transpiration from layer  'i'  (cm)
     fd.j  = depth factor for  layer 'i1
     WP.j  = wilting point water content  in layer 'i'  (cm)
     ET_  = potential evapotranspiration (cm)

                                     52

-------
    This equation states that the transpiration from any layer 'i1 is the
minimum of the available water in layer 'i1  or the demand remaining after
extracting available water from layers above 'i1  in the profile.

    The depth factor, f^, is internally set in the code.  It linearly
weights the extraction of transpired water from the root zone with depth.  A
triangular root distribution is assumed from the surface zone to the maximum
depth of rooting with the maximum root density assumed to be near the
surface.  This algorithm essentially views the plant as a pump and assumes
that it will expend the minimum energy possible in pumping.  As long as the
soil water is equally available, water closest to the surface meets this
criteria.

    Transpiration may also be limited by soil moisture availability.  The
potential rate may not be met if sufficient soil water is not available to
meet the demand.  In that case, TFAT modifies the potential rate by the
following equations:

       ETp = ETp          if SW > 0.6 FC                            (3-101)
       ETp = SMFAC ET     if WP < WS < 0.6 FC
       ETp = 0            if SW < WP

where

    FC = soil moisture content at field capacity (cm)
    WP = soil moisture content at wilting point (cm)
    SMFAC  = soil moisture factor

The SMFAC concept has been used in other similar water balance models (Haith
and Loehr 1979; Stewart et al. 1976) and is internally set in the code to
linearly reduce ET  according to the limits imposed in Equation (3-100).  If
pan evaporation input data are available, ETD is related to the input values
by

        ETp  = Cp  PE                                                  (3-102)

where

     PE  =  pan evaporation  (cm  day  )
     Cp  =  pan factor  (dimensionless)

The  pan factor  is constant for  a given  location and  is a function of the
average daily relative  humidity, average daily wind  speed, and location of
the  pan with respect  to an actively transpiring crop.

     In  the  absence of pan evaporation data,  ETp is  estimated  by

                                     53

-------
       ETp = 14000 L^(SVD)                                          (3-103)

where

    Lj = possible hours of sunshine per day, in 12-hour units
    SVD  = 0.622 SVP/(Rg Tabs) and
    SVP  = saturated vapor pressure at the mean absolute air
           temperature (mb)
    Rq = dry-air gas constant
         = absolute mean air temperature (°K)*
    The final term in the water balance equations that must be defined is
the percolation value, I.  The use of the SCS curve number approach for
runoff precludes the direct use of a Darcian model.  TFAT, like PRZM,
resorts to "drainage rules" keyed to soil moisture storages and the time
available for drainage.  Two options are included.  Although these options
are admittedly simplistic representations of soil moisture redistribution,
they are consistent with the objectives of TEEAM and its  intended uses.

    Option l--Percolation, I, in this option is defined in the context of
two bulk soil moisture holding characteristics commonly reported for
agricultural soils:  field capacity and wilting point.  Field capacity is a
somewhat imprecise measure of soil water holding properties and is usually
reported as the moisture content that field soils attain  after all excess
water is drained from the system under influence of gravity, usually at
tensions of about 0.3 bar.  The difficulty with this concept is the fact
that some soils will continue to drain for long periods of time, and thus
field capacity is not a constant.  Admitting the lack of  theoretical and
physical rigor, the concept remains as a useful measure of soil moisture
capacity and has been successfully used  in  a  number  of water  balance models
(Haith and Loehr 1979;  Stewart  et al .  1976).   Wilting point  is  a function
of both the soil  and plants growing in the  soil.   It is  defined as  the  soil
moisture content below which plants are  unable to extract  water, usually  at
tensions of about 15 bar.

    Field capacity and wilting  point are used operationally  to  define  two
reference states in each  soil  layer for  predicting  percolation.   If the
soil water, SW, is calculated to be in excess of  field  capacity,  then
percolation is allowed to remove the excess water to a  lower zone.   The
entire soil profile excess is assumed to drain within  one  day.   The lower
limit of soil water permitted is the wilting  point.   One outcome  of these
assumed "drainage rules"  is that the soil layers  below  the root zone tend
to quickly reach field capacity and remain  at that value.   When this
condition is reached, all  water percolated  below the root zone  will
displace the water within the lower soil layer simulated,  and so  on.   There
is no allowance for lateral water movement.  Water balance accounting  in

                                     54

-------
this manner should be most accurate for sandy soils in which water movement
is relatively unimpeded and is least accurate for clay soils (Stewart et
al. 1976).

    Option 2--The second option is provided to accommodate soils having low
permeability layers that restrict the "free drainage" assumed in
Option 1.  In the context of the field capacity reference condition, two
things may occur.  First, conditions may prevail that raise the soil
moisture levels above field capacity for periods of time because the water
is "backed up" above a relatively impermeable layer. Second, the excess
water may not drain during the one-day period assumed in Option 1.  To
accommodate these conditions, two additional parameters are needed.
Maximum soil moisture storage, (or porosity) QS, is added to represent
moisture contents under saturated conditions.  The drainage rate also must
be modified to allow drainage to field capacity over periods in excess of
one day (one time step).  This is accomplished by adjusting the end of time
step moisture content by
               (0   _ 0fc ) exp(- aAt) + 0fc>                        (3-104)

    where

    0   = soil layer water content (cm3 cm"3)
    0f  = water content at field capacity (cm  cm~3)
    a   = drainage rate parameter (day  )
    i   = the layer index

    In this equation, t and t+1 denote beginning and end of time step
values, respectively, and i is the soil layer index.  The value t* denotes a
value of time between beginning and end of time step.  The variable 0.
here denotes current storage plus any percolation from the next layer above,
before the occurrence of any drainage from the current layer.   Because
Equation (3-103) is solved independently for each layer in the profile,
there is a possibility of exceeding the storage capability (saturation  water
content, 0 ) of a low-permeability layer in the profile if a more permeable
layer overlies it.  At each time step, once redistribution is complete, the
model searches the profile for any  0. > 0  .  If this condition is found,
the model redistributes water back into overlying layers, as if the
percolation of additional water beyond that necessary to saturate the low-
permeability layer had not occurred.   This adjustment is necessary due  to
the nature of Equation (3-103) and the fact that these equations for each
layer are not easily coupled.  The difficulty in coupling the equations for
the entire profile arises from the dichotomy that one of two factors limits
percolation from a stratum in the profile:  either the rate at which that
stratum can transmit water, or the ability of the stratum below it to store
or transmit water.  This dichotomy would lead to an iterative (or at least

                                     55

-------
corrective) approach to the explicit solution of a system of equations
for o.., represented by Equation (3-103).  It should be noted, however, that
the value of a selected by this approach is only relevant if the
permeability of the soil materials, and not storage considerations in the
profile (i.e., the presence of a water table), is the limiting factor for
percolation of water.

    During those time periods when precipitation is occurring or water is
ponded at the surface, the hydraulic simulation in TFAT defaults to a third
algorithm in preference to options 1 or 2, which normally redistribute
water in the profile.  The use of this third algorithm is required by the
assumptions inherent in the Green-Ampt model which is utilized to simulate
surface water infiltration and ponding.  The model assumes a piston-like
downward movement of a sharply defined wetting front, behind which the soil
is completely saturated, and in front of which it remains at its initial
water content.  When these conditions no longer hold, the user-specified
option is again used for percolation calculations.

    Soil erosion—Removal of sorbed pesticides on eroded sediments requires
estimates for soil erosion.  The Modified Universal Soil Loss Equation
(MUSLE) as developed by Williams (1975) is used to calculate soil loss

       Xe = a (Vrqp)°'56K LS C P                                    (3-105)


where

    Xe  soil loss  (tons day  )
    Vr  =  volume of event  (daily) runoff (m  )
    qp  =  peak storm runoff  (m sec  )
    K   =  soil erodability factor
    LS  =  length-slope factor
    C   =  soil cover factor
    P   =  conservation practice factor
    a   =  units conversion factor

    Most  of the parameters  in Equation  (3-104) are .easily determined  from
other  calculations within  PRZM  (e.g., Vr),  and others  are familiar  terms
readily available from  handbooks.  However,  the  peak  storm  runoff value,
q_, can vary  widely.  A  trapezoidal  hydrograph  is  assumed  in TFAT with  a
stochastically determined  average  storm duration,  as  described  later  in
this  section  under the  discussion  of  surface water  ponding.   From the
assumed hydrograph shape and the  storm  duration,  a  peak  runoff  rate is
calculated.

    The enrichment ratio,  rom,  is  the  remaining  term  which  needs to be
defined to estimate  the  removal  of sorbed  pesticides  by erosion.  Because

                                     56

-------
erosion is a selective process during runoff events, eroded sediments
become "enriched" in smaller particles.  The sediment transport theory
available to describe this process requires substantially more hydraulic
spatial and temporal resolution than used in PRZM, leading to the adoption
of an empirical approach (Mockus 1972).  The enrichment ratio for organic
matter is calculated from
           om) = 2 + °'2 1n< W                                   (3'106)
in which AW is the area of the habitat.

    Surface water ponding—Surface  water ponding is the accumulation of
excess water at the soil surface.   Ponding occurs when the surface soil
layers are saturated and consists  of water which does not infiltrate,
evaporate, or runoff.  Pesticides  may enter ponded water through direct
application or washoff from plants; ponded water can therefore be an
important route of exposure for animals which drink or bathe in ponds.
Simulation of surface ponds involves modeling both the movement of water
through ponds and the fate of chemicals dissolved in ponded water.

     The movement of  water through ponds is modeled in the TFAT module using
a water balance approach to calculate changes in pond depth over time:
       DPt = DP*'1 + Pfc - Qfc - Efc - I*                              (3-107)
where
       DP  = the depth of ponding at time t (cm)
       P£  = the amount of precipitation occurring during a time step At (cm)
       Q   = the runoff during time step At (cm)
       E   = evaporation during time step At (cm)
       I   = infiltration during time step At (cm)

     Because ponding is an ephemeral process which depends on rapidly
 varying rates of infiltration, runoff, and precipitation, the water balance
 equation is solved using time steps smaller than the daily TFAT step (i.e.,
 hourly).  Daily values of precipitation and runoff are dissaggregated into
 values for smaller time steps by assuming representative shapes of the
 rainfall hydrograph and the runoff hydrograph.  Calculation of the
 individual components of the water balance is discussed in the following
 paragraphs.

     The TFAT module uses daily values of precipitation depth; use of these
 daily depths in pond water balance calculations requires knowledge of the
 timing of precipitation.  The duration of precipitation events is estimated
 using an empirical relationship:


                                    57

-------
       Td = a(PRECIP)b                                              (3-108)

where

       Td = rainfall duration (hrs)
       a = empirical constant (hr cm~b)
       b = dimensionless exponent
       PRECIP = the daily precipitation depth (cm)

    The coefficients a and b will vary depending on the type of storms
encountered in the region of interest.  Although a and b will in reality
also vary from storm to storm and seasonally, it is assumed that they are
constants for a given location.  Once the duration of the storm is
estimated, the precipitation amount for ponding time steps is calculated by
assuming uniform precipitation over the storm duration:


       Pt = (PRECIP/td)At,  0  <  t  <  Td                                (3-109)

          ' °                   *  >  Td

    where AT = the  ponding  time  step  (hours)

    The amount of runoff  occurring  during  ponding  time  steps  is obtained  by
integrating the runoff  hydrograph over each  time step:
        .   t+At
       Ql = J   q(t)dt                                               (3-110)
            t

where q(t)  is  the runoff  rate at  time  t (cm  hr  ).   For TFAT  runoff  an SCS
trapezoidal hydrograph  is assumed,  as  defined by the peak  flow  rate, the
storm duration, and the time  of concentration for  storms  (see Figure 3.5).
The time of concentration TC  (hours)  describes the rise and fall of  the
hydrograph and is a user  input  parameter which is  a function  of  the
topography of  the simulated drainage  area.  The  peak runoff rate  is
calculated from the total  runoff volume:


                                                                    (3-m)
where

        = the  peak runoff rate  (cm  hr  )
      Q = the  total runoff volume,  calculated from Equation  3-98 (cm)

Ta is the time of initial abstraction, or  the lag  time between  the start of
the storm and  the beginning of  runoff:
            0.2 T ,S
       T	S_                                                 (3-112)
       'a ~ PRECIP                                                  (      '
                                     58

-------
                      Runoff
                      (cm/hr)
                           V
                               Ta  Tc
                                               Time (hours)
                     Figure 3.5  SCS trapezoidal runoff hydrograph.
S in the above equation  is the  SCS  watershed retention parameter defined by
Equation (3-99).

    The infiltration capacity of  the  soil  is estimated using the Green-Ampt
equation.  This infiltration model, developed by Green and Ampt (1911),
relates the infiltration rate to  the  length of the soil wetting front.  The
soil is assumed to be saturated behind  the front and infiltration is driven
by gravity and capillary suction  in the unsaturated soil ahead of the
wetting front.  The infiltration  rate equation is then given by:
       AC          rDP + HfU
       f =K$[1+ 1   p  f)a]                                       (3-113)
where

       F  = the cumulative  infiltration depth  (cm)
       KS = the saturated hydraulic conductivity of  the  soil  (cm/hr)
       Hf = a capillary  suction parameter  (cm)
       a  = the available (i.e., air) porosity  (cm   cm"  )

The soil infiltration capacity computed by the  Green-Ampt model  starts at an
infinite rate and decreases with time to an asymptotic rate  equal  to  K_.
The Green-Ampt rate equation cannot be analytically  integrated to  obtain
infiltration amount unless  the depth of ponding is constant.  The  ponding
algorithm therefore assumes for each time  step  that  the  pond  depth is a
constant equal to the start of time step depth  for infiltration
calculations.  The integrated Green-Ampt equation under  these conditions  is:

                                     59

-------
       (Ft+1- Ft) = Ks At + Hln (H + Ft+1)  - Hln (H + Ft)            (3-114)

where

       H = (DP* + Hf )n
        4-          '
       F  = infiltration amount at the start of the time step (cm)
       F    = infiltration amount at the end of the time step (cm)

The integrated Green-Ampt equation is solved by second-order Taylor series
iteration for the end of time step infiltration capacity.   This infiltration
depth (F   )• is the amount which can potentially occur according to soil
moisture conditions.  The actual infiltration depth occurring during the
time step will be the minimum of the infiltration capacity and the  amount of
available pond water:

       jt = pt+1 _ Ft            Ft+l _ Ft < AW                     (3-115)
          = AW                   Ft+1 - F  > AW
where
       AW = available water (cm)
          = DP* + P  - Q*

    Thus, using the precipitation amount (P ) SCS runoff volume (Q ) and the
infiltration volume (I ), the pond water balance (Equation 3-107) is solved
for the depth of ponding at the end of each ponding time step within the
TFAT daily step.  As discussed previously, evaporation demand is met
sequentially by the plant canopy, ponded water, and soil water.  Therefore,
evaporation calculations are decoupled from the pond water balance so that
canopy evaporation can be used first to meet evaporation demand.
Evaporation is then subtracted from the pond depth at the end of each TFAT
daily time step.  While  it is recognized that this does not account for
variations in evaporation during the day, evaporation losses are usually
small during the precipitation events which produce ponds.

Chemical Fate and Transport in Ponded Water--
    Toxic chemicals in ponded water undergo a number of transport and
transformation processes.  Pesticides may advect into the soil column by
infiltration, volatilize through the air-water  interface, and degrade
through various chemical and biochemical reactions.  These processes are
summarized in Figure 3.6.  The total change in  chemical mass in ponds over
time may be written in differential form as:
         -  J    J     J   +J    J
     dt  ~ ~JA ' Jpv ' JDC + JL ' JAN

                                      60

-------
where
     pv
     JDC
     JAN
= total concentration of chemical  in  ponded water (g cm  '
= volume of ponded water (cm )
= mass flux of  chemical due to advection resulting from
  infiltration  (g day  )
= mass flux of  the chemical due to volatilization (g day
= mass flux flux due to chemical  decay (g day  )
= mass flux due to chemical loading (g day  )
= mass flux due to animal utilization (g day  )
                                                                    ')
Advection  is  simply the transport  of chemical into the  soil  as water
infiltrates  into the soil.  The  advective flux for a well-mixed pond is
given by:
    JA =  INF  A Cp
                                                              (3-117)
where
    INF   =  pond filtration rate,  computed from the Green-Ampt equation
            (cm day"1)
    A     =  surface area of pond  (cm )
                     VOLATILIZATION
                                       \
                                        1  DEGRADATION
                                ADVECTION
                    IMPORTANT PROCESSES:


                        1. VOLATILIZATION OF DISSOLVED PHASE.
                        2. CHEMICAL AND BIOLOGICAL DEGRADATION.
                        3. ADVECTION INTO THE SOIL MATRJX.



               Figure 3.6 Chemical transport and fate in ponded water.
                                    61

-------
    Volatilization results from concentration gradients at the air-liquid
interface.  The pond volatilization flux is modeled as a diffusion process
through boundary layers within the pond and in the overlying air.   The
diffusive flux through the water boundary layer is computed from the
concentration gradient between the center of the pond and the pond surface:
           -AD
           - A °
                w    0.5(DP)

where:

                                                               7     1
       DW  = the diffusion coefficient of chemical in water (cm  day'1)
       Cps = the vapor phase concentration of chemical at the air-liquid
           interface (g cm  )
       KH  = Henry's law constant for the chemical (cm  cm" )

This equation assumes that the concentration at the center of the pond
equals the average concentration, Cp.  The flux through the air boundary
layer is computed assuming a linear concentration gradient through the plant
canopy and zero concentration at the top of the plant canopy:

       Jpv

where

       Dfl =  the vapor-phase diffusion coefficient for the chemical
            cm2 day"1)
       ^CH = ^ne height of the plant canopy

Equating the two boundary fluxes and solving for the vapor phase concentration
at the pond surface:
                2 D  C
       Cps =    Da "  \ Dw
             DP   - + -
The flux out of the pond can then be computed by substituting the above
expression into the air boundary layer flux equation:
                (2 D  Cn)        A Da
       Jpv =      DaW  P2 Dw    -Z^ = A Kv °P                     (3-
              (DP ^- + — ^)     CH
in which          L£\\     ^

       KV =  a volatilization rate parameter (cm day  ) which will vary with
             the depth of ponding


     Chemical  degradation in ponds  may  occur  by  hydrolysis,  photolysis,
 biodegradation,  and a number of other  transformation processes.   A detailed

                                     62

-------
description of the chemical reactions occurring in ponded water is beyond the
scope of this model.  Degradation is assumed to be the same in both the
dissolved and adsorbed phases and is therefore modeled by a simple first-order
decay process:
       JDC = kp Cp V                                                (3-122)
where
    kp = lumped first-order decay constant (day" )

    Chemical fluxes due to loading and animal uptake are simulated in other
TEEAM modules and subroutines.  The pesticide spray and deposition model
accounts for the largest portion of the pesticide load.  Other loads may occur
due to leaf washoff.  Animal uptake losses are derived in the food chain
portion of TEEAM.

    The individual pond chemical fluxes due to advection, volatilization, and
decay are derived as discussed above.  Substituting these relationships into
the pond chemical mass balance equation and dividing by the area of the pond
yields the following expression:
       d(C DP)
       _J	 = .(INFIC  - K  C  - K (DP)C  - K.C  - L             (3-123)
         at       v'pvpppAp

where:

    K^ = first order animal uptake rate constant, computed in the APUM
         module (cm day" )
                                                p     1
    L  = total loading rate per unit area (g cm~  day"1)

If it is assumed that the depth of ponding DP is constant over each ponding
time step (the time step used in the pond water balance calculations), the
mass balance equation can then be written in terms of lumped first order
constants:
       dCn
       d/ = Kl Cp + K2                                             <3-124>


where:

       Kj = 1/DP(-INF - K  - K DP - K )                             (3-125)
       K^ = L/DP              P      A                              (3-126)

The above equation is integrated and solved analytically for the pond  chemical
concentration at the end of each ponding time step (during which the pond
depth is assumed to be equal to the start of time step depth):

                                     63

-------
        t+1 = e 1  fC t + _ — 1    _
              e     L   +       -
Cp+  is the concentration at the end of the time step,  and C  is the
concentration at the beginning of the time step.  K^ and K£ are constants over
individual time steps but are updated at the start of each time step to
account for changes in pond depth and infiltrate rate computed by the pond
water balance equation.  Note that the time step used to solve the pond
chemical mass balance equation is the same as that used for the pond water
balance equation.

Volatilization from the Soil Surface —
    Volatilization (i.e., vapor flux) of pesticide at the soil surface is
modeled in a manner similar to that described for volatilization from ponds.
The diffusive flux out of the soil is approximated using a linear
concentration gradient from the center of the top soil  layer to the soil
surface:


       JGD • V(cg,i - cg.sl                                       (3-128)
    Cq ^ = vapor-phase concentration in the surface soil layer (g cm  )
    Cg's = vapor-phase concentration at the soil-atmosphere interface
           (9 cm'3)                       '
    Dg   = the vapor diffusion coefficient in soil (cm  day  ) (see
           Equation (3-72))

This expression assumes that the concentration at the center of the top soil
layer is equal to the average concentration in the layer.  The vapor phase
concentration in soil is related to the dissolved phase concentration by the
Henry's law equilibrium relationship (Equation  (3-83)).  The flux through
the boundary layer of air above the soil is given by assuming a linear
concentration gradient through the plant canopy and zero concentration at
the top of the .plant canopy:
        JGD =  a  -^                                              (3'129)


 Equating the two boundary fluxes and solving for the concentration at the
 atmosphere-soil interface:
               2 D   C.
                        .
         Cgs = DD      '
                  - + 2D
                CH
 The  vapor  flux  boundary condition  at  the  soil  surface can then be defined  in
 terms  of the  top soil  layer vapor  concentration  (Cq  j) by substituting the
 above  expression into  the atmospheric flux  equation:
                                     64

-------
                 2 D  C  .     DA
         i
         J
          fin   ^n n        ;  7                                     v
                _x_a + 20      CH
                 ZCH     g

                    2 D  D  A

             " ( Va + 2 °g ZCH ] C«-l
Thus, the vapor flux boundary condition at the soil surface is calculated as
a function of the vapor concentration in the top soil layer and the height
of the plant canopy.  Note that if ponds are present, there is no vapor
phase at the soil surface and the volatilization flux out of the soil is
zero.

Volatilization Flux through the Plant Canopy--
    In pioneering work on this topic, Parmele et al. (1972) discuss a number
of micrometeorological techniques for calculating pesticide volatilization
flux from observed aerial pesticide concentrations.  Their procedures are
based on the assumption that the vertical diffusivity coefficient (Kz) for
pesticide vapor is analogous to the vertical diffusivity for water vapor,
energy, or momentum.  The pesticide volatilization flux can be computed by
Fick's first law of diffusion, as follows:

       Jz(Z) = - KZ(Z) (dP/dZ)                                      (3-132)

where

                                              2   1  '
    JZ(Z)   = pesticide flux at height Z (g m   s  )

    (dP/dZ) = pesticide concentration gradient in the canopy atmosphere
              (9 N~ )

    KZ(Z)   = the vertical diffusivity at the height Z (m2 s'1)

The value of KZ depends on the turbulent flow of the atmosphere into which
the pesticide vapor is dissipated.  Therefore, it is a function of the
prevailing meteorological conditions and not of any  physical or chemical
property of the pesticide.

    In order to apply these concepts, pesticide concentrations at  two or
more heights are required to estimate the pesticide gradient and the
subsequent flux.  For the estimation of vertical  diffusivity,  more extensive
meteorological  information is also required.  All  of these data requirements
pose significant limitations for a predictive modeling  approach consistent
with expected and current TEEAM users.
                                     65

-------
    To avoid supplying the model with such extensive meteorological data, a
relationship for KZ is derived in the following paragraphs as a function of
height within the canopy.  Then one need only to consider the pesticide
concentration gradient (or a suitable surrogate) in order to compute the
pesticide volatilization flux.

    Estimation of Kz(Z]--Meh1enbacher and Whitfield (1977) present the
following formula to compute KZ at various heights within the plant canopy:

       K2(Z) = Kz(ZCH) exp(4.0 ( ^- - 1.0))                        (3-133)
                                  CH
       K2(ZCH) = U* k (ZCR - D)/4»h                                  (3-134)
where
    KZ(Z)    = thermal eddy diffusivity at height Z  (m2 s"1)
             = thermal eddy diffusivity at canopy height  (m   s   )
             = canopy height (m)
    ZQ       = roughness length  (m)
    D        = zero plane displacement height  (m)
    k        = von Karman's constant, 0.41
    U*       = friction velocity  (m s'1)
    .        = stability function for sensible heat
    ((j ( )   = integrated momentum stability parameter as  a  function
               of 
    UCH      = W1'nc^ Ve1ocity at  the canopy height  (m  s   )

    For agricultural applications, the canopy  height  is  used as  a  reference
height for calculating U*.  The  user is required to input  the wind speed,  at
a height of 2 meters above the soil surface.   The  wind speed at  the canopy
height (UQ^) is computed based on the logarithm law:

                             Z"
       U -      =  —I - TTl—                             (3-136)
        measured     -.  / measured   — ^
                              o

    The friction velocity U* can be visualized as a characteristic of the
flow regime in the plant canopy compartment in which the logarithmic
velocity distribution law holds.  Rosenberg (1974) describes Z0+D as the
total height at which the velocity profile above the canopy extrapolates to
zero wind velocity.  The values for both Z0 and D can be estimated with the
following procedures and equations presented by Thibodeaux (1979).  For very
short crops (lawns, for example),  ZQ adequately describes the total


                                     66

-------
roughness length, and little adjustment of the zero plane is necessary
(i.e., D = 0).  D is assumed to be zero in the current code when ZCH is less
than 5 cm.  For tall crops, Z0 is related to canopy height
       log ZQ = 0.997 log ZCH - 0.883                               (3-137)

    In tall crops ZQ is no longer adequate to describe the total roughness
length, and a value of D, the zero plane displacement, is needed.  For a
wide range of crops and heights, 0.02 m < ZQH < 25 m, the following equation
for D has been presented by Stanhill (1969):

       log D. = 0.9793 log ZQH - 0.1536                              (3-138)

This equation results from a linear regression analysts based on the
published data for nineteen different crops with limited data measured for
the same crop at different growth stages.

    With estimates of ZQ and D, U* (friction velocity) can be estimated if
the values of the stability parameters (i|>  and h) are known.  These two
variables are closely related to Ri, the Richardson number, which is the
measure of the rate of conversion of convective turbulence to mechanical
turbulence.  It is defined as follows (Wark and Warner 1976):
            (g/T) (aT/aZ)
       Ri = - »                                           (3-139)
               (au / szr
where
                                       o
    g = acceleration of gravity (m sec  )
    T = potential temperature (°K)
    Z = elevation (m)
    U = wind velocity (m s  )

Potential temperature is defined as the temperature which a parcel of dry
air would acquire if brought adiabatically from its initial pressure to a
saturated pressure of 1000 millibars (Perkins 1974).  In application of the
model, the measured temperature is used in the Richardson number estimation
as suggested by Rosenberg (1974).

    The sign of Ri indicates the atmospheric condition,  and its magnitude
reflects the degree of the influence.   There are several  different formulae
for relating Ri to the atmospheric stability parameters;  for these purposes,
the sign of Ri  is of greater concern than its magnitude.   When Ri is larger
than 0.003, the atmosphere exhibits little vertical mixing, reflecting
stable conditions; when the absolute value of Ri  is less  than 0.003,  neutral
stability conditions exist (Oliver 1971);  and when Ri  is  less than -0.003,
convective mixing becomes dominant and  atmospheric conditions are unstable.


                                     67

-------
    To relate the atmospheric stability parameters to the  Richardson  number,
Thorn et al. (1975) proposed the following  formulas based on  the  work  by Dyer
(1974) and Dyer and Hicks  (1970):

For stable conditions

       4>h = m = 1 + 5.2 Ri                                          (3-140)

For unstable conditions

       *h = ""m2 = C1 - 16  R1 )

For neutral conditions

       4>h =   )  - 2 tan"                                   (3-143)

Under neutral conditions, 4*  = 0  and  the  equation is not used.

    In the application  of these procedures,  the calculations are performed
as follows:

    1) Evaluate Richardson number from temperature and wind velocity
       gradients.

    2) Determine stability condition  based on calculated Ri.

    3) Calculate 4>h and  , based on the  stability condition and associated
       equations (3-130), (3-131) or  (3-132).

    4) Calculate i|> , from equation (3-133).

    5) Calculate Z0 and D from canopy  height  using equations (3-127) and
       (3-128).

    6) Estimate KZ(Z), by applying equations  (3-125),  (3-124), and  (3-123).

    The resistance approach for the estimation  of volatilization flux  from
the plant canopy--The calculation of the  volatilization  flux from the  plant
canopy is based on a resistance-type approach using  the  values of Kz

                                      68

-------
discussed above.  For pre-plant pesticides, and time periods just after
emergence and post-harvest, transport by volatilization from plant surfaces
is much less than vapor phase transport by other mechanisms.  For those
conditions in which the plant leaves do not act as significant sources or
sinks for pesticide vapor, the resistances of the air for the whole plant
compartment can be estimated as follows (Mehlenbacher and Whitfield 1977):

       ZR = Rb(j + Rpc                                               (3-144)

       K,  | = —r\	                                                   ( j~ InD 1
              LJ

       Rpc = jCH  dz                                                (3-146)
             c[    z
where

    zR   = total vertical transfer resistance (day cm" )
    Rbd  = boundary layer resistance (day cm  )
    d    = thickness of the stagnant boundary layer (cm)
    D    = diffusion coefficient in air (cm  day  )
    Rpc  = plant canopy resistance (day cm'1)

The flux is calculated as follows:
       Jpc - ACgs/lR                                                 (3-147)
where
    Jpc = volatilization flux from plant canopy  (g day   )

    Cqs = pesticide vapor concentration at the soil-atmosphere  interface

The average concentration of pesticide vapor within the plant canopy is then
given by summing the vapor fluxes into (due to soil,  pond volatilization)
and out of the plant canopy:

            " CCAN + (JGD + °pv - Jp

where C-... is the concentration of vapor in the plant canopy at the start of
the daily TFAT time step (g cm"3) and Cp^jJ: is the concentration at the end
of the daily step (g cm~3).

Soil Temperature Simulation—
    Soil temperature is modeled in order to allow the program to calculate a
temperature gradient for the Richardson number.  The Richardson number is
required so that atmospheric stability information can be passed to the
FSCBG model.

                                     69

-------
    Although a number of good models are available to calculate an energy
balance at the soil surface to obtain the soil surface temperature and
subsequently calculate soil profile temperatures, it was thought to be
premature to go to this level of detail in TEEAM.  Therefore, a regression
type model is used to calculate the soil surface temperature of the form

       Tb = A + BTa                                                 (3-149)

in which

    Tb is the soil surface temperature (°C)
    Ta is the ambient air temperature (°C) and
    A and B are regression coefficients.

    Regressions on actual air and soil surface temperature data indicate
that the linear relationship is quite adequate to provide this prediction on
a monthly basis.  Over the course of a day, it is suspected that the
relationship would also hold.

Pesticide Granules—
    In addition to soil and aerial applications, pesticide may also be
applied to the soil surface in the form of pesticide granules.  Granules are
often applied as a means of controlling or slowing the release of pesticide
into the environment, and may consist of kaolin, biodegradable polymers, or
numerous other substances.  Pesticides are incorporated into the granules by
impregnation of preformed granules, coating of nonadsorbent granules, or
extrusion of a mixture of granule medium and pesticide.  The release of
pesticide from granules may occur by one or more of the following
mechanisms:

    •  Dissolution of the granule by rain or ponded water
    •  Diffusion and leaching of pesticide through the granule pore
       structure
    •  Leaching of pesticide from coated granule surfaces
    •  Volatilization of pesticide (i.e., vapor movement)

The relative importances of these release mechanisms will depend upon
properties of the pesticide (vapor pressure, solubility, credibility),
properties of the granule medium  (pore structure, credibility), the method
by which pesticide is incorporated into the granule, and environmental
conditions (rainfall, temperature).  Animals are exposed to granule-applied
pesticide by either direct ingestion of granules or through the release of
granule pesticide into soil and ponds.

    Because the release of pesticide from granules varies greatly depending
upon the granule formulation, there is as yet no general physically-based
model of the fate of granules in agricultural fields.  The release of

                                    70

-------
 pesticide  from  granules  in TFAT is therefore  modeled  using  a first order
 rate  equation.   The  change in granule  pesticide  concentration over time is
 then  given by:
         dC
           GR  = -KT.D Cco                                             (3-150)
          dt       GR  "GR

where

     CGR  = the  concentration  of  pesticide  in  granules  (g  g   )

     KGR  = a  9ranu1e rate  constant (days   )

Integrating  the  above equation  with  respect  to  time yields  the  form used  by
TFAT to  calculate granule concentrations  and release  rates:

       Cpn  =  CrD exp(-knDAT)                                        (3-151)
         uK    oK       oK

where

     C«R   = the concentration of pesticide in granules at the  start of the
             day  (g g"1)
     CGR   = the granule pesticide concentration  of  the end of  the day (g g  )

     AT    = the daily  TFAT time  step,  equal to one  day

The mass  of pesticide released  to the soil surface during the TFAT time step
is calculated as:

              t
       LPD - CnDMnD  (l-exp(-KrDAT)1                                 (3-152)
       uK    ul\ ur\           bt\

where

    LQR = the mass of pesticide released from granules (g)
    MGR = the total  mass of applied granules  (g)

The rate  constant KGR will obviously depend upon both the granule
formulation and environmental conditions.   To reflect the dependence of
release rates on moisture conditions, KGR is  estimated as a weighted average
of a fully saturated rate constant and a dry  rate constant:
where

       e = the water content in the top soil (cm  cm" )
      e  = the saturated water content in the top soil layer (cm  cm" )
                                     71

-------
    Kwet = the re^ease rate constant for granules immersed in water (days  )
    Kdry = t'ie re^ease rate constant for dry granules (days  )

Thus, the granule release rate constant KQR is recalculated at the start of
each TFAT daily time step to reflect current moisture conditions.  The rate
of release will range from a rapid "wet" rate when moisture levels are high
to a slower dry rate (i.e., due -to volatilization) during dry periods.  Note
that it is assumed that moisture conditions for the granule can be
represented by the degree of saturation in the surface soil layer.

3.3  PLANT GROWTH MODULE (PLTGRN)

3.3.1  Introduction

    Plant growth can be viewed as an ancillary ecosystem process which must
be simulated in order to predict the uptake and translocation of pesticides
or other xenobiotics from the soil, either to estimate exposure to
herbivores or to predict direct effects on plants themselves.

    Plant growth is difficult to describe mathematically, as it is a complex
function of the interaction between many chemical, physical, and climatic
factors, including moisture availability, nutrient availability,
temperature, radiation, soil texture, and plant physiological features.  The
change in plant biomass with these factors is necessary information for the
toxicant application module (for simulating the effect of plant canopy on
deposition), the soil fate and transport module (for simulating the effect
of roots on water movement), the plant contaminant translocation module (for
estimating the toxicant burden in plant biomass), and the terrestrial  areal
exposure module (for estimating the available plant biomass, both
contaminated and uncontaminated, for ingestion).

    In PRZM Release 1 there was a simple model which simulates plant growth
as a linear function of time.  Plant parameters, including canopy areal
development, plant biomass, and root depth, are assigned initial values,
usually zero, at emergence and take on maximum values at maturity.  At all
intervening times, plant parameter values are weighted by the fraction of
the growing period that has elapsed.

    This plant growth model has been replaced in TEEAM.  The motivation is
that, at a later date, environmental conditions and the presence of
toxicants which may affect plant growth can be accounted for.  This cannot
be done with the current model.  The plant growth module used is a
relatively simple model which is applicable to a variety of plants, both
annuals and perennials.  The model is based on the crop growth formulation
developed for the USDA's Erosion-Productivity Impact Calculator (EPIC).  The
documentation for EPIC is in draft form and the actual FORTRAN code is

                                      72

-------
unclear, with few comments describing its operation; as a result, the code
used in TEEAM was rewritten from previous articles describing EPIC  (Williams
et al. 1983; Williams et al. 1987).

3.3,2  Development of TEEAM Plant Growth Module as Adapted from EPIC

    The EPIC plant growth model predicts potential biomass increase as a
function of solar radiation, leaf area index (LAI), and hours of available
sunlight.  Actual biomass increase is the potential biomass increase reduced
by growth regulating factors:  temperature stress, water stress, and
nitrogen or phosphorus limitation. .The TEEAM version of this model
currently considers only temperature stress.

    The potential growth rate is defined by the equation:
        d(B_)
        -jf- = (BE) (PAR)                                          (3-154)

where

    Bp   = potential biomass (kg ha  )

    BE   = crop-specific parameter for converting energy to
           biomass (kg nr MJ'1 ha   day'1)
    PAR  = photosynthetic active radiation  (MJ  m~2)

    The photosynthetic active radiation is  computed once each day utilizing
the formula:

       PAR = 0.02092 (RA) {1 - exp[-0.65 (LAI + 0.05)]}             (3-155)

where

    RA   = solar radiation (ly)
    LAI  = field level leaf area index (decimal)

    The leaf area index is also computed at the start of each day.  It is a
function of the computed biomass and asymptotically approaches a maximum
value in the growing season after which it  starts to decline.  This is
expressed by the following relation:
                       (LAI  ) (WLV)
              WLV + 5512 exp [-0.000608 (WLV)]  » Bl - DLAI
       LAI =         .   R
                     1 - b1  ?
              IAT  f 	i_l^                   D  > ni_AI          (3-156^
                 o 11   DLAI^                  '  1   LM-Mi          \o-ijv)

-------
where
       mx
LAI
WLV"
LAIQ
Bl

DLAI
maximum potential  leaf area index (dimensionless)
aboveground  biomass  minus yield (kg ha   )
leaf area  index  when B^ equals DLAI (decimal)
dimensionless  expression of accumulated  biomass
(fraction)
fraction of  growing  season after which LAI  begins  to decline
(fraction)
This relationship  is plotted for the case of  B-^  <  DLAI in Figure 3.7.   As
the plot demonstrates,  LAI reaches 90% of LAImx  when WLV is approximately
4100 kg/ha.  This  relationship is utilized  for all  crops; the only crop-
specific parameters in  this relationship are  LAImx, the maximum leaf  area
index, and DLAI, the fraction of the growing  season when LAI begins to
decline.

    The variable B^ is  calculated as a function  of accumulated degree days:

                HU
   Bi -
                                                                       <3-157>
                         LAI/Ulmx-WLV/(WLV+5512»E(-608E-6*WLV))
    x


    N
            024
                                       (Thousands)
                              Above ground biomass minus yield


         Figure 3.7 Leaf area index as a function of above-ground biomass minus yield.
                                      74

-------
where

    HU   = heat unit (°C day)
         = (T-Tb) (1 day) if T (the average air temperature, °C,
           for the day) is greater than Tb (a user specified base
           temperature, °C)
    PHU  = parameter for crop specifying the (potential) heat units
           necessary for maturation (°C day)
      n  = number of days

    The value of WLV is computed daily using the relation:

       WLV = Bp - RWT - YLD                                         (3-158)

where

    RWT  = root biomass (kg ha'1)

    YLD  = crop yield (kg ha'1)

    Change in yield is computed with the relation:
where

    GK = ratio of total biomass to crop yield under favorable
         growing conditions (decimal)

    Root growth is computed with:

        diRWJl = d{jJEl (0.4 - 0.2 Bj)                               (3-160)

    Root depth is another variable necessary for the TFAT module.  It is
computed with the relationship:
                = 2.0 (RZ) (HU)/(PHU),  DPTH < RZ                   (3-161)

where

    DPTH = root depth (cm)
    RZ   = crop-specific maximum root depth (cm)


     Finally, the actual plant growth  is the potential biomass increase
reduced by the growth regulating factor:
                                     75

-------
                     REG                                            (3-162)
         \j u     vi \f

where

    B    = actual biomass (kg ha  )
    REG  = growth regulation factor  (fraction)

    REG is also used to reduce the rates of YLD,  RWT,  and DPTH  increase.

    In the TEEAM version, REG is only a function  of temperature stress:
                     T  - T 7
       REG = exp {n [  ° T  p}                                      (3-163)

where

    n  = temperature stress parameter for the crop (decimal)
    TQ = optimal air temperature for the crop (°C)
    T  = daily average air temperature (°C)

    The temperature stress parameter is computed  from user-supplied
temperatures representing the optimal growing temperature (TQ)  and a minimum
(base) temperature at which the plant can grow:

        „ .  ln(0.9)                                                n
            [Vll
where
    T  = the mean of the optimal temperature
         (TQ) and base temperature (Tb) (°C)
            2
    The EPIC plant growth model also includes the simulation of the weight
distribution of roots with depth.  This additional variable is not currently
simulated in TEEAM, but may be added easily if water stress is considered  in
future versions.

 3.3.3  EPIC Modifications for  Inclusion in TEEAM

    The model described above  is, in modified form, included as the TEEAM
 subroutine PLTGRN.  PLTGRN returns the following variables to TEEAM for
 subsequent calculations:  plant above ground biomass (kg m  ), root biomass
 (kg nf ), root depth  (cm), canopy cover (fraction), and canopy height  (m).
 Plant above ground biomass is  computed as a straightforward transformation
 of the computed total plant biomass minus the root weight:

                                    76

-------
       WEIGHT = (B - RWT)/10,000                                    (3-165)

    The root biomass used in subsequent TEEAM calculations is the live root
biomass calculated within PLTGRN converted from kg ha   to kg m  :

       ROOTWT = RWT/10,000                                          (3-166)

    The root depth is passed directly to TEEAM from PLTGRN since the units
are the same in both code sections (m).

    Canopy cover is computed from the following function (a logistic
function) of leaf area index (LAI):

    COVER = COVMAX/U.O + ex'p(4.61 - 3.07 LAI)]                     (3-167)

where:

    COVER  = fraction of ground covered by the plant
    COVMAX = crop specific maximum cover

    This function was selected since it approximated (Figure 3.8) the
relationship described in the PRZM documentation (Carsel et al. 1984)

    COVER = [2.0 - Erfc (1.33 LAI - 2.0)1/2.1                       (3-168)

where:

    Erfc = complementary error function

    Equation 3-167 is used rather than 3-168 since values for COVER are
required for values of LAI less than 1.0 (Equation 3-168 is not defined for
values of LAI less than 1.5).  LAI as computed within PLTGRN is a field
level value, e.g., the amount of leaf area in a field per area of the field,
and can have values less than 1.0.

    Canopy height is not computed as part of EPIC.  For the TEEAM plant
growth model, the EPIC rooting depth equation was adapted to simulate canopy
height over time:


                  = (KHGT) (REG) (HU)/(PHU)«   HGT < HGT            (3-169)
                                                        m
where

    HGT    = canopy height (m)
    KHGT   = crop-specific maximum rate of height increase (day  )
    HGTmx  = crop-specific maximum potential height (m)


                                     77

-------
  I

  X
  0
  o
                                                Q Williams equation (Error function)
                                                + TEEAM equation (Logistic function)
                                  Leaf area Index
                    Figue 3.8  Cover as a function of leaf area index.
3.4  PLANT CONTAMINANT TRANSPORT MODULE  (PLTRNS)

3.4.1  Introduction

    This submodel simulates contaminant  transport within plants of the
simulated ecosystem.  It is necessary  to simulate this process to determine
the quantity of contaminant which enters the plant biomass and to determine
the concentrations of contaminant within various  types of plant tissue.
These concentrations can then be used  in determining detrimental effects  to
the plants and to determine the concentrations in plant tissues which may
form part of the diet of the herbivores  and  omnivores- of the simulated
ecosystem.  To simulate the total concentration of contaminant within and on
plants, it is necessary to determine the rate of  uptake of the contaminant
through the root system and to determine both the amount of contaminant
which is deposited on and remains on the aboveground plant matter after
toxicant application to the ecosystem.

    Ultimately, it would be preferable to simulate the plant tissues which
might be preferentially eaten by the animals in the simulated ecosystem
(e.g., leaves, fruit, seeds, stems, and  roots).  At the current time,
however, the plant contaminant transport model only simulates aboveground

                                     78

-------
and root biomass concentrations.  There is the inherent assumption that all
plant material which is aboveground has the same contaminant concentration
regardless of the plant tissue types which might be involved.  (Similarly,
all root tissues would have an associated "root concentration" which would
be the same for all roots.)

3.4.2  Background

    A plant contaminant transport model (UTAB) being developed at Oregon
State University (Boersma et al., in press) was evaluated for its potential
use as a module within TEEAM.  In this evaluation it was determined that, in
its current state,  UTAB would be overly complex (requiring too many input
parameters and not  commensurate with the level of detail  of the plant growth
model) to be used in the TEEAM structure.   However,  at  a  later date,  it may
be appropriate to link this model into the TEEAM framework.

    Instead, a relatively simple model which has readily  available
parameters which determine the amount of contaminant being transported into
the roots and the amount passing from the roots to the  aboveground structure
has been implemented.  This model is based upon the  experimental  work of
Briggs et al. (1982) relating the rates of root uptake  and translocation to
the lipophilicity of the applied contaminant.

3.4.3  Development  of Module

    The model has two-compartments; the contaminant  flows into the roots
from the soil solution and flows into the aboveground biomass from the roots
(Figure 3.9).  The  root compartment is simulated as  a completely mixed
reactor.  The input to the roots is the soil solution concentration times
the flow rate reduced by a reflection coefficient (Rw)  representing the
resistance of the root surface to the chemical.  Within the roots, the
contaminant can degrade at a user-specified first order rate (x).  The
contaminant leaves  the root through the transpiration stream leading to the
aboveground biomass.  This rate of transfer is governed by a
root/aboveground biomass reflection coefficient (Rr).

    The aboveground biomass compartment is simulated as having two phases,
an aqueous and a nonaqueous phase.  The transfer between  these phases is
governed by a partition coefficient (Kp).   The contaminant can decay within
this compartment at the same rate as specified for the  roots.  The
contaminant can leave the aboveground biomass  to the atmosphere governed by
an aboveground biomass/atmospheric reflection  coefficient (Re).

    The mass balance equation for plant roots  is written  as:
          dC
       Mr dT = Rw  Cw ^r - Rr Cr <>t ' X Cr Mr

                                    79

-------
in which

    Mr = the total weight of plant  roots (wet basis) (g)

    Cr = the concentration of  contaminant in the roots (g g"1)

    RW = a reflection coefficient for contaminant transfer from  the  soil
         solution to the root  (decimal)

    Cw = the concentration of  contaminant dissolved  in  the soil solution
          (g cnf3)
                                            Q     1
    Qr = the flow into the root biomass  (cm  day  )

    Rr = a reflection coefficient for  contaminant transfer from the root
         to the aboveground transpiration stream (g  cm~3)

    Qt = the flow rate into the stem  xylem  (cm3 day"1)
                     i
    x  = the first order degradation  rate of the chemical in the plant
         tissue (day  )

The first term on the right-hand side  represents the gain to the roots  from
the soil solution; the second, the  loss  to  the aboveground biomass; and the
third, internal degradation.

    For the purpose at hand, the flow rate  Qr and Qj. are assumed to be
equal, indicating no change in water  storage within the plant.  These flow
rates are the product of a velocity and  a cross-sectional area within
various portions of the plant. The velocity is defined throughout the  plant
to be uniform and equal to TDET, the  evapotranspiration rate given by TFAT
             Roots:
                                               Above ground
                                               biomass:
                                               Non-aqueous phase
                                                        Transpiration
                                                             stream
           Figure 3.9  Schematic of plant contaminant transport module (PLTRNS),
                                     80

-------
(cm day  ).  Therefore, to maintain continuity of flow, Ar and At, the
respective cross-sectional areas of flow in the roots and plant tops, must
be equal.  Since TFAT simulates vertical transfers of water and contaminant
for a unit area (cm ) these areas must also take on that dimension.

    In reality, the velocity of water movement changes greatly in the plant
due to variation in cross-sectional flow area and losses/changes in storage
which actually occur.  In order to maintain simplicity, constant velocity
and flow areas have been utilized.

    At steady-state conditions and with the assumption of no internal decay,
Equation (3-167) reduces to:
     C    R
     r = r                                                        (3-171)
      w    r
This ratio is equal to the root concentration factor (RCF, with units of ml
(or cm3) of external solution divided by the mass of roots in grams) as
defined by Briggs et al . (1982).  Thus, the root/aboveground transpiration
stream reflection coefficient can be expressed in terms of RCF and the root
uptake reflection coefficient:
    Rr -                                                            (3-172)
    The assumption of no internal decay is valid for the above analysis
since, in the experimental determination of RCF, degradation had been
accounted for.

    Using the above relationship; the fact that Qr = Q^. = (TDET) A, where A
is the total area of the simulated ecosystem; and dividing both sides of
Equation (3-170) by Mr, the following relationship is obtained:
        dC    R  C  TDET ARC  TDET A
        _j: =  w  w _ _  w  r __  c                      (3-173)
        dt        Mr             RCF Mr     XLr                     (     '

    The ratio of Mr to A is the root biomass density simulated by the PLTGRN
module.  Cw and TDET are simulated by the TFAT module.  Given values of Rw,
RCF, and x , Equation (3-173) can be solved numerically.

    The mass balance for the contaminant in the aboveground biomass can be
written as:
           dC
where
       Mag -d   = Wt - ReVe - xCagMag
    M   = the total weight (wet basis) of the aboveground biomass (g)


                                    81

-------
    C,a = the concentration of  contaminant in the aboveground biomass
      9    (g g-1)

    Re  = the reflection coefficient for transfer to the atmosphere
          (decimal)

    Cj.  = the concentration of  contaminant in the transpiration stream
           (g cm'3)

    Qe  = the flow  of water into the atmosphere (cm3 day  )

    As with the root compartment, no change in water storage is simulated so
that Qt = Qe = (TDET) A.

    Solving Equation (3-174) for steady state and with similar reasoning as
previously described, assuming  no internal decay, produces:
     Ct   Rr
     r - r                                                         (3-175)
      r    e

The ratio of Ct to  Cw is defined by Briggs et al . (1982) as the
transpiration stream concentration factor (TSCF,  with units of ml  of
external solution divided by mass of water in the transpiration stream in
grams).

    Using this definition of TSCF and the previous definition of RCF,
Equation (3-173) can be rewritten:

        Rr   TSCF                                                    ,, 17,x
        R; = RCT                                                    (3-176)
Given this and the  relation obtained in Equation (3-171), Rg can be written
as a function of Rw:
              ^
        Re =                                                         <3-177)
Using this relationship, the assumption of no change in internal water
storage, and dividing Equation (3-173) by M   produces:

        dCart   R,, C,. TDET A   R  C, TDET A
          ag _  w  r _    w  t _    r                      /o
        dt        RCF M     '   TSCF M     " XLag                    ^J
                       ag             ag

    The only undefined component in the above equation is C^., the
concentration of contaminant in the aqueous phase (transpiration stream).  A
mass balance of the two phases within the aboveground biomass compartment
can be written as:
                                     82

-------
       Vag = Vt + ™naCna                                       <3-179>

where

    Vt  = volume of the aqueous phase (cm )
    m   = mass of the nonaqueous phase (g)
    C   = the concentration of contaminant  in the nonaqueous phase (g g~ )
     na

    By defining the following parameters, assumed to be constant for a
plant/contaminant combination:

    4>    = aboveground biomass water content (cm  g  ) [or equivalently, for
           water (g g  )]
    K_ = partition coefficient for plant, the ratio of the concentration in
           the nonaqueous  phase to the concentration in the aqueous phase
           (cm3 g'1)
    p    = ratio of the mass of the nonaqueous aboveground biomass (dry
           weight) to the  total aboveground  biomass (wet weight) (g g  )
           (essentially, the percent dry matter)

    Ct can be written in terms of Ca  using Equation (3-177):

       Ct = V^a + Vna^                                        (3-180)

    Given this relationship, Equation (3-172) can be solved numerically for
    In the plant translocation module the root and aboveground
concentrations C- and Caq are computed by a finite difference solution of
Equations (3-175) and (3-180).  A backwards difference scheme is used to
approximate the time derivatives:
        dcr
        dT - -Sfi^1                                             <3-181>
        dC     Ct+1 - C*
        _   . __                                              (3_182)

where the superscript t refers to values at the start of the daily time
step AT, and t+1 refers to end of time step values.  Equations (3-173) and
(3-178) are then written in finite difference forms:
C t+1 - C t
r r
AT
R C t+1TDET A
w w
M t+1
r
R C t+1TDET A
w r
RCF Mrt+1
r t+1
r
                                     83

-------
          an      an                          TDET A
          §9 _ §3
            _
                          RCFH            TSCFMag

All of the terms in the above equations are known at the end of the time
step except for C     and C     .   Because Equation (3-182) contains the
root zone concentration, Equation (3-183) is solved first for the new value
of C    .  C      can then be determined explicitly from Equation (3-184).
    r       ay

    The parameters TSCF (transpiration stream concentration factor) and RCF
(root concentration factor) are expressed as functions of Kow by
Briggs et al. (1982):

       log (RCF - 0.82) = 0.77 log K   - 1.52                       (3-185)
                                    ow

       TSCF = 0.784 exp - [(log KQW - 1.78)2/2.44]                  (3-186)

in which KQW is the octanal/water partition coefficient.

    Obviously, major biochemical processes are ignored in this model.
Descriptions of these processes are lumped into the empirical coefficients
such as TSCF and RCF.

3.5  TERRESTRIAL ANIMAL EXPOSURE MODULE  (APUM)

3.5.1   Introduction

     In  preceding sections, we have described the simulation of toxicant
movement within the  soil /pi ant/atmosphere systems.  These systems provide
the vectors for toxicant uptake by ecosystem fauna.  This section describes
the processes through which animals are  exposed to and accumulate the
toxicants.

    Once the toxicant concentrations  in  the environment are known, two
factors control accumulation:   (1) the presence of the animal in areas where
the toxicant is present and  (2) the uptake of the toxicant into the body of
the animal.  Unlike  plants, animals enjoy varying degrees of freedom in
their movement.  Therefore,  if  they are  capable of detecting the toxicant
they may avoid or  be attracted  to contaminated areas.  Even in the absence
of a toxicant, animals may move preferentially to various portions of their
habitat.   Factors  which affect  their  movement include loss of habitat,
predator avoidance,  seasonal ity, and  food search.  Even though their
movements  are certainly determined by environmental and species-specific
behavioral  factors,  there  is  also an  inherent degree of randomness in their
movements  within their  home range.
                                     84

-------
    The second factor in evaluating accumulation is the uptake of the
toxicant into the animal's body tissues.  Figure 3.9 demonstrates potential
routes of exposure.  The animal may ingest a number of contaminated
vectors:  water (i.e., ponded water at the soil surface), soil, living or
dead plant material, pesticide granules, or living prey.  In certain
species, vapor-phase toxicants or chemicals adsorbed to particulates may be
inhaled.  Other species may absorb chemicals through dermal contact with
contaminated soil or water.  The magnitude of uptake via these routes will
obviously be species- and chemical-dependent.

    Once the chemical is taken into the animal's body, it may accumulate in
various organs, metabolize, filter through the endocrine system (in higher
animals) or leave the body through elimination.  Depending upon the species,
elimination may occur through the skin or in excreted wastes.  Death of the
organism and subsequent decay of the body materials will ultimately return
the chemical to the soil.  Organ-specific accumulation and return to the
soil, via mortality and decay, are not considered in this model.

3.5.2  Module Development

    The terrestrial food chain model consists of three principal
components:   (1) animal movement, (2) animal feeding, and (3) toxicant
assimilation.  The animal movement component determines how often animals
come into contact with soil, prey, and other sources of toxicant
contamination.  The animal feeding component calculates the mass of food
ingested by animals, and the toxicant assimilation component determines how
much toxicant is assimilated into body tissues from ingested food and other
sources.  The model currently assumes that population levels are at steady-
state; an additional component can be added at some future time to simulate
birth and death rates.

3.5.2.1  Animal Movement—
    The location and movement of animals are major factors in determining
exposure.  Unfortunately, there appear to be few simulation models which
describe animal movements.  In the absence of more species-specific or
detailed models, a simple Markov model is used to move animals among
environmental compartments.  This type of model allows for a degree of
randomness  in the movement of the animals with respect to the toxicant.
Realistically, animals are not uniformly or continuously exposed and such a
model will account for this fact.

    The Markov model moves animals among habitats, or within a habitat
among soil horizons, based on the animal's current location and a
transition probability matrix.  Soil animals move among soil horizons,
while higher order predators move between habitats.  The transition
probability is simply the probability that the animal will emigrate to

                                     85

-------
location "b" given that it is currently in location "a".   Several  such
matrices could be input and used to vary the movement of  animals on a
seasonal or life-stage basis or in consideration of modification of
behavior triggered by various environmental  conditions.

    Mathematically, the state of such a system at time 't'  is given by

       pW = p^-1)?                                               (3-187)

where p is a vector describing the location or-distribution of the animal
population, the superscript 't1 denotes time, and the matrix P is a
transition probability matrix.

    The transition matrix, P, contains elements fab defined by

        fab  =  Prob  (Xt  = b|Xt_ra)                                    (3-188)

 In  other words, the probability that the organism  is  in  a  given location  (b)
 depends upon  its location  (a) during the previous  time step.  P is an m x  m
 matrix  of the following form, where m  is the  number of possible locations:
         P  =  f                                                      (3-189)
              •
              •

 '             ml          mm

 and  has  the following property


         z fab =  1                                                    (3-190)
         D
 (Haan  1977).

     The  output of the model  is a vector of probabilities  P^'  or  Pc^.
 The  ph represent the probability that an animal  is  in  a particular habitat
 (h)  and  the p_ refer to the  probability that the animal will  be in a given
 soil horizon (c) within a given habitat during a time  't1.  Conservation  of
 mass requires that

          j! Ph(t) - 1                                                (3-191)

 and

          I Pc(t) = 1                                                (3-192)
                                      86

-------
    This model can be used in one of two ways; either to describe the
random movement of a single individual or subpopulation moving en masse, or
to describe the distribution of a population in time.

    In the first mode of operation (en masse movement of animals) the
probability that the animal moves to each possible new location  is
determined from the previous location of the animal and its Markov
transition matrix.  A uniform random number between 0 and  1 is generated to
determine which habitat or soil horizon the animal (or animal subgroup)
actually moves to.  Although the location of the animal at any particular
time is selected randomly, the transition matrix ensures that the animal
will exhibit preferences for habitats and compartments with high transition
probabilities.

    The following example illustrates this first mode  of  operation.
Suppose there are three distinct habitats of interest:   a nesting area, a
contaminated feeding area, and an uncontaminated feeding  area.   The
transition matrix might look like the following:
       P =
a
1
1
2
3
0.50
0.70
0.25
b
2
0.25
0.25
0.25
3
0.25
0.05
0.50
                                                                    (3-193)
Reading the first row, the probability that the animal  will be in habitat 1
given that it was in habitat 1 in the previous time step is 0.5.  The
probability that it will be in habitat 2 given that it  previously was in
habitat 1 is 0.25 and so forth.

    Once the initial vector of the animal location (p'    ') is given
(habitat 1, 2, or 3) the matrix can be used to move the animal among
habitats.  Suppose the animal is initially located in habitat 1, so that

       p(t=0)  =  0                                                 (3-194)
                  0
    A uniform random number between zero and one is then chosen.  If the
number chosen were 0.89, the animal would move to habitat 3 since 0.89 is
>0.50 (fj_j_) and also >0.75 (f12 + fll)-  At tne next ^me steP> another
random number would be chosen and the animal would move based on the
transition probabilities from habitat 3.

    The elements of matrix P can be designed so that the animals may move
in a completely random fashion (for instance all the fab very nearly equal)
or completely deterministically (single non-diagonal f^ on each row being
unity).
                                    87

-------
    The second mode of operation is to use the transition matrix to
determine the distribution of a population among several habitats or soil
horizons through time.  Suppose the initial distribution among habitats is
given by:
              °-67
            = 0.20                                                  (3-195)
              0.13
In this example, 67 percent of the population is in habitat 1, 20 percent
is in habitat 2, and 13 percent is in habitat 3.  The distribution of the
population at the next time step is given by:
            » p(°)p                                                 (3-196)

    In general, the distribution at time t is given by:

       pW - p'*-1'?                                               (3-197)

As time becomes large (t * ») the population distribution reaches a steady-
state condition such that:

            - p^1*                                                (3-198)

    Although it is felt that accounting for randomness in animal  behavior
is important in determining exposure, it seems equally important  to be able
to simulate more deterministic behavior patterns which may result in
modification of exposure.  Such behavior might include the following:

    •  Aggregation of animals at ephemeral ponds for drinking or  bathing

    •  Saturated soil conditions which might cause earthworms to  migrate to
       the soil surface

    •  Modification of feeding caused by extreme precipitation or
       temperature

    An appropriate means of handling this within the framework of the
proposed model would be to supply special transition matrices which would
be used in conjunction with the occurrence of these environmental
conditions.  For instance, if the water content in a soil compartment goes
to saturation, then the special transition matrix would reflect a high
probability that earthworms would migrate to the surface of the soil.  If
the ambient air temperature is above a certain threshold, the transition
matrix might reflect a high probability that a terrestrial mammal might not
venture forth to feed.  In the current version of the model, only one
transition matrix can be input and behavioral responses to these special
environmental conditions cannot be simulated.

                                     88

-------
3.5.2.2  Animal Feeding—
    A wealth of information is available on food intake rates and food
preferences for many of the species of interest on a seasonal or life-stage
basis.  Some examples of the literature can be found in Section 6.7.3.   In
lower trophic  levels, animals may feed more or less continuously unless
they are in their nesting habitat.  Therefore, their total intake of food
and the mixture of items may be fairly constant (given unlimited
availability).  In the upper trophic levels, animals (carnivores, raptors)
may kill only intermittently.   In this  case the modeling  is more
involved.  Factors may include location of predator relative to  prey,  time
elapsed since the last kill  and probability of  prey capture.   In addition,
once a kill has been made the  predator  may devour  only  certain portions of
the prey.  In this case,  modeling of the concentrations of toxicants in
various organs of animals in intermediate trophic  levels  might be useful;
however, this is not available in this  version  of  TEEAM.

    The feeding model assumes  that an average daily food  intake  rate ul
can be established for each species.  Here the  superscript "i" denotes a
given animal group or population.  The  total daily intake is broken  down
into the uptake of specific prey by means of preference factors.  Thus the
specific daily consumption of  food item j by predator i under ideal
predation conditions is

       uj = 8J UJ                                                   (3-199)

    el  is a vector of food preference  factors  of  species i for  prey j,
assuming that the availability of prey  "j" is typical  of  that in an
unstressed ecosystem.  The  Bn- , must sum to unity.  For instance,  if the
predator is a passerine bird and its ideal diet consists  of 25% seeds (prey
1), 50% earthworms (prey  2), and 25% soil macroarthropods (prey  3),  then
the vector e1 would be

        ,   1  0.25
       6=2  0.50                                                 (3-200)
            3  0.25
    Predator/prey relationships can be  established in the model  through the
use of this vector.   An entry  of zero for any species "j" in the vector
indicates that species "i" does not prey on species "j."   Food items
currently available in the model include other  animals, plants,  soil,
ponded water, inhaled air, and pesticide granules.

    Successful  predation  in higher-order animals depends  upon two
factors:  that both predator and prey are brought  into  contact (i.e.,  are
in the same habitat)  and  that  the predator is successful  at capturing the
prey.   The expression of  this  success in the model  is accomplished by the
                                     89

-------
use of probabilities.  The daily consumption rate of food item j by
predator i then becomes:

       U1. = e1. UJ T pi,.  pV                                        (3-201)
        J    J  i  t  J/h   h
                  n
where

    U1.   = feeding rate of predator i on prey j (mg/day)

    ?]..  = probability that predator i catches prey j
     ?'.    given that both are in habitat h
    PuJ  = probability that predator i and prey j are
     n     both in habitat h                           ^
  ..The probabilities that predator and prey are both in each habitat
(pZJ) are determined from the Markov animal movement model.   The capture
probabilities pi
availability of
 T                                                   '
p./h are conditional probabilities which assume unlimited
rprey.
    Because it is unrealistic to allow animals to continue to feed after
ingesting lethal dosages of toxicant, the feeding model decreases total
uptake rates as a function of cumulative toxicant dosages:

    UJ = Kf(UJ)°                                                    (3-202)

where

    UJ.=othe current total uptake rate for animal group i (mg day~ )
    (UJ)  = the initial uptake rate for animal group i when no toxicant has
             been ingested
    Kf = uptake reduction factor computed as a function of cumulative
         dosage (0 < Kf < 1.0).

Dosage in this case is defined as the cumulative mass of toxicant ingested
per unit biomass.  In the current version of the model Kf is calculated as
a linear function of dosage based on the 10 percent and 50 percent lethal
dosages:

     K  -  Q     V'5 - -9)  i nln ,    (.5 - .9)   nt                 C
     Kf ~ 'y ~ (LD50 - LD10)LD1° + (LD50 - LD10) U                  (
where
    LD10 = the dosage of toxicant which is lethal to 10 percent of animals
            (g g'1)
    LD50 = the dosage of toxicant which is lethal to 50 percent of animals
     t      (g 9  }                                      1
    D  =   the cumulative toxicant dosage at time t (g g  )

                                     90

-------
This assumes that at the LD10 , uptake is reduced by 10 percent (i.e.,  10
percent of the animal group is dead) while at the LD50, uptake is reduced
by 50 percent.  At zero dosage the uptake factor K^ is set to 1.0 to
indicate that uptake is unaffected by toxicant ingestion.   In addition,  if
the calculated value of K^ is greater than unity  or  less than  zero, the
program defaults the value to the appropriate  limit.

    If a population model  which accounts  for mortality  due to.toxic effects
is ultimately coupled with this model,  the factors pin and e1 may be
altered depending upon the abundance of prey and  the use of the uptake
reduction factor (Kf) could be eliminated. In order for the  system to be
stable the condition

       MJ< > u! At                                                   (3-204)
        ^    J
must be satisfied, where M^ is the total  biomass  of  species j  during the
time step At.  Simply stated, the intake  rate  by  predator  "i"  of  prey "j"
cannot exceed the total biomass of prey "j."

3.5.2.3  Toxicant Assimilation--
    Toxicant assimilation within the organism  is  a complex process affected
by the following factors:

       • The biochemistry and physiology  of the  specific species  of
           interest

       • The properties of the chemical (e.g., lipophtlicity)

    Because of the complexity of the processes involved, the  modeling
approach must at this time be empirical.   Since  Corvallis  ERL personnel are
ultimately to provide models for the fate of  toxicants  within the animal
species of interest, the intent is to implement  a set  of very simple
algorithms in the prototype model which would  apply  to  idealized  species.
These simple models will ultimately have  default  parameter values which
will provide the capability to simulate broad  differences  among species in
toxicant assimilation and subsequent body burden.

    The mass of toxicant present in the biomass  of a group of animals will
vary over time due to assimilation, metabolic  degradation, and predation  by
other animals.  The mass balance equation for  toxicant  in  each individual
animal, subgroup, or population is therefore written mathematically  as:
            V^- = JI  -J]-Jl                                 <3-205>
                                    91

-------
where

    M^ = biomass of animal group i (mg)

    Cg = concentration of toxicant in animal  group i  (mg mg" )

    j] = toxicant assimilation rate (mg day  )
     a

    J^ = toxicant degradation rate for animal group i (mg day"  )

    JP = toxicant predation loss rate (mg day" )

    Toxicant assimilation will depend upon the rate at which toxicant is
ingested and the rate at which ingested toxicant is absorbed into the
tissues of the animal.  The rate at which toxicant is ingested  by animal
group i when it preys on food item j is given by the product of the feeding
rate U. (Equation 3-197) and the toxicant concentration in food items:
                      h                             i
where F^j is the rate of toxicant ingestion (mg day  ) by animal group i
preying on food item j and CJ is the toxicant concentration in food item j.
Ingested toxicant may then either be assimilated into the animal's tissues
or eliminated from the animal's body.  Toxicant assimilation is modeling
using an empirical efficiency factor representing the fraction of ingested
toxicant which is assimilated.  The toxicant assimilation rate for animal
group i then becomes:
                                                                    <3-207>
where o. . is the assimilation efficiency factor when animal group i eats
food item j.  The terms a1J must be greater than or equal to 0.0 but less
than or equal to 1.0.

    Metabolic degradation of toxicant within organisms is modeled as a
first-order decay process:

       Jd •      (C  H)                                            <3-208>
where K ^, is a first order rate constant for degradation of toxicant in
animal group i.  The predation loss term accounts for the removal of

                                     92

-------
pesticide mass when other animals prey on and ingest toxicant mass from
animal group i:
 Substituting the above rate expressions into the overall toxicant mass
 balance equation for animal group  i,
                                    j       h

    Organisms  are modeled  as subgroups, or populations depending upon the
 level of detail required in simulating animal movement.  Animals modeled as
 subgroups will move randomly en masse between locations as determined by
 the Markov  transition matrix; animals modeled as populations will be
 distributed among locations based on the product of  initial population
 distributions  and the transition matrix.  The toxicant mass balance
 equation is then written for each subgroup and population, resulting in a
 system of first order differential equations for the entire food chain.
 Movement probabilities  (Puj) are computed by the animal movement model, and
 a fourth-order Runge-Kutta integration scheme is used to solve the system
 of.mass balance equations  (Equation 3-210) for the mass of toxicant
 (Mo CM in  each subgroup or population at the end of each daily time step.

    The biomass of individuals and populations is assumed to be constant.
 Ultimately,  a  population model can be incorporated and the estimates of
 Mg at each  time step would be an output of that model.  The assimilation
 efficiency  factors o. .  can hopefully be derived from thermodynamic
 considerations or from  empirical approaches such as proposed by Jorgensen
 (1984).  They  are typically considered to be functions of the lipophilicity
 of the compound, fat content and metabolic rate of the predator, and the
 size of the particle or interaction of the chemical with the prey ingested
 (Tinsley 1979).  The assimilation efficiency is also a function of the
 uptake route (i.e., ingestion versus inhalation).

    Some food  items or  items ingested during the process of obtaining food
 can be handled as special  cases of Equation (3-206).  For instance, soil
 ingestion does not require a "capture probability" and only depends upon
 the product of the soil ingestion rate, the concentration of the soil in
 habitat "h", and the probability of being in habitat "h".  Uptake by
 inhalation  takes on a similar form.  Uptake in ingested water depends on


                                    93

-------
the presence of ponded water in addition to the factors  listed  above.
Direct ingestion of pesticide may also occur if the chemical  is deposited
in the form of granules.

    The Uj  for the food  items of soil animals at the base of the food
chain may be difficult to define individually.  These animals (earthworms,
arthropods, etc.) move between soil  horizons and are exposed  to pesticides
in soil, water, and ingested detritus.  The toxicant mass balance equations
for these organisms are therefore formulated differently than for the
higher-order animals:
        d(CM )n
        -dip- = WT - WW  - Jl                        <3-211>

In this model all of the  feeding forms are lumped into a single mass-
transfer coefficient,  kj, which describes the net uptake of pesticides
(C-pVj) from the soil.   Cj is the total pesticide concentration on a
total-volume basis and Vj is the total volume of the soil compartment.

    In order for the modeled concentrations in soil animals to approach
observed values in the literature, the values of ky and Kmet must be
carefully selected.  This is done in the model by making use of a
terrestrial bioconcentration factor, defined as the ratio of concentration
in the biomass (C0) to the concentration the soil (Cs>r
              C
        BCF = 7^                                                    (3-212)
              Ls
This ratio can be related to the coefficients Km ^ and kj.  If we assume
that the predation loss term J^ of the organisms is small compared to the
biomass M0, and that MQ is relatively constant, then Equation (3-211)  can
be rewritten as
        dC    kCV
        -d  ' -T     - KmetCo

The quantity Cj can be expressed as a function of Cs by making use of the
soil* chemical mass balance equation:

        CT - (PS + — ) Cs                                           (3-214)
                   Kd
where p  is the soil bulk density, e is the soil water content, and K^ is
the partition coefficient.

Equation (3-212) can then be written as:

         dCo = WpS + ^ Cs -                                   (3-215)
          dt           M
                                     94

-------
                 dC
At steady state, -rr^ ->- 0, and therefore

            (PS + £-) Cs
       kTVT        d     - K    C                                   (3-216)
Solving for the soil animal mass transfer coefficient:

       ,        o met o                                              /o oi-7\
       kT = - - -                                           (3-217)
            vT('.*S;>c.

At steady state the ratio C /C  is equal to the bioconcentration factor
BCF.  V-j- can be expressed as A AZ where A is habitat area and AZ is soil
compartment depth.  The term MQ/A is defined as the organism density,  PQ.
Substituting these relationships into Equation (3-217):
                p BCF
       kT = - ^ -                                            (3_218)
As long as BCF and Kmet are reasonably well known, ky can be defined from
Equation (3-218) so that soil animals attain a concentration of
approximately (CSBCF) at steady state.  This approach to parameter
estimation is similar to that described in Donigian and Dean (1985).  Review
of the literature on uptake and metabolism of chemicals by earthworms
presented in Roberts and Dorough (1985) indicates that in general data are
available to estimate coefficients in this way.

3.6  THE MONTE CARLO MODULE (MC)

    This section describes the Monte Carlo method used for uncertainty
analysis of the TEEAM model.  Given a set of deterministic values for each
of the input parameters, X^, X2 . . . Xp, the TEEAM model computes a number
of soil, animal, and plant output parameters Y^:

       Y! = g (X1§ X2, X3 . . . Xn)                                 (3-219)

    Application of the Monte Carlo simulation procedure requires that at
least one of the input variables, X^ . . . Xp, be uncertain with the
uncertainty represented by a cumulative probability distribution.  The
method involves the repeated generation of pseudo-random number values of
the uncertain input variable(s) (drawn from known distributions and within
the range of any imposed bounds) and the application of the  model  using
these values to generate a series of model  responses,  i.e.,  values of  Y^.
These responses are then statistically analyzed to yield the cumulative


                                    95

-------
probability distribution of the model response.   Thus,  the various steps
involved in the application of the Monte Carlo simulation technique involve:

    1) Selection of representative cumulative probability distribution
       functions for describing uncertainty in the relevant input variables.

    2) Generation of pseudo-random numbers from the distributions selected
       in (1).  These values represent a possible set of values for the
       input variables.

    3) Application of the model to compute the derived output(s).

    4) Repeated application of steps (2) and (3) to produce a sufficiently
       large sample of model outputs for frequency analysis.

    5) Presentation of the series of output (random) values generated in
       step (3) as a cumulative probability distribution function (CDF).

    6) Analysis and application of the cumulative probability distribution
       of the output as a tool for decision making.

    The input variables required by TEEAM can be broadly classified into two
different sets that exhibit different uncertainty characteristics.  These
are:

    •  Variables that describe the chemical, biochemical, and toxicological
       properties of the chemical of concern.  Examples of these variables
       include the octanol-water partition coefficient, acid, neutral, and
       base catalysed hydrolysis rate, soil-adsorption coefficient, Henry's
       Law Constant, etc.

    •  Variables that describe the environmental properties of the various
       media and impact the fate and transport of the pollutant within each
       medium.  Examples of these variables include the soil porosity,
       organic carbon content, metabolic degradation rates, etc.

    Uncertainty in the first set of variables primarily arises due to
laboratory measurement errors or theoretical methods used to estimate the
numerical values.  In addition to experimental precision and accuracy,
errors may arise due to extrapolations from controlled" (laboratory)
measurement conditions to uncontrolled environmental  (field)  conditions.
Further, for some variables semi-empirical  methods are  used to  estimate  the
values.  In these cases, errors in using the empirical  relationships  also
contribute to errors/uncertainty in the model  outputs.

    Uncertainty in the second set of variables may include both measurement
and extrapolation errors.  However, the dominant source of uncertainty in

                                    96

-------
these is inherent natural (spatial and temporal) variability.   This
variability can be interpreted as site-specific or with in- site variation in
the event that the model is used to analyze exposure due to the use of a
toxicant at a particular site.  Alternatively, it can represent a larger
scale (regional /national) uncertainty if the model is used to  conduct
exposure analysis for a specific chemical or ecosystem on a generic, nation-
wide or regional basis.  Note that the distributional properties of the
variables may change significantly depending upon the nature of the
application.

    Whatever the source of uncertainty, the Monte Carlo method requires that
the uncertainty in input parameters be quantified by the user.  This implies
that for each input parameter deemed to be uncertain, the user select a
distribution and specify the parameters that describe the distribution.  The
following sections describe the methods by which the TEEAM Monte Carlo
module generates random numbers and analyzes model outputs.

3.6.1  Description of Monte Carlo Parameter Distributions

    The Monte Carlo Module has the ability to generate data from a number
of probability distributions, including uniform, normal, log-normal,
exponential, and Johnson SB.  A description of each of these distributions
is provided in the following paragraphs, including parameters of the
distributions, equations for the probability and cumulative density
functions, and a brief discussion of the properties of each distribution.

Uniform Distribution—
    A uniform distribution is a symmetrical probability distribution in
which all values within a given range have an equal chance of occurrence.
A uniform distribution is completely described by two parameters:  1) the
minimum value (lower bound) A, and 2) the maximum value (upper bound) B.
The equation for the uniform probability density distribution of variable x
is given by:

    fu<*) ' TB^y                                                 (3-220)

where f(x) is the value of the probability density function for x.  The
cumulative distribution F(x) is obtained by integrating Equation (3-220).
This yields the probability distribution:
                                                                    (3-221)
where FU(X) is the probability that a value less than or equal to x will
occur.
                                     97

-------
Normal Distribution--
    The normal distribution refers to the well known bell-shaped
probability density function.  Normal distributions are symmetrical about
the mean value and are unbounded, although values further from the mean
occur less frequently.  The spread of the distribution is generally
described by the standard deviation.  The normal distribution has only two
parameters:  the mean and the standard deviation.  The probability density
function of x is given by:
     fn(x) = S     6XP I-O^1                              (3-222)
where Sx is the standard deviation, and mx is the mean of x.  The
cumulative distribution is the integral of the probability density
function:
     Fn(x) =  J   fn(x)dx                                           (3-223)
The above integration must be performed numerically, but tables of
numerically-integrated values of Fp(x) are widely available in the
statistical literature.

Log-Normal Distribution—
    The log-normal distribution is a skewed distribution in which the
natural log of variable x is normally distributed.  Thus, if y is the
natural log of x, then the probability distribution of y is normal with
mean my> and standard deviation Sy and a probability density function
similar to Equation (3-222).  The mean and standard deviation of x (mx and
Sx) are related to the log-normal parameters mv and Sv as follows:
                                              Jr      J
     mx = exp[my + 0.5(Sy)2]                                         (3-224)


     S* = mJ[exp(S^  - 1]                                            (3-225)

To preserve the observed mean and standard  deviation of x,  the  parameters
of the log-normal distribution (my and Sy)  are therefore selected  such that
the above relationships are satisfied.  Note  that  my and Sy do  not equal
the natural logs of mx and Sx respectively.  Log-normal distributions have
a lower bound of 0.0 and no upper bound, and  are often used to  describe
positive data with skewed observed probability distributions.
                                     98

-------
Exponential Distribution--
    The probability density function for an exponential distribution is
described by an exponential equation:
             exp(-x/rn )
     f (x) = 	—                                             (3-226)

where mx is the mean of x.  The cumulative distribution is given by:

     Fe(x) = 1 - exp(-x/mx)                                         (3-227)

The exponential distribution is bounded by zero; the probability density
function peaks at zero and decreases exponentially as x increases in
magnitude.

The Johnson System of Distributions--
    The Johnson system involves two main distribution types:  SB (Log-ratio
or bounded) and SU (unbounded or hyperbolic arcsine).  These two
distribution types basically represent two different transformations
applied to the random variable such that the transformed variable is
normally distributed.  The specific transformations are:

     SB:  Y = *n({gi£})                                             (3-228)

     SU:  Y = anfj^ + (1 + (^)2)°'5l                             (3

where:

    an   = natural logarithm transformation
    x    = untransformed variable with limits of variation from A to B.
    Y    = the transformed variable with a normal distribution

    Selection of a particular Johnson distribution for sample data  set  is
accomplished by plotting the skewness and kurtosis of the sample data.  The
location of the sample point indicates the distribution for the sample  data.

     For additional details of  the  Johnson  system of  distributions,  the
 reader is  referred to  McGrath  et al.  (1973)  and Johnson and Kotz  (1970).

 3.6.2  Uncertainty in  Correlated Variables

     In many  cases model  input  variables are  correlated due to various
 physical mechanisms.   Monte Carlo  simulation of such variables  requires not
 only  that  parameters be  generated  from the appropriate univariate
 distributions,  but also  that the appropriate correlations be preserved  in
 the generated  input  sequences.  The  Monte  Carlo module currently  has the
 ability to generate  correlated normal and  log-normal numbers.   The
 procedures used  are  described  in the following  paragraphs.

                                     99

-------
    The correlation coefficient is a measure of the linear dependence
between two random variables and is defined as:
      "x., -                                                          («*>
where:

      PX   = the correlation coefficient between random variables x  any y
      covfx.y) = the covariance of x and y as defined below
      CTY  aw = the  standard deviation for x and y.
      *»  y
The covariance of  x and y is defined as:
     cov(x.y) = E[(x-mx)(y-m )]
         +<*>
     =  I J  (x-mx)  (y-m ) fx y(x,y) dxdy                             (3-231)
         — CO
where
       E = the expected value
       mx, m  = the mean of the random variables x and y
       fv w(x,y) =  the joint probability distribution of x  and y.
Note that the linear correlation coefficient between x  and y  can  be
computed using
                 n
                 J  x,y, - n mxmy

     »x,y - —-~	5*                     <3-232>
                   9       *>    "   O       9
            ( .yx2 - nmx-2) (.y,2 - ™/)  )

    To generate  correlated  random  variables three  steps  are required.
First uncorrelated normally distributed  random  numbers  are generated.  This
vector is then transformed  to a  vector of  normally distributed numbers with
the desired correlation.   Finally,  the normally distributed numbers are
transformed to numbers with the  desired  distribution.

    The transformation of uncorrelated to  correlated normal numbers consists
of multiplying the uncorrelated  vector of  numbers  with  a matrix B:

    Y' = B e                                                        (3-233)

where

       e  =  the vector of uncorrelated, normally distributed random numbers
       B  =  an N by N matrix

                                    100

-------
        Y1 =  a vector of standard normal deviates of mean zero and standard
              deviation of unity

 The matrix B is related to the variance-covariance matrix S as follows:

     S = BBT                                                         (3-234)

 where BT is the transpose of the matrix B.  Since the normal variables Y1
 have means of zero and unit variances, the variance-covariance matrix is
/equivalent to the correlation matrix.

j.^f   Thus, if the correlation matrix S is known, B can be found from
 tquation (3-234) by using a Cholesky decomposition algorithm.  This
 algorithm will decompose a symmetric positive definite matrix, such as S,
 into a triangular matrix such as B.

     Having generated a vector of correlated normally distributed random
 numbers, the vector Y1 can be converted, through appropriate
 transformations, to the distribution of choice.  Thus for parameters X^ that
 have a normal distribution, the Y1  numbers are transformed as follows:

        X— m  J_ „ /Vl\                                                / "3 OOC\
      •i — Illy T "vV'-i/                                                \J~£jj)

 For parameters that follow the lognormal distribution, the following
 transformation applies:

       X. = exp[(Yl)  (olnf1) +yln§1l                                 (3~236)

 where

       v,   -is the  log mean of the  1   parameter.  .                  (3-237)
       °ln1  is the  log  standard deviation of the i   parameter       (3-238)

 Other distributions can be easily  incorporated into the analyses at a later
 time  when suitable  transformations from the normal distribution can be
 found.   It  is important to note  that  in using  this technique, the
 correlations  are estimated in normal  space so  if these correlations are
 estimated using actual data, the data should be transformed  to a normal
 distribution  before correlation  coefficients are estimated.

     For two correlated variables,  one with a normal distribution  (x2) and
 the other with  a log  normal  distribution  (xj), the following equation  is
 used  to transform  correlations to  normal  space (Mejia et  al.  1974):

               PX   x   [exp(a  2) - 1]*
       p       =—111         l                                        (3-239)
       yl'y2    °y1

                                      101

-------
where

     p      = the correlation coefficient between the two variables in the
      yl'y2   normal space
     p      = the correlation coefficient between the two variables in the
       1* 2   arithmetic space
     a   = the variance of y^ derived from Equation (3-224)

    If both X! and x2 are log-normally distribution then the correlation
coefficient in the normal space is transformed using Mejia et al (1974):

                 1                     Sx  Sx
     p      = ^-=	 ln{l + p      I	— 1}                  (3-240)
      yl'y2   Sv Sv           X1»X2    mx  mx
       •*•*"    JlJO           At     A <  Art

where the relationships between S  (S  ) and S   (S  ) are given by
Equations (3-224) and (3-225).    1   2       yl   y2

3.6.3  Generation of Random Numbers

    Having selected the distribution for the various input parameters, the
next step is the generation of random values of these parameters.  This
requires the use of pseudo-random number generating algorithms for Normal
and Uniform numbers.  There exist numerous proprietary as well as non-
proprietary subroutines that can be used to generate random numbers.   A
number of these are comparable in terms of their computational  efficiency,
accuracy, and precision.  The routines included in TEEAM have been checked
to ensure that they accurately reproduce the parameters of the distributions
that are being sampled.
                                     102

-------
                                  SECTION 4

                      MODEL  INSTALLATION AND EXECUTION
    This section describes how to load the TEEAM software into an IBM-PC
compatible computer from the supplied floppy disks, how to verify that the
software has been loaded properly, and how to execute TEEAM simulations.
The hardware and software required for installing TEEAM on an IBM-PC
computer are also discussed.  If a computer other than an IBM-PC compatible
is to be used, some INCLUDE files will have to be modified and the software
will have to be recompiled.

4.1  IBM-PC COMPATIBLE ENVIRONMENT REQUIREMENTS

4.1.1  Hardware

    An IBM-PC compatible computer with 640K memory, one floppy disk drive,
and at least 5 megabytes of available hard disk storage is the minimum
hardware requirement.  The floppy disk drive is necessary to download the
software and test data files as supplied.  Either a 360 KByte (DSDD) drive
or a 1.2 MByte (DSHD) drive can be used; the installation floppy disks can
be sent in either format.  A math coprocessor (8087, 80287, or 80387) must
also be present.

4.1.2  Software

    The TEEAM software was developed using an MSDOS 3.2 operating system.
Earlier versions of PCDOS or MSDOS which are able to recognize file
directories should also be compatible.


    The DOS supplied ANSI.SYS driver must also be available.  This driver is
installed by modifying the DOS specific CONFIG.SYS file (located in the root
directory) to contain the following line:

                           DEVICE  =  ANSI.SYS

where  is the location (directory) of the ANSI.SYS file.  This file
will normally be located in a directory containing other DOS specific files;
the location of this file (or even its presence) was determined by the
individual who originally configured the microcomputer.  If this file is not

                                     103

-------
available, it will  be necessary to consult with an individual  who  is
familiar with the MSDOS operating system to determine the steps  necessary to
have it available for the TEEAM software.

    The CONFIG.SYS file may also have to be modified to specify  how many
files are available to the TEEAM software.  It the statement 'FILES ='  is
not present or it is present but with a value less than 20 (e.g.,  'FILES =
10') then the statement 'FILES = 20'  (without the quote marks) should be
added to the CONFIG.SYS file.

    If the installation disks are 360K format, it will be necessary to have
the DOS Restore command available.

4.2  LOADING EXECUTABLE CODES AND TEST DATA FILES

    The executable code and test data files can be loaded onto the hard disk
of the target computer by taking the  following steps:

    1. Insure that the target hard disk does not have a directory  called
       TEEAM and that the hard disk is in the root directory.

    2. Put the TEEAM floppy disk labeled 'MASTER' in the floppy disk  drive.

    3. Make the floppy disk drive the default drive (e.g., type 'A:'  if the
       floppy drive is the A drive).

    4. Type INSTALL 
, where
is the source drive (e.g., A) and is the target drive (e.g., C). The spaces between the INSTALL command and the drive designators are required; colons should not be present in the drive designators. 5. If additional disks are required, you will be prompted to insert them in the floppy drive and press return. 6. After the executable code and test data files have been loaded, you will be asked if you want to load the TEEAM source "code. The source code is only necessary if you wish to make modifications to the code and it is not necessary to load these files (and thus, decrease the available storage on your hard disk) if you do not want to make modifications. The source code can always be obtained from the floppy disks if you require it at a future date. Using the INSTALL batch file is a convenience but is not necessary if you wish to copy the TEEAM files within your own directory structure. The INSTALL file creates a subdirectory (called TEEAM) on the hard disk, copies 104

-------
the executable code, some batch files, and run files into this directory,
creates several subdirectories within the directory TEEAM and copies the
test data files into these directories, and, optionally creates a
subdirectory of the source code and include files.

4.2.1  Executing Test Data Inputs

    A batch file is supplied which will execute all of the test data files
provided, one after another.  This file, TESTRUN.BAT is available within the
TEEAM subdirectory and can be executed by typing TESTRUN while within the
TEEAM subdirectory.


    These test data sets can be executed individually, if desired, by:
(1) copying the appropriate test run file (labeled TESTDAT.n, where n = 1 to
the number of test data sets) to the file name TMRUN.DAT, (2)  erasing the
file KECHO.PRN if present, and (3) then typing TEEAM (followed by a return)
to start execution.

4.2.2  Verifying Test Data Outputs

    Running the batch file TESTRUN.BAT will create output files with the
extension .PRN within the test data set subdirectories, (labeled TESTDAT.n,
where n = 1 to the number of test data sets).  Within these subdirectories
will be files with the same prefix but with the extension .VRF.  Each of the
.PRN files should be compared to the corresponding .VRF file either by using
an editor to visually compare the results or by using the DOS compare
utilities FC (DOS 3.2 or greater) or COMP (earlier versions of DOS).

4.3  GENERAL PROCEDURES FOR TEEAM EXECUTION

    This section contains descriptions of techniques and suggestions for
normal (operational) execution of TEEAM.  It is generally convenient to
locate all of the input files pertaining to a specific problem within their
own subdirectory.  The input files should be developed using the file
descriptions of Section 5.  It will normally be useful to start with the
test input data files provided and edit them to meet the specifications of
the problem at hand.

    The names chosen for the input files should be specific to the problem
to reduce the potential for confusing these files with those developed for
alternate problems - TEEAM enforces no restrictions on the names chosen,
either for the prefix or extension (DOS limits the prefix to 8 characters
and the extension to 3 characters).  It is good practice to choose an
extension which implies an input file (e.g., .IN or .DAT) or output file
                                    105

-------
(e.g., .OUT or .PRN).  The names of the files used and the path describing
where the files are located (the subdirectory) are defined in the file
TMRUN.DAT, the run control file (see Section 5.2.1).   The file TMRUN.DAT
must be in the TEEAM subdirectory.  When all of the necessary files are
complete, TEEAM can be executed by typing TEEAM followed by a return.

    The PATH statement in the TMRUN.DAT file can be used to specify the
directories or subdirectories where the model input and output files are
located.  When the PATH statement is used in the TMRUN.DAT file, the defined
path must end with the backslash character ('\').  The defined path will
pertain to all files which are defined after that path statement until
another path is defined.

    If the file subdirectory is a subdirectory within the TEEAM
subdirectory, the path defined should not begin with a backslash.  If the
file subdirectory is external to the TEEAM subdirectory, the -defined path
should begin with a backslash (this implies to DOS that the search for the
subdirectory should begin from the root directory).

    The path record can also be used to define files which are located on
different drives.  This is potentially useful to increase the execution
speed of multiple habitat and Monte Carlo simulations.  Both multiple
habitat and Monte Carlo simulations use scratch (unformatted) files for
temporary storage of data and these scratch files are opened with the last
defined path (note that a PATH record can be the last record of the file
records within TMRUN.DAT).  If the last path indicates that the files are to
be located on a RAM disk (please consult your DOS manuals for defining a RAM
disk) execution speed will be measureably increased.

4.4  MACHINE AND COMPILER DEPENDENCIES

    The code has been designed to be implementation independent where an
implementation would include computer make and type, input/output devices
used, and brand and version of FORTRAN 77 compiler used.  The model is
compatible with the DEC VAX running VMS, PRIME 50 series running PRIMOS,  IBM
hardware or compatible mainframes running OS/VS2 MVS, and IBM PC compatibles
running MS DOS.  Minimal knowledge of the particular operating system is
required of the user.

    There are a few exceptions to the rule of implementation independence
which were necessary to provide useful features for the user interface and
to follow EPA recommendations for FORTRAN code development.  Extensive use
has been made of INCLUDE files for defining COMMON variables and PARAMETERS
(see Section 9).  The use of INCLUDE files ensures that variables and
parameters common to all subroutines will be of the same size and type;
                                    106

-------
unfortunately, the format for INCLUDE statements is dependent on the FORTRAN
compiler used.  Currently, the only "work-around" for this compiler
dependence is to provide versions of the code for alternate compilers.

    An additional compiler-dependency is the definition of the minimum and
maximum values of single precision real numbers which do not result in
underflows and overflows.  These values are defined in a single INCLUDE file
and are provided to prevent "system crashes" or unintelligible system error
messages; these values are used to check the intermediate calculation of
variables and supply a warning to the user if an underflow or overflow
condition exists and to provide an organized exit from the software if the
condition is fatal.  Since these values are located in a single file, it is
a relatively simple matter to change the values provided to values which
would be more appropriate to the FORTRAN compiler being used.

    A screen management routine has been supplied to present the software in
operation.  If it is defined that an IBM PC (or compatible) computer is
being used, the screen will be updated without scrolling text off the top of
the screen.  If this IBM PC option is not set (in the same INCLUDE file as
above) before compiling, TEEAM execution information will be listed line-by-
line to the standard output device.
                                     107

-------
                                  SECTION  5

                         INPUT SEQUENCE DEVELOPMENT
5.1  OVERVIEW OF TEEAM INPUT DATA

    This section describes the contents and formats of the various data
files required to run TEEAM.  In general, all input to the model  occurs
sequentially from batch files.  These files may be created or edited from
sample input files which are supplied with the code using a text  editor on
the user's computer system.  A separate batch input file is required for
each of the following TEEAM modules if they are used in the simulation:

       • Execution Supervisor (EXESUP)
       • Aerial Spray (FSCBG)
       • Spray Grid Definition (GRDDEF)
       • Terrestrial Fate and Transport (TFAT)
       • Plant Growth and Translocation (PLTGRN and PLTRNS)
       • Terrestrial Animal Exposure (APUM)
       • Monte Carlo (MC)

    The Execution Supervisor data file selects the modules to be executed,
defines the names of the input files to be utilized, and is required for all
TEEAM runs.  The TFAT module data file defines the physical and transport
properties for each habitat and is also required for all TEEAM runs.  The
remaining modules (and their corresponding batch input files) may or may not
be used depending upon options selected in the Execution Supervisor file.
The modules FSCBG and GRDDEF compute the distribution of pesticide applied
during spray events.  The plant growth and translocation modules compute the
growth of plants and the transport of pesticides through the plant roots and
aboveground biomass.  The APUM module calculates the uptake and assimilation
of pesticide by animals ingesting water, pesticide granules, soil, plants,
and other animals or inhaling contaminated air.  Finally, the Monte Carlo
module generates random inputs to the various modules for Monte Carlo runs
of TEEAM; inputs for this module describe the probability distributions of
various model parameters.  Note that Monte Carlo simulations can only be
performed for one habitat simulation.

    Specific formats for each module and corresponding  input file are
described in Section 5.2 and in Tables 5.1 through 5.9.  Due to the length
of these tables, they are presented collectively at the end of this

                                    108

-------
section.  In the format descriptions each data line is referred to as a
"Record."  These records may consist of one line of data or may be repeated
as necessary (i.e., for each soil horizon, animal, or crop).  Records are
labeled for reference by the appropriate module acronym followed by a
sequential number.  For instance, the first data record in the TFAT module
is labeled "TFAT1."  Record descriptions in Tables 5.1 through 5.9 consist
of one line listing the record number and the names of the input variables
contained in the record, one line describing the required FORTRAN input
format, and several lines defining the input variables and their appropriate
units.

    Note that since several of the modules were developed from pre-existing
model codes, data input formats vary significantly from module to module.
TEEAM data records are usually formatted, meaning that each input variable
on a record occupies a field of specified length.  For instance, the format
(2F10.0) indicates that there are two data fields of length 10 each.
Character formats are also used extensively to read labels and option
flags.  For example, the format (A20) indicates that a character string up
to 20 characters long (including blanks) is read.  All character strings
read by the model are converted to upper case letters immediately after
input; the model therefore does not differentiate between upper case and
lower case labels and responses.

    A feature common to all TEEAM input files except for the FSCBG module is
the Comment line.  Comment lines may be inserted anywhere in the input data
files, and are indicated by the presence of three asterisks ("***") as the
first non-blank characters in the line.  These lines are ignored by the
model for computational purposes and allow the user to type in comments,
table headings, and other information useful in making the input file more
understandable.

5.2  DESCRIPTION OF INPUT FILES FOR TEEAM MODULES

5.2.1  Execution Supervisor

    The run control file which is the input to the execution supervisor is
the only input file which cannot have an optional name; it must be named
"TMRUN.DAT" and it must be located within the default directory.  This file
is used to define which options (modules) are selected for a simulation and
what the files are to be named that are used for input and output.
Additionally, several parameters which are required for simulation control,
such as starting and ending dates, are also defined by this file.
Figure 5.1 illustrates a sample TMRUN.DAT file.

    There are three major groups of input data:  1) options, 2) files, and
3) global parameters.  Formats for these data are shown in Table 5.1.   The

                                   109

-------
simulation trace level can also be defined with this file.  The simulation
trace level is used to display to the screen which subroutine TEEAM is in
during execution.

   All of the potential  options have default values and, if an option is
not specifically set, the default value will be assumed.  These defaults are
as follows:  TFAT=ON, AERIAL SPRAY=OFF, FSCBGOFF, PLTGRN=ON, APUM=ON, MONTE
       •**
           SnapTa TMRUN.DAT file
       *•* Opcions
       TFAT                     ON
       PLTGRN                   ON
       AERIAL 3PRAY             OFF
       FSCBG                    OFF
       APUM                     ON
       NKAB                     1
       MONTE CARLO              OFF
       ENDRUN
       ••* Files
       PATH                     \EPAB7\acnew\DATA\
       METEROLOGY               MET . SML
       FSCBG OUTPUT             CBGFOO
       FSCBG INPUT              SFSDATAXKPSPR.IN
       PLTGRN                   KPLNT.dat
       APUMIN                   NOCUTWRM . IN
       GRID DEFINITION         GRDDEF1.IN
       MCIN            .         0RANDU.IN
       TFAT                     PRZM.ONE
       «** separate  path for output  files for convienience
       PATH                     \EPAB7\MCNEW\
       APUMOUT                  NOCUTWRM. PRN
       ANIMAL TIME SERIES      KRSLTS.PRN
       CONCENTRATIONS          KCONC . PRN
       HYDROLOGY                KHYDRO.PRN
       PESTICIDES               kf Ips . prn
       TIME SERIES              KFLTS . PRN
       MCOUT                    MCOUT.PRN
       MCOUT2                   MCOUT2.PRN
       ENDFILKS
       •**  Global data
         010450        *   300450
            .75   0.000        2    36.2
       *** Daylight  hours for Tifton GA
           10.2    10.9     11.9    12.8    13.7     14.0
           13.9    13.5     12.2    11.2    10.4     10.0
       TRACE                    0

   Figure 5-1.  Sample  Execution supervisor input data file  (TMRUN.DAT)

                               110

-------
CARLO=OFF, and NHAB=1.  The options section must end with ANAME="ENDRUN"
(Table 5.1) even if all of the default values are to be used.

    Note that turning PLTGRN "ON" invokes both the EPIC plant  growth
algorithm and the plant transportation module PLTRNS.  If PLTGRN is "OFF,"
plant growth is not shut off; rather the plant growth algorithm in the
original version of PRZM is enabled.

    The file records of the execution supervisor input are used to assign
the user supplied names and locations of the files required for a simulation
to the appropriate TEEAM file.  If files are defined within these records
which are not necessary because of the options selected in the options
records (EXESUP1), the file name is ignored.

    The global parameters records are used to define certain environmental
and simulation control parameters which need to be defined for all habitats
and thus are not specific to each TFAT habitat.

    The trace level record is, in general, only necessary for  debugging
problems with a simulation.  If the trace level is set to a value greater
than one, the subroutine that TEEAM is currently executing during a
simulation, as well as the path of subroutine calls to access  that
subroutine, are displayed on the standard output device.  A trace can be
useful for debugging a simulation but, when used, can increase execution
time considerably.  If the trace level record is not included  in the EXESUP
data set, a default value of 3 will be assumed.  A value of 3  indicates that
subroutine calls three levels deep will be displayed.

5.2.2  Input Data for the FSCBG Module

    The FSCBG module was adapted from the FSCBG code (Dumbauld, Bjorklund,
and Saterlie 1980) for incorporation into the TEEAM model.  The purpose of
the FSCBG module is to calculate the top-of-canopy deposition  resulting from
an aerial or ground spray event.  The input format for the FSCBG module
(Table 5.2a) is similar to the original input requirements described in
Dumbauld, Bjorklund, and Saterlie (1980).  Comment lines cannot currently be
utilized in the FSCBG input file.

    Only the first record, the simulation title, uses a FORTRAN format.  All
of the other records use list directed input formatting.  In list directed
formatting, data values are entered on a line in columns 1 to  80.  All data
values are separated by a single comma (,).  If a data value is intended to
be omitted, a comma should be entered immediately after the preceding
comma.  A slash (/) may be used to indicate an end of line.  If a slash is
entered before all data values have been supplied, the remaining variables
on that line are left unitialized.
                                   Ill

-------
    The second record required by the FSCBG module is an array of option
integers (ISW).  The values which these integers can have and the resulting
actions taken are presented in Table 5.2b.  Most of the options present in
the original FSCBG code have been deleted for the TEEAM installation.  As a
result of this, the option numbers appearing in Table 5.2b are no longer
sequential but the options still available correspond to the original FSCBG
numbering system.

5.2.3  Input Data for the Spray Grid Definition Module

    The spray grid definition module uses the output from the FSCBG aerial
spray application module and computes the average top-of-canopy deposition
to each habitat.  To accomplish this task, the spray grid definition module
has to be given the correspondence between the FSCBG receptor grid (the
spatial coordinate system defining the area which could potentially receive
a spray application or drift from a spray application) and the individual
habitats.  Given this information the spray grid definition module is able
to calculate the average top-of-canopy depositions.


    The input file to the spray grid definition module defines three types
of data:  1) the location of habitats with respect to the aerial spray grid
(FSCBG receptor grid), 2) the number and dates of spray events, and
3) parameters defining how the top-of-canopy data (FSCBG output) should be
distributed to soil and foliage and within the plant canopy.  These data  are
defined in Table 5.3.

    The locations of habitats and the definition of the FSCBG receptor grid
are described by Records GRID1-GRID4.  This module will generate the FSCBG
receptor grid based pn the number of east-west and north-south grid points
selected in Record GRI02, the initial x- and y-coordinates, and increments
for x- and y-coordinates, (GDX and GDY, respectively) in Records GRID3 and
GRID4.  Note that each grid point generated represents the center of a
subgrid area of dimensions GDX by GDY.  These grid points are the locations
where FSCBG computes the top of canopy pesticide deposition for each subgrid
area.

    The locations of habitats are  defined by entering the  x-  and y-
coordinates of the southwest and northeast corners  of each  habitat on Record
GRID1.  These coordinates should be  within the FSCBG  receptor  grid.   It  is
preferable, but not necessary,  that  the  habitat  corner coordinates be
specified as the midpoint between  receptor grid  coordinates.   If  the
midpoints are used, the habitat area computed based  on the  coordinates will
be the same as the area computed by  summing all  the  subgrid  areas within  a
habitat.
                                     112

-------
    The data on number and dates of spray events and the parameters defining
how the FSCBG output should be distributed within the plant canopy are
passed on to the TEEAM execution supervisor for processing by the TFAT
module.

5.2.4  Input Data for the Terrestrial Fate and Transport Module (TFAT)

    The Terrestrial Fate and Transport Module (TFAT) uses two input data
files:  a meteorological data file and a parameter file.  The meteorological
data file contains daily values of precipitation, potential
evapotranspiration, temperature, solar radiation, and wind speed.  Formats
for this file are described in Table 5.4.  The parameter file contains the
.following four groups of data for each habitat describing the transport of
pesticide:

       • Hydrology and Crop Data        (Records TFAT1 - TFAT11)
       • Pesticide Application Data     (Records TFAT12 - TFAT19)
       • Soils Data                     (Records TFAT20 - TFAT27A)
       • Output Specification           (Records TFAT28 - TFAT30)

    Specific formats for each parameter data group are shown in Table 5.5
and are discussed below.  Note that the entire parameter data set for the
TFAT module should be repeated as a set for each habitat.

    The hydrology and crop data group consists of data describing the
runoff, erosion, infiltration, and crop conditions at the soil surface.
Record TFAT1 is a title used to identify input and output files for the TFAT
run, while Record TFAT2 is a title for the hydrology data group.  Record
TFAT3 contains data describing initial crop conditions, the area of the
habitat, and the number of time steps used by the ponding and infiltration
algorithms.  Note that if the FSCBG and GRDDEF modules are used, the area of
the field entered here should be the same as the area defined by the the
GRDDEF habitat spray grid coordinates.  Record TFAT4 consists of data
required by the ponding and infiltration routines, including Green-Ampt
parameters and initial conditions.  Records TFAT6 and TFAT7 contain surface
erosion parameters; note that TFAT7 is not required if erosion is not
simulated (ERFLAG = 0 on TFAT6).  Finally, Records TFAT8 through TFAT11
describe the growth and runoff properties of the various crops which are
simulated.


    The pesticide application data group (Records TFAT12-TFAT19)  consists of
data describing the timing, amount, and distribution of TFAT pesticide
applications.  These data are used to define application events (e.g.,
granular, direct to soil  or soil  incorporation)  for which off-site drift is
not an issue.  Note that additional applications may also occur as
calculated by the FSCBG module.   If a spray event in FSCBG is specified on

                                    113

-------
the same day as an application event in TFAT,  the spray application event
parameters will override the TFAT parameters.   Also included in this data
group are several pesticide volatilization parameters.   Record TFAT12 is a 1
to 80 character title for the pesticide application data group.  Record
TFAT13 contains the number of pesticide applications to be described in
Record TFAT14.  Record TFAT14 is repeated for  each application and contains
the date of application, the amount, the depth of incorporation, and for
granular applications, concentration of pesticide in granules.  The type of
application is selected by the parameters FAM  and KANOPT on Record TFAT15,
and the remaining input of application data depends upon the values of these
parameters.  If FAM=1, the application is incorporated, into the soil and no
further application data except Records TFAT18 and 19 are required.  If FAM
= 2 or 3, pesticide is applied to the plant foliage.  FAM = 2 indicates that
the user specifies on Record TFAT16A the fraction of each application that
is intercepted on plant foliage, while FAM = 3 indicates that the model will
calculate the interception of the application  based on the canopy filtration
parameter FILTRA on Record TFAT16B.  If FAM=4, pesticide is applied to the
soil surface in a granular formulation; release of pesticide from granules
is then estimated using the rate constants input on Record TFAT16C.

    Records TFAT17A through TFAT17D describe the penetration of pesticide
applications into the canopy, and are required only if FAM = 2 or 3.  Input
of these records will depend upon the value of KANOPT on Record TFAT15.  If
KANOPT = 1, the user specifies how each application penetrates the canopy on
Records TFAT17A and TFAT17B.  If KANOPT = 2, the model will calculate canopy
penetration based on data in Record TFAT17C.  Record TFAT17D is required for
KANOPT = 1 or 2, and describes the decay and washoff of pesticide on plant
foliage.

    The final two records in the application data group contain data needed
to describe the volatilization of the pesticide from soil and water.  Record
TFAT18 consists of the degradation rate constant for the pesticide  in ponds,
the diffusion coefficient in water, and the initial concentration  in
ponds.  Record TFAT19 contains the reference height for wind measurements,
the pesticide diffusion coefficient in air, and the initial atmospheric
concentration of pesticide vapor.

    The soils data group  (Records TFAT20 - TFAT26A) describes the properties
of each soil  horizon  in the habitat.  Record TFAT20 is a 1 to 80 character
title for  soils data.  Record TFAT21 contains data for the entire soil
column, including the number of soil layers modeled, the soil core depth,
and various option flags  for input of soil properties.  Record TFAT22A is
input only  if the model is to calculate the partition coefficient for the
pesticide  in  soil  (KDFLAG = 1).  Soil properties are then described for
various horizons in the soil column; soil properties for all layers within a
horizon are assumed to be uniform.  Record TFAT23  is the number of  horizons


                                     114

-------
NHORIZN, while Record TFAT24 describes the soil properties for each horizon
and is repeated NHORIZN times.  Records TFAT25A through TFAT25D are then
input depending upon the bulk density option (BDFLAG) selected on Record
TFAT21.  Finally, Records TFAT26 and TFAT27A are used to assign initial
pesticide levels, if applicable, to each soil layer.

    Output options are selected on Records TFAT28, TFAT29, and TFAT30.
Record TFAT28 is used to select the output frequency for hydrologic
summaries, pesticide summaries, and concentration profiles; these may be
printed daily, monthly, or annually for various numbers of soil layers.
Records TFAT29 and TFAT30 specify which variables are to be written out as
time series.  Labels used to select variables for time series output are
shown in Table 5.6.

5.2.5  Plant Growth and Translocation Location Modules

    The data required for both the plant growth (PLTGRN) and plant
translocation (PLTRNS) module, are defined in a single file.  The order of
parameter input is illustrated in Table 5.7.  These parameters have to be
provided for each crop within each habitat.  The number of crops within each
habitat is defined in the TFAT input file (see Table 5.5).  The order that
these crop parameters are read in is the same order as defined for the TFAT
module (in order of the succession of the crops).  Note that the planting
dates and crop succession information have been retained in TFAT.  Only
information specific to plant growth simulation is read in by the PLTGRN
module.

    If more than one habitat is being simulated, all of the parameters for
crops in the first habitat are defined, followed by all of the crops in the
second habitat, etc.  If the same crop appears in more than one habitat, it
must be redefined in each habitat in which it appears.

    Each record of this input file expects only one parameter to be
defined.  The first 24 spaces of each record are available for the user to
insert notes as to what the parameter is—this  field of 24 characters is
ignored by the software.   The only exception to this rule  is that the first
record (PL1) must have the label  'PLANT1  in the first 5 columns to indicate
that a new plant is being defined.

    The initial condition values,  records PL17  through PL27, are used if
some non-zero initial  conditions exist for plant biomass and for contaminant
within the plant biomass; otherwise,  these values can be set to zero (or,
equivalently, contain a blank field).
                                    115

-------
5.2.6  Input Data for the Terrestrial  Animal  Exposure Module (APUM)

    This section describes the format  and content of the input file  required
for the Terrestrial Animal Exposure Module (APUM).  APUM data records may be
subdivided into four data groups:

        •  Soil Animal Data (Records APUM1 -  APUM9)
        •  Higher Animal Data (Records APUM10-APUM21)
        •  Predation Data (Records APUM22-APUM26)
        •  Output Specifications (Records APUM27-APUM29)

    The specific contents and formats  for these data records are shown in
Table 5.8.  A general discussion of the types of data and user options for
each data group is provided in the following  paragraphs.

    The soil animal data group (Records APUM1-APUM9) describes the uptake
and movement of animals living in or on the soil in each habitat.  These
animals are exposed to chemicals by ingestion and by contact with soil, and
their uptake of pesticide is modeled using bioconcentration factors.  Soil
dwelling animals may move between soil horizons, but cannot move between
habitats.  They also may not prey on other animals.  The first Record
(APUM1) in this group contains a descriptive run title used to identify
input and output files.  Records APUM2 and APUM3 contain the number of soil
horizons and the number of soil animals respectively in each habitat.  Note
that number of soil horizons NSCOM must be less than or equal to the number
of TFAT module soil horizons in the corresponding habitat.  Records APUM4
through APUM9 contain specific uptake and movement data and are repeated for
each soil animal in each habitat.  Soil animals are  identified for output
and predation by the  label ALABEL read on Record APUM4.

    Movement of each  soil animal may be modeled by one of three options,
as specified by the variable ICOLD on record  APUM7.  If ICOLD = 0, the
animal is modeled as  a population distributed among soil horizons.  The
distribution of the population changes over time as determined by the
horizon-movement transition matrix for the animal.  This option requires
input of the transition matrix on Record APUM8 and the initial population
distribution on Record APUM9.  If ICOLD < 0,  the animal is modeled as a
population with a steady-state distribution.   The transition matrix on
Record APUM8 is then  not required, and the population distribution will
remain constant as input on Record APUM9.  Finally, if ICOLD > 0 the animal
moves randomly from horizon to horizon as determined from its previous
location and the transition matrix.  ICOLD in this case is the initial
horizon location of the animal.  This option requires input of the
transition matrix  (Record APUM8) but does not require Record APUM9.

    The higher animal data group (Records APUM10-APUM21) describes the
uptake and movement of animals which prey on  other animals and move between

                                    116

-------
habitats.  Modeling of these animals requires a specific breakdown of food
preferences, assimilation factors, and interhabitat movement parameters.
The first data Record APUM10 contains the number of higher animals
modeled.  The remaining records (APUM11-APUM21) contain specific uptake and
movement data and are repeated for each higher animal.  Record APUM11 is a
unique 1 to 20 character label identifying the animal for output and
predation.

    As in the case of soil animals, one of three options for modeling
movement between habitats may be specified using the parameter IHOLD on
Record APUM19.  If IHOLD = 0, the animal is modeled as a population
distributed among habitats as determined from the habitat transition
matrix.  This option requires input of the habitat transition matrix on
record APUM20 and the initial population distribution on record APUM21.  If
IHOLD < 0 , the distribution of the animal population is at steady state as
specified on record APUM21.  Under this option the habitat transition matrix
on record APUM20 is not required.  Finally, if IHOLD > 0 the animal moves
randomly as determined from its previous location and the transition
matrix.  IHOLD in this case is the initial location of the animal.  This
option requires record APUM20 but does not require record APUM21.


    The predation data group (Records APUM22-APUM26) describes the predator-
prey relationships between various animals.  Record APUM22 contains the
number of predator-prey pairs to be described in Records APUM23-APUM26.
Record APUM23 contains the label for the predator, and must correspond to a
previously defined higher animal label input on Record APUM11.  Record
APUM24 is the name of the prey and must correspond to a label  input on
Records APUM4 or APUM11.   Prey may consist of either soil  animals or higher
animals.  The remaining records describe the preference factor,  assimilation
efficiency and capture probabilities  for the predator-prey pair.   Note that
Records APUM23 through APUM26 must be repeated for each predator-prey pair;
if an animal has 3 prey,  3 sets of Records APUM23-APUM26 are required.

    The final data group  (Records APUM27-APUM29)  consists  of various output
options.  Record APUM27 is a 4-character label selecting the frequency at
which detailed dosage breakdowns are  written;  these may be written daily,  •
monthly, or annually.  The model  also has the ability to write out time
series of selected variables for plotting.  The variables  which are to be
written out are specified by labels input on Record APUM29 (see Table 5.8
for the labels corresponding to model  output variables).

5.2.7  Input Data for the Monte Carlo Module (MC)

    The Monte Carlo module uses one batch input file to specify the
distributions of variables and to select output options.   This section
describes the format of this input file and the available  user options.

                                    117

-------
Note that when Monte Carlo simulations are performed, only one TEEAM habitat
can be used in the system configuration.

    The Monte Carlo input file consists of data records and two types of
general utility records.  The first type of utility record is the comment
line, indicated by the presence of three asterisks ('***') as the first non-
blank characters in the line.  Comment lines may be inserted anywhere in the
data set.  The second type of utility record is the END line used by the
code to mark the end of specific data groups.  These records are indicated
by the word 'END1 in the first three columns of the input line, and should
be used only where specified in the following discussion.

    Monte Carlo input data are comprised of four data groups:

        •  Simulation control parameters (Records MC1-MC2)
        •  Input distribution parameters (Records MC3-MC4)
        •  Output options (Records MC5-MC6)
        •  Correlated variable input (Records MC7-MC8)

Data are read sequentially starting with Data Group 1 and ending with Data
Group 4.  Specific formats for each Data Group are shown in Tables 5.9  and
are discussed below.

    The simulation control  data group consists  of  two  records of  data
describing simulation options.   Record  MCI  contains the  (alphanumeric)  title
for the run and  is used  to  label  the output.  Record MC2 contains the  number
of Monte Carlo runs to be used  in the simulation.

    The input distribution  group  consists  of  one  line  of data  (Record  MC3)
for each model  parameter that is  to be  randomly varied.  The first  entry on
Record MC3 is a  label, of length  up to  20  characters,  used  to  identify these
parameters.   The labels  which may be used  are shown on Table 5.10.  The
second entry is  the array index INDX for variables which are arrays used to
indicate which array element is to be randomly  varied.   Variables
dimensioned by HORIZON in Table 5.10 require  the  soil  horizon  number  for
INDX, variables  dimensioned  by  CROP require the crop  number, while  variables
dimensioned by ANIMAL require the animal  number.   Variables dimensioned  by
NAPP require the application number.  Crops and animals  are numbered  in the
order in which they are input in the TFAT  and APUM data  files.   The
remaining data on Record MC3 consist of frequency  distribution  parameters
for the selected variables,  as  shown  in Table 5.9. After  a Record  is
provided for each desired random variable,  an END  card (Record  MC4) must be
supplied to mark the end of this data group.   Note that  by setting  the
distribution flag VAR(5) to 0 the user  can specify a  variable  as a
constant.  In this case, the mean value of the  variable  (VAR(l))  will  be
used in all  simulations.  This  option allows  the  user to randomly vary
parameters without extensive modification  of  the  input file.

                                    118

-------
    The output options data group specifies the statistical output options
for each variable to be written out.  Note that the TEEAM Monte Carlo module
calculates output statistics for variables which are time series' by taking
the maximum value of the variable averaged over a specified period.  Thus,
if an averaging period of N-days is input, the model will select the maximum
N-day average value from each Monte Carlo run for calculations of summary
statistics and cumulative distributions.  This moving average approach was
used because lethal dosage feeding studies are often conducted over a known
number of days.  Thus, if the user wishes to create a distribution of the
highest average daily dosage of chemical to the animal over a 10-day period,
then 10 days would be selected as the averaging period.

    If the maximum daily value of dosage or whole body concentration is
desired, the user can specify one day as the averaging period.  If the user
desires the average of the entire time series, then the length of the time
series would be chosen as the averaging period.  The output data group
consists of one line (Record MC5) for each output variable containing (1) a
character label up to 20 characters long identifying the output variable,
(2) an array index INDX indicating which array element is  to  be  written  out,
(3) a flag indicating if a cumulative distribution should  be  plotted  for
this variable (selected by supplying "CDF" here),  (4)  a flag  indicating  if
values of the variable are to be written out for each  Monte Carlo run
(selected by supplying the word "WRITE"  here), and (5) the number of  days
used in calculating moving averages.  The labels used  to identify variables
are shown on Table 5.8b.  Variables dimensioned by HORIZON in Table 5.11
require the soil  horizon number for INDX, variables dimensioned  by CROP
require the crop  number, while variables dimensioned by ANIMAL require  the
animal number. Variables dimensioned by NAPP require  the  application
number.  A statistical  summary table will be printed out for  all  variables
selected in this  data group. • An END card (Record  MC6) is  supplied to mark
the end of this Data Group after Record  MC5 is input for each output
variable.

    The correlated variable data group is used to  indicate which of the
input variables specified in Data Group  2 are correlated.   Note  that  only
variables with normal and/or log normal  distributions  can  be  correlated.
One line of data  (Record MC7) is provided for each pair of correlated
variables.  The first two entries on this record are labels identifying  the
two variables that are correlated.  These labels must  correspond to Monte
Carlo labels input on Record MC3.  The third entry on  these data lines  is
the value of the  correlation coefficient.  After a data line  is  supplied for
each correlated pair of variables, an END card (Record MC8) must be provided
to mark the end of the data group.
                                    119

-------
Table 5.1.  INPUT FORMATS FOR THE EXECUTION SUPERVISOR MODULE (EXESUP)


Variable             Description                                          Units


   OPTION RECORDS EXESUP1 — ANAME, STATS
                             FORMAT (2A24>

   ANAME             Name of simulation option which can be one of
                     the following list:

                     Value              Module or variable

                     'TFAT1              Terrestrial fate and transport
                     'AERIAL SPRAY1     TFAT interface to FSCBG
                     'FSCBG'            Spray application
                     'PLTGRN1           Plant growth and translocation
                     'APUM1              Animal uptake and movement
                     'NHAB1              Number of habitats
                     'MONTE CARLO1      Monte Carlo module
                     'ENDRUN1           End of option's selection

   STATS             Either "ON" or "OFF" for the modules or an integer
                     value in columns 25-35 for the number of habitats.
                     The following are the default values which are
                     assumed if the module status is not explicitly
                     defined.

                     Module             Stats

                     TFAT               ON
                     AERIAL SPRAY       OFF
                     FSCBG              OFF
                     PLTGRN             ON
                     APUM               ON
                     MONTE CARLO        OFF
                     NHAB               1
   FILE RECORDS EXESUP2 — ANAME, FNAME
                           FORMAT (2A24)

   ANAME             Label defining which input or output file is being defined.
                     The following values of ANAME can be used and must be present
                     if they are required.
                                    120

-------
Table 5.1.  INPUT FORMATS FOR THE EXECUTION SUPERVISOR MODULE (EXESUP)
            (continued)
Variable
Description
           Units
  Value

'PATH'
'TFAT1

'PLTGRN'


'METEOROLOGY1


'GRID DEFINITION1



'FSCBG OUTPUT1

'APUMIN'


'MCIN1


'APUMOUT1


'ANIMAL TIME SERIES'


'CONCENTRATIONS'



'HYDROLOGY'
When Required

never required but
is available as a
convenience in
defining file
location

TFAT is ON

PLTGRN is ON
always


AERIAL SPRAY is ON



AERIAL SPRAY is ON

APUM is ON


MONTE CARLO is ON
APUM is ON and
MONTE CARLO is OFF

APUM is ON and
MONTE CARLO is OFF

TFAT is ON and
MONTE CARLO is OFF
TFAT is ON and
MONTE CARLO is OFF
File Represented

Path for following files.
TFAT definition input file.

Plant growth and translocation
definition input file.

Meteorological data (time
series) input file.

Input file defining the spatial
relationship between FSCBG
output and TEEAM habitats.

Output of" FSCBG simulation.

Animal uptake and movement
definition (input) file.

Monte Carlo simulation
definition (input) file.

Animal model output
file.

Animal time series
output file.

TFAT contaminant
concentrations output
file.

TFAT hydrology output file.
                                   121

-------
Table 5.1.  INPUT FORMATS FOR THE EXECUTION SUPERVISOR MODULE (EXESUP)
            (continued)
Variable
               Description
          Units
Values

'PESTICIDES'
'TIME SERIES'
'ANIMAL TIME SERIES'
'MCOUT'

'MCOUT21
               When Required

               TFAT is ON and
               MONTE CARLO is OFF

               TFAT is ON and
               MONTE CARLO is OFF

               APUM is ON and
               MONTE CARLO is OFF

               MONTE CARLO is ON

               MONTE CARLO is ON
 ENDFILES1             Always - at end of
                       file definition to
                       indicate that all
                       necessary files have
                       been defined.
File Represented

TFAT pesticides output file.
TFAT, PLT6RN, and PLTRNS
time series output file.

Animal model time series
output file.

Monte Carlo output summary.

Monte Carlo output of
individual simulations.
FNAME
The name of the file defined by ANAME.
If the path of a previous PATH is to be
ignored the first character of FNAME
must be '@'.  If FNAME is a path (i.e.
ANAME was defined as 'PATH') it should
end with the character '\' (for MSDOS
microcomputers).
Global parameters record EXESUP3 ~ ISDAY, ISMON, ISTYR, IEDAY, IEMON, IEYR
                                    FORMAT (2X, 312, 10X, 312)

ISDAY   Starting day of month of simulation,
        e.g., 1 = 1st day of month.
ISMON

ISTYR

IEDAY
      Starting month of simulation, e.g., 2 = February.

      Starting year of simulation, e.g., 79.

      Ending day of month of simulation, e.g., 31 = 31st
      day of month.
                                    122

-------
Table 5.1.   INPUT FORMATS FOR THE EXECUTION SUPERVISOR MODULE (EXESUP)
            (continued)
Variable
        Description
Units
  IEMON

  IEYR
 Ending month of simulation, e.g., 12 = December.

 Ending year of simulation, e.g. 80.
  Global parameters record EXESUP4 — PFAC, SFAC, IPEIND, ANETD
                                      FORMAT (2F8.0, 18, F8.0)
  PFAC
  SFAC
  IPEIND
  ANETD
 Pan evaporation adjustment factor.  This factor
 is multiplied by daily pan evaporation to
 estimate daily potential evapotranspiration (ET).
 If daily air temperatures are used to calculate
 ET, any dummy number can be input for PFAC
 (e.g., 0.75).

 Snow melt factor.  Amount of melt per degree (°C)
 above freezing.  Values of snow factor are in
 the order of 0.45.  If snow melt is not to be
 calculated, enter 0.0 for SFAC.

 Pan evaporation flag.  If IPEIND = 0, pan evapora-
 tion data are read.  If IPEIND = 1, temperature
 data are read and used to calculate potential ET.
 If IPEIND = 2, then pan evaporation, if available,
 is used in the meteorologic file; if not,
 temperature is used to compute potential ET.

 Minimum depth to which evaporation is extracted
 year round (e.g., 20) for all habitats.
 Global parameters record EXESUP5 — DT
 FORMAT (6F8.0)
 DT(12)
Average length of daylight in each month.  A total
of 12 values (one for each month) are required
and are input in two lines in the file.
 Trace Record EXESUP5 — ANAME, STATS
                         FORMAT (2A24)
 ANAME

 STATS
The label 'TRACE1

The label 'OFF1  if a subroutine trace is not
desired or an integer value which is the
depth of subroutine nesting to be displayed.
If trace record  EXESUP5 is not present, TRCLVL
defaults to 3.
 decimal
  cm/°C
                                                                         cm
 hr/day
                                   123

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE
   Variable          Description                                   Units
Record FSCBG1-TITLE
               FORMAT (40A2)

    TITLE            A 1 to 80 character title for the
                     FSCBG simulation

Record FSCBG2--ISW
               FORMAT (- list directed -)

   ISW (30)          Array of FSCBG option flags.  Only
                     options 1, 2, 5, 6, 18, 19, 21 and 22
                     are available.  See Table 5.2b.

Record FSCBG3--IFWATR, WNGSPN, HGTCFT, DENLIQ
               FORMAT (- list directed -)

    IFWATR           Flag to identify if spray liquid base
                     is water. If this flag is set equal to
                     "1" or omitted, the program assumes the
                     theoretical drop evaporation equations
                     for water are to be used.  If this
                     parameter is set equal to "2", the
                     program assumes the spray liquid is not
                     water, but calculates a theoretical
                     evaporation rate requiring these
                     additional input parameters:

    WNGSPN           Aircraft wingspan or helicopter rotor diam.     m

    HGTCFT           Height of aircraft or ground sprayer above      m
                     ground

    DENLIQ           Density of drop liquid.  If this parameter    g cm"3
                     is omitted from the input data, the
                     program defaults to a density of 1.0 gram
                     per cubic centimeter.


Record FSCBG4--AIRMOL, AIRPRS,
               FORMAT (- list directed -)

    AIRMOL           Molecular weight of air.  If this            g moie'1
                     parameter is omitted from the input
                     data, the program uses 28.9644 as a
                     default value.


                                    124

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                    Units
    AIRPRS           Air pressure at the site altitude.  If          mb
                     this parameter is omitted from the input
                     data, the program uses 1013.25 millibars
                     as a default value.

***********************RECORD FSCBG4A1 IS REQUIRED*************************
                        if IFWATR = 1

Record FSCBG4A—RELHMO
                FORMAT (- list directed -)

    RELHMO           Average relative humidity above               percent
                     the canopy

***********************RECORD FSCBG5A IS REQUIRED*************************
                          if  ISW(l)  is equal to  1

Record FSCBG5A—ARCRWT, ARCRSP
                FORMAT (- list directed -)

    ARCRWT           Aircraft weight                                 kg

    ARCRSP           Aircraft or ground sprayer ground speed        ms~*


***********************RECQRD FSCBG5B IS REQUIRED*************************
                          if  ISW(l)  is equal to  0

Record FSCBG5B--WAKVEL
                FORMAT (- list directed -)

    WAKVEL           Wake settling velocity                         ms"1


***************RECORDS FSCBG6A1 AND FSCBG6A2 ARE REQUIRED******************
                          if  ISW(2) is equal to  0
                                   125

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
Record FSCBG6A1--DRPUPR
                 FORMAT (- list directed -)

  DRPUPR(20)         Upper limits for drop-size                     v"i
                     categories, for up to 20 categories.   These
                     values must be input in descending order of
                     diameter and the number of drop-size
                     categories is determined by the program
                     from the number of non-zero input values.
                     If ISW(2) is equal to "0" or omitted,
                     this array may be the mean diameter.

Record FSCBG6A2—DRPLWR
                 FORMAT (- list directed -)

  DRPLWR(20)         This parameter is an array specifying the      um
                     lower limit of each drop-size category
                     for the same number of drop-size categories
                     input to the array DRPUPR.  The lower limit
                     of any drop size category cannot be zero
                     ("0").  If the mean diameter is input to
                     the array DRPUPR, this array is omitted
                     from the input data.


***************RECORDS FSCBG6B1 TO FSCBG6B5 ARE REQUIRED******************
                           if  ISW(2) is equal to 1

Record FSCBG6B1—DRPUPR
                 FORMAT (- list directed -)

  DRPUPR(20)         Upper limits for drop-size                     vim
                     categories, for up to 20 categories.   These
                     values must be input in descending order of
                     diameter and the number of drop-size
                     categories is determined by the program
                     from the number of non-zero input values.
                     If ISW(2) is equal to "0" or omitted,
                     this array may be the mean diameter.
                                   126

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable
Description
Units
Record FSCBG6B2—DRPLWR
                 FORMAT (- list directed -)
  DRPLWR(20)
This parameter is an array specifying the
lower limit of each drop-size category
for the same number of drop-size categories
input to the array DRPUPR.  The lower limit
of any drop size category cannot be zero
("0").  If the mean diameter is input to
the array DRPUPR, this array is omitted
from the input data.
Record FSCBG6B3—DRPPCT
                 FORMAT (- list directed -)

  DRPPCT(20)         Array containing the fraction of the
                     total volume of material  for each drop-
                     size category subject to  evaporation.
                     Fracation from 0 to 1.  The default
                     value for each drop-size  category is 1.

Record FSCBG6B4—AIRPRS, AIRMOL, VAPMOL, RELHMO
                 FORMAT (- list directed -)
    AIRPRS
Air pressure at the site altitude.  If
this parameter is omitted from the
input data, the program uses 1013.25
millibars as a default value.
    AIRMOL
    VAPMOL


    RELHMO
                                             fraction
  mb
Molecular weight of air.  If this            g mole
parameter is omitted from the input
data, the program uses 28.9644 as a
default value.

Molecular weight of the vapor from the       g mole
evaporating drops (Default = 18.015).
                                                                        -1
                                                                        -1
Average relative humidity above the           percent
canopy.
                                    127

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
***********************RECQRQ FSCBG6B5 IS READ IF*************************
                   IFWATR is equal to 2  (and  ISW(2) = 1)

Record FSCBG6B5—DFUSIV, HETLAT, THERMC, BCONST, CCONST
                 FORMAT (- list directed -)
                                                                   ?      1
    DFUSIV           Diffusivity of evaporating vapor into       cm  sec
                     air at the drop temperature.  If this
                     parameter is omitted from the input
                     data and IFWATR equals "2," the program
                     calculates DFUSIV, assuming the liquid
                     is similar to water via the equation

         DFUSIV = .211 •((TD+273.16)/273.16)1'94-(1013.25/AIRPRS)

          where TD is the drop temperature approximated by
          the program.

    HETLAT           Latent heat of vaporization at the          cal  mole"*
                     drop temperature.  If this parameter is
                     omitted, the program calculates HF.TLAT,
                     assuming the liquid is similar to water,
                     via the equation

          HETLAT = 597.3 •(273.16/(Tn+273.16))A-VAPMOL

          where:

              A  = 0.107+3.67xlO'4-(Tn+273.16)
              TQ = drop temperature ( C) approximated by the
                   program.
                                                                    cal
    THERMC           Thermal conductivity of the vapor into      	.,
                     air at the drop temperature.  If this       sec cm   *
                     parameter is omitted from the input
                     data, the program calculates THERMC,
                     assuming the liquid is similar to water,
                     via the equation

          THERMC = A(1-(1.17-1.02(B/A))) VAPINF/AIRPRS

          where:
              A = 5.69xlO~ji+1.7xlO-7-DRPTMP
              B = 3.78xlO"b+2.0xlO-7-DRPTMP
              DRPTMP = calculated drop temperature (°K)
              VAPINF = calculated vapor pressure of vapor at infinity (mb)
              AIRPRS = barometric pressure (mb)
                                     128

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
    BCONST           Constant used in the equation that           decimal
                     describes the vapor pressure of the
                     non-water liquid as a function of
                     temperature via the expression

          vapor pressure (in Hg) = exp (BCONST-CCONST/TD)

          where Tn is the drop temperature (°C) approximated by
          the program.  The default value for BCONST is 21.07.

    CCONST           Factor used in the equation that                °K
                     describes the vapor pressure of the
                     non-water liquid as a function of
                     temperature (Default = 5249.9)


***********************RECQRD FSCBG6C1 IS REQUIRED************************
                          if  ISW(2) is equal to 2

Record FSCBG6C1—DRPPCT
                 FORMAT (- list directed -)

  DRPPCT(20)         Array containing the fraction of the         fraction
                     total volume of material for each drop-
                     size category subject to evaporation.
                     Fraction from 0 to 1.  The default
                     value for each drop-size category is 1.


******************RECORDS FCBG6C2 TO FCBG6C4 ARE REQUIRED*****************
                          if  ISW(2) is equal to 2

                     DAU, DBU, and DCU

                     Arrays of coefficients of the quadratic
                     equation that gives the drop diameter
                     in micrometers as a function of time
                     above the canopy.  There are a maximum
                     of 20 values for each array.  The order
                     of values in each array is in descending
                     order of drop size.  There are no default
                     values for these parameters.
                                    129

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
                     The number of drop size categories is
                     determined by the number of non-zero values
                     of DAU and DBU input.  These equations are
                     applicable from release time (T=0) up until
                     the fraction of material given by DRPPCT
                     has been evaporated.  When the fraction
                     of material given by DRPPCT has been
                     evaporated, the program will switch to the
                     coefficients EAU, EBU, and ECU.  Also, if
                     the quadratic given by DAU, DBU, and DCU is
                     not valid beyond a certain time (drop size
                     begins to grow with increasing time), the
                     program will switch to EAU, EBU, and ECU.
                     The DAU, DBU, DCU equation is -

             DROP = DAU(J) + DBU(J)-T + DCU(J)-T2

             where:

                 DROP is diameter in micrometers

                 J is the index over drop size categories

                 T = 0 through Tl, where Tl is the time at which -

             DAU(J)+DBU(J)-T1+DCU(J)-T12 = (DAU(J)3»(1-DRPPCT(J))«3333

                                   or

      DAU(J)+DBU(J)-(T1+DT)+DCU(J)»(T1+DT)2 > DAU(J)+DBU(J)-T1+DCU(J)'T12

Record FSCBG6C2—DAU
                 FORMAT (- list directed -)

    DAU(20)           First coefficient in equation to compute     ******
                     drop diameter in each drop-size category
                     above the canopy as a function of time.
                                    130

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
Record FSCBG6C3—DBU
                 FORMAT (- list directed -)

    DBU(20)           Second coefficient in equation to compute    urn/sec
                     drop diameter in each drop-size category
                     above the canopy as a function of time.

Record FSCBG6C4—DCU
                 FORMAT (- list directed -)
                                                                        o
    DCU(20)           Third coefficient in equation to compute     ym/sec
                     drop diameter in each drop-size category
                     above the canopy as a function of time.


**************RECORDS FSCBG6C5 TO FSCBG6C7 ARE REQUIRED*******************
                          if ISW(2) is equal to 2

                     EAU, EBU, AND ECU

                     Arrays of coefficients of the quadratic
                     equation that gives the drop diameter in
                     micrometers as a function of time above  the
                     the canopy after the fraction material given
                     by DRPPCT has been evaporated down to a
                     minimum of 5 micrometers.  There are a maximum
                     of 20 values for each array.  The order  of
                     values in each array is in descending order
                     of drop size.  If not input, the program keeps
                     the drop size constant after time Tl (DRPPCJ
                     material evaporated) given under DAU, DBU,
                     DCU.  The EAU, EBU, ECU equation is -

             DROP = EAU(J) + EBU(J)*T + ECU(J*T**2

             where:

                DROP is the diameter in micrometers.

                J is the index over drop size categories.

                T is the time in seconds from release time and is
                > = time Tl.
                                     131

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
Record FSCBG6C5—EAU
                 FORMAT (- list directed -)

   EAU(20)           Alternate first coefficient in equation        urn
                     to compute drop diameter in each
                     drop-size category above the canopy
                     as a function of time.   Used when liquid
                     within drop-size category had vaporized
                     to minimize size.

Record FSCBG6C6--EBU
                 FORMAT (- list directed -)

   EBU(20)           Alternate second coefficient in equation    ym sec"
                     to compute drop diameter in each
                     drop-size category above the canopy
                     as a function of time.   Used when liquid
                     within drop-size category had vaporized
                     to minimize size.

Record FSCBG6C7—ECU
                 FORMAT (- list directed -)
                                                                        _p
   EAU(20)           Alternate third coefficient in equation     vim sec
                     to compute drop diameter in each drop-
                     size category above the canopy as a
                     function of time.  Used when liquid
                     within drop-size category had vaporized
                     to minimize size.

********************THE REMAINING RECORDS ARE READ***********************
Record FSCBG7—NSOURC
               FORMAT (- list directed -)

    NSOURC           Total number of line sources (spray lines).
                     The program is capable of processing a
                     maximum of 60 line sources.  If this value
                     is input as "0" or omitted from the input
                     data, the program defaults to a value of "1",
                                   132

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable        Description                .                   Units
Record FSCBG8—Q
               FORMAT (- list directed -)

     Q(30)          This parameter specifies the spray           g m   or  .
                   emission rate for the line sources        gallon acre
                   in units of grams per meter, or
                   gallons per acre, depending on the
                   input parameter SWATH.   If the para-
                   meter SWATH is greater than zero,  the
                   program assumes Q is in gallons per
                   acre and the area sprayed is sprayed
                   in lines of an equal distance (SWATH)
                   apart.  If the spray lines are not
                   uniform or the area is not regular,
                   input Q in units of grams per meter
                   and set the parameter SWATH equal  to
                   "0" or omit SWATH from the input data.
                   If the parameter Q is input as "0"
                   or omitted from the input data, the
                   program defaults to a value of "1".
                   If any Q(2) through Q(NSOURC) are
                   omitted or zero, the program will
                   default the respective Q to the value
                   of the previous Q in the array.


Record FSCB69--SWATH, DELTAH
               FORMAT (- list directed -)

     SWATH          Distance between spray lines when Q is          m
                   input in units of gallons/acre.  If
                   Q is input in grams/meter omit SWATH
                   or set SWATH to 0.0 (Default = 0.0).

    DELTAH         Depth of each volume source for calculation     m
                   of concentration and/or dosage from gas
                   evaporated.  If there is no evaporation, this
                   parameter is not used.   (Default = 1.0).  If
                   >0.0, the program assumes this is the decre-
                   ment in height to use for gas calculations from
                   the release height to the ground (or the drop
                   stops evaporating or completely evaporates).
                   If < 0.0, the program assumes the absolute
                   value is the number of gas sources you desire
                   from the release height to the ground and
                   divides this height into the number of source
                   intervals specified.
                                 133

-------
Table 5.2a.  INPUT FORMATS FOR FSCB6 MODULE (continued)
   Variable          Description                                 Units
Record FSCBG10--TAU, TAUO
                FORMAT (- list directed -)

      TAU             Time to spray cloud stabilization.  If       sec
                     this parameter is input as "0" or
                     omitted from the input data, the program
                     defaults TAU to 2.5 seconds.

Record FSCBG11--SIGXYZ, DECAY, XLRZ, DELU
                FORMAT (- list directed -)

    SIGXYZ           Initial standard deviation of source          m
                     material distribution along the spray
                     line.  If this parameter is input as
                     "0" or omitted from the input data,
                     the program defaults SIGXYZ to WNGSPN/4.3.

     DECAY            Coefficient of time dependent exponential    sec
                     decay for the removal of material due to
                     chemical or physical processes.

     XLRZ            Lateral and vertical reference distance.      m
                     This parameter is normally calculated
                     by the program.  However, if XLRZ is
                     input greater than or equal to zero, the
                     input value is used.

     DELU            Wind-speed sheer above the canopy          m sec
                     (Default = 0).

Record FSCBG12--HM, THETA, DAREA, BETA1
                FORMAT (- list directed -)

      HM             Mixing layer height above ground.             m

     THETA            This parameter specifies the wind            deg
                     direction (direction from which wind
                     is blowing) in degrees, measured
                     clockwise from 0 degrees (north).

     BETA1            Ratios of Langrangian to Eularian time-    decimal
                     scales used in the correction factor
                     on the standard deviations of the
                     horizontal and vertical wind directions.
                     Suggested value, if used, is 1.0
                     (default is 0.0).

                                    134

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE (continued)
   Variable          Description                                   Units
Record FSCBG13--DX
                FORMAT (- list directed -)

    DX(60)           Source X (East-West) Coordinates — This        m
                     parameter is an array specifying the start
                     and end X coordinate of each line source.
                     Input the start followed by the end X
                     coordinate for each source.  Either end of
                     the line may be the start coordinate.  X
                     must increase from west to east.

Record FSCBG14—DY
                FORMAT (- list directed -)

    DY(60)           Source Y (North-South) Coordinates — This      m
                     parameter is an array specifying the start
                     and end Y coordinate of each line source.
                     Input the start followed by the end Y
                     coordinate for each source.  Y must increase
                     from south to north.

Record FSCBG15-PCTMAT
                FORMAT (- list directed -)

  PCTMAT(20)         Input these values in descending order of    fraction
                     drop-size categories.

Record FSCBG16--VSGAM
                FORMAT (- list directed -)

   VSGAM(2)          Array of two settling velocities that        m  sec"
                     specify the settling velocities that
                     break the curve of surface reflection
                     coefficients given by GAMA, GAMB, and GAMC
                     into three parts.   Settling velocities (VS)
                     greater than or equal to VSGAM(l) use
                     GAMA(l), GAMB(l),  and GAMC(l) to calculate
                     the surface reflection coefficient.  VS
                     values less than VSGAM(2) use GAMA(2),
                     GAMB(2), and GAMC(2) to calculate the
                     surface reflection coefficient.  VS values
                     less than VSGAM(2) use GAMA(3), GAMB(3),
                     and GAMC(3).  Default values for VSGAM are
                     provided only if VSGAM is omitted from the
                     input list.  Default values are VSGMC(l) =
                     0.04 and VSGAM(2)  = 0.012.

                                    135

-------
Table 5.2a.  INPUT FORMATS FOR FSCBG MODULE  (concluded)
   Variable
  Description
                          Units
Record FSCBG 17-GAMA, GAMB,  GAMC
                FORMAT (- list directed -)

GAMA, GAMB, GAMC - Arrays of coefficients of the quadratic equation that
gives the fraction of material reflected at the surface as a function of
the drop settling velocity.
                                           2
Fraction = GAMA(J) + GAMB(J) + GAMC(J)  • VS

where:  VS is the drop settling velocity in meters/sec
        (1,  if VS > VSGAM(l)
    J = (2,  if VS > VSGAM(2) and VS <  VSGAM(l)
        (3,  if VS < VSGAM(2)

Three coefficients of GAMA,  GAMB, and GAMC  each must be input.  The first
value of each array is for settling velocities greater than or equal to
VSGAM(l).  The second value  of each array is for settling velocities less
than VSGAM(l) and greater than or equal to  VSGAM(2).  The third value of
each array is for settling velocities less  than VSGAM(2).  The default
values for the three arrays  are:
GAMA(l) = 0.75,
GAMB(l) = -2.5,
GAMC(l) = 0.00,

   GAMA(3)
   GAMB(3)
   GAMC(3)
GAMA(2) = 0.83465302,
GAMB(2) = -6.9031391,
GAMC(2) = 57.4092560,
GAMA(3) ,= .91639996
GAMB(3) = -22.357124
GAMC(3) = 821.426510
  First coefficient to calculate
  fraction of material reflected at
  the surface as a function of drop
  settling velocity.

  Second coefficient to calculate .
  fraction of material reflected at
  the surface as a function of drop
  settling velocity.

  Third coefficient to calculate
  fraction of material reflected at
  the surface as a function of drop
  settling velocity.
                        fraction
                    fraction sec m
                    fraction
                                                                          -1
                                     136

-------
Table 5.2b.  EXPLANATION OF FSCBG OPTIONS (ISW VALUES)
ISW(l) - Wake settling velocity option.
         If = 0,   the program assumes the wake settling velocity WAKVEL is
                  being input
         If = 1,   the program assumes the wake settling velocity is to be
                  calculated from the input parameters - ARCRWT, AIRDEN,
                  WNGSPN and ARCRSP.
ISW(2) - Evaporation model  option
         If = 0,   The evaporation model  is not executed.   The program
                  assumes there is no change in drop size with time.

         If = 1,   The evaporation model  is executed.  The program
                  calculates the rate of change of drop size with time.

         If = 2,   The evaporation model  is not executed.   The program
                  assumes the user inputs the equations of the rate of
                  change of drop size with time.


ISW(5) - Print evaporation model (ISW(2)) calculations option.

         If = 0,   Evaporation model calculations are not printed.  This
                  includes all calculations under ISW(2).

         If = 1,   The evaporation model  calculations are printed in the
                  form of quadratic equations only.

         If = 2,   The same as 1 and calculations at approx. one meter
                  height intervals are printed.

         If = 3,   The same as 1, but calculations at all  height intervals
                  are printed.

         If = 4,   The same as 3, but additional tables giving the accuracy
                  of the regression are printed.

         If = 5,   The same as 4, but debug calculations are also printed.
                  Only use a few drop size categories due to the large
                  volume of print output.


ISW(6) - Dosage Model Option.

         If = 0,   The Dosage is not calculated.

         If = 1,   The Dosage is calculated and printed.

                                 137

-------
Table 5.2b.  EXPLANATION OF FSCBG OPTIONS (ISW VALUES)  (continued)



ISW(18)- Option to print dosage and or deposition by drop size category.

         If = 0,  Only the total  over drop size categories is printed.

         If = 1,  Dosage and/or deposition for each drop size category  is
                  printed.

         If = 2,  Dosage and/or deposition for each drop size category  as
                  well as the sum over drop size categories is printed.


ISW(19)- Option to print dosage and/or deposition by gas versus drop or by
         volatile versus non-volatile material

         If = 0,  Only total dosage and/or deposition is printed.

         If = 1,  The program prints the contribution to the total  dosage
                  from evaporated (gaseous) material and from non-
                  evaporated (drop) material separately.  Deposition is
                  non-evaporated material only.

         If = 2,  The program prints the contribution to the total  dosage
                  from evaporated (gaseous) material and from non-
                  evaporated (drop) material separately and summed.
                  Deposition is non-evaporated material only.

         If = -1, The program prints the contribution to the total  dosage
                  and/or deposition from volatile and from non-volatile
                  material separately.  The fraction of volatile material
                  is determined from the array DRPPCT,  which gives  the
                  fraction of material in each drop size category subject
                  to evaporation.

         If = -2, The program prints the contribution to the total  dosage,
                  and/or deposition from volatile and from non-volatile
                  material separately and summed.
                                      138

-------
Table 5.2b.   EXPLANATION OF FSCBG OPTIONS (ISW VALUES)  (concluded)
ISW(21)- Option to calculate drop dosage and/or deposition from evaporating
         drops using the new or old model.   The old model  assumed VS was
         constant with distance and used VS*X/U in the vertical term.   The
         new model uses (A+B*X+C*X**2)*X/U  in the vertical term.   This
         option is not applicable when ISW(2) = 0.

         If = 0,  The new model is executed.   Results  for  dosage and
                  deposition are returned as zero if the program calculated
                  maximum distance for deposition is exceeded or for
                  deposition, if negative deposition occurs.

         If = 1,  The new model is executed,  but the maximum  distance
                  provision is relaxed.

         If = 2,  The old model is executed.   VS rather than  A+B*X+C*X**2 is
                  used in the vertical term.   There are no maximum distance
                  restrictions.
ISW(22)- Option to produce deposition output  in units  of mass and length
         depending on ISW(9)  and  ISW(ll),  in  number of drops per square
         meter and in drop volume mean diameter.   This option is used  only
         for ISW(18)  > 0.   ISW(9) must not be entered  as zero if ISW(22)
         is  set >  zero.

         Remaining options are  set to the  following values in TEEAM:

                            ISW                  Value

                             3                     0
                             4                     1
                             7                     0
                             8                     1
                             9                     1
                             10                    0
                             11                    0
                             12                    0
                             13                    1
                             14                    1
                             15                    0
                             16                    0
                             17                    0
                             20                    0

         Explanation  of  these options is given in  the  FSRED2 subroutine of
         the TEEAM code.
                                    139

-------
Table 5.3.  INPUT FORMATS FOR THE SPRAY GRID DEFINITION MODULE

   Variable          Description                                   Units
***************************RECORD GRID1 REPEATS****************************
                              FOR ALL HABITATS
RECORD GRID1—XSW, YSW, XNE, YNE
              FORMAT (4F8.0)
      XSW             Southwest coordinate of habitat.                 m
      YSW             Southwest coordinate of habitat.                 m
      XNE             Northeast coordinate of habitat.                 m
      YNE             Northeast coordinate of habitat.                 m
                     (Note:  X must increase from west to east,
                             Y must increase from south to north)
RECORD GRID2--MXPNTS, MYPNTS
              FORMAT (218)
    MXPNTS           Number of grid points in E-W transect.
    MYPNTS           Number of grid points in N-S transect.
RECORD GRID3—X(l), GDX
              FORMAT (2F8.0)
     X(l)            Initial x-coordinate of FSCBG receptor
                     grid                                            m
      GDX             Increment between FSCBG receptor grid
                     x-coordinates                                   m
RECORD GRID4—Y(l), GDY
              FORMAT (2F8.0)
     Y(l)            Initial y-coordinate of FSCBG receptor
                     grid                                            m
      GDY             Increment between FSCBG receptor grid
                     y-coordinates                                   m
RECORD GRID5--NDOSE
              FORMAT (18)
     NOOSE           Number of aerial spray applications
                                  140

-------
 Table 5.3.   INPUT FORMATS FOR THE SPRAY GRID DEFINITION MODULE (continued)
    Variable          Description                                    Units
                                  QRID6  REPEATS****************************
                            FOR EACH APPLICATION

 RECORD GRID6— SAPD,  SAPM,  IYRPLY
              FORMAT (2X,  312)

     SAPD             Spray application day

     SAPM             Spray application month

     IYRPLY            Spray application year

 *************************R£coRD$ QRID7 TO GRID 10***************************
                           REPEAT  FOR  EACH HABITAT

 RECORD GRID7— SAM, SANOPT
              FORMAT (218)

      SAM             Pesticide application model  flag.  There
                      are two options:  SAM = 2 indicates the
                      user  supplies the fraction of the
                      application that goes to the canopy,
                      and SAM = 3 indicates the model
                      calculates the canopy application.

    SANOPT            Canopy penetration flag.  SANOPT =1
                      indicates user-supplied canopy penetration,
                      and SANOPT = 2 indicates canopy penetration
                     will  be computed.

***********************R£CORD GRID8A ONLY REQUIRED************************
                           IF SAM =  2, REPEAT FOR
                          EACH APPLICATION  (NDOSE)

RECORD GRID8A— SERCAN(I)
               FORMAT (F8.0)

   SERCAN(I)          The fraction of application  I which is       fraction
                     applicable to the crop canopy

***********************RECQRD GRID8B ONLY REQUIRED************************
                                 IF  SAM = 3

RECORD GRID8B--SILTRA
               FORMAT (F8.0)

    SILTRA           Filtration parameter for canopy application
                     model
                                   141

-------
Table 5.3.  INPUT FORMATS FOR THE SPRAY GRID DEFINITION MODULE (continued)
   Variable          Description                                   Units


********************RECORDS GRID9A1 AND GRID9A2 ONLY*********************
                     REQUIRED IF SANOPT = 1, REPEAT FOR
                          EACH APPLICATION (NDOSE)

RECORD GRID9A1—SDEPLV(I), SIS(I)
                FORMAT (18, F8.0)

   SDEPLV(I)          Number of deposition levels in canopy
                     for application I

    SIS(I)           Thickness of each deposition level              cm

************************RECORD GRID9A2 REPEATS FOR*************************
                           EACH  DEPOSITION  LEVEL
                              J = 1, SDEPLV(I)

RECORD GRID9A2--SRACDL
                FORMAT (F8.0)

 SRACDL (J,I)        Fraction of deposition occurring on          fraction
                     canopy level J for application I

***********************RECORD GRID9B ONLY REQUIRED************************
                                IF SANOPT  -  2

RECORD GRID9B--SDIS, SETA
               FORMAT (2F8.0)

     SDIS            Thickness of canopy deposition layers           cm

     SETA            Penetration model attenuation constant         cm

RECORD GRID10—SLDKRT, SEXTRC
               FORMAT (2F8.0)

    SLDKRT           Decay rate constant for pesticide on          days"1
                     plant foliage

    SEXTRC           Foliar extraction coefficient for
                     pesticide washoff per cm of
                     precipitation

-------
Table 5.4.  INPUT FORMATS FOR THE TFAT METEOROLOGY FILE
   Variable
Description
  Units
Record MET1—MM, MD, MY, PRECIP, PEVP,  TEMP,  SOLAR, WIND
             FORMAT (IX, 312, 5F10.0)
      MM
      MD
      MY
    PRECIP
     PEVP
     TEMP
    SOLAR
     WIND
Month (i.e., 1 for January)
Day of Month (i.e., 1)
Year (i.e., 84)
Daily precipitation amount
Daily potential evaporation.  If a value
of -99 is entered, the model will estimate
PEVP from the air temperature and daily
hours of sunshine for the month.
Air temperature
Solar radiation
Mean wind speed at the reference
height ZWIND (on Record TFAT19)
    cm
    cm
degrees C
 langleys
      -1
  cm  s
                                    143

-------
Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
            (TFAT)
   Variable          Description                                    Units
*********************y\L|_ TFAT RECORDS ARE REPEATED AS**********************
                          A SET FOR EACH HABITAT

Record TFAT1—TITLE
              FORMAT (A80)

     TITLE            A 1 to 80 character title for the
                     TFAT simulation.

Record TFAT2--HTITLE
              FORMAT (A80)

    HTITLE           This card provides a comment line
                     of 80 characters for the user to
                     input information regarding hydrology
                     parameters.

Record TFAT3--INICRP, ISCOND, NPTIME, AFIELD
              FORMAT (318, F8.0)

    INICRP           User specified initial crop number if
                     simulation date is before first crop
                     emergence date (see record TFAT10).

    ISCOND           User specified surface condition after
                     harvest corresponding to INICRP (either
                     fallow cropping, or residue, corresponding
                     to dimensionless integer of 1, 2 or 3).

    NPTIME           Number of time steps per day used in pond       days
                     water balance and chemistry calculations
                     (must be > 1).

    AFIELD           Area of the habitat (plan view).                 ha


Record TFAT4--KSAT, HFPOND, PDEPTH, TC, ARAIN, BRAIN
              FORMAT (6F8.0)

     KSAT            Saturated hydraulic conductivity of           cm hr"1
                     surface sorls.

    HFPOND           Green-Atnpt suction parameter.                   cm

    PDEPTH           Initial pond depth.                             cm
                                   144

-------
Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
            (TFAT) (continued)
   Variable
       Description
                                               Units
TC

ARAIN
Average runoff time of concenration
for the habitat.
Coefficient A in rainfall-duration curve:
hours

hr cm~B
                     duration = A(precip)B
     BRAIN


Record TFAT5-


     ASOIL



     BSOIL


Record TFAT6-


    ERFLA6
       Coefficient B in the rainfall-duration
       curve

ASOIL, BSOIL
FORMAT (2F8.0)

       Constant in relationship between air
       temperature and soil surface temperature:
       TSOIL = ASOIL + (BSOIL)(TAIR)

       Slope in the relationship between air
       temperature and soil surface temperature.
                                            degrees  C
ERFLAG
FORMAT
(18)

Erosion flag.  If erosion losses are
not to be calculated, ERFLAG = 0, otherwise
ERFLAG = 1.
Record TFAT7--USLEK,
              DO NOT
     USLEK


    USLELS


     USLEP


    AFIELD

Record TFAT8-
       USLELS, USLEP, AFIELD, (Only if ERFLAG = 1;
	 include this card if ERFLAG = 0).
FORMAT (4F8.0)

       Universal soil loss equation (K) soil
       credibility parameter (e.g., 0.15).

       Universal soil loss equation (LS)
       topographic factor (e.g., 0.14).

       Universal soil loss equation (P) supporting
       practice factor (e.g., 1.0).
       Area of field or plot.
NDC
FORMAT
                                                ha
                     (18)
                                   145

-------
Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
            (TFAT) (continued)
   Variable          Description                                   Units
      NDC             Number  of different  crops  used  in the
                      simulation  (minimum  of  1).

 Record  TFAT9--ICNCN,  CINTCP, AMXDR, COVMAX,  ICNAH, CN, USLEC, WFMAX
               FORMAT  (18,  3F8.0.I8, 3(1X,  13),  3(1X,  F3.0),  F8.0)
               NOTE:   One record  each must be read in  to match the
                      total number of crops (NDC).

     ICNCN            Crop  number.

     CINTCP            Maximum interception storage of  the crop.        cm

     AMXDR            Maximum active root  depth  of the crop.           cm

     COVMAX            Maximum areal coverage  of  the crop at         percent
                      full  canopy.

     ICNAH            Soil  surface condition  after crop harvest
                      (1 =  fallow, 2 = cropping,  3 = residue).

      CN              Runoff  curve number  for the antecedent
                      soil  water  condition II, for fallow, crop,
                      and residue fractions of the growing season
                      (e.g. 86, 78, 82).

     USLEC           Universal  soil loss equation cover
                     management factor.  Three values must be
                     entered in the same order as (CN), fallow,
                     crop,  and residue.   Values only are
                     required if ERFLAG  = 1.   Leaving them in
                     the input stream will have no effect if
                     ERFLAG = 0  (e.g.,  0.20)
                                                                       _2
     WFMAX           Maximum dry foliage weight of the crop at     kg m
                     full  canopy .   Only required if the
                     exponential  filtration model is used for
                     pesticide application (values of WFMAX will
                     not affect the simulation if FAM = 1,  2,
                     or 4.   See record TFAT15.

Record TFAT10—NCPDS
               FORMAT (18)

     NCPDS           Number of  cropping  periods in the simulation
                     (minimum of 1).   If three cropping-years of

                                 146

-------
 Table 5.5.   INPUT FORMATS  FOR  THE TERRESTRIAL FATE AND TRANSPORT MODULE
             (TFAT)  (continued)

    Variable           Description                                    Units
                     continuous corn are simulated, NCPDS = 3.
                     If two winter cover crops are in the middle
                     of the three years of corn, NCPDS = 5.
Record TFAT11—EMD, EMM IYREM,  MAD,  MAM, IYRMAT, HAD, HAM, IYRHAR, INCROP
               FORMAT (2X, 312, 2X,  312, 2X, 312, 18)
               NOTE:  One card each  must be read in to match the total
               number of cropping periods (NCPDS).
      EMD             Day of month of crop emergence (e.g., 20).
      EMM             Month of crop emergence (e.g., 4).
     IYREM            Year of crop emergence (e.g., 82).
      MAD             Day of month of crop maturation (e.g., 15).
      MAM             Month of crop maturation (e.g., 10).
    IYRMAT           Year of crop maturation (e.g., 82).
      HAD             Day of month of crop harvest (e.g.,  20).
      HAM             Month of crop harvest (e.g., 10).
    IYRHAR           Year of crop harvest (e.g., 82).
    INCROP           Crop number of  crop growing in current
                     period (e.g., 1).
**********************END OF HYDROLOGY AND CROP DATA**********************
Record TFAT12—PTITLE
               FORMAT (A80)
    PTITLE           This card  provides a comment line of 80
                     characters for  the user to input
                     information regarding pesticide parameters.
Record TFAT13--NAPS
               FORMAT (18)
     NAPS             Number of  pesticide applications
                     (minimum of 1).  If no applications  are
                     desired in TFAT set TAPP (next record)
                                   147

-------
Table 5.5.   INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
             (TFAT)  (continued)
    Variable          Description                                    Units


                     to zero for this one required application,
                     or set the date of application beyond the
                     end of the simulation 	 date.

Record TFAT14--APD, APM, IAPYR, TAPP, DEPI, C6RAN
               FORMAT (2X, 312, 3F8.0)
               NOTE:  One card should be entered for each application
               up to the number of applications (NAPS).

     APD             Day of the month of pesticide application
                     (e.g., 10).

     APM             Month of pesticide application (e.g., 5).

     IAPYR            Year of pesticide application (e.g., 82).

     TAPP            Total pesticide or granule application.        kg/ha

     DEPI            Depth of pesticide incorporation.               cm

     C6RAN            Concentration of pesticide in granules          g/g
                     (required only for granule applications,
                     FAM = 4).

Record TFAT15--FAM, KANOPT
               FORMAT (218)

     FAM             Pesticide application model.  There are
                     four options:  FAM = 1 indicates application
                     to soil only, FAM = 2 indicates the user
                     supplies the fraction of the application
                     that goes to the canopy, FAM = 3 indicates
                     the model calculates the canopy application,
                     and FAM = 4 indicates application of
                     granules to the soil surface.

    KANOPT           Canopy penetration option, KANOPT =1
                     indicates user-supplied canopy penetration,
                     and KANOPT = 2 indicates the model will
                     compute canopy penetration.

Record TFAT16A--PERCAN (I)
                FORMAT (F8.0)

                     Required only if FAM = 2; repeat for
                     each application (NAPS times).

                                   148

-------
 Table  5.5.   INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
             (TFAT)  (continued)
   Variable          Description                                    Units
  PERCAN (I)         The fraction of application I which          fraction
                     stays on the crop canopy.

Record TFAT16B-FILTRA
                FORMAT (F8.0)

                     Required only if FAM = 3 (model computes
                     canopy application).

                                                                    2  ka'1
    FILTRA           Filtration parameter for canopy              m   y
                     application model.

Record TFAT16C—GKWET, GKDRY
                FORMAT (2F8.0)

                     Required only if FAM = 4 (granule application).

     GKWET            Granule pesticide decay constant when         days
                     granule is immersed in water.

     GKDRY            Granule pesticide decay constant when         days'*
                     granule is dry.

**********************RECC)RDS TFAT17A AND TFAT17B ARE**********************
                       REPEATED FOR EACH APPLICATION

Record TFAT17A—NDEPLV(I), DIS(I)
                FORMAT (18, F8.0)

               Required only if FAM = 2 or 3 and KANOPT = 1.

  NDEPLV(I)          Number of deposition levels in caaopy for
                     application I.

    DIS(I)            Thickness of each deposition level.             cm

Record TFAT17B—FRACDL(J,I)
                FORMAT (F8.0)

                Required only if FAM = 2 or 3 and KANOPT = 1;
                repeat for each deposition level J for
                application I.

 FRACDL (J,I)         Fraction of deposition occurring on          fraction
                     canopy level  J for application I.

                                   149

-------
Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
            (TFAT) (continued)
   Variable          Description                                    Units
Record TFAT17C--BDIS, BETA
                FORMAT (2F8.0)

                     Required only if FAM = 2 or 3 and KANOPT = 2.

     BDIS            Thickness of canopy deposition layers.          cm

     BETA            Penetration model attenuation constant.        cm'1

Record TFAT17D--PLDKRT, FEXTRC
                FORMAT (2F8.0)
                Required only if FAM = 2 or 3.

    PLDKRT           Decay rate constant for pesticide on          days
                     plant foliage.

    FEXTRC           Foliar extraction coefficient for
                     pesticide washoff per cm of
                     precipitation.

Record TFAT18~KDPOND,DWAT,CPND
               FORMAT (3F8.0)

    KDPOND           Degradation rate for pesticide in             days'1
                     ponded water.

     DWAT            Diffusion coefficient of pesticide           cm2  day'1
                     in water.

     CPND            Initial concentration of pesticide            g cm"3
                     on ponded water (should = 0.0 if
                     there is no initial ponding).

Record TFAT19--ZWIND, DAIR, IH, CCAN
               FORMAT (4F8.0)

    ZWIND            Reference height for wind speed                  m
                     measurements in the MET data file.

     DAIR            Diffusion coefficient of pesticide           cm2  day"1
                     in air.
                                  150

-------
Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
            (TFAT) (continued)
   Variable          Description                                   Units
      KH             Dimensionless Henry's Law constant
                     for pesticide (i.e., dissolved-
                     vapor phase partition coefficient).

     CCAN            Initial concentration of pesticide in
                     the canopy atmosphere.                        g cm"3

*********************END OF PESTICIDE APPLICATION DATA*********************


Record TFAT20—STITLE
               FORMAT (A80)

    STITLE           This card provides a comment line of
                     80 characters for the user to input
                     information regarding soils properties.

Record TFAT21—CORED, UPTKF, NCOM2, BDFLAG, THFLAG,  KDFLAG, HSWZT
               FORMAT (2F8.0, 518)

    CORED            Total  depth of soil  core                        cm

    UPTKF            Plant  uptake efficiency factor;  UPTKF =0
                     indicates no plant uptake simulated,
                     UPTKF  = 1 indicates  uptake is simulated
                     and is equal  to  the  crop transpiration rate,
                     0
-------
Table  5.5.   INPUT  FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
             (TFAT)  (continued)
    Variable          Description                                    Units
    KDFLAG           Calculation flag for soil/pesticide sorption
                     partition coefficients; KDFLAG = 0 indicates
                     partition coefficients known and entered,
                     BDFLAG = 1 indicates partition coefficients
                     not known and will be calculated.

     HSWZT            Switch for soil hydraulics; HSWZT =0
                     indicates free draining soils, HSWZT = 1
                     indicates restricted draining soils.

Record TFAT22A—PCMC, SOL (Only if KDFLAG = 1, DO NOT include if KDFLAG = 0)
                 FORMAT (18, F8.0)

     PCMC            Calculation flag for model to estimate
                     pesticide soil partition coefficients.
                     There are three options: PCMC = 1,
                     PCMC = 2, and PCMC = 3.

      SOL             Pesticide solubility.  The units vary
                     according to the model (PCMC) selected;
                     PCMC = I4imole fraction; PCMC = 2.     i
                     mg liter  ' P*MC = 3» "iicromoles liter'1.

Record TFAT23—NHORIZ
               FORMAT (18)

    NHORIZ           Total number of soil horizons (minimum of 1).

*************REPEAT RECORDS TFAT 24-TFAT25D FOR EACH SOIL HORIZON***********

Record TFAT24—HORIZN, THKNS, BD, DISP, DKRATE, THETO, AD
               FORMAT (18, 6F8.0)

    HORIZN           Soil horizon number.

     THKNS            Soil horizon thickness                          cm

      BD             Soil bulk density (if BDFLAG = 0) and/or      g cm'3
                     mineral bulk density (if BDFLAG = 1) in
                     each horizon.

     DISP            Dispersion/diffusion coefficient for         cm2  day"1
                     each soil horizon.
                                  152

-------
 Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AMD TRANSPORT MODULE
             (TFAT) (continued)
Variable
DKRATE
THETO

AD
Description
Pesticide decay rate in the soil horizon
Initial soil water content in the horizon

Soil horizon drainage parameter, used only
Units
days' *
cm3 cm"3
-1
day
                     if HSWZT = 1, otherwise, the value is
                     ignored.

Record TFAT25A—THEFC, THEWP, KD, OC (Only if THFLAG = 0 and KDFLAG = 0)
                FORMAT (8X, 4F8.0)

     THEFC            Field capacity soil water content            cm3 cm"3
                     of horizon

     THEWP            Wilting point soil water content             cm3 cm~3
                     of horizon

      KD             Sorption partition coefficient for soil       cm3  g"1
                     horizon/pesticide combination

      OC             Organic carbon content of soil horizon.       percent
                     This value is also required if BDFLAG = 1.

Record TFAT25B—THEFC, THEWP, OC (Only if THFLAG = 0 and KDFLAG = 1)
                FORMAT (8X, 3F8.0)

     THEFC            Field capacity soil water content of         cm3 cm~3
                     horizon.

     THEWP            Wilting point soil water content of          cm3 cm"3
                     horizon.

      OC             Organic carbon content of soil horizon.       percent
                     This value is also required if BDFLAG = 1.

Record TFAT25C--SAND, CLAY, OC, KD (Only if THFLAG =* 1 and KDFLAG = 0)
               FORMAT (8X, 4F8.0)

     SAND            Percent sand in soil horizon.                 percent

     CLAY            Percent clay in soil horizon.                 percent

      OC             Organic carbon content of soil horizon.       percent
                     This value is also required if BDFLAG = 1.


                                   153

-------
Table  5.5.   INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
             (TFAT)  (continued)
    Variable          Description                                    Units


                               "                                     01
      KD             Sorption partition coefficient for soil       cnr g"1
                     horizon/pesticide combination.

Record TFAT25D--SAND, CLAY, OC (Only if THFLAG = 1 and KDFLAG = 1)
                FORMAT (8X, 3F8.0)

     SAND            Percent sand in soil horizon.                percent

     CLAY            Percent clay in soil horizon.                percent

      OC             Organic carbon content of soil horizon.      percent
                     This value is also reguired if BDFLAG = 1.

Record TFAT26—ILP, CFLAG
               FORMAT (218)

      ILP             Initial level of pesticide indicator.
                     Signals user to input an initial pesticide
                     storage.  ILP = 0, indicates no initial
                     levels input; ILP = 1, indicates initial
                     levels are being input.

     CFLAG            Conversion flag for initial pesticide levej^
                     input.  CFLAG=0, indicates input in mg kg  »
                     CFLAG = 1, indicates input in kg ha~*.  This
                     flag need not be assigned if ILP = 0.

Record TFAT27A--PESTR (Only if ILP = 1)
                FORMAT (8F8.0)

     PESTR            Initial pesticide level in each compartment
                     (up to NCOM2) as entered from Record TFAT22.

                     Input must be either in mg kg~* or kg ha  .

*****************************ENQ OF SOILS DATA*****************************

Record TFAT28—ITEM1, STEP1, LFREQ1, ITEM2, STEP2, LFREQ2, ITEM3, STEPS,
               LFREQ3
               FORMAT (3 (4X, A4 4X, A4, 18)

     ITEM1            Hydrologic output summary indicator. WATR
                     is inserted to call hydrologic summaries.
                     A blank is left for ITEM1 if hydrologic
                     summaries are not desired.

                                  154

-------
Table 5.5.   INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
             (TFAT)  (continued)
   Variable          Description                                    Units
     STEP1           Time step of output.  Three options are
                     available:  DAY for daily, MNTH for monthly,
                     or YEAR for annual output.

    LFREQ1           Frequency of soil compartment reporting.
                     Example:  LFREQ1 = 1, every compartment is
                     output; LFREQ = 5, every fifth compartment
                     is output.

     ITEM2           Pesticide output summary indicator.  PEST is
                     inserted to call pesticide summaries (of mass
                     migration).  A  blank is inserted for ITEM2 if
                     pesticide summaries are not desired.

     STEP            Same as STEP1.

    LFREQ2           Same as LFREQ1.

     ITEM3           Pesticide concentration profile indicator.
                     CONC is inserted to call pesticide concen-
                     tration profile summaries.  A blank is
                     inserted if concentration profiles are not
                     desired.

    STEP3           Same as STEP1.

    LFREQ3           Same as LFREQ1.

Record TFAT29--NPLOTS
               FORMAT (18)

    NPLOTS           Number of time series to be written to
                     plotting file (maximum of 7).

Record TFAT30—PLNAME,  MODE, IARG, CONST (Only if NPLOTS is greater
               than zero)
               FORMAT (4X, A4, 4X, A4, 18, F8.0)

    PLNAME           Identifier of time series.  Possible options
                     are listed in Table 5.6.
                                   155

-------
Table 5.5.  INPUT FORMATS FOR THE TERRESTRIAL FATE AND TRANSPORT MODULE
            (TFAT) (concluded)
   Variable          Description                                   Units
     MODE            Plotting mode.  Two options are available:
                     TSER provides the time series as output,
                     TCUM provides the cumulative time series.

     IAR6            Argument of variable identified in PLNAME.
                     Example:  INFL is specified which corresponds
                     to AINF within the FORTRAN program. AINF is
                     dimensioned from 1 to NCOM2.  IARG must be
                     specified to identify the soil compartment
                     (1 to NCOM2) reporting for AINF (IARG is
                     left blank for sealers).

     CONST            Specifies a constant with which the user can multiply
                     the times series for unit conversion, etc.  If left
                     blank a default of 1.0 is used.

***************************ENQ op JPAT DATA SET***************************
                                     156

-------
Table  5.6.  VARIABLE DESIGNATIONS FOR TFAT TIME SERIES FILES
Variable
Designation
(PLNAME)
Water Storage
I NTS
SWTR
SNOP
THET
Water Fluxes
PRCP
SNOF
THRF
INFL
RUNF
CEVP
SLET
FORTRAN
Variable

CINT
SW
SNOW
THETN

PRECIP
SNOWFL
THRUFL
AINF
RUNOF
CEVAP
ET
Description

Interception storage
on canopy
Soil water storage
Snow pack storage
Soil water content

Precipitation
Snowf al 1
Canopy throughfall
Percolation into each
soil compartment
Runoff depth
Canopy evaporation
Actual evapptrans-
Unlts

cm
cm
cm
cm cm"1

cm day"
cm day"1
cm day"1
cm day"1
cm day"
cm day"1
cm day"1
Arguments
Required
(IARG)

None
1-NCOM2
None
1-NCOM2

None
None
None
1-NCOM2
None
None
1-NCOM2
    TETD             TDET


 Sediment  Flux

    ESLS             SEDL


 Pesticide Storages

    EPST             FOLPST
            piratlon from each
            compartment

            Total dally actual         cm day
            evapotranspi rat1on
            Event soil loss
    TPST
   SPST
PESTR
SPESTR
                                                                 ,-1
                           Tonnes,
                             day'1
Foliar pesticide           g cm
storage

Total soil  pesticide
storage 1n  each soil
compartment

Dissolved pesticide        g cm
storage in  each soil
compartment
                                           -2
                                                              ,-3
                                                              ,-3
None




None




None


1-NCOM2



1-NCOM2
                                     157

-------
Table 5.6.  VARIABLE DESIGNATIONS FOR TFAT TIME SERIES FILES (continued)
Variable
Designation
(PLNAME)
Pesticide Fluxes
TRAP

FPDL

WFLX

OFLX


AFLX


DKFX


UFLX


RFLX

EFLX

Pesticide Fluxes
RZFX


TUPX


TDKF

FORTRAN
Variable

TAPP

FPDLOS

WOFLUX

OFFLUX


ADFLUX


DKFLUX


UPFLUX


ROFLUX

ERFLUX


RZFLUX


SUPFLX


SDKFLX

Description

Total pesticide
application
Foliar pesticide
decay loss
Foliar pesticide
^washoff flux
Individual soil
compartment pesticide
net diffusive flux
Pesticide advective
flux from each soil
compartment
Pesticide decay flux
in each soil compart-
ment
Pesticide uptake flux
from each soil compart-
ment
Pesticide runoff flux

Pesticide erosion flux


Net pesticide flux
past the maximum root
depth
Total pesticide uptake
flux from entire soil
profile
Total pesticide decay
flux from entire profile
Units

g cm"?
day"1
g cm"2
day'1
g cm"2
day'1
g cm" 2
day-1

g cm"2
day-1

g cm"2
day'1

g cm'2
day'1

g cm:2
day'1
g cm"2
day'1

g cm"2
day"1

g cm'2
day"1

g cm'2
day'1
Arguments
Required
(IARG)

None

None

None

1-NCOM2
•?•

1-NCOM2


1-NCOM2


1-NCOM2


None

None


None


None


None

                                    158

-------
Table 5.7.  INPUT FORMATS FOR THE PLANT GROWTH (PLT6RN) AND PLANT
            TRANSLOCATION (PLTRNS) MODULES
Variable                        Description                     Units


 ****************  RECORDS  PL1-PL27 *************
                         ARE REPEATED FOR EACH CROP
                      WITHIN EACH HABITAT AND FOR ALL
                                  HABITATS

 RECORD  PL1  —  ABUFR
               FORMAT  (A5)

     ABUFR             The  label  'PLANT' to  indicate that
                      a  new  crop  type definition follows.


 RECORD  PL2  —  PHU
               FORMAT  (24X,  F10.0)

     PHU              The  potential  heat  units required          °C  day
                      for  the crop to mature.


 RECORD  PL3  —  BE
               FORMAT  (24X,  F10.0)

     BE                Crop parameter for  converting         kg ha"1 langley
                      energy to biomass.


 RECORD  PL4  —  LAIMX
               FORMAT  (24X,  F10.0)

     LAIMX             The  maximum leaf area  index for the        decimal
                      crop.


 RECORD  PL5  —  FLAI
               FORMAT  (24X,  F10.0)

     FLAI              The  fraction of the growing season         decimal
                      when leaf area index  starts to  decline.

 RECORD PL6 -  FGK
               FORMAT  (24X,  F10.0)

     FGK              Harvest  index  -  the ratio  of  total          decimal
                      biomass  to  crop yield.
                                 159

-------
Table 5.7.  INPUT FORMATS FOR THE PLANT GROWTH (PLTGRN) AND PLANT
            TRANSLOCATION (PLTRNS) MODULES (continued)
Variable
                   Description
Units
RECORD PL7 — RZ
              FORMAT (24X, F10.0)
    RZ
        Potential root depth.
cm
RECORD PL8 — TBASE
              FORMAT (24X, F10.0)
    TBASE
        Base temperature below which no
        growth occurs.
RECORD PL9 — TOPT
              FORMAT (24X, F10.0)
    TOPT
        Optimum growth temperature.
RECORD PL10 — KHGT
               FORMAT (24X, F10.0)
    KHGT
        Canopy growth rate.
m day
                                                                      -1
RECORD PL11 — HGTMX
               FORMAT (24X, F10.0)
    HGTMX
        Maximum canopy height.
m
RECORD PL12 -


    RW



RECORD PL13 -


    LAMDA
- RW
  FORMAT (24X, F10.0)
        Reflection coefficient for contaminant
        transfer from the soil solution to the
        root.
- LAMDA
  FORMAT (24X, F10.0)
        Degradation rate of contaminant
        within the plant.
decimal
day
                                                      -1
                                 160

-------
Table 5.7.  INPUT FORMATS FOR THE PLANT GROWTH (PLTGRN) AND PLANT
            TRANSLOCATION (PLTRNS) MODULES (continued)
Variable
                   Description
Units
RECORD PL14 - KOW
               FORMAT (24X, F10.0)
    KOW
        Octanol/water partition coefficient
        for the pesticide.
decimal
RECORD PL15 ~ KP
    KP
  FORMAT (24X, F10.0)

        Ratio of the concentration of the
        contaminant in the nonaqueous phase
        to the concentration  in the aqueeefs
        phase within the plant.
on
                                                                   *   -
RECORD PL16


    RHONA
  RHONA
  FORMAT (24X, F10.0)

        Ratio of the mass of the nonaqueous
        aboveground biomass (dry weight) to
        the total aboveground biomass
        (wet weight).
9 9
   -1
RECORD PL17


    XPOT
  XPOT
  FORMAT (24X, F10.0)

        Potential crop biomass.
kg ha
     -1
RECORD PL18 — XYLD
               FORMAT (24X, F10.0)
    XYLD

RECORD PL19 -


    XRWT
        Crop yield.

- XRWT
  FORMAT (24X, F10.0)

        Root biomass.
kg ha
                                                                      -1
kg ha
     -1
RECORD PL20 ~ XRWS
               FORMAT (24X, F10.0)
    XRWS
        Mass of sloughed roots.

                    161
kg ha
                                                                     -1

-------
Table 5.7.  INPUT FORMATS FOR THE PLANT GROWTH (PLTGRN) AND PLANT
            TRANSLOCATION (PLTRNS) MODULES (continued)
Variable
                   Description
Units
RECORD PL21 — XRWL
               FORMAT (24X, F10.0)
    XRWL
        Mass of live roots.
kg ha
                                                                     -1
RECORD PL22 — XDPTH
               FORMAT (24X, F10.0)
    XDPTH
        Depth of roots.
cm
RECORD PL23 — XB
               FORMAT (24X, F10.0)
    XB
        Actual  crop biomass.
 kg  ha
                                                                      -1
RECORD PL24 — XHGT
               FORMAT (24X, F10.0)
    XHGT
        Crop canopy height.
 m
RECORD PL25 -


    FBI


RECORD  PL26  -


    XCR
- FBI
  FORMAT (24X, F10.0)
        Dimensionless expression of
        accummulated thermal time.
  XCR
  FORMAT (24X, F10.0)
        Contaminant concentration within
        plant roots.
 decimal
                                                                 9 9
                                                       -1
 RECORD  PL27  -  XCAG
                FORMAT  (24X,  F10.0)
     XCAG
        Contaminant concentration within the
        aboveground plant biomass.
                                                                 g g
                                                                    -1
                                  162

-------
Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE (APUM)
   Variable          Description                                   Units
Record APUM1—RLABEL
              FORMAT (A80)

    RLABEL           A 1 to 80 character run title for the
                     animal data set.
****************************p£QORDS APUM2-APUM9****************************
                       ARE REPEATED FOR EACH HABITAT
                                (NHAB Times)


Record APUM2—NSCOM(IH)
              FORMAT(I5)

   NSCOM(IH)          Number of soil horizons through which
                     soil animals move in habitat IH (must
                     be less than or equal to the number of
                     TFAT horizons.
Records APUM3— NSUB(IH)
               FORMAT (I 5)

   NSUB(IH)          Number of soil animal groups in habitat IH
           ARE REPEATED FOR EACH SOIL ANIMAL GROUP IN HABITAT IH
                              (NSUB(IH)  Times)
RECORD APUM4— ALABEL(IL)
              FORMAT(A20)

  ALABEL(IL)          A unique 1 to 20 character label  identifying
                     the soil animal  IL.
RECORD APUM5-PDENS,  CO, BCF,  KMET,  LD10,  LD50
              FORMAT(6F10.0)

    PDENS            The population  density (biomass/unit area)    g
                     of the animal group corresponding to the
                     label  ALABEL in Record APUM4.

                                    1G3

-------
Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE
            (APUM) (continued)
   Variable          Description                                   Units
      CO             The initial  concentration of pesticide in    mg mg"1
                     the tissues  of an organism population.

     BCF             The bioconcentration factor for the animal     ratio
                     population in soil.

     KMET            Metabolic degradation rate constant for       days"
                     pesticide in the tissue of the organism
                     population.

     LD10            The 10-percentile lethal dosage of pesticide, mg mg

     LD50            The 50-percentile lethal dosage of pesticide, mg mg

                     Note that lethal dosage information is
                     defined on a weight basis of the affected
                     organism, not per weight of ingested food
                     material.
RECORD APUM6--LD10L, LD10U, LD50L, LD50U
              FORMAT(4F10.0)

     LD10L            Lower confidence bound on the LD10            mg  mg

     LD10U            Upper confidence bound on the LD10            mg  mg

     LD50L            Lower confidence bound on the LD50            mg  mg

     LD50U            Upper confidence bound on the LD50            mg  mg


RECORD APUM7—ICOLD, NMOVE
              FORMAT(2I5)

     ICOLD            Soil animal movement flag.
                       =0, animal moves as a population, with
                           the distribution among soil horizons
                           computed from the soil movement
                           transition matrix

                       <0, the animal group is modeled as a
                           population with a steady state
                           distribution among soil horizons.

                                   164

-------
Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE
            (APUM) (continued)
   Variable          Description                                    Units
                       >0, the animal group moves enmasse by
                           random particle tracking.  ICOLD is
                           then the initial soil horizon
                           location.

     NMOVE           The number of times per day that the          days"
                     animal population is redistributed.


RECORD APUM8—(PMVC(ICO, ICN), I6N = 1, NSCOM(IH)
              FORMAT(SFIO.O)
              Repeat for each habitat ICO.


PMVC(ICO, ICN)       The soil horizon movement transition matrix  fraction
                     for the animal, containing the probability
                     that the animal will move to each horizon
                     ICN if it is in horizon ICO.  Not required
                     for steady-state populations (ICOLD <0).


RECORD APUM9~(PLHAB(IC), 1C = 1, NSCOM(IH))
              FORMAT(SFIO.O)

   PLHAB(IC)          The fraction of the population contained     fraction
                     in each horizon 1C.  Not required if
                     the animal is modeled by enmasse particle
                     tracking (ICOLD >0).


RECORD APUM10--NLEVEL
               FORMAT(I5)

    NLEVEL           The number of higher animal groups (those
                     which prey on other animals and move
                     between habitats).


***************************p^QORDS APUM11-APUM21***************************
                    ARE REPEATED FOR EACH HIGHER ANIMAL
                               (NLEVEL  Times)
                                   165

-------
Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE
            (APUM) (continued)
   Variable
       Description
   Units
RECORD APUMll--ALABEL(IL)
               FORMAT(A20)
  ALABEL(IL)
       A unique 1 to 20 character label
       identifying the higher animal group IL.
RECORD APUM12--MO, CO, UF, KMET, LD10, LD50
               FORMAT(6F10.0
      MO

      CO


      UF


     KMET



     LD10

     LD50
       The total biomass of the animal group.

       The initial concentration of pesticide       mg mg
       in the tissues of the organism group.

       The total daily feeding rate for the
       animal population, per unit body weight,

       The metabolic degradation rate                days
       constant for pesticide in the organism
       population tissues.

       The 10 percentile lethal dosage.             mg mg

       The 50 percentile lethal dosage.             mg mg
        -1
mg mg'^day"1
       -1
        -1

        -1
RECORD APUM13—LD10L, LD10U, LD50L, LD50U
               FORMAT(4F10.0)
     LD10L

     LD10U

     LD50L

     LD50U

RECORD APUM14-


    BETA(l)



   UALPH(l)
       The lower confidence bound on the LD10,

       The upper confidence bound on the LD10.

       The lower confidence bound on the LD50.

       The upper confidence bound on the LD50.
-BETA(l), UALPH(l)
 FORMAT(2F10.0)
       Preference factor for feeding on soil,
       as a fraction of the total feeding
       rate UF on Record APUM12.

       Assimilation efficiency for uptake of
       pesticide from soil.
                        166
   mg mg

   mg mg
-1

-1

-1

-1
   mg mg
 fraction
 fraction

-------
Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE
            (APUM) (continued)
   Variable
Description
 Units
RECORD APUM15--BETA(2), UALPH(2)
               FORMAT(ZFIO.O)
    BETA(2)
   UALPH(2)
Preference factor for feeding on
plants, as a fraction of the total
feeding rate UF.

Assimilation efficiency for uptake of
pesticide from plants.
fraction
fraction
RECORD APUM16-BETA(3), UALPH(3)
               FORMAT(2F10.0)
    BETA(3)
   UALPH(3)
Preference factor for feeding on            fraction
granular pesticide pellets, as a
fraction of the total feeding rate UF.

Assimilation efficiency for uptake of       fraction
pesticide from granules.
RECORD APUM17—BETA(4), UALPH(4)
               FORMAT
    BETA(4)


   UALPH(4)
Daily rate of ingestion of ponded       liter mg"1 day
water, volume per biomass.

Assimilation efficiency for uptake of       fraction
pesticide from ponded water.
                                                                           "*
 RECORD  APUM18~BETA(5),  UALPH(5)
                FORMAT(2F10.0)
    BETA(5)


    UALPH(5)
 Daily air inhalation rate,  volume per   liter mg'l day""*
 unit biomass.

 Assimilation efficiency for uptake of       fraction
 pesticide from air.
 RECORD APUM19—IHOLD, NMOVE
               FORMAT(2I5)
     IHOLD
 Habitat  movement  flag
   =0,  animal moves  as  a  population,

             167

-------
  Table  5.8.   INPUT  FORMATS  FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE
              (APUM)  (continued)
     Variable           Description               .                    Units


                           with the population distribution
                           computed using the habitat
                           transition matrix

                       <0, animal is modeled as a
                           population with a steady-state
                           population distribution.

                       >0, the animal group moves enmasse
                           by random particle tracking.
                           IHOLD is then the initial
                           habitat location number.

     NMOVE            The number of times the animal population    days"1
                     is redistributed per day.


RECORD APUM20—(PMVH(IHO, IHN), IHN = 1, NHAB)
               FORMAT(SFIO.O)

 PMVH(IHO IHN)        The habitat transition matrix,              fraction
                     containing the probability that the
                     animal will move to each habitat IHN
                     if it is in habitat IHO.  Record
                     APUM20 is repeated for each habitat
                     IHO, and is not required for
                     steady-state populations (IHOLD <0)

 RECORD APUM21~(PHAB(IH), IH = 1, NHAB)
               FORMAT(SFIO.O)

   PHAB(IH)          The fraction of the population              fraction
                     initially in each habitat IH.  Not
                     required for animals moving enmasse
                     by random tracking (IHOLD >0).


 *************************END OF HIGHER ANIMAL DATA************************
Record APUM22--NPRED
               FORMAT (15)

     NPRED           The number of predator-prey pairs
                     to be input in the following records.
                                    168

-------
 Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL  EXPOSURE  MODULE
             (APUM) (continued)
    Variable
 Description
  Units
                                   /\puM23-APUM26***************************
                          ARE REPEATED  NPRED TIMES
Record APUM23--PRED
               FORMAT (A20)
     PRED
A 1 to 20 character label identifying
the predator.  PRED must be a
predefined higher animal label (ALABEL
in Record APUM11).
Record APUM24—PREY
               FORMAT (A20)
     PREY
A 1 to 20 character label identifying
the prey.  Prey must correspond to a
predefined animal label (ALABEL in
Records APUM 4 or APUM 11), and may be
either a soil animal or a higher animal
Record APUM 25—BETA, FALPH
                FORMAT (2F10.0)
     BETA
     FALPH
The preference factor for the predator      fraction
PRED feeding on PREY, as a fraction
of the predator's total feeding rate
UF on Record APUM12.

The predator's assimilation efficiency      fraction
for pesticide from the prey.
Record APUM 26—(PEAT(IH), IH = 1, NHAB)
                FORMAT (8F10.0)
   PEAT(IH)
The probability that the predator
captures the prey in each habitat,
given that both animals are in the
same habitat.
fraction
**************************END Qp PROBATION OATA***************************
                                   169

-------
Table 5.8.  INPUT FORMATS FOR THE TERRESTRIAL ANIMAL EXPOSURE MODULE
            (APUM) (continued)
   Variable          Description                                   Units
 Record APUM27—ASTEP
               FORMAT  (A5)

     ASTEP           A character  label  specifying  the
                     time step  for  printing  detailed
                     dosage  breakdowns.
                          =  'DAY',  daily  step
                          =  'MNTH1.  monthly  step
                          =  'YEAR1,  annual step


 Record APUM28--NPLTS
               FORMAT  (15)

     NPLTS           The number of  time series variables
                     to be written  out  for plotting and
                     data manipulation.

*****************************(^^p^^j RECORD A29*****************************
                                 NPLTS TIMES
Record APUM29--TSLBL, TSNAME
               FORMAT (A5, A20)

     TSLBL            The label for the variable to be plotted:    (mg nig"1)
                          = 'CORG', concentration of
                             pesticide in the organism
                          = 'DOSE', cumulative pesticide          (mg mg"1)
                             dosage
                            'LD10', 10 percent Lethal dosage      (mg mg'1)
                            'LD50', 50 percent Lethal dosage      (mg mg'1)
                             LD10L', lower bound on LD10          (mg nig"1)
                             LD10U', upper bound on LD10          (mg mg'1)
                            'LD50L', lower bound on LD50          (mg mg~})
                            'LD50U', upper bound on LD50          (mg mg'1)
= '
    TSNAME           The 1 to 20 character label identifying the
                     animal for which variable TSLBL will be written
                     out (must correspond to a previously defined
                     animal name, Records APUM4 or APUM11)


*****************************£^[) Qp APUM QATA*****************************


                                 170

-------
Table 5.9.  INPUT FORMATS FOR THE MONTE CARLO MODULE (MC)
   Variable
        Description
Units
RECORD MCI —
   TITLE
TITLE
FORMAT (A80)
        A 1 to 80 character descriptive title
        for Monte Carlo Data.
RECORD MC2 —
   NRUN
NRUN
FORMAT (15)
        The number of Monte Carlo runs to be
        performed in the simulation.
**********************£NO Qp SIMULATION CONTROL DATA***********************
RECORD MC3 — PNAME, INDX,  VAR(l), VAR(2), VAR(3), VAR(4), VAR(5)
              FORMAT (A20,  110,  5F10.0)
              Repeat for each Monte Carlo input variable.

   PNAME      The 1 to 20 character name identifying the input
              parameter to be varied.  Currently available
              parameter names are shown  in Table 5.10.

   INDX       For parameters PNAME which are arrays, the array
              index to be varied (i.e. the soil horizon number).
   VAR(l)      The mean value for the input variable
              distribution.
   VAR(2)      The standard  deviation.
   VAR(3)      The minimum value  of the input variable.
   VAR(4)      The maximum value  of the input variable.
   VAR(5)              A flag indicating  the  type of distribu-
                      tion  to be used in generating values for
                      the input  parameter PNAME.   Options are:
                                                     See Table 5.10
                                                     See Table 5.10
                                                     See Table 5.10
                                                     See Table 5.10
                      1
                      2
                      3
                      4
                      5
                      0
            Normal
            Log Normal
            Exponential
            Uniform
            Johnson  SB
            Constant
                                      171

-------
Table 5.9.  INPUT FORMATS FOR THE MONTE CARLO MODULE (MC)  (continued)
   Variable           Description                                      Units
RECORD MC4 — END
              FORMAT (A3)

   END                The Characters 'END', used to mark the
                      end of input of Record MC3.


**********************END OF INPUT DISTRIBUTION DATA**********************


RECORD MC5 — SNAME (1), INDX, SNAME(2), SNAME(3), NDAYS
              FORMAT (A20, 110, 2A20, 110)
              Repeat for each Monte Carlo output variable.

   SNAME (1)          The 1 to 20 character name of the
                      variable to be written out 1n statistical
                      summaries.  Available variable names are
                      shown 1n Table 5.11.

   INDX               For variables SNAME(l) which are arrays,
                      the array Index to be written out.

   SNAME(2)           A flag indicating whether a cumulative
                      frequency plot is to be printed for the
                      output variable (indicated by the
                      characters  'CDF').

   SNAME(3)           A flag indicating that values of the
                      variable are to be written out for
                      each Monte Carlo run (indicated by
                      the characters 'WRITE').

   NDAYS              The number of days in moving average              days
                      periods  (used for statistical output
                      of variables which vary in time).


RECORD MC6 ~ END
              FORMAT (A3)

   END                The characters 'END', used to mark the
                      end of Record MC5.
 ****************************END Qp OUTPUT DATA****************************


                                   172

-------
Table 5.9.  INPUT FORMATS FOR THE MONTE CARLO MODULE (MC)  (concluded)
   Variable
Description
Units
  RECORD MC7    — NAME1,  NAME2,  CORR
              .  FORMAT (2A20,  F10.0)
                Repeat for each pair  of correlated variables.

     NAME1              The name  of the first correlated variable.
                        NAME1  should  correspond to an input PNAME
                        on Record MC3.

     NAME2              The name  of the variable correlated with
                        NAME1.  NAME2 should  correspond to an
                        input  PNAME on  Record MC3.

     CORR               The value of  the correlation coefficient
                        between NAME1 and NAME2.

     RECORD MC8  — END
                   FORMAT  (A3)
     END
  The characters 'END',  used to mark the
  end of Record MC7.
  **************************^Q  gp MONTE  CARLO  DATA***********************
                                     173

-------






















CO
LU
_J
CO
<
1— 1
ae
<
a.
*— i
3
ae
O
UJ
i
x
g
P
LU
a
1—4
0
t—
a
LU
CO

ai
a







0>
5 ja
ae id
gfe '
Lu >









0)
3





i-l CO
1 1
vi E
S °
•o 01











•
a*
4-»
1
fe i?
0. i-
(/>
ai c
o> ai
id -o
•»- .*
•d i—
% J
•^ .r"
o o
CO d





z z
o o
ivl rvi
I-M i— i
ae ae
§ §
a a
< CO





s
Z s
•-I >-
 0)
& %
•° o






•
^—
o c •
v> o> a>
01 VI
c s  .c
at a.
ai .0
•o -o
i- 4J 01
u c >
i- ai i—
4-> i-O
VI O VI
a; i- vi
o. <*-•«-
 .0
>> t- fc.
id 4J O
U L. VI
a> id TJ
a a. id



~
o
INJ
*— i ^^
ae z
s g
LU ae
1— 0
2 £
:*: a
a *i




Li
I-H
O
"*
LU LU
1— Lu
2 S
^ °
5 §
LU l-H
Q 1-
1— 1
1— (-
co ae
LU <
a. Q.
= =

i— i
i i
i 1/1
S
•o






•
•  o.
<4- U-
ai o
u
fli
(V
4-> 4->
o id
z *-
> 5?
*o a>
CO O





a.
g
% ii
ae  CO
aT 5
ae —i




_
Lu
LU
8 5-
_j S
Lu LU
LU Q
06 u-
1— co
O LU
§ 5;
^ £
S 3
CO CU
s =
_. C
"7 °
-.-
01 -U
"B g
O .
w <*-



•
4->
C
(U
.^
u
•*- -a
It- C
a> 3
o o
u s-
D>
o ai
i- >
4-> O
i- .a
+j id
Id C
O. i- •
s- 4-> id
0) C i-
-M 0) J-
«l t Ml
*^ W
s §id
•o OE
C S- -l->
id a> c
4-> -t-» *
o id •—
030.





a.
£• §
S S
o <
1 1
*£ oe





h- =
ae ae
<: LU
^ 5
ae 2
LU
5 i
5 1
0 O
i =S
L3 §
P <



i









VI
§
Ol
o-
id •
VI
o *->
4-> C
id
VI r—
3 O.
o
0) C
3 i-
o-
s§
g.3
•— +J
id
<4- V.
0 •»->
o ai
i- O
4-> C
id o
oe u






x_s
a.
0
oe.
CJ
*~^
Q.
^







O
I
O^
 o o
s_ id i- t-
4^* C. 4^ 4J
c id id
0) • J- J-
(J U1 4-> 4->
C r— C C
o id at a>
u E u u
O t- C C
t- C 0 0
ca id o u





j
£
MH
Z
Lu

• «f—
>> o
-»-> a.
S S1
a. ••-
id 4->
o i—
.^
? *
0) O)
T- .C
U- 4J
4-> it-
id o
4-> 4->
C C
a> ai
j-> +J
c c
8 8
» t
>_ >»
0) (U
J-> -t->
« |
r— r—
"5 "5
<-O CO




f if
M ivi
i-^ »-*
gae
o
•X. 1.
O D.
LU 3
UJ LU
:c ^
I- i—







s
t- F-
•— " 3E
3 S
s °-
U CD
3 P
UJ — J
1— 1 1— t
U. 3
s =
(V
^H
|
I/I
1 -





|


s~
(0
•^
"o
14-

0
14-
4->
id
4->
VI
C
o
u

Ol •
4-> ai
id -o
o
as
U VI
01 ai
a o.







i
3
Q.





s
LU
i
>-
s
LU
O
ae
<
i— i
o
Lu
s
^
' ,-(
_ 1
o 2
w JC
•>s a.
o -*
•
(.
o
o.
id Q)
> "O

Ol *O
^ 43
O 1/1
^ 0)
4-> Q.
VI
S. ^

» (ri
V> *w
£ 21
id
^j *
C 4-> C
(0 *r~ O
+3 C i-
VI 3 4^
C id
0 (-0
o «-
2 0.
id vi a.
i— vi «
id
vi E ^j
~£ rt^
C 4J (-
ai o o
Z 1— M-
•






a.
o_
§
aT
= <
^ i—






=
— LU
^ 5
.< ae
z o
s r
CO U
^ *" '
ae —i
Z CU
LlJ Ct
z <
s =



s
Iw
•o


0
•1—
+J
id
u
•f"
Q
a.
(d
a>
^j

o
•^
j—

c

^
•^

id
^«
*3 VI
•-» (-
Ol U
jc u
1- 0





Q^
o.
»^
o
a.
^
1—4






S
O
•
a.
a.

-------






I/I
•^
c
3















O)
•o
o o
c L
o **
s °-
>""' t.
Itl «J
m Q
r^j
^^
>
£
o.
i— i
3

•a;
*•*
LU
J—

1

u. oi
Z Of 0
LU f--.-
o ot v.
»-< o id
Lu >
O
t—

O
LU
=
_J
LU

I
UJ
^p
Q_ *

^~

«
o •—
f^
• ™*
in

0)
ja






i
i











.
Q.
O
u

1

Id
t-
o
1.
ja
E
C
0)
^
3
O
O









^
Q.
O
oe.
u
<— i
z













s
^E
tj

2
O
_J
<
u_
s






1 1









•
O.
Q. 2
O O
(J -Q
Ol
oi -u
C (A
i- Ol
§ £
os 5
t. u
o o
<4- M-
j- s-
01 01
E E
C C
0, o.
J- i.
3 3
u <_>
I/) (/)
<-> CJ
 CO









x-» <-s
% %
g g
e\T PO
z z
U 0











B
z
o
s
Z h-
U «/>

a. •>

(-> 3C
S 3



C
o

U

s_
tl_





1-
0

u
0
u

^^
01
u
1
<£i
2!
Q.
J"
*i«
_E
"c

V.
J= r-^
^•5
3C «/l












^
=3








a
LU
^^
LU
Of.
LU
U.
UJ
Of.
OL.

^j
O

s



C
0
•r*
U
id
(.
14^





s.
o
(4-
J-
o
o
id
(^
Ol

1
O)
n-
C
Q.
I/I
"id
E
•^*
(0
•
V (/»
J= "C
o> id
•T" ^~
:c ex












z
3








s
LU
O
LU
Of.
LU
U.
LU
Of
Q.

\—

3
0.
s
<— I
1
^

•o

i— 1
1
CT)
E

J-
Ol

•—





c
•1—
T3

tfl
"id
_E
id
s-
^
O)
•r™
•C-
U
!c u
* S
^•^ id
id 2
O) -O
+J C
id o
Of. Q.












z
^3






-
LU
Of.
en
^£
HH
^^
Z
H^
oe
o

a
o
a.
s
r-H
1
^

•o

<— 1
C7>
E

k,
ai

'-




ai
Id
^
c

in
Id
E
id
i.
0)
.?
•*=
u

s
^J
id
01 •
•4-> V.
id i-
of id








^
^
<
Z
1
=3






s
H~*
z
o

K~
^f
^j
^f
^^
^r
t—t

C£
i— i
^
s



C
O
*^*
u
id
i-
n-





s-
o
14-
L.
O
U
id
14-
ai
^
O)
O)
t! wi
a. a)
^1

E Di
i-o
•^
I- U
Ol i-
J= 4J
O) tf>
•f— (U
z o.








,_^
in
<

|
•^






s
LU
U
i
LU
U_
LU
Of
a.

LU

^3
^£
£
CD
s



C
O
•r—
U
id
s-
14—










ai •
j= >>
O) 0)
••- s-
-C CX
ai TJ
jc ai
4-> •«->
14- U

i- C
O -r-
(J Ol
id £
ai s-
c °
ai
J- r—
£ E
a. id





>_
LU
Of
a.
_;
i
»— i
S,
LL.
LU
CO





01 ^r
!o =
> 01
Ol t-
$- ^^
Q. C ^H
O «-H

O — 1 3
LU a.
0) CO <
E 4
id — i s-
z < O




r— 1
1
I/I
(Q
•a











j-
o

ai
id
s-
o •
•r- r—
-*-> id
id E
id c
t. id
|*c

o ai
i- T5

O O
id T*
4-> VI
a> ai
Z a.








^^ „
5J
5
§
H-
UJ
"^f








s
O
CO
UJ
o

o
1— 1
_l
o

*5*
UJ

=

1— 1

>^
(O
T3
rH
5Z*
g

o>
E






cn
id
• rt1
i 	 O
id
E i-
S*£
i- £ "c
o »o
ai vi
0) S r-
« >,^

o>ja id
1- «4- V.
•o o 01
o) c cn
<4- O 1-

r- -!->
id o J-
4-> id o
O I- <4-









^.^
^
s
§
U-







s
LU
i
en
^g
i_^
O
LU
LU
U_

_l

|


• T3
PO C
o ai id
Z J=
4-> cn
•o Q.
i- ai o
o s- s-
O i- 0
01 3
a: o-

O 01
z >-. c vi
>-< Z OJ
< c •—
i- O i-
O >>1- «4-
id id
t- -o u •»->
01 ai f- «
E o "a.

c vi id =
C Q.
C 01 01 <
O E JZ
N i- -4-> T)
T- -0 C
i- .u id
01/13

£ id "~ t
O i- O
in s- 4-> ai
id j=
ai > t- •>->
J= 01
+•* QJ (A C
r- 3 O
3 ^1 QJ im*
O. 2 JZ 3
-j^S-

I- 3 O- «
VI t- >>
30. a?
o a. j:
ai i- a. 4->
5 °ij=
ai u
£5^2
*r" 2
cr u a> c
ai *^ c »i—
(-30
Z OI*V> 01
Si 81
ogSoj
I/I 4->
>» >> 01
J3 J3 r— C
T> TJ id "*~
oi at t- -a
c c i- ai
O O  O)
m vi .a
C C E
i^Ci
i- t- a>
T3 T3 J3 O)
i/i m 3 id
oi ai c

jQ ^3 r" ^^
*0 *O 10 fO
•^ *^* B C
I. J. ^_ ^_
<0 (0 C C
> > cO rd


,,
£
o
= Z
175

-------
























CO
LU
—I
3
t— t
QC
§

g
i—
0
3
|
i
X

LL.
I-H
1-
LU
O
I-H
O
r—
0
CO
CO
«_J
LU
^
X
<
CO
1— 1
LO
(U

id

1/1
4J

— »
^j












_



o.
'^
*••
in
a












ai
< xi
ac «
Sfe
u. >









"a!
XI
id
~*




•—*
^
id
•o









*
s-
0)
4J»
a>

(0
Is
CX

 "O




•
r—
•^
O
c
•^
O>
•0
•r*
u

• 4J
4-> ai
1- Q.
in
C S-
ai o
T3
0)

r~ Id
s «•

*O 0)
CO O



~
o
M
z" Q£
S g
»-H ^^
^C IjJ
O h-
O* ^
CD O




_
(— =
1— 1 LU
CO 1—
LU S
a
3 5
3 LU
co a
_, ,_
O LU
co a.

i
o>

f
u







c •
a) a>
ai io
2 id
4-> f
OJ O.
XI
^3
4-> 0)
C >
 r—
1- 0
U CO
<»- T-

OT3
° 5
C
OT3
T- 0)
4-> XI
T- 1-
4-> 0
J- in
id-o
a. >




•
• in
4-^ 4-^
C C
a> id
"o "o.
iT c
a> ^~
O 0)
(j -Q
•^
C 0
0 -^
i- 4->
4-> I/I
u ai
Q) n
r—
HI *O
0)

o Id
o s-
•^ id"
*0 OJ
co o






oT
^ §
CL. <_>
O '*^
S a
3e" 5
ae — i




LL.
LU
S 5-
_, 5
LL. LU
LU a
ae
I— co
O LU

^ ^
o 5
co a.
s s

1
C7>
rv-i
^e
U

1 *
C
a>

o
•r-
M-
14-
O)
o
(J
C
0

4^
.^
4J
id
a.

ai
id
2

"o

U
o







ST
g

1





h-
<
Q.
ac
LU
I
O

{3
p
c
o
^
u
id
s.


4_>
C
id
^_
CX
•o
c
2
o>
0)

o
X)
id
c
•T«

C
-u
c •

o id
u T.
O Qj
id id
2 E






CL.
g
O
g
o
ac




_
LU
^
3
a
1
oc
0
LU
i



i
i








in
i

sr .

4-> C
id
«/) f—
=> a.
Sc
3 i-
O"

15
id
II- S-
04.
O 0)
1- (J
4-> C
id o








^^
Q.
o
oc
a.
^






s
O
1
cr

g'
o



i



V

•i—
u

•i» 4J QJ
o i/> *o
I/I Ol -r-
a. u
S- f-
0 E 4->
M- I/I in
•T- QJ
s- c a.
o id
4-> D>i—
O J- f-
cd 0 0
14- 
coo
O 4->
i- 0
4-> i- C C
id 4-> o o
s- id T- f-
c « id
ai •> v. s-
(J tO 4J 4->
C r- C C
o id at ai
u E u u
o ••- c c
i- C O O
co id o u






"~?
^
•x.

LL.
CO










_
LL.
O
CO
s

<— l
1
in
^
id
•o






^
o
U)
fd
s-

c •
O I/I
•T— r^
4-> id
id E
^3 •*"
id c
«- id
aT c
•O-r-
0 0)
•^- T3
r— -r"
O U
XI t-
id 4->
4-> in
Q) Q)
x a.





"^
ff_
i— i
§
r—
X
*£





s
a
g
o
LU
a
o
o
S
LU
F
ro ro
i i
o o

ro co
e g
c
>> ^
4-> CX

O Ol
id c
CX *t~
id 4.
o i—
•o '3
(U 01
l|_ ^_>

4-* 4^
id id

4-* 4->
C C
QJ AI
4-> 4^
C C
O O
(J (J
J_ J_
ai a)
4-* 4-*
id id
2 2

"o "o
CO CO




z z
o o
>— i •— i
ae. ae.
o o
£. S.
(_) Q-
U_ 2
LU LU
1— 1—






s
>- =
t; ^
(_) 1— H
< 0
Q- Q.
O O
a i— i
-i l-
LU —I
1— 1 H- 1
Lu 2
s =
176

-------



4->
"c
3














_^
O)

i— f..
4-> Q

^j *j^
^-* c*
oo h
^ "
| o
oe
<
>
i—

i—
o
3
u
LU
1-
i
^.
Li_ 0)
t— < JO
z oe id
LU l-i-
0 Of. S-
•-• o id

o
I-
LU
OO
_l
LU
§

LU
i
^4 -g
LO

O)

id
I—


i— i
•» "*"" "r" '""' *""^ "*""
•^ ^ I I I (J o en O> O
E ^ I- I- ^~
O ^*^ ^^ 14— ^ ^ ^.
O) 0)
•^ "r-
o c
a. o
« at i- »- J-
> -a 4-» • o o »-
i- id a. 4- 14- ajo
ai o o o ji r- ««-
-01- i- •a.s-s. i- c id
1-.U r— Q.OOO O i- J= fc-
O 00 Q. O V. 4-> J-> S- C O
i-ai a. i-o-oo o ~o f- ••->
4->o.ido cuidid o
(^ 01 4^ *( ^^ oo oo id
O) 4- CU 3 C oo i— ^ u_
a. o -c ot-aiai a> «a  (J U E E Ol
$-ld i— 5 $- C CI-T- 0
oai jc id^idai a> c c c
4- J- U 4- O> J= J- t- Id Id QJ
id -i— u u i-
4-> • JZ S- V. S- 4- 4- t- S- O)
C4->C2 oooai a) a) ai 4- •
Idf-O 4- 4- 4- t- I- JC J= O) 00
4->Ci-c a. Q. en en s- a)
(/) 3 ^J ^— ^ t. ^ "P- 1— CLr—
O ^ O ^^ J3 JO <"^ — ^ ^ C
oa)i-id EEe>— i— J= J= i— id
f a. CCCE E i-.i- e en
S oo a. c T- T- JE i- JE f-
i— in id id ai ai ai c c s a; J c ai
id T~ > > > id id 4-* id T^
(ri E ft) r~M • ^ t. V- • 4^ id 4-* T—
J= 3 00 13 13 3 S- S- 00 Id J (d S- U
j-15 3 jci-^jzca)-oa)«J=4J
S+J I. O) O oo oo oo Cfii- cnid 4->C -(-> J- cr> oo
o o jc o cj o o f~O i-r— (do idi— T- d)
x i— >t- t— o oo oo oo i oo no. oca. a£id i o.




x— . 9 LO
^•^ ^L ^^ ^^ ^— ^ •* *
Q. Q. Q. Q. Q. —1 — 1
a. < o o o < <
< z ac ac a: Z Z
Z 	 C_> (_><_> fc_ M
«— • >- • • « z z
CL Q i-H CVJ <*» < <
C^ C^ ^•^** ^^x s^^* *i^»* w^
33 ^C ^C -E 21 21 y 2^ Z Z ZI
i^ (— H-I O C_J O Z3 ^ 3 3 O


_
— l^J ~
LU h- LU
= = 1— < <->
SLUS SLU^GCZ
(—(—>• LU u ae LU
Z — a. z o u- LU z _i Q.
h-*H ^J O_ ^^ ^J 21 O 2C
z a. _i _ioo£*-i 
ss s = = ss s = s s
177

-------















f- s
~v
•o
"u
§
0
to
UJ
_l
%
ac.

^»

^_
Q.
H-
O
O

ac
u
UJ
-^
r-t
z
U.
*— 1
h-
UJ
Q
4—4
0
1—

a
to
to
_i
UJ

^^
i

UJ

i
to

t
T— 1
•
in

Ol

id
t—


i/i
c














c
o
•F—
Q.
°£
O
I/I
a
















at
ae id
I — r-

O id
u_ >















*"-"
QJ
id
~—







c
o
(J
10







t.
O) •
J= >>
010)
£ Q.
a> TJ
f 0)
it- U

O i-
t! 0)
id J=
O) J-
c*t£
QJ
(_ r™
 ^
-8

QJ t
a. c^"
O--I

o — i =3
uj a.
OjCO «£
10 — J ^
z < o
»— i
i
«?
•o
I— 1
1
o>
E
i1





(/)
^
* 40
r~ T3
(O
^ & .
C Q.'— -
id >,
t- J= 'E
O O) O
It- *^
«5I^
10>,1
I- -Oi-
0 C
g>jQ 10
1- It- J-
•o o ai
01 C' Ol
U. O i-

•— 4->
10 CJ $-
2t£










i
i— i
3C
^f
NH>"
u_
^J








•
UJ
^.
(^
QC
C0
1—4
O
Ul
UJ
u.

1
^c
6
1—
s

.
1
10
Ol






,
O)
Ol
id

"o
u-
•g
*
a.
o
i/i
id
E
ai
•o

"o
1—
^_>
(/I
HI
a.









I—

a.
0
u.










—
UJ
a
I-H
y
,}__
^O
UJ
a.

Of
«t
I-H
O
u.
=


1
id
2





t-
-*

•»"• C

^
01
i- 3
O i— CO
i- i-  00 !_
c at o
at t- u
(J i- O)
Sic

-S^x
•t- . Q
(J 00 Z
1- t. 1-4
4-> 0)
CO >> 1_
ai id o
Q. i— It-








^^
Ul
^^
1^
to
UJ
Q.












UJ
a
I-H
t \
»— 4

UJ
a.

i
o




B





.
o

id

4.)
C •
ai
o
u
Ol
•o
•r*
O
•f—
4->
01
a.
u
o>
JZ
o.
I/I

J
<











z

t^J










=
H-
UJ
a.
CJ
^^
Of.
UJ
X
Q.
00
f^\
^
^f
=


t— 1
Ol
01






I/I
^J
Q
o
i-
c
•^
1

I
c
Ol
L>
O
U
ai
•o
•t—
o
•I—
^_l
ai
a.












ac
X















—
u
o
0

1—
8
ac
=


•— i
01
Ol






Ol

Q
^1
id
c
"~
1

10
c
s.
Coo
o+->
°g
•on.
^_
CJC
i-C

wo
a. en











CD
5
X










s
O

o
o

o

o
UJ
^»
o
ca
«c



i-H
1
Ol
0>



•
r^
i
•r-
c
10

C

Ol
•o
u
I/I
8L

o
§
^3
2
+J
C
Q)
y
§
o









I
^^
<£
o
g ^








=
^r
»-H
Z

^
o
u

oo
LU
a.

ci
O
s


1
01





10
10
E

°C
id
J.
O
oj-
ai
If
o
•o
5
T>

I/I
ai
a.

^
^_
"S
o








^^
1
 a.
•o <
i- 01
8t- 13
i- C
QJ 3 10
ce. o-
X <
a z o)
4-4 Z 4->
V- C

i- r- "3
ai 10 a.
J3 C C
§o *^ •
•^ ^
C I/I Ol QJ
C I- J3
c oi 10 E
0 E 3
N i- >> C
i- -O O)
S- JE C
O I/I 4-* O
^^^ 43
^•S.^ 5
°T •£ "^
i/i w % *•"•
10 Q.
ai > c a.
JE 1- «
4-> 0)
r- i. 01
4-> i- Ol J=
3 J= -O 4-»
0. 3 t-
C O 4->
1 •! "3
i- a> o.
0 01 J= C
^i^ -
fcS5 2
VI
3 Q.-O 4-
O 0) Ol
01 i- (- 171
jz o ai 3
4-> XI
H> e  C 4->
T-
30,0, a.
C-51" 3
O" > >l C >>
J3 J3 10 J3
c— i— VI -O
•0 10 a. oi
ceo c
0 0 S- 0
T- -r- tJ T-
l/l I/I (/I
c c c

E E i- E
i- i- Ol i-
^3 "O XI • X?
E VI
i/i vi 3 at vi
ai oi c r— ai
S xi — £ xi
10 10 10 10
i- i- E 10 i-
fc. J. i- 4-> S-
10 10 C 10 10
:> > 10 -o >


• •
UJ
o
z
178

-------
                                  SECTION  6

                            PARAMETER ESTIMATION
6.1  INTRODUCTION

    In this section, guidance for estimating model  parameters is given.  Not
all parameters are mentioned.  Values of those parameters which are simply
switches for options are described in sufficient detail in Section 5.
Similarly, parameters which specify options for output, etc. (i.e., plotting
file variable name designations) are dealt with only in Section 5.  Those
parameters which require special knowledge to estimate are included in this
section.  These include parameters such as dispersion coefficients,
adsorption coefficients, and other rates and constants.  Guidance for some
parameters is unavailable.  In general, the user is given specific guidance,
where it is possible, with references to documents  which may contain
additional information.  Parameter distribution information, as well as
information on mean values, is given, where possible, for use in Monte Carlo
simulation.  The guidance is arranged, according to computational module, in
the following sequence:

       • FSCBG
       • GRDDEP
       • TFAT
       • PLTGRN
       • PLTRNS
       • APUM

6.2  FSCBG PARAMETERS

    The FSCBG model can be used to simulate either  aerial applications or
ground spray events.  In the form installed in TEEAM, FSCBG should be
relatively easy to use.  Many of the parameters have default values and,
unless specific information is available to the contrary, the user is
encouraged to take advantage of these values.  Default values are indicated
in Table 5.2a in Section 5.  Values for parameters  of aerial applications
which the user may want to estimate for his particular application are
covered first.
                                    179

-------
6.2.1  Aerial  Spray Application

    The aerial application module requires  the  following groups of
parameters:

       • Meteorological
       • Source input and
       • Spray distribution

6.2.1.1  Meteorological  Inputs—
    A number of meteorological inputs are required  for  the  FSCBG model.
However, as stated above, only a few need be entered  by the user.  Guidance
available to the authors for estimating these parameters is given below:

    SIGAD and SIGEP standard deviation of the wind  azimuth  angle and
standard deviation of wind elevation angle—These values are set internally
for various atmospheric  stability regions.   Stability regions are derived
from the Richardson number computed in TFAT. They  are  not  required as  user
inputs.

    HM--surface mixing layer height—The mixing layer height is that  height
to which a parcel of air at the surface, after  heating, will continue to
rise until its temperature equals the local atmospheric temperature.  The
extent to which this mixing takes place varies  diurnally and from season to
season.  Typical values  which have been used include  those  found in Table  6-
1 for six different meteorological regimes.

    WSOCAN—wind speed above the canopy—This parameter is  no longer
required by FSCBG as a TEEAM module.  Daily wind speeds are read from the
meteorological file.

    THETA--wind direction--The user should  enter a  typical  value of wind
direction for the location of concern.

    RELHMO--re1ative humidity above the canopy—A typical  value  for  the site
should be used.  It is entered as a percent and should  be  obtainable  from
local meteorological information.

    AIRPRS—barometric pressure—The user should use  a  value typical  for  the
location of the ecosystem.

    AIRTPO—air temperature above the canopy—This  is no  longer  required  as
an input parameter when FSCBG is used as a TEEAM module as it is read daily
from the meteorological  file.
                                     180

-------
TABLE 6-1.  FSCBG PARAMETER  SPECIFICATIONS  FOR  FOUR  METEOROLOGICAL  REGIMES
Wind Speed
Location
Marine



Florida

Time of
Morning
Early
Early
Late
Late
Early
Late
mi/h
1.5
3.0
3.0
6.0
4.5
11.0
m/s
0.67
1.34
1.34
2.68
2.00
5.00
Hm
(m)
115
115
315
415
100
200
(deg)
6.5
6.5
15.0
10.0
5.0
8.0
3
(deg)
2.2 1
2.2
5.0
3.3
1.672
2.67
1
  Dumbauld et al. 1980
  Rafferty and Dumbauld 1980
3 GA and GE refer to SIGAD and SIGEP referred  to  in  this  section.
6.2.1.2  Source Inputs--

    Q--source strength—The source strength Q (grrf ) can be calculated by
knowing the desired application rate liquid density and swath width as
follows:
               Q(g nT1) = 0.935  (DENLIQ)  (SWATH)  (Q(gal ac'1)
(6-1)
in which
    Q(g m  ) is the application rate in g m
    DENLIQ is the density of the formulation (g cm" )
    SWATH is the swath width (m)
    Q is the application rate (gal ac  )
    0.935 is a conversion factor
    If Q is known in Kg/ha or Ib/ac, the Q(g m *) can be calculated by

                   Q (g nT1) = 0.1 (SWATH) QfKgha'1) or              (6-2)

                                    181

-------
                   Q (g nT1)  = 0.112 (SWATH)  Q(lb ac"1)               (6-3)

    HGTCFT—spray release height—The spray release height depends  upon  the
aircraft type and the canopy  height of the crop being sprayed.   Typical
release heights are 1.5 to 6  m for row crops  and 15 to 20 m for forests.

    ARCWGT—aircraft weight—Aircraft weight  varies according to the
model.  Table 6-2 gives aircraft model specific parameters for a number  of
aircraft.

    WNGSPN—aircraft (or rotor diameter)  wingspan--See Table 6-2.

    ARCRSP—aircraft speed—See Table 6-2.

    SIGXYZ--source dimension—This source dimension is typically estimated
from the visible width of the spray cloud behind the aircraft.   Under the
assumption that the spray cloud is Gaussian in shape and that the visible
edge of the cloud represents  the point at which the concentration is  one-
tenth of that of the cloud centroid, the source dimension is found by
dividing the visible width by 4.3.  A typical value for the visible width is
1.5 times the wingspan (Dumbauld et al. 1980). Thus,

                          SIGXYZ = 1.5 WNGSPN/4.3                    (6-4)

    Input values used for a Bell G-3 helicopter were 15.1 to 18.6 m
(Dumbauld et al. 1977).  However, based on equation 6-4,  values would appear
to be much smaller.

    WAKVEL—wake settling velocity—The wake  settling velocity is computed
according to
                                  8g W   (10~J)
                               W =  I  a  \                           (6-5)
                                  *  Pa  b  Va
in which

    Wa  is the weight of the aircraft (ARCWGT) (kg)
    g is gravitational acceleration (9.8 m s~ )
    Pa  is the air density (g cm"3)
    b is the aircraft wingspan or rotor diameter (WNGSPN) (m) and
    Vg  is the aircraft speed (ARCRSP) (m s"1)

    DECAY—coefficient of exponential decay of sprayed material--Unless the
user  has specific  information  concerning the decay of the material during
the spray event, this coefficient should be set to zero.  Unless the
chemical decays rapidly by photolysis, it is doubtful that significant decay
occurs  during the  relatively short residence time of the chemical in the
air.
                                     182

-------
            TABLE 6-2.  AIRCRAFT SPECIFIC MODEL PARAMETERS1
Aircraft
Ayres
52R-600 Thrush
52R Turbo Thrush
52R-1820 Bull Thrush
Air Tractor
AT-400
AT-301
Bell
Jet Ranger III (206B)
G-3
Cessna
Ag Husky
Ag Truck
Bellanca (formerly Eagle)
Eagle DW-1
Me lex (Pezetal)"
M18 Dromander
Miller
WH-12E
.*
Erstrom
280F
F-28F
Schweizer/Hughes
300C
500C
Schweizer/Aq Cat
6-164B
G-164B Turbine
Piper
Brave 400
Weatherly
620
620 TP
Weight
(Kg)
3140
3900
4550
3545
3320
1455
1050
2000
1910
2410
4200
1410
1180
1180
930
1100
3190
3190
2180
2680
2770
Wing span or
Rotor Diameter
(m)
13
13
13
13.5
13.5
10
11.3
12.6
12.6
16.5
17.7
10.5
9.6
9.6
7.7
12.8
12.8
11.6
11.6
12.3
Typical Working
Speed
(ms-1)
50
60
60
64
62
10-25
22^
53
50
44
70
26
35
35
11
11
50
57
55
55
57
Source:  Agricultural Aviation 1984,  except where otherwise indicated
Source:  Dumbauld, Rafferty and Bjorklund 1977

                                 183

-------
6.2.1.3  Spray Distribution--
    The two principal parameters which the user must enter are the drop size
distribution and mass percentage of the material in each drop size
fraction.  A typical distribution is shown in Table 6-3.  Upper and lower
drop size limits (DRPUPR and DRPLWR) are shown in the first two columns.
The fraction of material in each drop size fraction (PCTMAT) is given in the
third column.  The drop size distribution will obviously depend upon the
nozzle size.  Table 6-4 from Akesson and Yates (1976) shows the drop size
range and typical uses for various nozzle sizes.  Table 6-5 from the same
reference gives drop size distribution and cumulative percent by volume for
each size range.
      FABLE 6-3.  TYPICAL FSCBG SPRAY DROP SIZE DISTRIBUTION PARAMETERS
Drop Upper
Limit (m)
DRPUPR
1420
1020
840
742
667
577
514
456
351
302
253
200
169
135
100
51.6
Drop Lower
Limit (m)
DRPLWR
1020
840
742
667
577
514
456
351
302
253
200
169
135
100
51.6
20.0
Fraction of Material
In Drop-Size Category
PCTMAT
0.01
0.02
0.03
0.04
0.10
0.10
0.10
0.20
0.10
0.10
0.10
0.04
0.03
0.02
0.009
0.001
 Source:  Dumbauld et al. 1980
                                     184

-------




IO
UJ
CL.
>-
1—
UJ
—1
ivi
r~l
o
z
5
O£
CL,
10
C/>
^
o
1—4
Q£
<
Lk.
o

UJ
CO
^
a
Ul
a
UJ

o
o
UJ
Q£
a
Z

*
£
rf
2
^_
as
Ul
§
o
Ul
Of
UJ
i
^^
X
1
CL.
<
UJ
z-
s
Ul
NJ
to
CL.
o
Ct
1-^
>-
2
Q.
«/>


^*

i

•»—
VI
am
ai 4->
a ai
01
^3 U. ^"*.
V 4->
4-1 O C
is s
t- ,-< fc.
4-> 01
VI C CL.
UJ t— *~*



CM *-s
ai T3
H* f
id
cc E
ai
IM VI
^ C
(O O
a. u
O f—
o-£.



VI
ai
S1
vi id
0) OS
IM Q)
IM t-
O 3
Z I/I
VI
•5 t
(J O_
a. -g
>> c
I-? id




at if.
a. o
>,
P c
o
I- T-
Q. §
01 f- 4->
N S- I/I
•«- O >>
IO I/I (/»
ai
I?0 $
J- -0 »-
a. c a.
CO « CO
i- i *
O O vi
4-> 4-> .c at
U VI 4-> 4->
a) o)  I. a. t- 01
O c
C If- •— QJ •»-
i- id E -a
H- J- 3 i-
• 0 3 •— (J
VI 4-> O --
C r- •— > 4->
•2S333
4-> 4-1 i- O T3
id c i- f— id
u o 01
f- o id •»-> i-
r- id o
Ot"O TJ f-
o. c c -a
id id id 01 >>
VI r—
r— r— VI 3 i-
O O 4-> S-
v, «- o •• id
o 4-> o> vi E
j- c vi c T-
01 O C Q) (-
< O -r- 01 O.




in
CM
i-t
f*
id
4-»
VI
VI
01




in
CM
i-t
id
4->
VI
I/V
0)
r—



CM~~
C
J3
*"
ccS
gg"
•a -a o
4->
mm
o *-« o
O CM CM
co o^



VI I/I
oi i.
•— at
N N
N ••—
l/l O E
•— c o
O 4-»
vi c id
o id
J- <*- >>
at i-
«E3
8*2
IScT,
3SS

c
f- 4-> -c in
id 4-1 s. c
. T- O Q>
in - 3 <»- 01
t— r- O
Id O VI r- .C
O k. •— 3 +->
t- 4-> id T) <•- id
E c OT- ai a.
Q> o -^ a. v,
-c u E « a *->
O 01 «- (J
v- ^ o ai
01 O U "O I/I I/I
•0 4-> C r- C
T- o f- id id T-
<_> Q) O
T- > >, ..^
4-> VI 4-> C Id
vi id 0) i- o fc.
tt) Q) Ol <-> i- 3
a. j- « t- jj 4->
Id VI X Id r—
4-» 1 O O "O 3
VI 01 T3 4-> Id (J
01 Ol 1- -r-
i- i- 3 3 Ol (.
o id o o 01 01
U. r- i— r— T3 Id






O
CO
0
4->
§






s
ro
o
4J
o
o
t— 1



^_^
fsi
C
•r-
-^
JQ
CCO
ggs
"o
in in -n
o ^r
O 10 in
CO 0 — '



I/I VI
at t-
•— ai
N IM
N •»"
O E
C O
4->
vg"
^•"J?
t- -a id
a. e 4->
vi a o
g^^o
i- 0 C
U.O id


•— Q)
id 01
i- id
3i5
•— >
3 O
0 (J
T—
J- TJ
OIO
id 0
>,w
4-> 01
TJ a>
•^^: >»
Ssi5
4-> VI in
1 r- VI
3 id OJ
££8
^|C
f— .C V,
< 0-^






o
a» »
0
4-»
O
h*.






8
^r
o
4->
O
8




cg^
c
A
c """
go
LO
•°o
IO 4->
«•
1 O
«o ro
Q^ 	



I/I VI
ai s-
i— O)
N N
N •«-
0 E
C O
V> 4->
>> c id
id id
$L^$
•"•25
I«S
•5 c -o
« o c
Z<_> id

•a
ai -c
4-> Ol
0 3
zs
4J O
VI J=
at 4->
t. *>
C Q
"4- a) c
°-iv,
VI ••-
 S-
4J id a)
VI U > r-
o) i- o id
O.H- O i-
•^ 4-1
O VI 4-» C
i- in c ai
x id id vi
O •— i— VI
t— o a. a>






CO
en
0
4->
in
CO



-C 0
+^ ^*
•^
3 a.
8v,
vo ai
^
O T-
4-> 4->
O "5 O
O T3O
T id CM




CM
C
^
J3
Is
•°o
_x



CO
ai
N
N
O V>
C 01
VI >
>> c •*•
id « 4->
t- >t- i-
a. T3
-•g^
S10*,
l. OJ a
act.
o o a.
u o vi




VI 1 Ol
in >> c
id x T-
•— o c 3 l-
o c f- o «
i oi .c v- a>
T3 J= 4-> Ol C
QJ 0.1-
4-> 3 VI C
u vi id o
i- id • -r- vi
i- t/i -C 4-> a.
+-> .c i- o id o
in u 01 3 u i-
a> 3 ^: in o u
1_ « 4-> r—
O VI HI
• in c -o <—
o o> -o o c ja
t- T3 C -r- Id T-
X t- Id 4-> 4->
o u id c a.
4-> T- in 4-> o ai
JD -a T- in cj
•— s- n- g id in
r— O) O-^- O) 3
«t x: id i— in in






CO
at
0
4->
in
Ol



4-> 0
__ .%
3*^
^ 9"
O 3
Sv,
r-H QJ
>
0-r-
4-1 4->
o -oS
o -a o
co lain



.c
Q.
4^i"E^
*B-S5
3 O J- -Q
O ID O) •—
•a >
coo
co id in
O -C 4->
JJ BS O
O 4J
4-> (/) -*
in u o
v oi id ro
a •— js — •


>>
ff
E S-
3 a.
E vi

c -o
t- C
E id
jz in
4-> ai
•»- i— V,
3 NO)
N >
in o •*-
>>*-> C 4->
id •*- i-
»--^ 4->"0
a. i- o) -o
co -a T id

o c

4J O
•/J
4-* Id
in u- u a)
T> O O) in
t- •>-»
O) U id JD en
r— Id O> 3 C
4J >, id °3
id x • o
•— o 01 in t-
o C -C CL en a.
> O) 4-> O O
c .c s_ •*- »-
o ex c u o u
• a> in ti-
•o in in r— c o
a> a> s- -O o
4-> -o 01 i- i- ai
o f- jz 4-> 4-> a.
i- u 4-> a. id >>
t- f- O 01 4-> 4J
4-* -Q O *^
in J- T3 Vt E "O
ID QJ C U -r- C
cc. .c id in F— a





2
i
0
o
CM





^
*-«
o
4->
o
§





w
u»
0)

^-"* JZ^*
--i
£S £
e+J
f- (*
w v>
o id i.
Z 4-> id


in
ai
E "N
g^l
0

>c
i ,- o
7^ r=
O C
j: j- a)
^J ^J r^
* I!
in 3
>) 4-> 4->
« i«- t
«- *- 3
Q. S- O
CO T3 — 1






4->
JC
CTI •
•»- I/I
Ol -k->
J= CD
i->
4->
SO9
a
if-
o
0 4->
• *— 1
§«•
1- 0
Ol Ol
t- -o i.
4-> CO
V> 3 «t-
l_
i- T3 Ol
« 0) I/I
in o
01 US -C
f HI 4J
*" *0) J=
5 »• t
T- r— 3
3 a o
i- «-
&- u i
0 £ g
- id O
c 1
3 01
S — u
? £ I1
U
M *r~ r^
18 « "o
T3 V. **f-
o> 01 o
4-> t-
10 • Ol U
O I/I t- f~
•2 o. 3 a:
•O O 4J
c S a i.
i- T3 I- O
01  3
10 f- O> 4^
i- Ol C Ol
U Ol
J- T3 • If.
•*— r— ^Z
« 3 a. o
^ 2 E S
C 7 **>
O U)
irt +^
§^ 5 "
5 °«n£
**— •» 01
vi in >> >
O >> 4-> Ol
Q. Si 1- •—
» O. O 01
Irt I/I r— 3
oi oi-o
•— 01 > -^
N VI VI
g 5 1 t
C f-
J- S 4J
. 0> «-
0 *»••«-
0 S "H ^
E -C 2 *
O 4-* C VI
4-> i- 5 -M
t/» i o v»
>., '^ d
IO "O 4-*
, E "- Z
g S 8 ""
Q. 0) O C
i/> a »-i •"- 1
^< e\j 1*1 w



































S
2
VI
01
4->
a
>-
1
a
o
VI
I/I
01
Jf
<
• •
Ol
u
L.
3
S
185

-------
TABLE 6-5.  DROP SIZE DISTRIBUTION OF AEROSOLS AND SPRAYS,  CUMULATIVE PERCENT

                                  BY VOLUME1
Drop Size
(niiA
M"1/
1-5
5-10
11
10-15
15-20
20-40
40-60
60-80
86
80-100
100-120
120-140
130
140-180
180-200
200-220
220-240
240-260
260-280
280-300
278
300-350
350-400
400-450
460
450-500
500-600
600-700
700-800
900
800-1000
Fine Coarse Fine Medium
Aerosols Aerosols Sprays Sprays

5 0.1
45 0.4 0.1
50
77 2.0
97 2.0
100 12.0 0.1
35.0 5.0 2.0

50.0
59.0 15.8 6.0


50.0
100.0
81.0 17.0




100.00 46.0
50.0




92.0




100.0
Coarse
Sprays







0.01


0.1

0.4


3.0

7.0


14.0

24.0
36.0
46.0
50.0
55.0
74.0
88.0
96.0

100.0
Very Coarse
Sprays







0.001




0.1




5.0









25.0


50.0
100.0
  The volume median diameter (underscored) is that size of drop which divides
  the total  volume of drops found exactly in half; that is, 50 percent of the
  volume is  in drops above that size and 50 percent below.

Source:  Akesson and Yates (1976).
                                     186

-------
    ISW (2)—evaporation model  option—The user is encouraged to use
ISW (2) = 0, with which no evaporation will be assumed.   If the use of the
evaporation calculations are essential the user is encouraged to use ISW (2)
=1.  In this mode, the model  performs drop evaporation  computations.  The
user should avoid option ISW (2)  = 2 unless he has very  specific data about
drop evaporation in the spray over time.

6.2.2  Ground Spray Applications

    The FSCBG (Version  1.0) computer program is not designed for direct
modeling of ground vehicle spray releases.  However, the program can be used
to model some types of  ground vehicle spray releases if  care is used in
selecting the model inputs.  The FSCBG (Version 1.0) program assumes spray
material is released along a line from a moving vehicle  and after a short
period of time material falls to the surface subject due to meteorological
and gravitational forces.  During ground vehicle spray operations spray
material is often directed lateral to the vehicle movement or is sprayed to
the rear of the vehicle.  The FSCBG (Version 1.0) program does not account
for the effects on drop material due to the forced spray trajectory.
Therefore, to model these applications, parameters associated with the spray
must be modified to account for the displacement of the  drops, the distance
over which they are subject to evaporation, and the distance over which they
are subject to meteorological forces.  Further, the FSCBG (Version 1.0) code
cannot model direct impaction of material due to a forced spray.  FSCBG
(Version 1.0) input parameters that require special attention for ground
vehicle spray releases  are given below for two general modes of
application.  Obviously, for these applications, the  aircraft wake model
should be turned off.

6.2.2.1  Spray Perpendicular to Vehicle Movement—
    The first mode of application assumes nozzles are mounted on a spray
boom perpendicular to the movement of the vehicle and spray material is
directed to the back of the vehicle in the plane of travel.  If the spray
nozzles are directed down, FSCBG (Version 1.0) should not be used, because
the program cannot account for impaction due to the forced spray.

    NSOURC--This parameter specifies the number of spray lines in the
direction of vehicle travel.  Ideally, the number of lines is given by the
number of nozzles mounted on the boom.  However, if there is sufficient
overlap of the spray from adjacent nozzles, a single line can be used.
Sufficient overlap requires that the width (diameter) of the spray plume is
roughly twice the distance between nozzles at or before the spray material
reaches the apex of the spray trajectory.

    Q—This parameter specifies the rate of application of the spray mixture
in grams per meter of travel.  If each nozzle is being modeled as a single
                                     187

-------
line, Q is the amount of material from a single nozzle.  If nozzles are
grouped into a single line, Q is the sum of the material from the number of
nozzles used.

    SIGXYZ—This parameter specifies the standard deviation of the spray
material distribution.  If each nozzle is being modeled as a single line,
this quantity is approximated by the width (diameter) of the spray plume at
the apex of the spray trajectory divided by 4.3  However, if nozzles are
grouped into a single line, the width is the distance from the outer edge of
the first nozzle plume to the outer edge of the last nozzle plume.

    XLRZ--This parameter is greater than zero and is approximated by the
horizontal distance to the value specified by SIGXYZ above.

    DRPUPR, DRPLWR—These parameters give the upper and lower statistical
bounds of the drop diameter for each drop size category.  These values are
obtained from the manufacturer of the spray nozzle, known distributions, or
published studies such as Measurement of Drop Size Frequency from Nozzles
Used for Aerial Applications of Pesticides in Forests  (USDA Forest Service
October 1984).  However, information obtained from aircraft nozzle studies
should be used with caution because of the great differences between ground
vehicle and aircraft spray, such as the much greater speed of the
aircraft.  Values must be entered in descending order of size.  Multiple
categories of the same size, gaps between size categories or a single size
category will generally produce erroneous results.  The minimum drop
diameter that can be entered is 5 micrometers.

    HGTCFT--This parameter specifies the effective release height of the
spray material.  The FSCBG (Version 1.0) program does  not contain the
physics required to directly model the forced spray trajectory.  Because
evaporation, some dispersion, and trajectory modification due to wind speed
and direction occurs prior to the apex of the spray trajectory, the specific
height to use is ill-defined.  The height to use is somewhere between the
distance along the trajectory from the apex to the surface and the entire
length of the trajectory.  Generally, the height would  be that distance  over
which meteorological forces dominate the spray plume.

    DX, DY--These parameters specify the start and end  point of each spray
line.  If a  single line is being modeled, use the center of the vehicle.   If
each nozzle represents a source, use the relative location of each nozzle.

6.2.2.2  Spray Parallel to Vehicle Movement--
    The second mode of application assumes nozzles are mounted on  a spray
boom parallel to the movement of the vehicle and spray material  is directed
perpendicular to the vehicle plane of travel.  If spray nozzles are directed
down, FSCBG  (Version 1.0)  should not be used, because  the program  cannot
account for  impaction due  to the forced spray.

                                     188

-------
    NSOURC—This parameter specifies the number of spray lines in the
direction of vehicle travel.  This parameter in this case would generally be
one unless multiple passes were made by the vehicle.  This assumes that each
drop size category travels the same distance along the spray trajectory.  In
actual practice, however, the larger drops will travel further due to
momentum effects.  If the distance of travel for specific groups of drop
size categories is known, the problem could be broken into multiple FSCBG
program runs, one for each group of categories and the results manually
summed over the number of runs used.

 .••  Q—This parameter specifies the rate of application of the spray mixture
in grams per meter of travel.  This value would be the sum of the quantity
sprayed by all nozzles.

    SIGXYZ—This parameter specifies the standard deviation of the spray
material distribution.  This quantity is approximated by the width
(diameter) of the spray plume from a single nozzle at the apex of the spray
trajectory divided by 4.3.  If multiple runs are made over groups of drop
size categories, the value used is determined for the specific drop size
categories.

    XLRZ—This parameter is greater than zero and approximated by the
horizontal distance to the value specified by SIGXYZ above.

    DRPUPR, DRPLWR--These parameters give the upper and lower statistical
bounds of the drop diameter for each drop size category.  These values are
obtained from the manufacturer of the spray nozzle, known distributions, or
published studies such as Measurement of Drop Size Frequency from Nozzles
Used for Aerial Applications of Pesticides in Forests, USDA Forest Service,
October 1984.  However, information obtained from aircraft nozzle studies
should be used with caution because of the great differences between ground
vehicle and aircraft spray, such as the much greater speed of the
aircraft.  Values must be entered in descending order of size.  Multiple
categories of the same size, gaps between size categories or a single size
category will generally produce erroneous results.  If multiple program runs
are being made over groups of drop size categories, the sum of the mass
fractions (PCTMAT) must equal one over all runs.  The minimum drop diameter
is 5 micrometers.
    HGTCFT--This parameter specifies the effective release height of the
spray material.  The FSCBG (Version 1.0) program does not contain the
physics required to directly model the forced spray trajectory.  Because
evaporation, some dispersion and trajectory modification due to wind speed
and direction occurs prior to the apex of the spray trajectory, the specific
height to use is ill-defnned.  The height to use is somewhere between the
distance along the trajectory from the apex to the surface and the entire

                                     189

-------
length of the trajectory.  Generally, the height would be that distance over
which meteorological forces dominate the spray plume.

    DX, DY—These parameters specify the start and end points of each spray
line.  The spray material is sprayed out along a line parallel to the
movement of the vehicle.  In this case, use the best estimate, based on the
force and angle of the spray and meteorological conditions.

6.3  GRDDEP PARAMETERS

    The GRDDEP module provides a linkage between the FSCBG model, whose
output is defined on a two-dimensional grid, to TFAT habitats, which cannot
make use of spatially distributed information.  The function of GRDDEP is to
compute average deposition values and map spray deposition information into
TFAT habitats.  It requires inputs which map TFAT habitats onto the FSCBG
grid.  Habitats may only be quadrilateral in shape.  The important
parameters for the user's consideration are described below.

6.3.1  XSW, YSw, XNE, YNE—Habitat Coordinates

    The user enters the southwest (x,y) coordinates and northeast (x,y)
coordinates of each habitat.  Note that the FSCBG grid is oriented north-
south, although the swath lines may be oriented at any angle.  The x
coordinates must increase from west to east and the y coordinates must
increase from south to north.  GRDDEP uses these coordinates to compute
habitat area for deposition averaging calculations.  The user should
exercise caution that these areas match the habitat area entered in the TFAT
input stream.  (The program does an internal check to insure that these
areas match within ±10%.  The habitat area is  left in the TFAT input stream
in the case that FSCBG and GRDDEP are not used.)  GRDDEP computes this area
by finding the FSCBG grid points which lie within the specified habitat
quadrilateral, and then summing the areas assigned to each grid point.  As
these areas are computed by multiplying the X  axis increment by the Y axis
increment, the user should locate XSW, YSW, XNE and YNE at the centroid of
an area defined by FSCBG grid points.  By doing this, the area defined by
the coordinates, the area computed by GRDDEP and the habitat area read for
TFAT habitats will be equal.

    In GRDDEP, the user also specifies the number of spray applications and
how the deposited spray will be distributed between soil  and canopy, and
within the canopy.  (The simulation results of distribution within the
canopy are not used by TEEAM, however, the parameters to  simulate this
effect must be entered.)  The user is directed to the description of these
parameters in Section 6.4.1.3.
                                    190

-------
6.4  TFAT PARAMETERS

    With a few exceptions, the TFAT module has the same input formats and
parameter requirements as PRZM (Carsel et al. 1984).  This section contains
guidance for the original PRZM parameters (6.4.1) as well as additional
parameters required by TFAT.  New groups of parameters include:

       • Infiltration and ponding (6.4.2)
       • Pond chemistry (6.4.3)
       • Volatilization (6.4.4)
       • Granular Pesticide Fate (6.4.5)

    Note that some of the original PRZM parameters (e.g., SFAC, PFAC, ANETD,
    etc.) are read in the TEEAM Execution Supervisor input file.  These
    parameters are discussed here however.

6.4.1  Original PRZM Parameters

    TFAT relates pesticide fate in the upper soil to temporal variations in
hydrologic, agronomic, and pesticide chemical factors.  A minimum of
generally accessible input is required for successful use of TFAT.
The module does utilize some parameters, however, that users may find
difficult to obtain or calculate.  The following section describes these
parameters and provides detailed procedures for estimating or obtaining the
required values.  Parameters appear in the same general order that they
appear in the input file.  Options are available in the program to directly
estimate several parameters (THEFC, THEWP, BD, and KD) when related
information is supplied by the user.

6.4.1.2  Hydrology Parameters--

    NPTIME—number of time steps for ponding computations—A value of 24
hours is recommended.

    SFAC and PFAC--snow factor and pan factor—When the mean air temperature
falls below 0.0 °C, any precipitation that falls is considered to be in the
form of snow.  When the mean air temperature is above 0.0 °C,  however,  the
snow accumulation is decreased by a snowmelt factor, SFAC.

    The mean air temperature is read from the meteorological  file and
provides a value for (T).  The snowmelt factor, SFAC, for site-specific
analyses can be obtained from Linsley et al. (1975).  The mid-range of  their
values is 0.46 cm day  .  The calculated snowmelt is used to estimate the
antecedent moisture condition and subsequently the runoff caused by the
snowmelt.  The snow factor would be applicable only to those areas in which
temperatures are conducive to snowfall and accumulation.


                                    191

-------
    The pan factor (PFAC)  is  a dimension!ess  number  used  to  convert  daily
pan evaporation to daily potential  ET.   The pan  factor  generally  ranges
between (0.60-0.80).   Figure  6.1  illustrates  typical  pan  factors  in  specific
regions of the United States.

    ANETD—soil evaporation moisture loss during fallow,  dormant  periods—
The soil water balance model  considers  both soil evaporation and  plant
transpiration losses  and updates  the depth of root extraction.  The  total ET
demand is subtracted  sequentially in a  linearly  weighted  manner from each
layer until a minimum moisture level (wilting point)  is reached within each
layer.  Evaporation is initially  assumed to occur in the  top 10 cm of the
soil profile with the remaining demand, crop  transpiration,  occurring from
compartments below the 10-cm  zone and down to the maximum depth of
rooting.  These assumptions allow simulation  of  reduced levels  of ET during
fallow, dormant periods and increased levels  during  active plant  growth.
Values for (ANETD) used to estimate soil evaporation losses  are provided in
Figure 6.2.

    The values for ANETD in Figure 6.2  are only  applicable for  soil
hydraulics option 1,  the free drainage  model, and would not  be  appropriate
for use with hydraulics option 2, the limited drainage  model.  The limited
drainage model allows more available soil water  and, hence,  more  ET
extraction.  If drainage option 2 is selected, it is recommended  that ANETD
be set to equal 10 cm.  Calibration may be required  if  results  are not
consistent with local water balance data.

    DT—average day time hours for a day in each month—The  values of DT are
used to calculate total potential ET using Hamon's  Formula if daily  pan
evaporation data do not exist. Values  of DT  for latitudes 24 - 50°  north of
the equator are provided in Table 6-6.

       Values for DT are determined by:


           Step 1.  Finding the approximate degree latitude  north  of  the
                   equator for the  agricultural  use  site  under
                   consideration.

           Step 2.   Inputting  the 12 monthly  numbers  under the  degree
                    latitude  column  into the  parameter  file  (e.g., 42° north
                    latitude).

                   9.4, 10.4, 11.7, 13.1, 14.3, 14.9,  14.6,  14.0, 12.3,
                    10.9,  9.7, 9.0
                                     192

-------
ro
O)
S-
S-
o
ro
O)
o
S-
S-
o
o
 O
 O)
 t-
 s_
 o
 o
 (O
 s-
 o
 Q.
 03
 >
 01

 c
 ro
Q-
 QJ
 S-

-------
                            E    E     E     E
 o
to
i—
 I
o
                                   o
 I
m
 I
o
                                               m
                                               K)
                                                 I
                                               o
                                              I
                                                        CO
                                                        o
                                                        to
                                                        S-
                                                        o
                                                        Q_
                                                        (O
                                                        >
                                                        O)
                                                        o
                                                        in
                                                        CD
                                                        CO
                                                        o>
                                                        o
                                                        ti-
                                                        CD
                                                        TCi
                                                        O)

                                                        3
                                                        CD
194

-------







o
CM
OO
UJ
a
£
5
s§
s^
§UJ
2*
^z
~£
gi
<=o
,0
-£ in
gP
«t
•
*o
UJ
_l
g-















J-
0
-t->
id
o-
UJ
«t-
o
JZ
+J
8

ai
-O
c
l/>
O)
•o
3
4->
•»-»
*











"

£

*»•

«a-
§


CO

CO

CO

CO

8

CM

CM
*!•
CM
JC
4->
i
«•
CO
r^
oo
Cn
oo
CM
en
TT
en
O
O
fv
en
a\
o\
o
o
CM
o
co
o
ir>
o
vo
o
r*.
•
0
i— i
•
 >-H
o r^
O t-l
CM r^
O rH
co r»»
O ^H
>* r~-
O <-i
lf> OO
0 ^-1
(O 00
O rH
P^ OO
O .-H
00 00
O «-H
w o>
0 T-l
o -*
T-H CO
rH O>
i— 1 I-H
»-H O»
»-l »-H
CSJ C7>
• •
T-l «— 1
T-l I-H
£ C
a> id
u. x
UD
CO
^1-
CO
CO
CO
00
CM
T-<
CO
t— *
CO
o
CO
a>
CM
00
CM
ao
CM
r-».
CM
rx
CM
VO
CM
(0
•
CM
T-l
•
a.
 tr>
at u>
^- lf>
r^ «•
*T l«
O CM
in in
co en
«• ^3-
CM f^
^r «•
T-I in
«• ^r
a\ v
CO ^~
1^ CM
CO ^~
r~ O
co «•
tr> en
CO CO
^r oo
CO CO
r-. *o
CO CO
T-l ^~
• •
co co
i-H <— 1
b S
•d 3
Z >-3
ur>
tn
CO
in
O
in
o
in
vo
*r
^~
»*
CO
^~
I-H
*r
o
«•
en
CO
1^.
CO
in
CO
«J-
CO
CO
•
CO
I— 1
>>
r—
•-»
vo
v
«•
^>
CO
«r
CM
KT
O
T
0V
CO
00
CO
IV.
CO
VO
CO
in
CO
in
CO
^~
CO
CO
CO
Pv
•
CO
i-H
0>
3.
CO
CM
CO
CM
CO
CM
CO
CM
CO
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CM
CSJ
i-H
CM
i-H
CM
I-H
•
CM
T— 1
•
•M
a.
»
in en
o oo
VO VO
o en
r-. co
o en
oo in
o en
en r*.
o en
0 00
i-* en
0 O
T-H O
T-l T-l
T-l O
CM in
«-« o
CM «•
T-l 0
CM in
i-H O
CM VO
T-» O
«T f~
T-l O
*r oo
• •
T-H O
i-H i-H
» •
** >
8 £
o
00
CO
CO
in
co
00
oo
o
en
CM
en
^~
en
VO
en
CO
en
o
o
i-H
O
CO •
O CO
en
i-H
^- •
d Q.
•1
0)
m TD
o 5
T-H ».
<_>
• •
O)
u
• fc.
U 3
0) O
O (/)
195

-------
    USLEK,  USLELS.  USLEP,  USLEC soil  erosion—Universal  Soil  Loss
Equation—The role  of erosion on pesticide  loss  decreases  with  decreasing
chemical affinity for soil.   The total  mass of pesticide loss by this  means
for most highly soluble pesticides will  be  quite small.   If  the apparent
distribution coefficient is  less or equal to 5.0,  erosion  can be neglected
(i.e., the  erosion  flag ERFLAG can be set to zero).   For a compound  having a
distribution coefficient greater than 5.0,  erosion losses  (and  subsequent
pesticide loss) should be  estimated and the erosion  flag set (to one)
accordingly.

    Soil characteristics,  climatic conditions, agronomic practices  and
topography  contribute to the potential  erosion rate  from a field.   During an
erosion-producing runoff event, soil  particles and aggregates are  carried
across the  field.  These aggregates consist of coarse,  medium,  and  fine
particles,  with the fine particles (sediment) carried the  greatest  distances
across the  field.  Sediment  is the principal carrier of sorbed  pesticides.
The Universal Soil  Loss Equation (USLE) developed by USDA  is a  simple  method
used to determine erosion  losses. The USLE  is most accurate  for long-term
average erosion losses.

    The soil loss equation used in TFAT uses the modification described by
Williams (1975).  The Williams modification replaces the R (rainfall
erosivity)  term with an energy term.   The energy term enables the  estimation
of event totals for erosion  from the field. The  modified universal  soil loss
equation (MUSLE) requires  the remaining four USLE factors  with  no
modifications.

    TR—storm duration—This parameter is no longer required.

    USLEK--SOJ1 erodibility  factor--USLEK  is a  soil  specific parameter.
Specific values for various  soils are obtainable from local  Soil
Conservation Service (SCS) offices.  Approximate values  (based  on broad
ranges of soil properties) can be estimated from Table  6-7.

    USLELS--s1ope length and steepness  factor--USLELS is a topographic
parameter and is dimensionless.  Values for LS can be estimated from Table
6-8.

    USLEP—supporting practice factor—USLEP is  a conservation  supporting
practice parameter  and is  dimensionless.  Values range  from  0.10  (extensive
practices)  to 1.0 (no supporting practice).  Specific values for P  can be
estimated from Table 6-9.

    USLEC—cover and management factor—USLEC is a management parameter and
is dimensionless.  Values  range from 0.001  (well managed)  to 1.0  (fallow or
tilled condition).   One value for each of the three  growing  periods  (fallow,


                                    196

-------
          TABLE  6-7.   INDICATIONS OF THE GENERAL MAGNITUDE OF THE
                     SOIL/ERODIBILITY FACTOR, Ka
                                              Organic Matter Content
    Texture Class                     
-------
cropping, residue)  are required.  Specific  local values  can be computed from
Carsel  et al.  (1984)  or obtained  from the  local SCS office.  Generalized
values  are provided in Table  6-10.

     CINTCP--maximum crop interception—The crop interception parameter
(CINTCP) estimates  the amount of  rainfall  that is  intercepted by a fully
developed plant  canopy and retained on the plant surface, cms.   A range of
0.1  - 0.3 cm for a  dense crop canopy is  reported USDA  (1980).   Values for
several  major  crops are provided  in Table  6-11.
          TABLE 6-8.  VALUES OF THE EROSION EQUATION'S TOPOGRAPHIC FACTOR,  LS,
                FOR  SPECIFIED COMBINATIONS OF SLOPE LENGTH AND  STEEPNESS*
                                     Slope Length  (feet)
Slope    25
               50
               75
               100    150    200    300    400   500   600    800   1000
0.5
1
2
0.07
0.09
0.13
0.08
0.10
0.16
0.09
0.12
0.19
0.10
0.13
0.20
0.11
0.15
0.23
0.12
0.16
0.25
0.14
0.18
0.28
0.15
0.20
0.30
0.16 0.17
0.21 0.22
0.33 0.34
0.19
0.24
0.38
0.20
0.26
0.40
3
4
5

6
8
10

12
14
16

18
20
25

30
40
50

60
 0.19   0.23
 0.23   0.30
 0.27   0.38

 0.34   0.48
 0.50   0.70
 0.69   0.97

 0.90   1.3
 1.2   1.6
 1.4   2.0
 1.7
 2.0
 3.0

 4.0
 6.3
 8.9
 2.4
 2.9
 4.2

 5.6
 9.0
13.0
                     0.26
                     0.36
                     0.46

                     0.58
                     0.86
                     1.2

                     1.6
                     2.0
                     2.5

                     3.0
                     3.5
                     5.1

                     6.9
                    11.0
                    15.0
 0.29
 0.40
 0.54

 0.67
 0.99
 1.4

 1.8
 2.3
 2.8

 3.4
 4.1
 5.9

 8.0
13.0
18.0
 0.33
 0.47
 0.66

 0.82
 1.2
 1.7

 2.2
 2.8
 3.5

 4.2
 5.0
 7.2

 9.7
16.0
22.0
 0.35
 0.53
 0.76

 0.95
 1.4
 1.9

 2.6
 3.3
 4.0

 4.9
 5.8
 8.3

11.0
18.0
25.0
 0.40
 0.62
 0.93

 1.2
 1.7
 2.4

 3.1
 4.0
 4.9

 6.0
 7.0
10.0

14.0
22.0
31.0
 0.44
 0.70
 1.1

 1.4
 2.0
 2.7

 3.6
 4.6
 5.7

 6.9
 8.2
12.0

16.0
25.0
 0.47
 0.76
 1.2

 1.5
 2.2
 3.1

 4.0
 5.1
 6.4

 7.7
 9.1
13.0

18.0
28.0
 0.49
 0.82
 1.3

 1.7
 2.4
 3.4

 4.4
 5.6
 7.0

 8.4
10.0
14.0

20.0
31.0
                                                           0.54
                                                           0.92
                                                           1.4

                                                           1.9
                                                           2.8
                                                           3.9

                                                           5.1
                                                           6.5
                                                           8.0
        0.57
        1.0
        1.7

        2.1
        3.1
        4.3

        5.7
        7.3
        9.0
 9.7   11.0
12.0   13.0
17.0   19.0

23.0   25.0
12.0   16.0   20.0   23.0   28.0
a Values given for slopes longer than 300 feet or steeper than 18X are extrapolations
  beyond the range of the research data and, therefore,  less certain than the others.
  (Control  of Water Pollution from Cropland, Vol. I,  A Manual for Guideline
  Development.  U.S. Environmental Protection Agency, Athens, GA.  EPA-600/275-026a.)

                                       198

-------
    AMXDR—active crop rooting depth—PRZM requires  input of the maximum
active crop  rooting depth  (AMXDR), in centimeters, for  the simulated  crop
(or the deepest root zone  of multiple crop simulations)  measured from the
land surface.   Generalized information  for corn, soybeans, wheat, tobacco,
grain sorghum,  potatoes, peanuts, and cotton are provided in Table  6-12.  If
minor crops,  such as mint, are simulated,  or site specific information
alters the generalized information, consulting with  Carsel et al. (1984) or
the Cooperative Extension  Service in the  specific locale is advisable.
                   TABLE 6-9.  VALUES OF SUPPORT-PRACTICE FACTOR, PJ
    Practice
1.1-2
     Land  Slope  (percent)

2.1-7       7.1-2      12.1-18
                                                (Factor P)
18.1-24
  Contouring (Pc)            0.60        0.50        0.60        0.80        0.90

  Contour Strip
    cropping (P  c)°
      R-R-M-M,               0.30        0.25        0.30        0.40        0.45
      R-W-M-M               0.30        0.25        0.30        0.40        0.45
      R-R-M-M               0.45        0.38        0.45        0.60        0.68
      R-W                   0.52        0.44        0.52        0.70        0.90
      R-0                   0.60        0.50        0.60        0.80        0.90

  Contour listing or
    ridge planting (Pcl)     0.30        0.25        0.30        0.40        0.45

  Contour terracing (Pt)c   d0.6//n      0.5//n      0.6//n      0.8//n      0.9//n

  No support practice        1.0         1.0         1.0         1.0         1.0
  a Control  of Water Pollution From Cropland, Vol. I, A Manual  for Guideline
    Development.  U.S. Environmental Protection Agency, Athens, 6A.
    EPA-600/2-75-026a.

  b R = rowcrop, W = fall-seeded grain,  0 = spring-seeded grain, M =  meadow.
    The crops are grown in rotation and  so arranged on the field that rowcrop
    strips are always separated by a meadow or winter-grain strip.

  c These Pt values estimate the amount  of soil eroded to the terrace channels
    and are  used for conservation planning.  For prediction of off-field sediment,
    the Pt values are multiplied by 0.2.

  d n - number of approximately equal-length Intervals Into which the field
    slope 1s divided by the terraces. Tillage operations must be parallel to the
    terraces.

                                       199

-------
TABLE 6-10.  GENERALIZED VALUES OF THE COVER AND MANAGEMENT FACTOR, C,
            IN THE 37 STATES  EAST  OF THE  ROCKY MOUNTAINS3'13
Line Crop, Rotation, and Management0
No.
Product
Level
High |
Jvity
Mod.
C value
Base value: continuous fallow, tilled up and down slope
Corn
1 C, RdR, fall TP, conv (1)
2 C, RdR, spring TP, conv (1)
3 C, RdL, fall TP, conv (1)
4 C, RdR, we seeding, spring TP, conv (1)
5 C, RdL, standing, spring TP, conv (1)
6 C, fall shred stalks, spring TP, conv (1)
7 C(silage)-W(RdL, fall TP) (2)
8 C, RdL, fall chisel, spring disk, 40-30% re (1)
9 C(silage), W we seeding, no-till pi in c-k W (1)
10 C(RdL)-W)RdL, spring TP) (2)
11 C, fall shred stalks, chisel pi, 40-30* re (1)
12 C-C-C-W-M, RdL, Tp for C, disk for W (5)
13 C, RdL, strip till row zones, 55-40% re (1)
14 C-C-C-W-M-M, RdL, TP for C, disk for W (6)
15 C-C-W-M, RdL, TP for C, disk for W (4)
16 C, fall shred, no-till pi, 70-50% re (1)
17 C-C-W-M-M, RdL, TP for C, disk for W (5)
18 C-C-C-W-M, RdL, no-till pi 2d & 3rd C (5)
19 C-C-W-M, RdL, no-till pi 2d C (4)
20 C, no-till pi in c-k wheat, 90-70% re (1)
21 C-C-C-W-M-M, no-till pi 2d & 3rd C (6)
22 C-W-M, RdL, TP for C, disk for W (3)
23 C-C-W-M-M, RdL, no-till pi 2d C (5)
24 C-W-M-M, RdL, TP for C, disk for W (4)
25 C-W-M-M-M, RdL, TP for C, disk for W (5)
26 C, no-till pi in c-k sod, 95-80% re (1)
Cotton6
27 Cot, conv (Western Plains) (1)
28 Cot, conv (South) (1)
Meadow
29 Grass & Legume mix
30 Alfalfa, lespedeza or Sericia
31 Sweet clover
1.00

0.54
.50
.42
.40
.38
.35
.31
.24
.20
.20
.19
.17
.16
.14
.12
.11
.087
.076
.068
.062
.061
.055
.051
.039
.032
.017

0.42
.34

0.004
.020
.025
1.00

0.62
.59
.52
.49
.48
.44
.35
.30
.24
.28
.26
.23
.24
.20
.17
.18
.14
.13
.11
.14
.11
.095
.094
.074
.061
.053

0.49
.40

0.01


                                 200

-------
 TABLE 6-10.  GENERALIZED VALUES OF THE COVER AND MANAGEMENT FACTOR,  C,
       IN THE 37 STATES EAST OF THE ROCKY MOUNTAINS4'13  (continued)
                                                           Producti
Line Crop, Rotation, and Management0
No.
Base value: continuous fallow, tilled up and down slope
Sorghum, grain (Western Plains)6
32 RdL, spring TP, conv (1)
33 No-till pi in shredded 70-50£rc
Soybeans6
34 B, RdL, spring TP, conv (1)
35 C-B, TP annually, conv (2)
36 B, no-till pi
37 C-B, no-till pi, fall shred C stalks (2)
Wheat
38 W-F, fall TP after W (2)
39 W-F, stubble mulch, 500 Ibs re (2)
40 W-F, stubble mulch, 1000 Ibs re (2)
41 Spring W, RdL, Sept TP, conv (N&S Dak) (1)
42 Winter W, RdL, Aug TP, conv (Kansas) (1)
43 Spring W, stubble mulch, 750 Ibs re (1)
44 Spring W, stubble mulch, 1250 Ibs re (1)
45 Winter W, stubble mulch, 750 Ibs re (1)
46 Winter W, stubble mulch, 1250 Ibs re (1)
47 W-M, conv (2)
48 W-M-M, conv (3)
49 W-M-M-M, conv (4)
Level d
High J
C value
1.00 1

0.43 0
.11

0.48 0
.43
.22
.18

0.38
.32
.21
.23
.19
.15
.12
.11
.10
.054
.026
.021
• ~-j
Mod.

.00

.53
.18

.54
.51
.28
.22













This table is for illustrative purposes only and is not a complete
list of cropping systems or potential  practices.  Values of C differ with
rainfall pattern and planting dates.  These generalized values show
approximately the relative erosion-reducing effectiveness of various crop
systems, but locationally derived C values should be used for conserva-
tion planning at the field level.  Tables of local  values are available
from the Soil Conservation Service.

Control of Water Pollution from Cropland, Vol. I, A Manual for Guide-
line Development.  U.S. Environmental  Protection Agency, Athens, GA.
EPA-600/3-75-026a.

Numbers in parentheses indicate number of years in the rotation cycle.
No. (1) designates a continous one-crop system.
                                 201

-------
   TABLE 6-10.   GENERALIZED  VALUES  OF  THE  COVER AND  MANAGEMENT FACTOR,  C,
         IN THE 37 STATES EAST OF THE ROCKY MOUNTAINSa'b (concluded)

  High level  is exemplified  by long-term yield averages greater than
  75 bu. corn or 3 tons grass-and-legume hay; or cotton management that
  regularly provides good stands and growth.
e Grain sorghum, soybeans, or cotton may be substituted for corn in
  lines 12, 14, 15, 17-19, 21-25 to estimate  C values for sod-based
  rotations.
Abbreviations defined:
B     -  soybeans                      F  -  fallow
C     -  corn                          M  -  grass & legume hay
c-k   -  chemically killed             pi  -  plant
conv  -  conventional                  W  -  wheat
cot   -  cotton                        we -  winter cover

Ibs re     -  pounds of crop residue per acre remaining on surface after
              new crop seeding
% re       -  percentage of soil surface covered by residue mulch after
              new crop seeding
70-50% re  -  70% cover for C values in first column; 50% for second column
RdR        -  residues (corn stover, straw, etc.) removed or burned
RdL        -  all residues left on field (on surface or incorporated)
TP         -  turn plowed (upper 5 or more inches of soil inverted,
               covering residues)
                                     202

-------
             TABLE 6-11.  INTERCEPTION STORAGE FOR MAJOR CROPS
  Crop                         Density                       CINTCP (cm)
Corn
Soybeans
Wheat
Oats
Barley
Potatoes
Peanuts
Cotton
Tobacco
Heavy
Moderate
Light
Light
Light
Light
Light
Moderate
Moderate
0.25
0.20
0.0
0.0
0.0
0.0
0.0
0.20
0.20
- 0.30
- 0.25
- 0.15
- 0.15
- 0.15
- 0.15
- 0.15
- 0.25
- 0.25
    CN--runoff curve number—The interaction of hydrologic soil group (soil)
and land use and treatment (cover)  is accounted for by assigning a runoff
curve number (CN) for average soil  moisture condition (AMC II) to important
soil cover complexes for the fallow,  cropping,  and residue parts of a
growing season.   The average curve  numbers for  each of the three soil cover
complexes are estimated using Tables  6-13 through 6-17.  The following steps
provide a procedure for obtaining the correct curve numbers.  Corn planted
in straight rows will be used as an example.

         Step 1.  From Appendix  B (Carsel et  al.  1984) find the hydrologic
                  soil  group for the  particular soil  that is in the area
                  under consideration (Appendix B  from this  reference
                  contains a listing  of  soil  groups and their  hydrologic
                  soil  cover classification).  There  are  four  different  soil
                  classifications (A,  B,  C, D) and  are in the  order of
                  decreasing percolation  potential  and increasing  slope  and
                  runoff potential.   Soil  characteristics associated  with
                  each  hydrologic group  are as follows.

                  Group A:   Deep  sand, deep loess,  aggregated  silts,  minimum
                          infiltration of  0.76 -  1.14  (cm hr'1)


                                    203

-------















to
LU
^
to

o
UJ
z


LU

t—

Z

§
g

— J
2
i—

***^
^j
t— i
3


g

^r

cc
o
^
^
°
i
1
g


*
CM
T-l
1


UJ

9
i-












^j i f
M-S&
O •— 0
o. .-^
o> w c O
C > i 	 •
id •»•• i %
Of 4-> O
(J O
< at.

  T-
4) C i—
> 0 3
J- Z 0
10 *•-''
£

>> O "o>
4-> i- C
CLi- 1*- -r-
O J- 4->
t. 3 V) C

ZO O.




0» Ex->
O O O>
C i- C
O. O> H- •<-
O en 4->
i. i. 1/7 C
O O) >> >
3r o rc
g'jT c
4J E-J5
c o i—
10 ^- 3
i— TJ
a. +~-



ai
> M
i- axd
4-> t*~ 4-> C
10 O 10 O

C I/) OO 4->
flj fl> (J
Vt 4-> i. 3
££•%"§
Ol Z O.
Of





&
s-


_^
a s
0 0
»O CO



^ 5
.U
CD
*""* Sv?
*— I




*^-v y—s
5^?
CO OO 00 ffy
2? "•— 'L0 "s— *
«^-*O *•— 'O
«-H r-H
LO LO
CSj »i-( *
U  Q ^J Q
Q. CL

l/l +J VI •*-*


O O O
CO OO CO

1 1 1
o o o
,~H ,-H i«H
r- * rM ^*






&n tn LO
i i i







-— *to o
tO »-l 4->^—
^-« csi *—
ID *-H
C CO
•r- ^3 QJ
t >s C
CX O A Z3
^f ^j ^^ ^^






z z
H- 1 1— *

• •
— 1 Z -J I
_ O -10
«t LU < OO
HH Z »MI Z


U1
c
10
OJ
I- >)
0 0
U OO



o
o
CO



VI
^~

o
"^
*o



O ^"^
1— t
^^w^^^^CXj
^- LO «>^
CM O «-H
•
t-l Vf> O>
-^ 3

4.) «
O. C X

oo ^ *— ^


§

1
o
>— 1
1-1






in
i
in






S 8
o in aj
« — ^i-H 3

LO O
fj
^ ^^ J^
Z^ —






<
o

*
(/I O£
z«c
X M
H- 
8



3
in
t— i



3
^J

O
^



^—-1
*-« o
LO s—.
CO ^^CO
"*~*\O ^O
O r-l —
CM — '
O
• in CM
o •-<
O) .

c Q.

4^ T3 CO


CM
CM
1
8
CM






LO
i— 1
1
in



•
^*
CM
Q '*-'
4-1
™S
CSJ CM >
IO LO
— ^fH e\j o
^^ • •
f**^ O^^* ^C
ci^oi






«c->
oo'cTz"
i^ Z Z




4->
10
QJ




in
in
I-H



•g
W

rv?
*v?
*T*



o

^•^
^ ^*^
CM r-.
— -CM
I— 1
I— 1
«
4^ •
Q.4-1
01 U
IOO


O
I—I
1
o
in
1-1






LO
*?
in




^_
Itf
S-°
in»—
J3«— 'CM
W ^
U. i— t •**'
o •— «-•
4^ *^
« >>
S <":e






"So
« i— <

VI *
l-l UJ
I**-
b *
o >-
—1 Z


tn

5
10
4->
£



S




_g
^^



LO
CNJ


0

^MV^^MV
CM O\
«Mm
^^^
CO UO
•— 1 1— <

« •
Ol U
3 0>
^C C3


LO

«— 1
0
in
«— i






in
T— 1
1
in






o
LO ^"^
LO
LO
"*~ tt)
« C
^•^






ec. —i
o «t

• «
«>e<
o t— >
<: < o
3 O Z



M
4->
3
s
Ol
CL



s
o
CO



ja
^~



o
CM



O
4J

^•v**"^
oo r»»
-H CM

T-l I-H

^ •
t— 4->
3 U



O
in
«— i
o
CM
r— 1




•»- «/> CT
id C
•O -r-
4-»i— ^

i— u. 
^CMx-s^,
*^^LO ^H
C >>LO O
^3,;^

^^ ^ C
—5^0






a?
)—

0t
o<
00 >
CJ ^*
z ^



o
U
U
JD
o



o
ro
1
in
«— i



3
^J

t/^
M^




^^
. — .^J- O »-i
CM CM CM
00 CO
»-l v^ . U
O O.O
i— 1 CM Ol
00 O
>) • 4->
i— > LU
3 O ^C ^*^^ i ^
t~y ^ CO LO
• IO CO
X OOO CM CO
^> 4J KX s,^'^^


O

i-H
0
CM




>

LO
r-l
1






5 cr
CM
S'-'CM
CO
^^ ^— %m *™i
^2|?r
•^ >>

Z ^ r^
»^ »
p ^52 2






UJ
Z

«
OO
X
f-



E
cl
"5 ?
C5 tO







«
P^.
,__!

>^

•f—
y

•
(/i
U

c
•r—
4_>
I/)
0)

s_
10
3C
1
CT)
C
•r-
C
10
Q_

^~
10
3
13

.
UJ
«t
•o
•T1*
Ol
z
"S
10
oo
Q£

g
H
17
C^l

10
E
5
U
«
S
CL.

&
i.
t i


*
*—**
•o
UJ


*
3C

*
•o
•r*
^
J3
t-
S
U
204

-------
TABLE 6-13.  RUNOFF CURVE NUMBERS FOR HYDROLOGIC SOIL-COVER COMPLEXES*
             (ANTECEDENT MOISTURE CONDITION II, AND Ia = 0.2 S)

Land Use
Fallow
Row crops





Small
grain




Close-
seeded
legumes
or rota-
tion
meadow
Pasture
or range




Meadow
Woods


Farmsteads
Roads
(dirt)c
Cover
Treatment
or Practice
Straight Row
Straight Row
Straight row
Contoured
Contoured
Contoured and terraced
Contoured and terraced
Straight row
Straight row
Contoured
Contoured
Contoured and terraced
Contoured and terraced
Straight row
Straight row
Contoured
Contoured
Contoured and terraced
Contoured and terraced



Contoured
Contoured
Contoured








Hydrologic
Condition
_ _ —
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Good
Poor
Fair
Good
Poor
Fair
Good
Good
Poor
Fair
Good
	
	
	
Hydrologic Soil Group
A
77
72
67
70
65
66
62
65
63
63
61
61
59
66
58
64
55
63
51
68
49
39
47
25
6
30
45
36
25
59
72
74
B
86
78
78
79
75
74
71
76
75
74
73
72
70
77
72
75
69
73
67
79
69
61
67
59
35
58
66
60
55
74
82
84
C
91
85
85
84
82
80
78
84
83
83
81
79
78
85
81
83
78
80
76
86
79
74
81
75
70
71
77
73
70
82
87
90
D
94
91
89
88
86
82
81
88
87
87
84
82
81
89
85
85
83
83
80
89
84
80
88
83
79
78
83
79
77
86
89
92
(hard surface)^
a Soil Conservation Service, USDA.  SCS National Engineering Handbook,
  Handbook, Section 4, Hydrology.  1972.
  Close-drilled or broadcast.
c Including right-of-way.
                                  205

-------
            TABLE 6-14.  METHOD FOR CONVERTING CROP YIELDS TO RESIDUE3
t.
Cropb
Barley
Corn
Oats
Rice
Rye
Sorghum
Soybeans
Winter wheat
Spring Wheat
Straw/Grain
Ratio
1.5
1.0
2.0
1.5
1.5
1.0
1.5
1.7
1.3
Bushel
Weight
(Ibs)
48
56
32
45
56
56
60
60
60
 a Crop residue = (straw/grain  ratio)  x  (bushel weight  1n  Ib/bu)  x  (crop
   yield 1n bu/acre).

 b Knlsel,  W.G. (Ed.).   CREAMS:   A  Field-Scale Model  for Chemicals,
   Runoff,  and Erosion  from Agricultural  Management Systems.   USDA,  Conservation
   Research Report No.  26, 1980.
             TABLE 6-15.  RESIDUE REGAINING FROM TILLAGE OPERATIONS8
            .                                                     Residue
     Tillage0                                                   Remaining
    Operation                                                      (<)
    Chisel Plow                                                    65
    Rod weeder                                                     90
    Light disk                                                     70
    Heavy disk                                                     30
    MoIdboard plow                                                 10
    Till plant                                                     80
    Fluted coulter                                                 90
    V Sweep                                                        90


a Crop residue remaining = (crop residue from Table 10) x (tillage factor(s).

b Knlsel, W.G. (Ed.).  CREAMS:  A Field-Scale Model for Chemicals, Runoff,
  and Erosion from Agricultural Management Systems.  USDA, Conservation Research
  Report No. 26, 1980.
                                       ?,06

-------
                 Group B:  Shallow  loess,  sandy loam, minimum infiltration
                          0.38 - 0.76  (cm hr'1).

                 Group C:  Clay loams,  shallow sandy  loam,  soils low  in
                          organic content,  and soils usually high in clay,
                          minimum infiltration 0.13 -  0.38  (cm hr~*).

                 Group 0:  Soils that  swell  significantly when wet, heavy
                          plastic clays,  and  certain saline  soils, minimum
                          infiltration  0.03 - 0.13 (cm hr~*).

                     If the soil series  or  soil properties are not known,
                 the hydrologic soil group can be estimated from Figure
                 6.3.
    Table 6-16.  REDUCTION IN RUNOFF CURVE NUMBERS CAUSED BY CONSERVATION
                        TILLAGE AND RESIDUE MANAGEMENT3
Large
Residue
Cropb
(Ib/acre)
0
400
700
1,100
1,500
2,000
2,500
6,200
Medium
Residue
Cropc
(Ib/acre)
0
150
300
450
700
950
1,200
3,500
Surface
Covered
by Residue
(*)
0
10
19
28
37
46
55
90
Reduction
1n Curve
Numberd
(*)
0
0
2
4
6
8
10
10
a Knlsel, W.G. (Ed.).  CREAMS:  A Field-Scale Model for Chemicals,  Runoff,
  and Erosion from Agricultural Management Systems.  USDA, Conservation Research
  Report No. 26, 1980.

b Large-residue crop (corn).

c Medium residue crop (wheat, oats, barley, rye,  sorghum, soybeans).

d Percent reduction 1n curve numbers can be Interpolated linearly.  Only
  apply 0 to 1/2 of these  percent reductions to CNs for contouring  and terracing
  practices when they are  used 1n conjunction with conservation tillage.

                                    207

-------
    TABLE 6-17.  VALUES FOR ESTIMATING WFMAX IN EXPONENTIAL FOLIAR MODEL
Crop
Corn
Sorghum
Soybeans
Winter
wheat
Y1elda
(Bu/Ac)
110
62
35
40
Bushel4
dry wt.
(Ibs/Bu)
56
56
60
60
Straw/Grain
Ratio
1.0
1.0
1.5
1.7
Units
Conversion
Factor
1.1214 x 10"4
1.1214 X 10'4
1.1214 x 10'4
1.1214 x 10'4
WFMAX
1.38
0,78
0.59
0.72
10-year average
                   Care must  be exercised,  however in use of this
                figure.  Considerable spatial  aggregation was made in
                order to develop the generalized map over such a large
                area.  Where  possible,  development of more highly resolved
                data is preferable.

       Step 2.   From Table  6-13 find the  land  use and treatment or
                practice that is to  be  simulated (e.g.,  row crops,
                straight row).

       Step 3.   From Table  6-13 find the  hydrologic condition of the soil
                that is to  be simulated (e.g.,  good).

       Step 4.   From Table  6-13 find the  curve  number for antecedent
                moisture condition  II for the  site selected.  Example:
                Hydrologic  group = A, treatment practice is straight row,
                land use is row crops,  hydrologic condition is good.  The
                curve number  for the cropping  season is  67.

       Step 5.   Follow the  same procedure for the fallow portion of  the
                growing season  using only the hydrologic soil  group.
                Example:  Hydrologic soil  group A,  land  use fallow,  curve
                number for  condition II is  77.

       Step 6.   The  post-harvest or  residue portions  of  the year requires
                numbers that  reflect the  extent of  surface cover after
                harvest.  This can be quite  variable and  in many, cases may
                                  208

-------
                                                           Q_
                                                           UJ

                                                           E
                                                           O
                                                           S-
                                                           CO
                                                           O.
    o   •
    c: tn
    d) CM
    cno
    s-  i
    d> CM
    E 00
    LU  I
       00
    T3 --^
    C O
                                                           O 4-  l£>
                                                           S-  S-   I
                                                           en  co    O
                                                           O -r-
                                                           S-  3  co
                                                          -a     •!-
                                                          1—  CO  >
                                                          •r- -r-  S-
                                                           O  CJ  O
                                                           CO  O  C_>
                                                               CO
                                                           O)  CO   «
                                                           O <  >,
                                                          •i-     O
                                                           >  co  c:
                                                           s-  CD    c:
                                                           O -r~  O
                                                           
                                                           o
                                                          CJ
                                                               s-  o
                                                               o  s-
                                                               CLQ-
                                                               Q.
                                                           O
                                                          oo
                                                              co

                                                               O
        d)
        E

        o
        S-

        >
        c:
E CO LU

4J  S-   -
CO  O CO
d) 4-
       ZD
S-  d)
O T3
                                                               d)

                                                               O
                                                           ta
                                                           s- -a
                                                           en i—
                                                           n3 di
                                                          CO
                                                           a>
                                                           S-
                                                           CD
        d)
        to
        c
        o
        a.
        co
        OJ
209

-------
         require considerable judgement.   Under  "average"
         conditions a value  set to the mean  of the  fallow  and
         growing period  numbers (from steps  4, 5)  is  appropriate.
         In the example  case, this number will be  the mean of 77
         and 67, or 72.

Step 7.  The curve number input sequence  is  now  written as

                   77     67     72

            Additional guidance for management practices

                 Pesticides  are being increasingly used in
            conjunction  with conservation practices to reduce
            erosion and  runoff.  Most notable among these  practices
            is the use of conservation tillage.   The  idea  is to
            increase the soil surface residue and  hence reduce
            erosion and  runoff by increasing infiltration.  The
            curve numbers developed in steps 1-7 assume
            conventional practices and must  be further modified to
            reflect the  changes in management.   Both  the fall and
            growing season numbers must be modified.   For  purposes
            of this example, assume the corn is  produced by using
            chisel plows rather than the  conventional tillage
            assumed above.  The following steps  now apply.

Step 8.  From Table 6-14 find the straw/grain ratio for corn, which
         is 1.0.

Step 9.  From Table 6-14 find the bushel  weight  of corn, which is
         56.

Step 10. From Table 6-12 find bushel/acre yield  of corn, which is
         110.

Step 11. Multiply straw/grain ratio * bushel weight *" bushel
         weight/acre = crop residue produced by  the crop.   For
         corn, 1.0 x 56  x 110 = 6160.

Step 12. From Table 6-15 find the tillage practice desired for the
         crop use site (e.g. chisel plow).

Step 13. Multiply the crop residue determined in step 11 by the
         tillage factor  from step 12 to determine residue
         remaining, i.e, 6160 x 0.65 = 4004.

-------
         Step 14. From Table 6-16 find the reduction in curve number for
                  AMC II, crop curve number produced from residue remaining
                  after harvest determined in step 12.   For corn at
                  4000 pounds per acre, a 10% reduction in curve number is
                  produced.

         Step 15. Determine the curve number for antecedent moisture
                  condition (AMC) II.  From Steps 1-5, AMC II was 67.  67
                  * 0.10 = 6.7, which is rounded to 7.0.  The modified curve
                  numbers are 67 - 7 = 60 and 77 - 7 =  70.

         Step 16. The post-harvest curve number must also now be reduced by
                  averaging the fallow and growing season numbers, that is,
                  70 and 60 to yield 65.

    COVMAX--maximum area! crop coverage--TFAT estimates the ground cover as
the crop grows to some maximum value, COVMAX, by calling the PLTGRN
routine.  COVMAX is passed through TFAT to this module.  The maximum area!
coverage (COVMAX) determines the fraction of ground covered by the crop and
thus influences the mass of pesticide that reaches the  ground from an
application event.  For most crops, the maximum coverage will be on the
order of 80 to 100 percent.

    WFMAX--maximum foliar dry weight--If the user chooses to have the model
estimate the distribution between plants and the soil by an exponential
function, then WFMAX must be specified.  WFMAX must be  entered here whether
PLTGLN or the original PRZM plant growth model is used.

    The maximum foliar dry weight, WFMAX, of the plant  above ground (kg m~2)
is the exponent used in the exponential foliar pesticide application
model.  WFMAX for several major crops is given in Table 6-17.  Estimates for
other crops will require yield information that is available from from USDA
crop reporting service.  WFMAX is computed by finding the product of columns
2, 3, and 5, and by multiplying this number by the straw/grain ratio (col.
4) plus 1.0.  The straw/grain ratio defines the amount  of straw associated
with the final grain product.  Both the straw and grain should be accounted
for to determine the maximum weight. Thus, the straw-to-grain ratio should
have (1.0) added to it when used to compute WFMAX.  An  example is provided
for barley.

         Step 1.  Yield, bushel dry wt., and straw/grain ratio for barley
                  are 42.0, 48.0, and 1.5, respectively.

         Step 2.  WFMAX = Bu/Ac * Lbs/Bu * (straw/grain ratio + 1.)  *
                  conversion factor to yield (kg m~2) for TFAT input.

-------
         Step 3.   Conversion  factor  =  2.47 Ac  *  1  ha * 0.454  kg  =
                                          ha    104m2      Lbs

                                      1.1214 x 1(T4.
         Step 4.   WFMAX = 42.0  * 48.0  *  (1.5  +  1.0)  *  1.1214  x  10"4, which
                  equals 0.56.

    EMD, EMM. IYREM,  MAD. MAM,  IYRMAT. HAD, HAM.  IYRHAR—cropping
information for emergence, maturity, and  harvest—Generalized cropping
information including date of emergence  (EMD, EMM,  IYREM), maturity  (MAD,
MAM, IYRMAT), and harvest (HAD,  HAM, IYRHAR)  for  eight major  crops  including
corn, soybeans, wheat, tobacco,  grain  sorghum,  potatoes, and  peanuts are
provided in Table 6-12.  Simulations involving  minor crops such as mint,  or
where site specific information alters the general  practices  provided,  may
require consultation  with Carsel et al.  (1984)  or the  local Cooperative
Extension Service. The user should note  that if  the PLTGRN option  is on,
the crop maturation information is not utilized.

6.4.1.3  Pesticide Parameters--
    Pesticides can be applied directly to the soil  surface, the plant
canopy, or to both.  Two modeling problems arise  when  one considers  this.
First, the initial distribution of the applied  pesticide between plant
foliage and the soil  surface must be estimated.  Second, the  remaining
foliar deposited pesticides then become  available for  degradation
(photolysis) or removal (volatilization,  washoff).   Recall that two  options
are available for distributing  the applied pesticide (the FAM parameter).

    TAPP—total pesticide application—The total  pesticide application  per
event is entered in terms of kg-active ingredient (a.i.) ha   .   Typical
application rates are included  on the  product's registration  label.
According to Smith (1988, personal communication) the  coefficient  of
variation for granular application is  40  to  70% for spray application.

    DEPI—depth of incorporation—This variable is  only needed  if  soil
application of chemical is specified  (i.e.,  FAM = 1).   Typical  incorporation
depths are 5-10 cm.  If soil injection is being simulated, user should  be
aware that injection  below 15 to 20 cm is difficult to achieve  and
represents an approximate upper limit  of  incorporation depth  (Matthews
1979).  Representative values for several soil  application methods  are  given
in Table 6-18.

    BETA--canopy penetration attenuation constant—The attenuation
coefficient, 6, depends mainly upon the  leaf  area index (LAI)—the higher
the LAI, the higher will be the value  of &.   Uk and Courshee  (1982)  report
                                    212

-------
     TABLE 6-18.  PESTICIDE SOIL APPLICATION METHODS AND DISTRIBUTION
 Method of
 Application
Common Procedure
Distribution
DEPI
 Broadcast
 Disked-in
 Chisel-piowed
 Surface banded
 Banded incorporated
Spread as dry granules
or spray over the whole
surface

Disking after broadcast
application
Chisel plowing after
broadcast
Spread as dry granules
or a spray over a fraction
of the row

Spread as dry granules
or a spray over a fraction
of the row and incorporated
in planting operation
Remains on the      0.0
soil surface
Assume uniform     10.0
distribution to
tillage depth
(10 cm)

Assume linear      15.0
distribution to
tillage depth
(15 cm)

Remains on soil     0.0
surface
Assume uniform      5.0
distribution to
depth of incor-
poration (5 cm)
an attenuation coefficient of 0.018 cm *  for a cotton crop having an LAI of
3 to 4 and 0.042 cm"-*- for a cotton crop having an LAI of 6.

    FILTRA—initial  foliage to soil distribution—The filtration parameter
(FILTRA) relates to  the equation for partitioning the applied pesticide
between the foliage  and ground (this applies when FAM = 3).   Lassey (1982)
suggests values in the range of 2.3 - 3.3 m  kg  .  Most of  the variation
appears to be due to the vegetation and not the aerosol.  The user should
note that the value  of FILTRA entered here only applies to pesticide
application events specified in the TFAT  habitat input sequence.  If
application events are specified from the GRDDEP module, then a
corresponding value  of this parameter would be entered there, given that the
user wishes to invoke the exponential deposition model.

    FEXTRC--fo1iar washoff extraction coefficient—Washoff from plant
surfaces is modeled  using a relationship  among rainfall, foliar mass of
pesticide, and an extraction coefficient.  The parameter (FEXTRC) is the
                                      213

-------
required input parameter to estimate  the  flux  of  pesticide  washoff.   Exact
values are varied  and  depend upon the crop,  pesticide  properties,  and
application method.  Smith and Carsel (1984) suggest 0.10 is  suitable for
most pesticides.

    PLDKRT--fo1iar disappearance rate constant—The degradation of
pesticides on plant surfaces is modeled by a simple first-order rate
expression.  This  is a very chemical  specific  parameter that  must  be
measured.  Typical values for selected pesticides are  provided in  Table 6-
19.

    If the user has monitoring data which shows the degradation of plant
foliar concentrations  with time, then the coefficient  can be  calibrated to
cause the simulated concentrations to mimic the observed data.  The
exponential decay  model is given by:
                          C . C^-                                    (6-6)

       in which C is the simulated  concentration
                CQ is the initial concentration
                FEXTRC is the first order decay rate  (day -1)  and
                t is the time elapsed  since application  (days)

In linear form the equation is


                          In (C/r  ) =  -(FEXTRC)t                     (6-7)
                                Lo
Therefore, the coefficient FEXTRC  is the slope of the plot of  the  natural
log of the normalized concentrations (C/C0) vs. time  in  days.

    KD--pesticide soil-water distribution coefficient—The user can enter
directly the distribution coefficient  or the model will  calculate  a value
given other pesticide properties.   If  the parameter KDFLAG is  set  to a valu*
of 0, then direct data input is made as the parameter KD.  If  KDFLAG is set
to 1, however, additional information  is required.

    PCMC, SOL--options for use in estimating distribution coefficients from
related input data—The fate of pesticides in soil arrd water is highly
dependent on the sorptive characteristics of the compound.  Sorptive
characteristics affect the physical movement of pesticides significantly.
The sorptive properties of pesticides  generally correlate well  with the
organic carbon content of soils. The  carbon content  of  most soils decreases
with depth.

-------
    The TFAT module allows the user  to  estimate an organic carbon partition
coefficient for  the pesticide from one  of three models  based on water
solubility.  The KQC is subsequently multiplied by organic carbon to obtain
the partition coefficient.  The three models are:
       PCMC1     Log KQC = (-0.54 *  Log  SOL)  + 0.44

         where   !<„ = organic carbon  distribution coefficient
                 SOL = water solubility,  mole fraction
                                                  (6-8)
      TABLE 6-19.  DEGRADATION RATE CONSTANTS OF SELECTED  PESTICIDES ON FOLIAGE3
        Class
               Group
Decay Rate (days'1)
   Organochloride
   Organophosphate
               Fast
(aldrin, dieldrin, ethylan,
heptachlor,  Hndane,
methoxychlor).

               Slow
(chlordane,  DDT, endrin,
toxaphene).

               Fast
(acepate, chlorphyrifos-methyl,
cyanophenphos,  diazinon, depterex,
etMon, fenitrothion, leptophos,
raalathion, methidathion, methyl
parathlon, phorate, phosdrin,
phosphamidon, quinalphos, allthion,
tokuthion, triazophos, trlthlon).

               Slow
(azinphosmethyl, demeton, dlmethoate,
EPN, phosalone).
  0.231 - 0.1386
                                                               0.1195 - 0.0510
  0.2772 - 0.3013
                                                                0.1925 - 0.0541
Carbamate

Pyrethroid
Pyridlne
Benzole add
Fast
(carbofuran)
Slow
(carbaryl)
(permethrin)
(plchloram)
(dicamba)
0.630
0.1260
0.0196
0.0866
0.0745

- 0.0855



  a Knlsel, W.G. (Ed.).   CREAMS:  A Field-Scale Model for Chemicals, Runoff, and
    Erosion from Agricultural Management Systems.  USDA,  Conservation Research
    Report No. 26, 1980.

                                      215

-------
       PCMC2     Log  KQC  =  3.64  -  (0.55  *  Log  SOL)                     (6-9)

         where   SOL  =  water  solubility, milligrams  liter"1

       PCMC3     Log  KQC  =  4.40  -  (0.557 * Log SOL)                   (6-10)

         where   SOL  =  water  solubility, micromoles  liter"

    These models are selected by  setting  PCMC to  values of 1,  2,  or 3,
respectively.   These methods were selected because  of referenced
documentation  and provisions for  direct use with  the most commonly reported
physical  pesticide parameter, water solubility.   The three models used in
TFAT for  estimating  partitioning  between  soil and water are limited to
specific  types  of pesticides.  These equations are  best used for  pesticides
having melting  points  below  120 °C.  Solubilities above these  temperatures
are affected  by crystalline  energy and  other  such physical properties.  The
three models  are not appropriate  for pesticides whose solubilities are
affected  by crystalline  energy  or other physical  properties, and  would have
a tendency to  overestimate the  partitioning between soil and water.  Of the
three models,  the first  model is  for true equilibrium of completely
dispersed particles  of soil/water concentrations  less than 10.0 g 1  .  The
second and third models are for soil/water concentrations greater than
10.0 g I"1 and for short equilibrium periods of 48 hours or less.  For
applications, the first model  would be the most appropriate.
    Some pesticides having properties amenable for use with the water
solubility models are provided in Table 6-20.   The pesticide solubility,
SOL, must also be input.  Units must be consistent with the model chosen.
Table 6-20 also provides pertinent values for the selected pesticides.

    The user should be aware that an organic carbon partition coefficient is
also expected by the plant translocation model.  The values selected for use
in the two respective parts of the program should be consistent, unless the
user has information to the contrary.

    KD user-specified distribution coefficients—A useful relationship
exists between the octanol-water distribution coefficient and the organic
carbon distribution coefficient.  This relationship can be used when
measured soil distribution coefficients are not available, or the pesticides
posses crystalline energy properties that would preclude the use of any
water solubility models.

    The octanol-water distribution coefficient can be used for calculating
distribution coefficients for pesticides that possess monomer-associated
properties for solubility in water.  Karickhoff et al. (1979) proposed a
relationship between KQW and KQC given by

                                  7.16

-------
 ^


o r;
JS

!ss
(E S S
C 5 ,-
II

8

LU
?
o o —  B •o
•go    £ —


01 4- i  ~ c
« o +-  o o
o QC. in  vt rsi
                    .

                    2*

               i i i
      £  o

       t

      l«

      LL.  O
           O — C 1- O  E
           u> n — — IN  ~
                      O»
                      E
                     i
                                                                 o>     M- o>
                                                                                          •*- Ol     •*- O)
                                O O   K> O


                                   l"   O C3
                                                         in . o o o m
                                                                                            — 2s o Q —
                                                                                            C4 fO O fO *~ ^ ^ V ^— ^— *•• *— OO CO P* CN    tf\ O*
                                                                  o o o ° ° ° ° ° ° ° °

                                                                  •JJ^^xxxxxxxx


                                                                  fO — S^**^

                                                                  flJ-'lN0* — lftm-*» —
                                                                                      ja x> o o a   a o
                                O* m 10 KI r*.   eg *   v«>
                                                                           in oo   —
                                                                  —    IOUMK1
                                X X X  X X
                                                                                 XX      X X X X    X
                                           XX   XX    X   XX   X    X   XXXX   XX
                              inpinnnoo
                                "NOKir^
                                IN — r-
                                  So r- —' m o KI — IN — o   oopiN—*inpo22'~<>'2lo2C
                                  r^     CM»riN   o« — in   — o     — KI in O g   mor»goo
                                  "S—     o"              •£   — * 2.        %   v —     -8
                                  ™  L.    B   • *•    X  4)  (0 t. Q A   A    >&••(/>      L.   "^ B     — J3
                                     =    t *-•»-$    O  -o  O O u i_   I    c o (. 6 w?    o.   IDIO     £ L.
                                     c  c 6 3 c i   jr  —  £ — c o   —    « i. x Z o    ucOO^-t-5«c
                                           •— —   ^^    ••• — .... — ^   ..  —. — —• ^    O *T  P '  ^ff fff f? * ^  ^-
                                                                               9    —*•-*-•*- (nSo
-------









ft
$
^
Ul
>
a
z JS
u ^
= ^ UJ
8 9 o
o ~;
Hi
°" ° z
L" K
" t i
UJ * K-
 o a
Lu _ Q
ii?
a S 5
|||

S o §
° ° g
S —
S2 f-
o
7
UJ
m








0

S

2 , 4-
« i - i
•o O £
* 4- • •
i ± £ 28
— — 01
O — e H- O


c




0
U

J-
«
>.
ID
•o

0




I
O
10
(_
«4-



5

«


1
2
u
0
2
u
0
u
0
u
0

—
f


1


lem 1 ca 1
u


01 0 »- 01
„ 10 10 « >0
r* l<\ — O —
o — o o o
r r r i* i°
— CM — K> i-

o ^ o o S
'"" 0 "~ — —
-gj* =*Hfts*g


o o o o o o o o o o o
£ s si s 3 1 s 1 s £
CM 	 CM ,ft (S •«• 	 CN



s*«£ $R*£s£5


X


X X


x xxxx


XXX XX



^ Q Q OO — O O Ol O Ift O
IO Q » — . >.
C C O .C £
ID — C 4- 4-
— > >• C 0 0
SOU t- 0 E e
— O 4- »• 1
+->.— £ — 0 t- W 0.
«J X .C 4- .C UOOO4-S
— iD4-ooi/>a.3.c»-o 0 0 X N — .C L.
+-».». — 4- S O «3 T3 Q. 0.
§
2 S S . = =
S fe S ? S 2 5§g
T O) C — •O O — .C — T)
u c i_ — > i_ u> to 4- o — y
u.u.u.u.0 OOC3I — —



at

§'

3
•
•
in


?
o
X
in


J3
5








X






m



c
L.
C

l_
3
e
-*


*"
CM
5
r
Ov
CM
5
in


I
O
S
r*


^n
§











X



in



c
O
+-

in
I5 gs
Ift

m o> to in
it T i
0 O O O

IO 1ft K> CM



ss si





X





XX X



•- in
§O* r^ 0
in o
•i*


fe 0
— L. (A
£ « O
— 0 Q_ £
||- |f
4- 4- 4- 4- C
til £
fe §
— u u
X >• 0 C U
E X Q. O 3
5 o i — o
XXX Z








385
3550000
Olft
"~ "x
in
*o
Y
o
5 8



fumesate
hate
U

§0
e
t. «
z o



<0 10
i i
i i
« CM

0 CM
I'
"x
CO
CN
f
O
8 5
o —



£a


X

X






X X


Ift
o
X ^~
00 CM
CM


§
« L.
X O
O a
c
xamyl
arathio
o a.


M- ^ 0) 0)
-o- o o- -
s is 3
1 1 K»

o q
OO
oo o> ^ ch vo
•- CM CM CM
3S|8°-8


ft IO Ift Ift CO Ift
i T i i i i
o o o o o o

^ "> •»»«•»



sissss








X XXXX


X



in
O O O O O O
Kt m oo o o


c
I - - ^ -
o L. a. o. *o c
4- O O O c x
I t k k i k
fe _ fe
C 0 £ — £
a 4- u c u
L «) «J C — 0
a. a. a. a. a. a.

-------















fe
I
-J
LU

0
LJJ J?
« Lb
rx * Z
o Q 5
« _ i-
^ K §
if 3 _,
OT § 0
Si:
Ps
.J < <
LU * 1-

* g 8
<3 *"* £
^ (/> S
I«s
|||
§ § 1
— 5

 f;
7
•0
LU
_J
co
*










i
« 0
•*- o
$ 0
_~

O in
— 1 4- >.
4- C c O 0
O O — Q "D
•o O a* —
 IM

3 ^
Z E

5
O!  «> 0> 10
o — *r rg — *t
o o o o So
i i i i
in  fvi m ITI
O O O O

e> "~ KI
§8 ° ^ § § |N" S
^_ JO ^ §
m K» o
00 V
3 S -ooooo 5 SSoK,g
CO CM
iiiii iifTiTi
ooooo ooooooo
RS255SS2 Is^l
_Ki-«-oinm -(MM-oo


•OAAJ3OAAO .QT3.O.O-O

»o in in
• • ••••trt* •
assss .SR-sss
X XX



X


X

XXX XX X

X XX XXX

":
CNOtOOOOOO ^ Q O Ki O
00 CsJ


cO c c c
JX — — — 0
CUCX— — — 9 £
O-*->-+-O^«Jw ECU.
a.N4>L.ONL.l-C. «J * O
«— o«— 53— — o^z
CV L. C .C C — H- M- O — ID U
il^lll^liillJ

» 3-0
1 « C o S I t
a. L. c — » t> — « o
— 0 Q. O 0 O) C3OC£ —
c •»- OL. « — ^ 10 — aO a. .c
3OO93O0O L. (/) O O (-.





























*
*"*•
2
2 »
• . c *•* \D
0 O *
fe Z 5 -
u -
I/) O « —
§b >. —
H- A O
£ O O> 3
3 O O O
WE —
• — c
~. " ^- — O
c v> *-* y> 4-
— Z <- Z - o
§Z « X >. e
— 4- ~» C 4-
— o a o
4- X » >- Q. L.
— 4- 4- E O.
4- — • — 8
L. — O — O 0.
co — c — o
a. J3 o A en u.
3 0 3 CO
— — —
o y> u o .= -c
— u> z 10 n mm
6 « Q 0 — 4^
•DO *^ O 3 U
eg 6 O. CD
o — m —
»- — lf> — O JI
|E "" • ^ " t

JCIIO)'-^ HO) "« ^
o •*-- w ^* Z
L, 4- —• •*- — » O
5> -C > £ >s .* 3
•+- O) >. •»- O) e 15 tn
« — a? ~ Za> 5 —
^ -- » — — * H -o
-t- .O — J3 C -
3I.A 31.1O C -

•+- V) 3 O 0)3 — C
u w) u x a a
(ft 6 O E 4> . u Z
§0. — i_ a — trtl K\ —
0.0 
-------







&
£

-J
Ul
*0
z £
UJ ?
SSu,
srs
U. 5 M
i£|
y =! _,
« § 0
S 3 "
oSE
£ * 
G "J !-
m t- 5
3 < 5
UJ h"
"* «S z
U. S
o ~~ o
*•* (/» w
3 B£
Z . f «
+• o +*
o c
8 i i
« u. -
8e a.
O 3
C L. 
5 * ^
O C C
u> o a
•o •»- in
JS ?
5 is
o -o
0 •• C
£ c a
•»- — 4-
V>
-5 § g
Jf "^ c
Q O O
o » i.
S S 5
= 3
5 £ *
.So
•O 4- O
— O —
o — +-
— 3 X
A O —
t_ — (0
o5 ,5 .<


s
§
UJ
•o
c
to
•t
•^
1
(i
•.
trt
«]
u
1
5
b
«
i
^ i
a a.
0 „.
C/> K*
^>
•a vo
"o
"• 8
<
.. z
5 •*-
^ i
g .
-U
0
t_
• a
£ S
c S
0
+- c
s °
5 i;
>
» s
— c
— o
0 0

1. 13
*•
Ul •
4 ?
O *

U) (A
*I
°o
£ s
£
§ i
F— o
•t- +-
Q U)
a. >»
— «
(A
(A -t-
5 I
* Q)
CO C
OJ «
• (D
CD U
* 3
ac •»-

* 3
.C O
(A —
J™
§
u









a
in
p*
I
04
V.
£
UJ
4-
I
|
4)
>
O
•^
O
c

0
•^
3
cn

i

(0
3
i
a
•

o
>
•o*
c
ID
!
u
?
"^
I
4-
3
1
L.
O
>-
a
X
*o


o
t_
4-
1
Ol



-------
              log  KQC  =  1.00  (log  KQW)  -  0.21                          (6-11)


         where  KQW = octanol-water distribution coefficient (cm3 g"1)
                KQC = organic carbon distribution coefficient (cm3 g"1)

    Carbofuran is a pesticide that exhibits crystalline energy relationships
and its apparent distribution coefficient should be estimated using its log
KQW, which is 2.44.  Substituting into the Karickhoff equation

            log KQC = 1.00 (2.44) - 0.21 = 2.23

                KQC = 102'23 = 169.8

For a soil with 0.5% organic carbon the K^ of the pesticide is

            v  _ v    percent organic carbon                        ,,-
            Kd = Koc          100                                   (b
            K  _ 169.8 (0.5) _ n fl,
            ^d ~     100     ~ U'B5

This compares to an estimated KJ of 2.68 using the PCMC1 water solubility
model.  Selected pesticides having properties suitable for use with the
octanol water distribution model by Karickhoff are provided in Table 6-21.

    DKRATE--degradation rate constants—The processes that contribute to
pesticide disappearance in soils are varied and depend on environmental
factors as well as chemical properties.  Unfortunately, only rarely are
process-specific rate constants (e.g., hydrolysis) reported for the soil
environment.   In most cases, a lumped first-order rate constant is
reported.  A first order model using a lumped decay rate is used in TFAT.
Although such an approximation is imprecise, most modeling efforts follow
the same approach and many pesticides appear to behave similarly.  For
example, Nash (1980) found that disappearance of many compounds was highly
correlated to a first order approximation with r  > 0.80.

    The dissipation rate of pesticides below the root zone is virtually
unknown.  Several  studies have suggested the rate of dissipation decreases
with depth; however, no uniform correction factor was suggested between
surface/sub-surface rates.  First order dissipation rates for selected
pesticides in the root zone are also tabulated in Table 6-21.

    UPTKF--plant uptake of pesticides--The plant uptake efficiency factor,
or root reflection coefficient (UPTKF) provides for removal of pesticides by
plants and is a function of the crop root distribution and the interaction
of soil, water, and the pesticide.  Several approaches to modeling the
uptake of nutrients/pesticides have been proposed ranging from process
models that treat the root system as a distribution sink of known density or

                                  221

-------
TABLE 6-21.  OCTANOL WATER DISTRIBUTION COEFFICIENTS (LOG KQW) AND SOIL
                 DEGRADATION  RATE  CONSTANTS  FOR  SELECTED  CHEMICALS
Chemical Name
                                          Degradation Rate
                                          Constant (days'1)
Reference
Alachlor
Aldicarb
Altosid
Atrazine
Benomyl
Bifenox
Bromacil
Captan
Carbaryl
Carbofuran
Chloramben
Chlordane
Chloroacetic Acid
Chloropropham
Chloropyrifos
Cyanazine
Dalapon
Dial if or
Diazinon
Dicamba
Dichlobenil
Dichlorofenthion
2,4,-Dichlorophenoxy-
acetic Acid
Dichloropropene
Dicofol
Dinoseb
Diuron
Endrin
Fen i troth ion
Fluometuron
Linuron
Malathion
Methomyl
Methoxychlor
Methyl Parathion
Monolinuron
Monuron
MSMA
Nitrofen
Parathion
2.78
0.70
2.25
2.45
2.42
2.24
2.02
2.35
2.56
2.44
1.11
4.47
-0.39
3.06
4.97
2.24
0.76
4.69
3.02
0.48
2.90
5.14

2.81
1.73
3.54
2.30
2.81
3.21
3.36
1.34
2.19
2.89
0.69
5.08
3.32
1.60
2.12
-3.10
3.10
3.81
0.0384
0.0322

0.0149
0.1486
0.1420


0.1196
0.0768

0.0020

0.0058

0.0495
0.0462

0.0330
0.2140
0.0116


0.0693


0.0462
0.0035

0.1155
0.0231
0.0280
02.91

0.0046
0.2207

0.0046


0.2961

- 0.0116

- 0.0063
- 0.0023



- 0.0768
- 0.0079

- 0.0007

- 0.00267


- 0.0231

- G.0067
- 0.0197
- 0.0039


- 0.0231


- 0.0231
- 0.0014

- 0.0578

- 0.0039
- 0.4152

- 0.0033


- 0.0020


- 0.0046
a
a

a
a
a


a
a



d

c
d

a
a



d


d
d

a
c
a
a

a
a

d


a
                                222

-------
   TABLE  6-21.  OCTANOL WATER DISTRIBUTION COEFFICIENTS  (LOG KQW) AND SOIL

                   DEGRADATION RATE CONSTANTS FOR SELECTED CHEMICALS

                                    (CONCLUDED)
                             Degradation Rate

Chemical Name                Log Kowb       Constant (days  )    Reference
Permethrin
Phorate
Phosalone
Phosmet
Picloram
Propachlor
Propanil
Propazine
Propoxur
Ronnel
Simazine
Terbaci 1
Terbufos
Toxaphene
Trifluralin
Zineb
2.88
2.92
4.30
2.83
0.30
1.61
2.03
2.94
1.45
4.88
1.94
1.89
2.22
3.27
4.75
1.78
0.0396
0.0363


0.0354
0.0231
0.693
0.0035


0.0539


0.0046
0.0956
0.0512

- 0.0040


- 0.0019
- 0.0139
- 0.231
- 0.0017


- 0074



- 0.0026

e
a


a
d
d
d


a


e
a
a
    a Nash, R. G. 1980.  Dissipation Rate of Pesticides from Soils.
Chapter 17.  IN CREAMS:  A Field Scale Model for Chemicals, Runoff, and
Erosion from Agricultural Management Systems.  W. G. Knisel, ed.  USDA
Conservation Research Report No. 26.  643pp.

    b Smith, C. N.  Partition Coefficients (Log KQW) for Selected
Chemicals.  Athens Environmental Research Laboratory, Athens, GA.
Unpublished report, 1981.

    c Herbicide Handbook of the Weed Science Society of America, 4th ed.
1979.

      Control of Water Pollution from Cropland, Vol. I, a manual for
guideline development, EPA-600/2-75-026a.

    e Smith, C. N. and R. F. Carsel.  Foliar Washoff of Pesticides (FWOP)
Model:  Development and Evaluation.  Accepted for publishing in Journal of
Environmental Science and Health - Part B.  Pesticides, Food Contaminants,
and Agricultural Wastes, B 19(3), 1984.
                                   223

-------
strength to empirical  approaches  that  assume  a relationship  to  the
transpiration rate.   Oejonckheere et al.  (1983)  reported  the mass of  uptake
into sugar beets for the pesticides aldicarb  and thiofanox for  three  soils
(sandy loam, silt loam,  and sandy clay loam).   Mass  removal  expressed as a
percentage of applied  material  for aldicarb on sandy loam, silt loam, and
clay loam ranged from  0.46-7.14%, 0.68 -  2.32%,  and  0.15  - 0.74%,
respectively.  For thiofanox,  2.78 - 20.22%,  0.81 -  8.70%, and  0.24 2.42%
removals were reported for the  respective soils.  The amount of uptake was
higher for sandy soils and increased with available  water.   Other reviews
have suggested ranges  from 4 -  20% for removal by plants.

    The procedure adopted for TFAT estimates  the removal  of  pesticides by
plant uptake based on  the assumption that uptake of  the pesticide  is
directly related to the transpiration  rate.   Sensitivity tests  conducted
with PRZM indicate an  increase  in the  uptake  by plants as the root  zone
depth increases, and as the partition  coefficient decreases.  For  highly
soluble pesticides and for crop root zones less than 120 cm, the model
simulates total uptake within the range reported by  Dejonckheere et al.  For
highly soluble pesticides and for crop root  zones of greater than  120 cm,
values of greater than 20% were simulated.  For initial estimates  a value of
1.0 for UPTKF is recommended.   If more than  20 - 25% of the  pesticide is
simulated (to be removed by plant uptake), UPTKF should be calibrated to a
value less than 1.0.

    The value of UPTKF is utilized even if the PLTGRN module is on.  The
mass flux resulting from the value of  UPTRF  selected is used as input to the
plant translocation model (see Section 6.5).

    CORED--thickness of soil column—The user will want to enter a value so
that the biologically active portion  of the  soil is  simulated.   For field
crops and grasses, this will be 100  to 200 cm.

    DISP--dispersion coefficient--The  dispersion or "smearing out"  of the
pesticide as it moves down in the soil profile is attributed to a
combination of molecular diffusion and dispersion.  The transport equations
solved in PRZM also produce truncation error  leading to a purely
mathematical or numerical dispersion.   For this reason the  DISP parameter
must be evaluated in light of both "real" and "numerical" components.

    A number of sensitivity simulations using PRZM have been performed  to
investigate the impact of model parameters other than DISP  on the apparent
dispersion. From these simulations,  the following guidance  is offered.

       •  A spatial step or compartment size of 5.0 cm will mimic observed
         field  effective  dispersion quite  well and should be used as  an
         initial  value.
                                    224

-------
       • No fewer than 30 compartments (parameter NCOM2) should be used.

       • The DISP parameter should be set to 0.0 unless field data are
         available for calibration.

    APD. ARM. IAPYR—pesticide application—The use of PRZM requires the
establishment of a pesticide application procedure.  The user should follow
the two steps described below in establishing representative application
dates:

       • establish an application period window covering the range of
         possible application dates

       • adjust the application dates within the window so that application
         does not occur on a day immediately before, during, or immediately
         after a rainfall event (pesticides are not normally applied to a
         field with high moisture content or under conditions where the
         efficacy would be diminished).

6.4.1.4  Soils Parameters—
    The amount of available moisture in the soil is affected by such
properties as temperature and humidity, soil texture and structure, organic
matter content, and plant characteristics (rooting depth and stage of
growth).  The moisture remaining in a soil after "gravity drainage" has
ceased is known as field capacity.  The moisture content in a soil below
which plants cannot survive is called the wilting point.  The wilting point,
which varies among specific soils is influenced by colloidal material and
organic matter, but most soils will have a similar wilting point for all
common plants.

    The PRZM model simulates soil water retention in the context of these
bulk soil properties.  Drainage of "excess water" is simulated as a simple
daily value or as a daily rate.  Most specific model parameters can be input
directly by the user and some can be internally estimated given certain
related soil properties as inputs.

    THEFC. THEWP—moisture holding capacity—Field capacity (THEFC) and
wilting point (THEWP) are required as user inputs.  Often these soil-water
properties have been characterized and values can be found from soils data
bases.  Where such data are not available, one of the three estimation
methods given below can be used.  Method one requires the textural
properties (percent sand, silt, and clay), organic matter content (%), and
bulk density (g cm"3) of a specific soil.  Method two utilizes a soil
texture matrix for estimating soil water content if only the sand (%) and
clay (%) contents are known.  Method three provides mean field capacity and
wilting points if only the soil texture is known.


                                   225

-------
Method 1 (also done within the code if THFLAG = 1)

   The regression equation from Brakensiek and Rawls (1985)  to is used
estimate the matric water potential for various soils:

 GX = a + [b x SAND(%)]  + (c x CLAY(X)] + [d x ORGANIC  MATTER(%)] +
     [E x BULK DENSITY (g cm"3)]


where       0X   = water retention cm3 cm~3 for a given matric
                   potential (field capacity = -0.33 bar and
                   wilting point = -15.0 bar)
            a-e  = regression coefficients

   Step 1. From Table 6-22 find the matric potential for field capacity
           and wilting point (-0.33 bar and -15.0 bar).

   Step 2. For each matric potential, find the regression coefficients
           (a-e) that are required in the Rawls and Brakensiek equation
           (e.g., for -0.33 potential, coefficients a-e are 0.3486,
           -0.0018, 0.0039, 0.0228, and -0.0738).

   Step 3. For any given soil (example:  Red Bay Sandy Loam where sand
           (%), 72.90; clay (%), 13.1; organic matter (%), 0.824; and
           bulk density (g cm~3), 1.70) solve the equation for the
           -0.33 and -15.0 potential.  For this example, THEFC = 0.170,
           THEWP = 0.090.

Method 2

   Use Figure 6-4 for estimating the field capacity and Figure 6-5 for
estimating the wilting point of any soil, given the percent sand and
clay.

     Step 1. Example:  Red Bay Sandy Loam (field capacity).  Find the
             percent  sand across the bottom of Figure 6-4 (i.e., 73.0)

     Step 2. Find the percent clay of  the soil along the side of the
             triangle (i.e., 13.0).

     Step 3. Locate the point where the two values  intersect on  the
             triangle and read the field capacity,  THEFC = 0.17.

     Step 4. Follow Steps 2-4 for wilting point using Figure 6-5.
             THEWP =  0.09.
                              226

-------
    Method 3

       Step 1. Use Table 6-23 to locate the textural  class of the soil  of
               choice.

       Step 2. After locating the textural  class,  read the mean field
               capacity and wilting point potentials  (cm3 cm"3), to the
               right of the textural  class.  Example:  Sandy loam.  The mean
               field capacity (THEFC) and wilting  point (THEWP) potentials
               are 0.207 and 0.095, respectively.

    Guidance for estimating distributional  properties for THEFC and THEWP is
given in Tables 6-24 and 6-25.   These tables show  the arithmetic means  and
coefficients of variation for Hydrologic Groups A, B, C and D soils with
depth.  Also shown is the type  of distribution which  is most appropriate.
    TABLE 6-22.  COEFFICIENTS FOR LINEAR REGRESSION EQUATIONS FOR PREDICTION
                   OF  SOIL WATER  CONTENTS AT SPECIFIC MATRIC POTENTIALS4
Matric
Coefficient
-0.20
-0.33
-0.60
-1.0
-2.0
-4.0
-7.0
-10.0
-15.0
Intercept
a
0.4180
0.3486
0.2819
0.2352
0.1837
0.1426
0.1155
0.1005
0.0854
Sand
b
-0.0021
-0.0018
-0.0014
-0.0012
-0.0009
0.0007
-0.0005
-0.0004
-0.0004
Clay
(*)
c
0.0035
0.0039
0.0042
0.0043
0.0044
0.0045
0.0045
0.0044
0.0044
Organic
Matter
/ QJ \
\ /
d
0.0232
0.0228
0.0216
0.0202
0.0181
0.0160
0.0143
0.0133
0.0122
Bulk
Density
(g cm'3)
e
-0.0859
-0.0738
-0.0612
-0.0517
-0.0407
-0.0315
-0.0253
-0.0218
-0.0182
R2
0.75
0.78
0.78
0.76
0.74
0.71
0.69
0.67
0.66
    a Rawls, W. J., U.S. Department of Agriculture, Agricultural Research
Service, Beltsville, MD. Personal Communication.
                                     227

-------
   100-1
                    0.55
                          0.50
                                0.45
       0.5% Organic matter
       0.0% Porosity change
                                     0.40
                                          0.35
                                              0.30
                                                    0.25
                                                          0.20
                                                               0.15
                     I   1   I   '    I   '   I
                    20    30     40    50
 I
60
                                                                     0.10
T  T   I1   I   '   I
70     80    90    100
                                  Sand  (%)
   Figure 6.4  1/3-Bar soil moisture by volume.  (Provided by Dr.  Walter J. Rawls,
             U.S. Department of Agriculture, Agricultural  Research Service,
             Beltsville, Maryland.)

Jury (1985) indicates overall  CV for wilting point water content  (15  bar
tension) to be lower, at 24 percent.  He also indicates that  the most
appropriate distribution for static soil properties such as these  is  the
normal.

    Correlation coefficients between fixed capacity and wilting point have
moderate to high values ranging from 0.64 to 0.85  (Carsel et  al.  1988).

    BD--bu1k density and field saturation—Soil bulk density  (BD)  is
required in the basic chemical transport equations of TFAT and is  also used
to estimate moisture saturation values.  Values for BD can be input
directly.  When such data are not available for the site of interest,
methods have been developed for their estimation.  Two methods are provided
for estimating BD of various soils.  Method one requires the  textural

-------
properties (percent sand, clay, and organic matter).  Method  two  uses  mean
bulk density values if only the soil texture is known.  The following  steps
provide procedures for estimating bulk density.

Method 1  (Also done within the code if BDFLAG  = 1)

   A procedure from Rawls (1983)  is used to estimate  bulk  density for
any given soil, provided the percent sand, clay,  and  organic  matter
contents are known.  Example:  Marlboro fine sandy  loam—sand 80.0%,
clay 5.0%, and organic matter 0.871%.  Using the  following equation:
   100
                        0.40
                              0.35
 0.5% Organic matter
0.0% Porosity change
                                    0.30
                                          0.25
             10    20    30    40    50    60    70    80    90    100
                                                0.20
                                                        0.15
                                                             0.10
                                                                     0.05
                                  Sand  (%)
    Figure 6.5  15-Bar soil moisture by volume.   (Provided by Dr. Walter J.  Rawls,
              U.S.  Department of Agriculture,  Agricultural  Research  Service,
              BeUsville, Maryland.)
                                    229

-------
                   TABLE 6-23.  HYDROL06IC PROPERTIES BY SOIL TEXTURE3
Range of
Textural Properties
(Percent)
Texture
Class Sand Silt Clay
Sand 85-100 0-15 0-10
Loamy 70-90 0-30 0-15
Sand
Sandy 45-85 0-50 0-20
Loam
Loam 25-50 28-50 8-28
S1lt Loam 0-50 50-100 0-28
Water Retained at
-0.33 Bar Tension
on3 cm"3
0.091b
(0.018 - 0.164)c
0.125
(0.060 - 0.190)
0.207
(0.126 - 0.288)
0.270
(0.195 - 0.345)
0.330
Mater Retained at
-15.0 Bar Tension
cm3 cm"3
0.033b
(0.007 - 0.059)c
0.055
(0.019 - 0.091)
0.095
(0.031 - 0.159)
0.117
(0.069 - 0.165)
0.133
                                               (0.258 - 0.402)        (0.078 - 0.188)

Sandy Clay    45-80      0-28     20-35            0.257                 0.148
  Loam                                         (0.186 - 0.324)        (0.085 - 0.211)

Clay Loam     20-45     15-55     28-50            0.318                 0.197
                                               (0.250 - 0.386)        (0.115 - 0.279)

Silty Clay     0-20     40-73     28-40            0.366                 0.208
  Loam                                         (0.304 - 0.428)        (0.138 - 0.278)

Sandy Clay    45-65      0-20     35-55            0.339                 0.239
                                               (0.245 - 0.433)        (0.162 - 0.316)

Silty Clay     0-20     40-60     40-60            0.387                 0.250
                                               (0.332 - 0.442)        (0.193 - 0.307)

Clay           0-45      0-40     40-100           0.396                 0.272
                                               (0.326 - 0.466)        (0.208 - 0.336)


    a Rawls, W. J., D. L. Brakensiek, and K. E. Saxton.   Estimation of Soil  Mater
Properties.  Transactions ASAE Paper No. 81-2510, pp.  1316 - 1320.  1982.

    " Mean value.

    c One standard deviation about the mean.
                                        230

-------
    TABLE 6-24.  DESCRIPTIVE STATISTICS AND DISTRIBUTION MODEL FOR FIELD
                       CAPACITY (PERCENT BY VOLUME)
Original Data
Stratum
(n)
Class A
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class B
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class C
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class D
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Sample
Size

52
50
42
39

456
454
435
373

371
362
336
290

230
208
178
146
Mean

11.8
9.6
7.3
7.1

19.5
18.8
18.7
17.5

22.4
22.8
22.7
22.2

24.1
26.1
25.0
24.1
Median

9.4
8.1
5.9
5.8

19.1
18.8
18.7
17.5

22.5
23.2
22.9
21.3

24.2
26.3
25.6
24.4
s.d.

9.2
7.9
5.8
5.0

8.3
7.4
7.1
7.6

7.8
7.8
8.6
8.9

9.1
9.3
8.2
8.1
cv
(*)

78
82
79
70

42
39
39
43

35
34
38
40

38
36
33
33
Distribution Model
Transform

In
In
In
In

SU
SU
w
SU
\j
su

su
w
su
u
su
*J
su

su
w
su
w
su
\i
su
Mean

2.25
1.99
1.73
1.73

0.316
0.311
0.298
0.288

0.363
0.369
0.368
0.359

0.387
0.419
0.403
0.390
s.d.

0.65
0.73
0.73
0.71

0.13
0.12
0.11
0.12

0.12
0.12
0.13
0.13

0.14
0.14
0.13
0.12
s.d. = standard deviation
cv   = coefficient of variation

Source:  Carsel et al. (1988)
                                    231

-------
   TABLE 6-25.   DESCRIPTIVE STATISTICS AND DISTRIBUTION MODEL FOR WILTING
                         POINT (PERCENT BY VOLUME)
Original Data
Stratum
(m)
Class A
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class B
0.0-0.3
0.3-0.6
0.6-0.9
0.6-1.2
Class C
0.3-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class D
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Sample
Size

118
119
113
105

880
883
866
866

678
677
652
582

495
485
437
401
Mean

4.1
3.2
2.9
2.6

9.0
9.4
9.1
8.6

10.8
12.2
12.2
11.8

14.6
16.9
16.6
15.7
Median

3.1
2.3
2.1
1.9

8.7
9.3
8.9
8.4

10.4
12.1
11.9
11.5

13.8
17.0
16.3
15.1
s.d.

3.4
2.4
2.3
2.3

4.0
4.3
4.4
4.6

5.1
5.6
6.0
5.7

7.6
7.3
7.4
7.6
cv
Distribution Model
(%) Transform Mean

82
75
81
87

45
46
48
53

48
46
49
48

52
43
45
48

In
In
SB
VJ
SB

SU
SU
SU
w
su

su
w
su
su
su

su
su
\f
su
su

1.83
0.915
3.32
3.43

0.150
0.156
0.151
0.143

1.63
0.202
0.201
0.194

1.26
0.277
0.271
0.257
s.d.

0.64
0.71
0.88
0.92

0.066
0.071
0.072
0.076

0.62
0.091
0.096
0.092

0.76
0.12
0.12
0.12
s.d. = standard deviation
cv   = coefficient of variation

Source:  Carsel et al. (1988)
                                    232

-------
              BD =
	100.0
%OM   +   100.0 - %OM
OMBD          MBD
(6-13)
    where    BD  = soil bulk density,  g cm"3
             OM  = organic matter  content of soil, %
             OMBD  = organic matter bulk density of  soil,  g cm~3 = 0.224
             MBD = mineral bulk density, g cm"3

             NOTE:  MBD must be entered if BDFLAG =  1.

       Step  1.  Locate the percent  sand (80.0) along  the bottom of Figure  6-
                6.
   100-i
    90-
_o
o
    Figure 6.6  Mineral  bulk density (g cm  ).  (Provided by Dr. Walter J. Rawls,
              U.S. Department of Agriculture, Agricultural Research  Science,
              Beltsville, Maryland.)
                                   233

-------
       Step 2.  Locate the percent clay (5.0)  along the side of Figure 6-6.

       Step 3.  Locate the intersection point  of the two values and read the
               mineral bulk density (1.55).

       Step 4.  Solve the Rawls equation for BD (e.g.,  1.47).

    Method 2

    Step 1. Use Table 6-26 to locate the textural  classification of the
            soil.

    Step 2. Read mean bulk density for the general soil texture.  Example:
            Sandy loam.   The mean bulk density is  1.49 g cm  .

    Table 6-27  shows distributional properties of  bulk density
information. The information given is categorized by  Hydrologic Soil Group
(A, B, C, D).   The most  appropriate distribution for this property is the
normal (Jury 1985).  Jury indicates  slightly lower CVs, on the order of 9
percent.

    OC--percent of soil  organic matter—Guidance on estimating percent
organic matter  is found  in Table 6-28.  Information is categorized by
Hydrologic Soil Group and by depth.  Also shown are coefficients of


         TABLE 6-26.  MEAN BULK DENSITY (g cnT3) FOR FIVE SOIL TEXTURAL
                                CLASSIFICATIONS3
Soil Texture
Silt Loams
Clay and Clay Loams
Sandy Loams
Gravelly Silt Loams
Loams
All Soils
Mean Value
1.32
1.30
1.49
1.22
1.42
1.35
Range
0.86
0.94
1.25
1.02
1.16
0.86
Reported
- 1.67
- 1.54
- 1.76
- 1.58
- 1.58
- 1.76
     aBaes, C.  F.,  Ill  and R.  D.  Sharp.   1983.   A Proposal  for Estimation of
 Soil Leaching  Constants  for Use  in Assessment  Models.   J.  Environ.  Qual. 12(1)
 17-28.

                                     234

-------
        TABLE 6-27.  DESCRIPTIVE STATISTICS FOR BULK DENSITY (9 CM"3)
Stratum
(m)
Class A
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class B
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class C
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class D
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Sample
Size

40
44
38
34

459
457
438
384

398
395
371
326

259
244
214
180
Mean

1.45
1.50
1.57
1.58

1.44
1.51
1.56
1.60

1.46
1.58
1.64
1.67

1.52
1.63
1.67
1.65
Median

1.53
1.56
1.55
1.59

1.45
1.53
1.57
1.60

1.48
1.59
1.65
1.68

1.53
1.66
1.72
1.72
s.d.

0.24
0.23
0.16
0.13

0.19
0.19
0.19
0.21

0.22
0.23
0.23
0.23

0.24
0.26
0.27
0.28
cv
«)

16.2
15.6
10.5
8.4

13.5
12.2
12.3
12.9

15.0
14.5
14.2
14.0

15.9
16.0
16.3
17.0
 s.d. = standard deviation
 cv   = coefficient of variation

 Source:  Carsel et al. (1988)
 variation for each soil group and depth.  Carsel et  al.  (1988)  determined
 that the Johnson SB distribution provides the best fit to  this  data.

    Rao and Wagenet (1985) and Nielsen et al.  (1983)  have reported that
these values are often normally distributed.  Carsel  et al. (1988) have
noted that organic carbon is weakly correlated with field capacity and
wilting point water content with correlation coefficients ranging from 0.1
to 0.74.  Strength of correlation decreases  with depth.

    AD—soil water drainage rate (for HSHZT  =  1)—The HSWZT flag indicates
which drainage model  is invoked for simulating the  movement of recharging
water.   Drainage model 1 (HSWZT = 0)  is for  freely  draining soils; drainage
                                    235

-------
model 2 (HSWZT =1) is for more poorly drained soils.  For soils with
infiltration rates of more than 0.38 cm hr~l (associated with SCS hydrologic
soils groups A, B, and some C), setting HSWZT = 0 is recommended.  For soils
with infiltration rates of less than 0.38 cm hr"1 (associated with groups D
and some C) setting HSWZT = 1 is recommended.

    The drainage rate parameter (AD), required when HSWZT = 1, is an
empirical  constant and dependent on both soil type and the number of
compartments to be simulated.  Although there is limited experience using
this option, an analysis was performed to determine the best value for AD
   TABLE 6-28.  DESCRIPTIVE STATISTICS AND DISTRIBUTION MODEL FOR ORGANIC
                         MATTER (PERCENT BY WEIGHT)
Stratum   Sample
(m)        Size
          Original  Data
                             Distribution  Modela
Mean
Median
s.d.
                                                 'var
(X)
Mean
s.d. = standard deviation
cv   = coefficient of variation
a Johnson SB transformation  is used for all cases  in this table

Source:  Carsel et al.  (1988)

                                     236
s.d.
Class A
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class B
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2
Class C
0.0-0.3
0.3-0.6
0.3-0.9
0.9-1.2
Class D
0.0-0.3
0.3-0.6
0.6-0.9
0.9-1.2

162
162
151
134

1135
1120
1090
1001

838
822
780
672

638
617
558
493

0.86
0.29
0.15
0.11

1.3
0.50
0.27
0.18

1.45
0.53
0.28
0.20

1.34
0.65
0.41
0.29

0.62
0.19
0.10
0.07

1.1
0.40
0.22
0.14

1.15
0.39
0.22
0.15

1.15
0.53
0.32
0.22

0.79
0.34
0.14
0.11

0.87
0.40
0.23
0.16

1.12
0.61
0.27
0.21

0.87
0.52
0.34
0.31

92
114
94
104

68
83
84
87

77
114
96
104

66"
80
84
105

-4.53
-5.72
-6.33
-6.72

-4.02
-5.04
-5.65
-6.10

-3.95
-5.08
-5.67
-6.03

-4.01
-4.79
-5.29
-5.65

0.96
0.91
0.83
0.87

0.76
0.77
0.75
0.78

0.79
0.84
0.83
0.88

0.73
0.78
0.82
0.86

-------
over a range of soil types on which agricultural  crops are commonly grown.
Each of three soil types was tested with  a  constant soil  profile depth (125
cm).  The profile was divided into a variable  number of compartments and the
optimum value of AD for each soil/compartment  combination was obtained.

    The analysis was performed by comparing the  storage of water in the soil
profile following the infiltration output from SUMATRA-1  (van Genuchten
1978).  This model was used as "truth"  because field data were lacking and
SUMATRA-1 is theoretically rigorous.  The amount  of water moving out of the
profile changed by only 1-2% over the range  of  compartments tested (15 -
40) for the three soils evaluated.  Calibrating  PRZM by comparison was
accomplished and estimates of AD calculated.   Suggested values of AD for
clay loam, loamy sand, and sand as a function  of  the number of compartments
are given in Figure 6-7.
                 2.8-
                 2.4-
             I

             -o  2'°'
             o'
                 1.6-
                 1.2-
                    15
20
35
        25       30

Number of compartments
            Figure 6.7  Estimation of drainage rate AD  (day ) versus number
                      of compartments.
                                    237

-------
 6.4.2  Infiltration and Ponding

     KSAT—saturated hydraulic conductivity—This parameter represents the
 limiting infiltration rate when the soil  column is saturated and suction
 pressure is not longer important.   KSAT depends upon soil mineralogy,
 texture, and degree of compaction.   Ranges of values for various
 unconsolidated materials are shown  in Table 6-29.  Note that these values
 are given in ms   whereas TEEAM requires units of cm hr  .  KSAT has also
 been correlated with SCS Hydrologic Soil  Groups (Brakensiek  and  Rawls 1983):
 ranges of values for each soil  group are  shown in Table  6-30.  '

     For Monte Carlo simulation, it  has often been observed that  the
 distribution of hydraulic conductivities  in both surface soils and
 groundwater aquifers is approximately log-normal (Freeze 1975; Willardson
 and Hurst 1965). Jury (1985) gives  values for the coefficient of variation
 of 120 percent.  Table 6-31 gives distributional properties  for  various soil
 textures.

     HFPOND--suction parameter--HFPOND represents water movement  due to
 suction in unsaturated soils, and has units of length  (cm).  As  in  the case
 of KSAT, HF pond has been correlated with SCS Hydrologic Soil Groups
 (Brakensiek and Rawls 1983); ranges of values for each soil  group are shown
 in Table 6-29.

    TABLE 6-29.   REPRESENTATIVE  SATURATED  HYDRAULIC CONDUCTIVITY  RANGES
                         FOR SEDIMENTARY MATERIALS
Material
                       Saturated
                       Hydraulic
                      Conductivity
                        (ms-1)
Material
                  Saturated
                  Hydraulic
                 Conductivity
                      -1
Clay
Silty
Sandy
Silty
Sandy
Silt
Silt
Loam
Sandy

clay
clay
clay loam
loam sand

loam
10
10
10
10
10
10
10
-12 _
-12 _
-11 _
-10 _
-9 _
-9 _
-9 _
io-9-
loam
10
-8 _
lO-9
io-9
1Q-8
io-7
icr6
io-6
10'6
io-6
IO-7
Very fine sand
Find sand
Medium sand
Coarse sand
Gravel and sand
Gravel
Sandstone
Limestone
Shale
10
10
10
10
10
10
10
10
10
-7 _
-6 _
-5 _
-5 _
-5 _
-5 _
-6 _
-7 _
-7 _
10
10
10
10
10
10
10
10
10
-4
-3
-3
-2
-2
-2
-3
-4
-4
* Excluding cavernous limestone.
Source:  Adapted from Todd (1970).
                                    238

-------
       TABLE 6-30.  VALUES OF GREEN-AMPT PARAMETERS FOR SCS HYDROLOGIC
                                  SOIL  GROUPS
SCS
Hydro "log ic
Soil Group
A
B
C
0
Saturated Hydraulic
Conductivity KS
(cm/hr)
1.0 - 10.0
.60 - 1.0
.20 - .60
.005 - .20
Suction
Parameter HF
(cm)
10
10 - 20
15 - 10
20 - 150
  Source:  Adapted from Brakensiek and Rawls (1983).
    TC--time of concentration for runoff--This parameter represents the time
it takes for runoff to travel from the hydraulically most distant part of a
catchment (in this case, habitat) to the point of discharge.  In hydrograph
analysis, TC is the time from the end of rainfall excess to the point on the
falling limb of the hydrograph where the recession curve begins (Mockus
1972).  TC is a physical characteristic of the catchment, and will depend
upon the shape of the catchment, ground slope, drainage density, and surface
roughness.  A number of methods for estimating TC are described in the Soil
Conservation Service Hydrology Handbook (Mockus 1972).  The following
empirical equation described in the SCS Handbook can be used for areas of
less than 2000 acres:
                  1  -».«O/r .  1\.
                  l^Z* - (L+l) -                                (6-14)
                      1900 YU'D
where
    TC = Time of concentration (hours)
     a = Hydraulic length of the catchment (ft)
     S = Watershed retention parameter

    CN = SCS runoff curve number for the catchment
     Y = Average catchment slope (percent)

Other guidance for estimating TC can be obtained from local flood control
districts and Soil Conservation Service Offices.
                                     239

-------
         TABLE 6-31.  DESCRIPTIVE STATISTICS FOR SATURATED HYDRAULIC
                        CONDUCTIVITY (KSAT) (cm hr'1)
Soil Type
Clay**
Clay Loam
Loam
Loamy Sand
Silt
Silt Loam
Silty Clay
Silty Clay Loam
Sand
Sandy Clay
Sandy Clay Loam
Sandy Loam
Hydraulic Conductivity (KSAT)
x s CV
0.20
0.26
1.04
14.59
0.25
0.45
0.02
0.07
29.70
0.12
1.31
4.42
0.42
0.70
1.82
11.36
0.33
1.23
0.11
0.19
15.60
0.28
2.74
5.63
210.3
267.2
174.6
77.9
129.9
275.1
453.3
288.7
52.4
234.1
208.6
127.0
n
114
345
735
315
88
1093
126
592
246
46
214
1183
 *  n = Sample size, x = Mean, s = Standard deviation, CV = Coefficient of
    variation (percent)
 ** Agricultural  soil, less than 60 percent clay
 Source:  Carsel  and Parrish (1988).
    ARAIN,  BRAIN—constants in the rainfall-duration curve—ARAIN and BRAIN
are used to relate rainfall amount to rainfall  duration.   These parameters
depend upon storm patterns, topography,  and season.   Estimates of these
constants can be obtained from rainfall-duration data by  plotting the
logarithms  of rainfall  amount (cm) against the  logarithms of duration
(hours).  BRAIN can then be estimated as the slope of the best fit line to
these data, while ARAIN can be estimated as the inverse logarithm of the
intercept.   Dean (1979) found the values of these parameters shown below for
winter and  summer storms in Watkinsville, GA:
                                     240

-------
         Season   ARAIN     BRAIN R2

         Winter   3.25      1.470.599
         Summer   1.61      1.300.525
For instance, for a typical summer storm at Watkinsville, if the depth is
1.5 inches (3.81 cm), the average duration is

       DUR = 1.61 (3.81)1'30
           = 9.2 hours

These values are probably representative for convective-type summer storms
and frontal -type winter storms occurring in most parts of the U.S. east of
the Rocky Mountains.

6.4.3  Volatilization and Pond Chemistry

    PAIR— vapor phase diffusion coefficient--The diffusion coefficient is
defined by Pick's first law as the proportionality between the chemical flux
and the spatial gradient of concentration.  Although in theory DAIR will be
chemical and temperature dependent, Jury et al . (1983) found that the vapor
diffusion coefficient showed little variation among pesticides and
recommended a value of 4,300 cm2 day'1 for all pesticides.  Thibodeaux and
Scott (1985) calculated values ranging from 3900 to 7800 cm2day"1 for
12 benchmark chemicals exhibiting a broad spectrum of characteristics.
Included among these were DDT and chlorpyrifos with diffusion coefficients
of 4000 and 3900 cm2 day"1 respectively.  Diffusion coefficients can also be
estimated as a function of temperature and molecule size using techniques
described by Lyman et al . (1982).

    DWAT--mo1ecular diffusion coefficient in water--DWAT represents the rate
of diffusion of dissolved chemical through water, and is a function of
temperature, viscosity,  and molecule size.   Water diffusion coefficients are
generally on the order of four orders of magnitude smaller than vapor
diffusion coefficients.   Lyman et al. (1982)  present a number of estimation
methods but recommend the Hayduk-Laudie equation due to its simplicity and
relative accuracy:
                            _5
             DWAT =  - U4 - -                                (6-15)
where
                             (
                               D
    nw = the viscosity of water (centi-poise)
    Vg = the Lebas moal volume of the chemical  (cm  molecule")

Vg can be estimated as described by Lyman et al.  (1982).

                                    241

-------
    KH--Henry's law constant—Henry's law constant,  used here in
dimensionless form, relates the equilibrium concentration of pesticide vapor
to the concentration of the water dissolved phase.  Table 6-32 is a
compilation of a number of experimentally determined values of KH for
various chemicals.  If KH has not been measured for a chemical, it can be
estimated as the ratio of the saturated vapor density to the solubility in
water:
                                 C
                            KH = -^-                                (6-16)

where

    Cy = saturated vapor density for the pesticide (mg a  )
    CL = water solubility of the pesticide (mg a~ )

    KDPOND--decay rate constant for pesticide in ponds--KDPOND is a lumped
first order rate constant used to represent chemical transformation
processes such as microbial degradation and hydrolysis.  This term does not
represent volatilization or infiltration losses since these processes are
modeled explicitly by the TFAT ponding algorithm.  At this time, there are
few data on decay of chemicals in ponds, and KDPOND should only be used to
model losses which cannot be explained by volatilization or infiltration.

6.4.4  Granular Formujations

    GKWET and GKDRY—wet and dry decay rate constants for granular
pesticide--GKWET and GKDRY are rate constants which determine the release of
pesticide from granules applied to the soil surface, and have units of days"
 .  The values of these parameters will vary depending upon the chemical
properties of the pesticide and the structure of the granules.  As an
initial guide for these parameters, Table 6-33 lists overall measured rate
constants for granules under various lab and field conditions.  GKWET is the
rate constant for granules immersed in water, and reflects moisture-
dependent processes such as diffusion, leaching, and biodegradation.  This
parameter can be estimated from plots of immersed granule pesticide
concentration against time.  If the half-life of pesticide in immersed
granules is known, the wet rate constant can be estimated as follows:
            GKWET = f                                              (6-17)
where t]/2 is the time in days at which 50 percent of the pesticide mass has
been released from the granules.  Half-lives and decay rate constants for
water- immersed granules can gften be obtained from the pesticide
manufacturer, since laboratory immersion tests are commonly performed on new
granule formulations.  The dry rate constant GKDRY is a lumped parameter
which is intended to reflect release processes occurring when granules are
dry, such as volatilization.  Volatilization rate constants can be estimated

                                     242

-------
TABLE 6-32.  ESTIMATED VALUES OF HENRY'S CONSTANT FOR
                 SELECTED PESTICIDES

Compound
Alachlor
Aldrin
Anthracene
Atrazine
Bentazon
Bromad 1
Butyl ate
Carbaryl
Carbofuran
Chlopyrifos
Chrysene
Cyanazlne
DDT
Diazinon
Dlcamba
Dleldrin
Diuron
Endrin
EPTC
Ethoprophos
Fenitrothion
Fonofos
Heptachlor
Lindane
Llnuron
Malathlon
Met homy 1
Methyl Parathion
Metolachlor
Metrlbuzln
Monuron
Napropamlde
Parathion
Permethrin
Picloram
Prometryne
Simazine
Terbufos
Toxaphene
Trial! ate
Trichlorfon
Trifluralin
2, 4-D (acid)
2, 4, 5-T (acid)
References:
A Donigian et al. (1986)
B Spencer et al . (1984)
Henry's constant
(dimensionless)
1.3E-06
6.3E-04
4.4E-05
2.5E-07
2.0E-10
3.7E-08
3.3E-03
1.1E-05
1.4E-07
1.2E-03
4.7E-05
1.2E-10
2.0E-03
5.0E-05
3.3E-08
6.7E-04
5.4E-08
1.8E-05
5.9E-04
6.0E-06
6.0E-06
2.1E-04
1.7E-02
1.3E-04
2.7E-06
2.4E-06
4.3E-08
4.4E-06
3.8E-07
9.8E-08
7.6E-09
7.9E-07
6.1E-06
6.2E-05
1.9E-08
5.6E-07
1.3E-08
1.1E-03
2.3E+00
7.9E-04
1.5E-09
6.7E-03
5.6E-09
7.2E-09

C Jury et
D Schnoor

Ref.
A
D
D
A
A
C
A
A
A
A
D
A
C
C
A
C
C
D
C
C
B
A
D
B
A
B
A
A
A
A
C
C
C
A
B
C
A
A
A
C
B
A
A
B

al. (1984)
et al. (1987)
                       243

-------











l/»
LU
'
^3
Z
1
i
QC
Ll_
LU
a
1—4
U
1— >
t*
LU
a.
°
LU
LU
ee.

Q£
2
i—

h^
Z
8

LU
Jj
"*
a
LU
ae
1

00
(ft
1
to
LU
_j
5
i—

































O)
u
1










>l 1 "l»
1— 1
v»
J3
13

^_,
C
o
0
ut
c
o
•r-
-M
•^
•a
o
°



u
•a
•r-
o

*•>
a*
a.
tO *O to to
00 00 OO CO
ov o^ ot 01
«— 1 T— 1 «— 1 r-t

c c c c
i i i i
•a 10 id «
.c .c .c .c
0 U 0 0
T? "S "S T?

8-5 1 -S
^J *^B ^J
• c id c
•a o -t-> o
cu c u
« o
VI -O O -O
4) (U
C 10 C VI
ft) O 41 O
CLr™ CL r—
O U O U
c c c c
••- 1- ••- a) •»- a»
O) t- J-
•o -o »- -o 3 -a 3
0) O) 3 Ol 4-> 0) 4->
4-< •(-> 4-> 4-> VI 4^ I/I
id  4->
id id a*   at
01 4^
4^ ^3
^^ ^~ ^1
c_i ^3 ai ^—
alS-^ a)
E ^~* t*— f~
<4-
•o • •

ai id 
a. CL ecu
J. fc. i- 


















^•^t ^^^ ^^«s
oo oo oo
i— i <— i i— i
~_» ^- s_»
I/I VI V)
C C C
*r— *r- »r—
XL JX J^

t~ "^ *T™ '
222


VO
CO OO CO
tO  •«-
•f— «^ r—
e E J.
B tt *^^
~ o 8
8t- a t. ^r t.
a» o cui-i a>
LO 4-> <-l 4-> 1 4->
i id i id o 
CO C LO C --I «=
C C C
i- -a i- -o T- -a
Q) Q) ft)
at vi ai vi ai 1/1
•a v- ~o i- xi *-
•r- Q) *r~ QJ *r- ft)
4^ f- 4-> "r- 4^ *r-
VI VI I/I
Q. CO CL VI CL  ai ai
<«- •— <4- i— 14- r—
030303
o> 
-------
by comparing release rates in open and closed containers (Chapman and
Chapman 1986).   Release rates in open containers will include volatilization
of pesticide vapor, whereas volatilized vapor will be trapped in closed
containers.

6.4.5  Soil Surface Temperature Regression Coefficients

    In order to estimate soil surface temperatures from air temperatures
from the equation

        Ts = A  + BTa                                                (6-18)

the regression  coefficients A and B must be given values.  Guidance is
provided from two data sets in the literature.  Baver et al. (1972) present
the data of Smith (1932), which show the monthly variation of air and
surface soil temperatures at a 6-inch depth.  Representative data are also
given by Gibbs  et al. (1980).  The data of Gibbs et al. represent 10-year
averages (1969-1978) for plots having bare soil and sod cover as surface
treatments.  Regression of soil temperatures on air temperatures give the
coefficients shown in Table 6-34.  Coefficients of determination (R )
indicate an excellent linear fit for both data sets.   Intercept (A) values
range from 0.6° to 14.5°C, whereas slope values (B) ranges from 0.82 to
1.5.  Warmer surface soil temperatures are associated with warmer average
air temperatures.  Therefore, higher values of the intercept should be used
in areas having higher air temperatures.  The effect of a cover on the soil
is to damp temperatures fluctuations by providing insulation.  Therefore,
soil surface temperatures will be warmer under a vegetative cover when air
temperatures are cold and cooler when air temperatures are warm.  Wetter
soils will tend to be less subject to variation than dry soils because water
increases the soil's heat capacity and thermal conductivity (Baver et al.
1972).  Therefore, higher-valued intercepts and slopes for the regression
equation would  be more appropriate for better-insulated soils.

6.5  PLTGRN PARAMETERS

    The plant growth module (PLTGRN) of TEEAM is an adaption of .the EPIC
plant growth model (Williams et al. 1988).  The parameters to define plant
growth can, for the most part, be estimated using the guidance provided in
the EPIC documentation (Williams et al. 1987).  Typical ranges (as suggested
by Williams et  al. 1987) for these plant parameters are presented in Table
6-35.

    The EPIC documentation includes the parameters for defining 69 crop
types.  Parameters for defining 11 of these crop types are provided in Table
6-36.  The values listed in this table are useful starting values.  When
used in a specific modeling scenario, it is recommended that the values be
adjusted (calibrated) to represent observed conditions.

                                    245

-------
     TABLE 6-34.  EXAMPLE REGRESSIONS OF SURFACE SOIL TEMPERATURE ON AIR
                                 TEMPERATURE
Location
California
A B R2
14.5 1.5 0.97
Ta
44
 Geneva, NY
Bare Soil
Soil Cover
0.6
1.6
1.1
0.82
0.98
0.97
13
13
6.6  PLTRNS MODULE

    The required inputs for the plant translocation are few, and with the
exception of a few coefficients, can be derived from readily available
information.  Generalized discussions and review of much of the available
literature on adsorption and translocation of pesticides (herbicides) in
plants can be found in Ashton and Crafts (1981).

6.6.1  RW—Root Reflection Coefficient

    Nash (1974) reviewed the literature on uptake of pesticides by plants to
that date.  He concluded that, below molecular weights of 500, plants tend
not to discriminate between organic molecules except on the basis of
polarity, and the single most important factor in determining uptake appears
to be solubility.  Therefore, one might suspect that the reflection
coefficient would vary proportionately with  solubility or inversely  with
KQW.  This observed effect may be due to decreased availability of dissolved
phase chemical due to adsorption onto soil  materials.  McFarlane et  al.
(1987) concluded that the uptake rates of bromacil, phenol,  and nitrobenzene
were quite different, even though their KQW values are comparable (1.49  to
2.02).  Increases in the rate of transpiration stimulated uptake of  bromacil
and nitrobenzene, (indicating the utility of the transpiration-based uptake
model) but had no effect on the uptake of phenol.  However,  the log  of the
average rate constant for uptake was highly inversely correlated (R2=0.98)
with the Kow of the chemical.  This suggests that chemicals  with low KQW
should have reflection coefficients closer to unity and chemicals with
higher KQW would have coefficients closer to zero.  This is  consistent with
chromatographic theory of transport of organic chemicals in  media containing
water and an immobile organic phase and with the observations of Briggs
et al. (1983).  More discussion for this parameter can be found under the
UKTKF parameter (Section 6.4.1.3).
                                    246

-------
6.6.2  LAMDA--Deqradation Rate of the Contaminant Within the Plant

    Briggs et al. (1982) measured the degradation of some 0-
methylcarbamoyloxime chemicals and substituted phenylureas in plant
shoots.  First order degradation constants (day~*) were reported.  These
values are shown below:
             TABLE 6-35.  PLANT GROWTH PARAMETERS, TYPICAL RANGES
Variable
PHU
BE
FGK
TOPT
TBASE
Description
Potential heat units
Biomass-energy ratio
Harvest index
Optimal temperature
Base temperature
Units
°C-day
kg/ha MJ
decimal
°c
°c
Typical
Range
a
10-50
0.01-0.95
10-30
0-12
 LAIMX


 FLAI



 KHGT

 HGTMX

 RZ
(minimum temperature
 for plant growth)

Maximum leaf area            decimal
i ndex

Fraction of growing          decimal
season where leaf area
starts to decline
Plant height increase rate  m  day

Maximum crop height             m

Maximum root depth             cm
                                  -1
0.5-10.0


0.4-0.99



    b

0.1-3.0

 50-300
 a  Can be estimated for the plant as the sum of the average daily
    temperature minus the plant base temperature (TBASE) over the period
    from planting to harvest.

 b  Best estimate is maximum height divided by length of period from
    emergence to crop maturity.
                                    247

-------
             Chemical                                        Rate (day l)
     Oxamyl 0-Methylcarbamoyloxime                               0.39
     Benzaldehyde 0-Methylcarbamoyloxime                         0.48
     4-Chlorobenzaldahyde 0-Methylcarbomoyloxime                 0.36
     3,4-Dichlorobenzaldehyde 0-Methylcarbomoyloxime            0.39
     3-Phenoxybenzaldehyde 0-Methylcarbomoyloxime                0.29
     4-Phenoxyphenylurea                                         0.69

     A striking feature of these values is  their similarity.   All  of the
 other substituted phenylureas tested  had degradation rates too slow to
 measure over the time scale of the experiments  (24-48 hours).  The same was
 true for aldicarb and aldicarb sulfone (aldoxycarb).   Menzie (1980) provides
 an annotated bibliography of papers which  may contain quantitative
 information about the degradation of  pesticides in  plants.  Fletcher et al.
 (1985) also describes the PHYTOTOX database.  This  database  contains
 information from over 3500 publications concerning  the effects of herbicides
 on plants and may also contain quantitative  information on the degradation
 of chemicals in plants.

               TABLE 6-36.  PLANT GROWTH PARAMETERS, CROP SPECIFIC  VALUES

                          Grain
       Soybean  Alfafa  Corn Sorghum  Wheat  Barley  Oats  Sunflower Cotton Peanut  P1ne
BE
FGK
TOPT
TBASE
LAImx
FLAI
HGTmx
RZ
25
0.31
25
10
5
0.9
1.5
200
20
0.25
20
4
5
0.9
1.25
200
40
0.50
25
8
5
0.8
2.5
200
35
0.50
27.5
10
5
0.8
1.5
200
47
0.42
15
0
8
0.8
1.2
200
35
0.42
15
0
8
0.8
1.2
200
35
0.42
15
0
8
0.8
1.2
200
25
0.31
25
10
5
0.75
2.5
200
17.5
0.50
27.5
12
5
0.85
1.0
200
20
0.42
25
13.5
5
0.75
2.0
200
11.5
0.76
25
2
5
0.9
20
150
6.6.3  KOW—Octanol Water Partition Coefficient

    The octanol water partitional coefficient is widely reported for many
chemicals, e.g., Rao and Davidson 1980).  Either KQW or solubility  is
required data for the registration of most chemicals and.should be  readily
available.  If solubility, and not K  , is available, relationships  in  Lyman
et al. (1982) can be used to estimate values of the latter.  Values  selected

                                     £48

-------
should be consistent with similar coefficients (Koc, Kd) in other parts of
the model.

6.6.4  KP—Partition Coefficient

    The partition coefficient used in TEEAM describes the partitioning
between the water of the transpiration stream and the nonaqueous organic
phases of the aboveground plant biomass.  The partition coefficient should
be estimated using available data or, .alternatively the KQC of the compound
and the organic carbon content of the aboveground biomass.  Koc may be
estimated from solubility (see Section 6.4.1.3) or from KQW (Lyman et al.
1982).  The organic carbon content (OC) will vary among plant species and
varieties; however, a reasonable approximation of the OC can be made
assuming that the makeup of the vascular plant is similar to that of algae
use in which the fraction of carbon by weight is approximately 0.36.  The
partition coefficient is estimated by multiplying the KQC by the organic
carbon content (OC).

6.6.5  RHONA—Ratio of Dry Weight to Het Weight

    A reasonable assumption, in the lack of more exact data, is that the
living plant is approximately 60 percent water.  Assuming a weight of the
dry matter of 0.2 g cm"3, then the dry weight to wet weight ratio is about
15 percent.

6.7  APUM MODULE

    Guidance for estimating parameters of the APUM module is categorized
into three areas:

       • Species Abundance
       • Animal Movement
       • Feeding, Uptake, and Depuration

6.7.1  Species Abundance

    TEEAM does not utilize or simulate the number of individuals within the
ecosystem; rather, it makes use of the biomass of each species.  Biomass of
each species remains constant over the period of the simulation in the
current version of the model.  Estimates must be provided for each
species.  If it is desirable to obtain a sense of variability of predicted
body burden concentrations, etc., in a species caused by differences in
movement, feeding, or other characteristics of subgroups of the population,
the user may divide the species population into several subgroups.  In this
case, biomass estimates must be provided for each subgrouping.  The user
must also differentiate between the biomass of each species present in each
habitat of the ecosystem which are not habitat mobile.

                                    249

-------
    The most logical way to specify the biomass of each species is to begin
with the end-of-the-foodchain species.  Assume that this individual  (or pair
of individuals) represents the biomass of that species in its territory.
Once the territorial area is known, the biomass of the lower trophic-level
species can be estimated by multiplying the area (L2) by an appropriate
biomass density for each species of concern (M L~2).  Species of concern
could be any of the species which form part of the foodchain for the end-
point species.  Examples of these organisms in the soil and their
prey/predator relationships are shown in Table 6-37.

    Densities for some of the more important species are discussed in the
following sections.

6.7.1.1  Earthworms—Table 6-38 taken from Lofty (1974) shows biomass
estimates for earthworms in soils under a variety of land uses.  Biomasses
range from 1.6 to 287 g m~2.  Higher values are associated with grasslands,
while lower values tend to be associated with woodland and arable land.

6.7.1.2  Enchytraeids—White or pot worms are common larger oligochaetes
which tend to inhabit the top 3 inches or so of the soil if the humidity is
fairly constant.  Their numbers undergo wide swings seasonally (Smith
1966).  Odum (1971) gives density values of 1-3 g nT2 for pasture soils and
7 g m~2 for mor-type soils of Denmark.

6.7.1.3  Microarthropods—These consist primarily of mites and
springtails.  The most numerous are the mites.  Odum (1971) indicates
biomass of 2 to 5 g m~2.  Harding and Studdard (1974) give estimates ranging
from 0.03 to 3.2 g m   for springtails and 0.3 to 5.4 g m   for mites.

 6.7.1.4   Macroarthropods—Macroarthropods  include primarily woodlice,
 millipedes,  termites, fly larvae,  and beetles.  Edwards  (1974) gives a
 biomass  density of  2.1  g m    for woodlice  in grassland of a British wood.
 Edwards  (1974) gives population estimates  for millipedes ranging from 2.3 to
 300 m  .   Using an  assumed  average weight  of 3 mg per  individual, this
 translates to  biomass densities of 0.007 to 0.9 g m"2.   An average for
 woodland  is  probably on the  order  of  0.3 g m   .  Beetles and fly larvae are
 estimated to have  biomass densities on the order of  1.5  g m    (Edwards
 1974).   Termites are estimated  to  have biomass densities of 5  to 50 g nf2.

 6.7.1.5   Molluscs—Mason  (1974) gives  information on the population biomass
 of  snails living in a beech  wood.  Twenty-one  species  were reported, with
 their total  biomass being on the order of  0.7  g m~2  dry  weight.  Estimates
 ranging  from 0.08  to 8.2 g  m~2.  A reasonable  estimate for grasslands
 appears  to be  3 to  7.5  g m~2.
                                     250

-------
   TABLE 6-37.  EXAMPLES OF MICROPHYTIC FEEDERS AND OF CARNIVORES WHICH ACT AS SECONDARY
                  AND TERTIARY CONSUMERS WITHIN OR ON TOP OF THE SOIL
                                                  Carnivores
Microphytlc
Organi sm
Springtails:
Mites:
Protozoa:
Nematodes:

Feeders
Microflora
Consumed
Algae
Bacteria
Fungi
Fungi
Algae
Lichens
Bacteria
and other
microflora
Bacteria
Fungi
Secondary Consumers
Predator Prey
Mites: Sprinqtails
Nematodes
Enchytraeids
Centipedes: Springtails
Nematodes
Snails
Slugs
Aphids
Flies
Moles: Earthworm
Insects
Tertiary Consumers
Predator Prey
Ants: Spider
Centipedes
Mites
Scorpions
Centipedes; Spiders
Mites
Centipedes
(other)
Beetles: Spiders
Mites
Beetles
(other)
Termites:
Fungi
Source:  After Brady 1974.
Note:  In the current version of TEEAM,  soil dwelling organisms may not prey on one
      another.  Degree of uptake is specified through the use of bloconcentration
      factors.
6.7.2  Animal  Movement

    Animal movement input data include  the frequency of movement,  movement
transition matrices,  and initial or  steady state population distributions.
Because these  data are highly species-specific, it is not possible to
provide detailed  steps for parameter estimation.  However, some general
guidance  is provided  in the following paragraphs regarding the types of  data
available and  the interpretation of  these  data for TEEAM input.

6.7.2.1  NMOVE—The Frequency of Animal Movement--
    This parameter is used in the model to specify how often the animal's
location is recomputed within each day.  NMOVE will  probably depend upon how
often the animal  feeds during the day.  For instance, animals which feed
during the early  morning and late afternoon might move between their nesting
locations and  feeding locations four times each day.   This parameter could
also be limited by the animal's rate of movement and  the distance  between
locations.

-------
6.7.2.2  PMVC--The Soil Horizon Transition Matrix for Soil Animals--
    This matrix contains the conditional probabilities that a soil animal
will move to each soil horizon given its initial horizon location.  It is
input only for soil animals which do not move between TEEAM habitats and is
not required for animals modeled using a steady-state population
distribution.  The transition matrix for soil animals will often depend on
environmental conditions such as temperature and moisture.  For instance,
earthworms will generally move deeper into the soil column in response to
extreme surface temperatures (Lee 1985).  When soil horizons become
saturated, earthworms move towards the surface to breathe.  Earthworms may
also exhibit preferences for certain soil types with high organic matter
contents.  The transition matrix should be formulated to represent these
environmental factors and should also reflect the relative locations of the
various soil horizons, their thicknesses, and the travel time of the animal.

6.7.2.3  PMVH--The Habitat Transition Matrix for Higher Animals—
    This matrix contains the conditional probabilities that an animal  will
move to each habitat given its initial habitat location.  It is not required
for soil animals and for animals modeled using a steady-state population
distribution.  The habitat movement matrix should reflect both quantitative
animal movement data and the user's qualitative understanding of the
animal's movement between nesting and feeding habitats.  Table 6-39 lists
habitat information for mallard ducks in Nebraska, and is an example of the
types of data from which habitat transition matrices can be derived.  In
general, these data indicate that mallard ducks prefer to feed in the grazed
corn stubble habitat.  However, distances from the nest to the various
      TABLE 6-38.  POPULATIONS OF EARTHWORMS IN DIFFERENT HABITATS
                            Site
No. m
                                                      -2
gm
                 -2
Arable land
Arable land
Arable land with dung
Fallow soil
Orchard with grass
Orchard with grass
Pasture
Pasture
Under pig litter
Pseudotsuga mor
Mixed woodland
Quercus woodland
Pinus woodland
Bardsey Island, U.K.
Herts., U.K.
Herts., U.K.
U.S.S.R.
Cambs., U.K.
Holland
N. Wales
Westmorland, U.K.
U.S.A.
N. Wales
N. Wales
Hants., U.K.
Hants., U.K.
287
18
79
18.5-33.5
848
300-500
481-524
389-470
960
14.0
157
184
40
76
1.6
39.9
4.6-8.4
287
75-122
112-120
52-110
272
4.7
40
68
17

-------
 feeding habitats  indicate that the ducks  might utilize  the nearest  habitat
 at times.   Information on the times of day at which these animals feed in
 each habitat could be used to further refine the transition matrix.

 6.7.2.4  PHAB--The Distribution of Animal Populations—
     This array contains the fraction of the animal  population located  in
 each habitat (or  soil horizon for soil animals).  It is not required  for
 animals modeled by random subgroup movement.  For populations which  are not
 at steady  state,  PHAB serves as an initial condition for movement
 calculations and  might be used to specify the distribution at the start of  a
 season.  For steady-state populations the input distribution remains
 constant throughout the simulation, and should reflect  the long-term  average
 population distribution.   For instance, if the locations of mallard ducks
 were assumed to be constant, the percent  use data in Table 6-39  could  be
 used as the population distribution.

 6.7.3  Feeding, Uptake, and Depuration

     The user must specify feeding and chemical uptake and depuration  rates
 in order to simulate biomagnification in  the terrestrial  food chain.   The
 specification of  these rates is somewhat  different  for  animals living  in or
 on the  soil  versus higher tropic level animals.

      TABLE  6-39.   USE  OF  VARIOUS  HABITATS  BY MALLARD DUCKS  IN NEBRASKA
                                              Percent      Distance
Habitat                    Availability (%)     Use     from Nest (km)
Corn Stubble
Cultivated Corn Stubble
Plowed Corn Stubble
Grazed Corn Stubble
Feed lot
16.8
2.7
5.1
72.1
0.4
4.6
2.3
1.1
75.9
16.1
3
3.1
1.8
3.8
3.6
Source:  Jorde et al. 1983.
6.7.3.1  Total Food, Water, and Air Intake Rates-
    Food, water, and air intake rates must be specified for each species as
appropriate.  Soil dwelling organisms do not require input of water or air
consumption rates.

                                     253

-------
    Feeding rates have been reported for many of the more common organisms
that live in the soil.  Shown in Table 6-40 are mean intake rates,  range of
intake rates, and food assimilation efficiencies.   Although assimilation
efficiencies are not utilized directly by the model  for soil  dwelling
species, intake rates should be adjusted downward  by multiplying the gross
intake rates by the assimilation efficiency.  It should be recognized that
for these organisms, metabolic rates and therefore food intake rates are
affected by temperature.  Lower soil temperatures  are associated with lower
food intake rates.

    For avians, higher food intake rates, per gram of body weight,  are
associated with smaller species.  Welty (1962) gives food consumption rates
for these species:

       • European Robin (16 grams)        14.7 gg'1 day'1
       • Song Thrush (89 grams)            9.8 gg'1 day'1
       • Quail (170 grams)                 8.8 gg'1 day'1

Dorst (1974) states that avians weighing between 100 to 1000 g consume 5 to
9% of their body weight per day, while those weighing 10 to 100 g consume
from 10 to 30% of their body weight per day.  Little information was found
in the literature concerning the assimilation efficiency of food items by
avians.  Farner et al. (1972) give some information on the absorbability of
fatty acids in the digestive tract of chickens.  These data may be
representative of the assimilation efficiency of food items and pesticides
in the diet of avians:

         Fatty acids                        4  -  95%
         Monoglycerides                    41  - 100%
         Triglycerides                     76  -  96%
         Hydrolyzed Triglycerides          67  -  93%

    Welty (1975) also gives information on the water consumption rates of
birds.  These rates also vary on a function of size with higher utilization
rates per body weight also associated with smaller species as follows:

       • Wren (12 grams)                   0.37 gg'1 day'1
       • Junco (21 grams)                  0.16 gg'1 day"
       • Whippoorwill (40 grams)           0.07 gg'1 day"1
       • Quail (144 grams)                 0.04 gg"1 day"1

Welty (1975) also notes that ground doves that consumed 0.1 gg'1 day"1 of
water at normal temperatures, consumed 0.3 gg'Vday'1 at temperatures of 30
to 40°C.

    Farner et al, (1972) give a regression for the respiration rate (R) of
birds in ml min'1 as a function of body weight (W) in kg.

                                     254

-------

00
5
I-H

o
0
C3
2
_J
^J
LU
a
i— i
0
00
1
o
00
0
t— 1
>
ae
0
LL.
00
LU
i— l
O
LU
O
HH
LU-
LL.
LU
O
h-H
1—
1
-«J
1— 1
HH
00
oo
^c
a
z
00
LU

eg
111
U-J
O
LU
CO
r^
"™








(U
0
(U

OJ
14-
cu
Of.


c *- ^*
o
issimilati
:fficienc>
ej_ LU

T— 1
id o>
•0 0)
rH Id
t OC
CT
CT
(U
id
a;
id
c
,, f»
•a
2:

















E
(O
E

u
0
LO
00
1— 1
,rH rH
-M
n-  a»
_i _j _j



00
c\]
CO
I— (


CO
O vO
O 0
-M •»->
rH ^~
C3






«— '
O

TJ
"^
U

J_
•g
13
^«— X^"
(/)
ai 4->
•r- CO
U O)
(U t-
0. 0
CO 14-

J= C
id id
c o
 c a.
E id O
i. > $-
O Id 3
5 CO LU
JC 	
^«)
^_
id
LU















VO
+1
o
LO


O
O
O
rH
O
o




LO
CsJ
0
O

















u
3
r—
r~
Q
Z




^r ^r
en en
rH rH
IO CO
S. S-
id id
3 2
LU LU



s s
1 3


o
C3
LO
O
O
O





rH CO
0 0
O O















(/)
(U
TJ (U
0) U
a. i-
•r- i—
r- -O
i— O
*r~ O
S. 3




•3-  1
°
O




LO
CM
O CO
O O


CO
a>
I—
id -M
t- (U
0) (U
-f-> CD
a.
O i-
0) O
"o -a
c_> c
id
j-
O (U
'^ id
C l-
id id
CO c —
a> id
4-> J- >>
E -4-> LL!

(U -r-
H- a




!^
~C. rH
id
•o
c id
•r~ +J
•a 4->
J- 3
id 4J
m co



s


8
C3
VO
0
o






VNJ
O











,*-•*
CO
ai
'i


10
•J3
 J-
c id
•r- 4->
-a -i->
t» ^
z: co



i


VO
O
0
00
rH
O
0





^f
rH
O





*—*
(/I
^~
•^

-------
         log R = log 284 + 0.77 log W                               (6-19)

    No information was found concerning the trapping efficiency of
pesticides in the avian lung.

6.7.3.2  Food Preferences--
    Food preference factors must be established for each species
simulated.  A generalized food web for soil dwelling species was shown in
Table 6-37, based on information found in Brady (1974).  For the purposes of
this model, the uptake of pesticides by soil-dwelling organisms is based on
biocentration factors.  Soil-dwelling organisms do not feed on others.  In
reality, earthworms and millipedes feed on detritus and soil.  Snails and
slugs will feed on both detrital material at the soil surface and also on
living plant matter.  Centipedes may feed on snails and slugs and other
centipedes.  Beetles may feed on mites and other beetles, while ants may
feed on centipedes and mites.  Avians are allowed to feed on other organisms
and might feed on ants, centipedes, beetles, earthworms, snails and slugs,
mites, millipedes and plant matter, directly.  The user has the perogative
as to which species will be simulated and what their food preferences will
be.

    An abundance of food preference information is available for most common
avians.  Food preferences may vary widely with season, sex, habitat, and
availability of food items.  Frequently, for females, consumption of animal
material in the diet increases during nesting season.  Some general guidance
for passerines (esp. T. migratorius), mallard ducks and bobwhite quail is
given in the following discussion.

    The American robin (T. migratorius) has a diet which consists, on the
average, of 60% plant matter and 40% animal matter (Bent 1949).  The plant
consists chiefly of wild berries and fruits; seeds do not comprise a large
portion of the  diet.  During winter, the diet consists almost exclusively
of plant material.  Animal materials consist of beetles, caterpillars,
hymenoptera, flies, and grasshoppers.   According to Bent, an eastern robin's
diet was observed to consist of the following:
                                    256

-------
         Food Item                         Frequency of Occurrence (%}
         Plants:                                        81.5

           Bar berry                                    61
           Sumach                                       29
           Coral berry                                   4.5

         Animals:                                       93.5
           Beetles                                      82.5
           Millipedes                                   38.5
           Ants                                         27.0
           Cutworms                                      9.5
           Sowbugs                                       6.5
           Wireworms                                     4.0
           Flies                                         3.0
           Cockroaches                                   1.5

Earthworms are also taken from pastures arid lawns.

    Western robins in Utah between April and July were observed to have the
following food preferences:

         Food Item                         Percentage in Diet

         Alfalfa weevil                                  14
         Cutworms                                        23
         Click beetles                                   11
         Earthworms                                       9
         Flies                                            6
         Dung beetles                                     6
         Ground beetles                                   4

Beetles made up 54% of the diet in April, but only 13% from November to
April.

    Swanson et al. (1985) give the data on food preferences for mallard
ducks shown in Table 6-41.  Food preferences of mallards wintering in south
central Nebraska (Jorde et al. 1983) were even more dominated by plant
matter:
                                  257

-------
     TABLE 6-41.  PROPORTION BY VOLUME (%) OF PLANT AND ANIMAL FOODS
       IN THE ESOPHAGI OF MALLARDS COLLECTED DURING SEASONS OF
                1974-80 IN SOUTH CENTRAL NORTH DAKOTA
Food
Total Animal
Gastropoda
Insects
Coleoptara
Lepidoptera
Crustacea
Oligochaeta
Total Plant
Seeds
Vegetation
Roots/tubers
Male
n = 39
37.6
6.3
16.8
0.5
—
11.3
—
62.4
56.4
6.0
4.1
Nonlaying
Female
n = 41
37
4.5
22.6
2.5
1.5
7.5
—
63
58.5
4.5
3.9
Laying
Female
n = 37
71.9
16.4
27.1
4.8
2.8
12.9
11.8
28.1
24.8
3.3
2.8
          Source:  Swanson et  al.  (1985)
         Food Item

         Plant:
           Seeds
             Corn
             Milo
           Polygonum spp.
           Vegetation
             Lemna minor

         Animal:
           Mollusks
           Insects
Percentage by Weight

              97

              46
               2
              11

              16
               3
               3
    In general, information on the habits and feeding preferences  of ducks
may be found in the following references:

Belrose, F.C.  1976.  Ducks, Geese and Swans of North America.   Stackpole
    Books.  Harrisburg, PA.  543 pp.

                                  258

-------
Martin, A.C. and P.M. Uhler.  1939.  Food of Game Ducks in the United States
    and Canada.  U.S. Dept. of Interior, Fish and Wildlife Service Research
    Rpt. No. 30.  308 pp.

McAtee, W.C.  1918.  Food Habits of the Mallard Ducks of the United
    States.  U.S. Dept. of Agriculture Bull. 720.  35 pp.

6.7.3.3  Depuration Rates--
    Depuration or clearance rates vary widely with chemical and species.
While higher clearance rates would appear to be associated with higher
metabolic rates and hydrophilic compounds, exceptions appear in the
literature.  A summary of the clearance rates found in the literature for a
wide range of species and compounds is shown in Table 6-42.  Wide variation
is found for the carbanate insecticides, even within the same species.
Clearance rates for the more hydrophobic chlorinated hydrocarbons are, in
general, lower, with the exception of clearance rates for dieldrin
calculated from data for wood thrushes reported by Jeffries and Davis
(1968).  The more hydrophobic hexachlorobenzene apparently cleared slugs
faster than 2,4-D.

    Clearance rates can be calculated for soil dwelling species using the
bioconcentration factor concept.  The concentration in a soil dwelling
species can be modeled using:
                 ' KtCtVt - K mCoMo
in which Kt is the uptake rate from the soil,
         Cj- is the total soil concentration,
         Vj. is the volume of soil having concentration, C^.,
         Km is the metabolic clearance rate, and
         C0 is the concentration of pesticides in the organism.

or
           dC    KfC.V.
           dT = -lip - KmCo                                      <
assuming that the organism mass M0 does not change over time.  At steady
state:
           KC M
           -Tp - Vo                                            <6-22'
Rearranging terms:
           Ktvt   co
                •                                                   (6-23)
                                   259

-------

LU
C£
1—

Q£
LU
1—
i— i
i

LU
3:
^«—

z
»— *
a
z
=3
0
u_

co
UJ
o
III
UfcJ
a.
CO
_J
«^
^fc
1— 1
z

CO
o
s
^
z
1— 1

CO
LU
a
o
i— i
CO
O-
co
o
H- I
l*v*
^£
^>

g
u_

co
LU
1
i
^^
c^
t—t
l__
i
~^i
O-
III
LLJ
a


*
CM
<5j"
1
VO

LU
_J
CO


o
i-
o
a

•a
c
id

-M
t-
a>
J3
o
0£







id id
co *r
00 O
0 0



LO l»x
o
CM







CU
T3
•r™
X
O
<4_
•^
to c
id
JO J-
S- 3
0 0
r— Id
< 0












E
S-
0
2

i-
id
LU






























id i
mid
o vo



r-. i-x
O 0
CM CM






^
id
X
O

T^j
c
id
c c
(O (O
s- u

o o
J- i.
id id
c_> o


















































d
LOId
O LO
0 ^H



r> r~.
0 0
CM CM














C C
tO (v
>- 4-
3 3
0 0
i- S-
id id
t 3 C 1




CO
•r-
i-
-t-> id
co -a

S- CU
ai o
I— U_

« •
— 1 LU

















*~~*
LO
CO
»— 1
x^

CU
CU
_J





id >
^f
t
LU
CO
1^



0
CO
en
A
^"














C
*
T3
"cu
•r—
O
















































d id
^f ^r
i i
LU LU
co cr>
VO LO



O 0
P**- CO
CO O
ft ft
r- o
LO




w
-a

x
o
a.
cu
s-
o



OJ 0
3: a























CO
CO
en
t— i


o^
c
•T"
f*^
LU

"^
c
id
(U
3
CT
id
•T*^






JO

8S
o o



VO O
t-H O
«r o
ft
o
8
^




cu
c
cu
N
£=
cu
J3
O
s-
o

a u
i «
•a- x

CM :E

















3

CO
^__^
CO
VO
r- 1
N»*

CO
T"

id
a

^^
c
id

CO
cu
•r—
^
j|
*l
CU









JD
CO
,_,



O
co

«t
«^-














c

•a
"cu

a










Si
CO
3
$_
J=
1—

•§
O
3



«o"
00
o^
i-H


*
r~~
id

+j
cu

E
2
1 *
(/)
^
o






id
^O
CM
O
O
0



O
CO

M
^1














C

"cu

0









f—
f~m
3
O

O>
C

i.
cu




'•d*^—. j?x-v
^* CO ^" CO
OO • CO •
en co en co
T— 1 . I— 1 •

c «
• vo • vo
r— r—
id c id c
0 O
-4-> -t- -»-> -i-
CU -4-* 
O O
i— CU r— CU
CU CO CU CO

U CU O  CO +•> CO






id id
id LO r-^
co coid r-v
0 0 0 O
0000


















^^
CU LO
C CM
id»-i
T3
$- $-
O O
X JC J=
cu u u
LU 1- >, Id
a -t- x i-
o s: o «*














cu

v/
o
id
$-
0
















































«
ai
u
c
C
cu

(U
J-
0)
cu -M
(j
c c
O) -r-
ai 

id jo
260

-------
    Since the bioconcentration factor BCF is defined as the concentration in
the organism over that in the soil:
           M.
            ±± = BCF                                                (6-24)
            mo
    Furthermore, the term Vt/MQ can be written as AZ/P  where AZ is the
depth of soil having concentration Ct (m) and p  is the organism density
(§m  )> so that 6-24 can be rewritten as:
           K.AZ
           r±- = BCF                                               (6-25)
           Vo
    Using 6-25, and knowing the intake rate, the metabolic clearance rate,
Km, can be chosen to yield field observed bioconcentration factors.
Bioconcentration factors for a variety of pesticides in earthworms range
from less than 1 to 10 with the norm being about 5.  Bioconcentration
factors for slugs are somewhat higher ranging from 1 to 100 (Thompson and
Edwards, 1974).  The data of van Gestel and Ma (1987) give a range of
bioconcentration factors for various chlorinated phenols in earthworms of
0.4 to 5.3.  No bioconcentration factors were found in the literature for
other soil animals.

    Some  specific references on accumulation and clearance rates for avians
appear below.  Not all of these references have been reviewed by the authors
of this report.

Bailey, S., P.J. Bunyan, B.D. Rennison and A. Taylor.  1969.  The Metabolism
    of 1,1-di  (chlorophenyl)-2,2-di-chloroethylene and l,l-di(p-
    chlorophenyl)-2-chloroethylene in the Pigeon.  Toxicol. Appl.
    Pharmacol.  14:23.

Baldwin, M.D., J.W. Crawford, D.H. Hutson and D.L. Street.  1976.  The
    Metabolism and Residues of (  C) Endrin in Lactating Cows and Laying
    Hens.  Pesticide Sci.  7:575-594.

Charnetski, W.A.  1976.  Organochlorine Insecticide Residues in Ducklings
    and Their Dilution by Growth.  Bull. Environ. Contam. Toxicol.  16:138-
    144.

Cummings,  J.G., M. Eidelman, V. Turner, D. Reed, K.T. Zee and R.E. Cook.
    1967.  Residues in Poultry Tissues from Low Level Feeding of Five
    Chlorinated Hydrocarbon Insecticides to Hens.  J. Assoc. Off. Anal.
    Chem.  50:418-425.

De Vos, R.H., J. Bruman and A.B. Engel.  1972.  Residues of Organochlorine
    Pesticides in Broilers from Feed Fortified with Known Levels of These
    Compounds.  Pesticide Sci.  3:421-432.

-------
Fries, G.F., R.J. Lillie, H.C. Cecil and J. Bitman.  1977.  Retention and
    Excretion of Polychlorinated Biphenyl Residues by Laying Hens.  Poultry
    Sci.  56:1275.

Haseltine, S.D., M.T. Finley and E. Cromartic.  1980.  Reproduction and
    Residue Accumulation in Black Ducks Fed Toxaphene.  Arch. Environ.
    Contain. Toxicol.  9:461-471.

Heinz, G.H. and R.W. Johnson.  1979.  Elimination of Endrin by Mallard
    Ducks.  Toxicology.  12:189-196.

Hicks, B.W., H.W. Dorough and R.B. Davis.  1970.  Fate of Carboforan in
    Laying Hens.  J. Econ. Entomol.  63:1108-1111.

Kan, C.A. and J.C. Jonker-Den Rooyen.  1978.  Accumulation and Depletion of
    Some Organochlorine Pesticides in High-Producing Laying Hens.  J. Agric.
    Food Chem.  26:935-940.

Stadelman,  W.J., B.J. Liska,  B.E. Langlois,  G.C. Mostert and  A.R.  Stemp.
    1965.  Persistence of Chlorinated Hydrocarbon Insecticide Residues  in
    Chicken Tissues and Eggs.   Poultry Sci.   44:435-437.

Stikel, L.F.,  W.H.  Stikel,  R.D.  McArthur and D.L.  Hughes.   1979.   Chlordane
    in Birds:   A Study of Lethal  Residues and Loss Rates.   jjn:   Toxicology
    and Occupational  Medicine.   W.B.  Deichman (organizer),  p.  387-396.
    Elslevier/North Holland,  N.Y.

Stikel, W.H.,  J.A.  Galyen,  R.A.  Dyrland and D.L. Hughes.  1973.   Toxicity
    and Persistence of Mirex in Birds.  In:   W.B.  Deichman (ed.)  Pesticides
    and the Environment:  A Continuing Controversy,   p.  437.  -
    Intercontinental  Medical  Book Corp., NY.

Stikel, W.H.,  L.F.  Stikel,  R.A.  Dyrland and D.L. Hughes.  1984a.   DDE in
    Birds:   Lethal  Residues and Loss Rates.   Arch. Environ. Contam.
    Toxicol.  13:1-6.

Stikel, W.H.,  L.F.  Stikel,  R.A.  Dyrland, and D.L. Hughes.   1984b.  Arochlor
    1254 Residues in Birds:  Lethal Levels and Loss Rates.  Arch. Environ.
    Contam. Toxicol.   13:7-13.

6.7.3.4  LD50, LD10 - Lethal  Dosages-
    Little information was retrieved for lethal dosages for either soil
dwelling organisms or avians.   Lee (1985) provides an extensive table of LD
values of various chemicals to earthworms.  Unfortunately, the units given
are kg a.i. ha   and are incompatible with the units required by TEEAM.  The
information necessary to compute an LD50 based on dosage per gram of the

-------
predator are not available in Lee.  However, the reference given for the
data, Ruppell and Laugh!in (1976), may contain more information.  Lee does
give some LD50 information for some fungicides (benomyl, carbendagin,
thiabendazole).  The LD50 for these chemicals is reported as 10 yg/worm.
Given an individual biomass of about 1 gram, the LD50 would be 10 yg g    or
10   yg mg~ .  The original reference for this data is Wright (1977).

    The LD50, LD10 parameters are important to the operation of the TEEAM
code in order to produce reasonable body burden concentrations in upper
trophic level species.  The population of the species is not modeled, and
therefore mortality due to pesticide accumulation of individals of the
species cannot be accounted for.  Because of this limitation, the species
will accumulate pesticides, based on the difference between intake and
depuration rates, approaching a quasi-equilibrium level.  This level may be
much greater than normally would be observed because lethal effects would
eventually limit the intake.   Therefore, the LD values are used to adjust
the intake rates so that unreasonable values of body burden concentrations
are not simulated.  In lower trophic level species (i.e., soil dwelling
animals) this is not a problem because the depuration rates would normally
be calculated from the bioconcentration factor (BCF).  Therefore, the
correction is not applied to soil dwelling animals.

    Lee (1985) also reports on the work on Hill et al. (1975) who determined
lethal dietary concentrations to bobwhite quail, Japanese quail, ring-necked
pheasants and Mallard ducks to be 311 to 1869 yg g'1 for DDT, 37 to 169
yg g   for dieldrin and 92-480 yg g   for heptachlor.  Barker (1958)
reported poisoning of robins to be associated with dietary concentrations of
53-204 ug g  .  Collett and Harrison (1968) found that^evels of DDT taken
as food by blackbirds and thrushes were 13 to 29 yg g~ .  Again, the units
of these LC50s are incompatible with the LD50 required by the model.  The
model LD50 is the 50 percent mortality lethal dosage for unit weight of the
bird.  Information required to convert to these units (i.e., the food intake
and weight of the birds) is not available in Lee but may be available in the
original references.

6.8  SENSITIVITY OF TEEAM TO INPUT PARAMETER VALUES

    A sensitivity analysis was'performed on TEEAM to determine which input
parameters have the strongest effect on model outputs.  The results of this
analysis give insight into the effects of uncertainty and should provide
some guidance to the user on where to concentrate data collection efforts.
The methodology and results used in the TEEAM sensitivity analysis are
summarized in the following paragraphs.

-------
6.8.1  Sensitivity Analysis Approach

    The base data set for this sensitivity analysis  is the  peanut  field
ecosystem discussed in Section 8.  The target  species is  the American
Robin.  These birds are exposed to pesticides,  in  this example,  through
ingestion of soil and water, inhalation of vapors, and predation on
earthworms.  The base simulation has a duration of thirty days,  with
pesticide applied to soil on the first day of  simulation.

    The sensitivity of this system to input  parameters was  evaluated by
running TEEAM in its Monte Carlo mode for 500  runs.  Probability
distributions for input parameters were assumed based on  best  estimates  of
their range of uncertainty; Table 6-43 lists the various  input parameters
and their assumed probability  distributions.   The  analysis was  performed  for
two cases:  1)  uncorrelated inputs  and  2)  correlated  inputs.  For case  2,
correlations were initially assigned  as  high  (.7),  medium  (.5), or  low  (.2),
and then adjusted by trial  and error  until  a  positive-definite  correlation
matrix was found.  The correlation  matrix  is  shown  in Figure 6.8.  Two
•utput parameters were evaluated:   1) maximum 5-day average  dosage  to the
•arget species  and 2)  maximum  5-day average concentration  of pesticide  in
the target species.

    The TEEAM Monte Carlo simulation  of  the robin-peanut  field  ecosystem
provided 500 independent realizations of  dosage and concentration output
derived from randomly selected input  data.   Sensitivity  of these  model
outputs was then evaluated by  stepwise  regression  analysis.   In stepwise
regression, the importance of  each  input  variable  is evaluated  by comparing
the sum of squares due to the  regression  to the sum of  squares  due  to
residual errors, as quantified by an  F-statistic.   In the  first step,  the
input variable with the highest correlation coefficient  is added  to the
regression model and tested for significance.  Other variables  are  brought
into the regression and tested for  significance until no  more variables can
be added to the regression model.   This  results in a subset  of  input
variables which contribute significantly  (at a specified  confidence level)
to the regression equation which best estimates model outputs.   A 95%
confidence level was used for  all  F-statistic significance tests  of TEEAM
variables.

6.8.2  Sensitivity Analysis Results

    Table 6-44 lists the variables  which were most significant in
determining pesticide dosage to the target species.  Also listed  are the
individual correlation coefficients.   These variables collectively explain
96% of the variance in pesticide dosage.   The three most important  variables
in this case were the pesticide application rate,  the bioconcentration
factor for earthworms, and the total  daily feeding rate for the  robins.


                                     264

-------
Table 6-43.  INPUT PARAMETERS USED IN SENSITIVITY ANALYSES AND THEIR
             ASSUMED DISTRIBUTIONAL PROPERTIES
Parameter
Soil Bulk Density
Adsorption Partition Coefficient
Wilting Point Water Content
Field Capacity Water Content
Soil Hydraulic Conductivity
Decay Rate in Soil
Decay Rate on Foliage
Henry's Law Constant
Pesticide Application Rate
Root Reflection Coefficient
Decay Rate in Plants
Octanol Water Partition Coefficient
Runoff Curve Number
Bioconcentration Factor for
Earthworms
Metabolic Degradation Rate in
Target Species
Total Feeding Rate
Soil Preference Factor
Pond Water Ingestion Rate
Air Inhalation Rate
Units Distribution
g cm
cm3 g-1
cm3 cm'3
cm3 cm"3
cm hr
days'
days'1
cm cm
kg ha"1
Dimensionless
days"1
cm3 g-1
Dimensionless
Dimensionless

days'1

g g'1 day'1
gg-1
1 mg day
1 mg'1 day'1
Normal
Log Normal
Normal
Normal
Log Normal
Log Normal
Log Normal
Normal
Normal
Normal
Log Normal
Log Normal
Normal
Normal

Log Normal

Normal
Normal
Normal
Normal
Coefficient
of
Mean Variation
1.45
30.0
0.10
0.20
4.40
0.02
0.02
5.8E-5
7.50
0.50
0.02
1052
78.0
7.14

1.20

0.10
0.05
5E-7
0.0043
0.10
1.20
0.25
0.15
1.25
1.00
1.00
0.29
0.27
0.20
1.00
0.80
0.05
0.20

1.00

0.10
0.10
0.10
0.10
These variables would be expected to have a strong linear effect on dosage
since they determine the mass of pesticide available and the mass of
pesticide taken up into the foodchain.  Other parameters such as soil bulk
density, partition coefficients, and pesticide decay rates determine the
persistence of the pesticide in soil.  Note that in this case the principal
exposure routes for the target species were through soil ingestion and
predation on earthworms.  Correlations between input parameters had little
                                     265

-------
i
i

a
o
i
w

X
3
i
s
i
i
Si
3
*
s
2
!
1
£
•^
S
1
i
Si
i
i
i
i
i


S
i
i
g

g

i
g
g
s
I

I
g
I
i
s
i
i
i
i
i

s
I
I
i
1

g

s
I
1
1
g

1
s
I
Ullll« COEF/SOIL
i
i
i
1
i

s
1
1
I
g

S

3
|
I
I
g
*~
I
1
I
S
1.
i
i
i
i
i

i
g
i
i
1

3

3
|
|
I
3
=
3
I
i
i
UJ
1
i
i
i
s

I
i
i
i
§



i
i
i
i
3
"!
3
I
1
i

c
01
c
-o
O)
•D
X
S-
(O
E
c
o
•I—
fO
o>
S-
o

•
CO
O)
s-
a-
U-

J6G

-------
Table 6-44.  SIGNIFICANT PARAMETERS CONTROLLING PESTICIDE DOSAGE TO THE
             TARGET SPECIES
                                                  Correlation
Parameter                                         Coefficient
Pesticide Application Rate                           0.76
Bioconcentration Factor for Earthworms               0.36
Total Feeding Rate                                   0.36
Soil Bulk Density                                    -0.22
Pesticide Decay Rate in Soil                         -0.20
Soil Preference Factor                               0.10
Adsorption Partition Coefficient                     0.08
effect on model sensitivity, since the list of significant parameters was
identical for both the correlated and uncorrelated cases.
    Table 6-45 lists the variables which were most significant in
determining pesticide concentration in the target species.  As would be
Table 6-45.  SIGNIFICANT PARAMETERS CONTROLLING PESTICIDE CONCENTRATIONS
             IN THE TARGET SPECIES
                                                  Correlation
Parameter                                         Coefficient
Metabolic Degradation Rate                           -0.52
Pesticide Application Rate                            0.24
Bioconcentration Factor for Earthworms                0.15
Total Feeding Rate                                    0.18
Pesticide Decay Rate in Soil                         -0.09
Soil Bulk Density                                    -0.06
                                   267

-------
expected, many of the same parameters which contribute  to  dosage also
contribute strongly to concentration.  However,  in this case the metabolic
degradation rate was the most important parameter, indicating that most of
the ingested pesticide was broken down and decayed by metabolic processes.

    In this simple ecosystem, a few variables were observed to control  the
output of the model.  In an actual ecosystem, it is likely that the dosage
to, or concentration in, the species of interest would  be  controlled by a
larger set of factors.  However, the results of  this analysis do appear to
be rational.  For instance, the most important factors  controlling these
outputs are those which affect the major pathways of exposure; in this  case,
direct ingestion of contaminated soil and soil fauna.  Obviously, pesticide
application rate, bioconcentration factors, feeding rates, soil ingestion
rates, decay rates in soil, and adsorption partition coefficients are the
principal parameters controlling these exposure  pathways.   Therefore, the
model user should speculate on the major pathways of exposure, and
concentrate parameter estimation efforts on those which most directly
control concentrations in important media and the uptake of pesticide from
these media to the species of interest.
                                     268

-------
                                  SECTION 7

                                MODEL OUTPUT
7.1  INTRODUCTION

    Each TEEAM module produces some form of output.  In general, this output
consists of echoed input, tabular event-based or periodic summaries of
ecosystem status, or time series files which are Lotus compatible and can be
exported for plotting.  The following sections describe the output which is
or may be produced from each module.

7.2  INPREA

    The INPREA module produces output which echoes the input provided by the
user in a readable format.  The input contained in the files having the
following logical unit numbers:

       • KRUN (Simulation control)
       • KPGDEF (Plant growth parameters)
       • KPZDEF (TFAT parameters)
       • KMCIN (Monte Carlo parameters)
       • KADIN (Animal model parameters)
       • GRDEF (grid/habitat map and spray application control parameters)

is output to the single KECHO file.  An example of the echoed input of each
of these files is shown in Figures 7.1 through 7.4.

7.3  FSCB6

    The FSCBG model produces an output to a scratch file each time the model
is called from within the daily loop of the execution supervisor (EXESUP).
It produces an output of the type shown in Figure 7.5.  This output echoes
new input conditions as well as providing a tabular output of the pesticides
deposition values at each receptor grid point.  Output from each spray
application event is stacked sequentially in the output file.

7.4  TFAT

    The TFAT module potentially produces four output files.  These files
contain the following:

                                   269

-------
*  TERRESTRIAL FATE ANO TRANSPORT MODULE   *
*              HABITAT 1                 *
This run  was made at **CURRENT DATE WAS NOT FOUND**
HA8ITAT 1
for DIAZINON application to peanuts,  Tifton. GA
TFAT1
SIMULATION START DATE (OAY-MONTH-YEAR)           1 APR., 50
SIMULATION  END  DATE (DAY-MONTH-YEAR)          31 RAY., 50
-HYDROLOGY  AND  CROP  PARAMETERS
                                                     -TFAT2
HYDROLOGY AND SEDIMENT RELATED PARAMETERS
PAN COEFFICIENT FOR EVAPORATION                    0.7500
FLAG FOR ET SOURCE  (0=EVAP,1=TEMP,2=EITHER)              2
DEPTH TO WHICH ET IS COMPUTED YEAR-ROUND (CM)        36.20
MONTHLY DAYLIGHT  HOURS
   «ONTH     DAY  HOURS
   JAN.       10.20
   APR.       12.80
   JULY       13.90
   OCT.       11.20
SNOW MELT COEFFICIENT  (CM/OEG-C-OAY)               O.OOOOE+00
INITIAL CROP NUMBER                                    0
INITIAL CROP CONDITION                                  1
MONTH
FES.
MAY.
AUG.
NOV.
DAY HOURS
10.90
13.70
13.50
10.40
MONTH
MAR.
JUNE
SEP.
DEC.
DAY HOURS
11.90
14.00
12.20
10.00
HABITAT AREA (HA)
                                   4.000
 Figure  7.1.   Example of a  portion  of  the TFAT  input  echo.
                                    270

-------
               Plant growth and translocation  rradule,  habitat   1
*  Plant number  1
*
*  Plant growth parameters:
*
* Heat units fop plant to nature (C day) ............................... 0.210E+Q<
* Rate of conversion of radiation to bicaass ................................ 20.0
* Maxieuft leaf area index [[[ 5.00
* Fraction of season for decline of LAI .................................... 0.750
* Ratio of total bioaass to  crop yield ..................................... 0.420
* Maxiwiat root depth (») [[[ 60.0
* Miniwiis growth teaperature (C) ............................................ 13.5
* OptimuB growth temperature (C) ............................................ 25 . Q
* Plant height growth coefficient (1/day) ................................... 1.00
* Maxinus canopy height (n) ................................................ 0.900
*
*  Plant translocation paraseters:
*
* Reflection coefficient for transfer from the soil to the root (dec.) ...... 1.00
* First order decay rate (I/day) ....................................... O.OQOE+QO
* Octan-1-ol/water partition coefficient ............................... 0.105E+04
* Aboveground bioaass water  content (g/g) .................................. 0.850
* Concentration in nonaqueous /concentration in aqueous (ct5**3/g) ........... 315.
* Ratio of the nonaqueous to the total aboveground bioaass (g/g) ........... 0.150
*
*  Plant growth initial conditions:
*
* Potential bioi»ass (kg/ha) ............................................ 0.100E-02
* Yield (kg/ha) [[[ 0 . 1006-02
* Root bioi-ass (kg/ha) ................................................. 0. 100E-02
* Roots sloughed (kg/ha) ............................................... 0.100E-02
* Live roots (kg/ha) [[[ 0.100E-02
* Root depth (CM) [[[ Q.100E-02
* Actual bioisass (kg/ha) ............................................... 0.1QOE-02
* Plant height (a) [[[ 0.100E-01
* Accumulated thermal  tiise ............................................. O.OOOE+00
*
*  Plant translocation  initial conditions:
*

-------
   ANIMAL PESTICIDE UPTAKE AND MOVEMENT MODULE
 **************************************************

  THRUSHES AND EARTHWORMS IN  GEORGIA PEANUT FIELD ECOSYSTEM             APUM1
 «*****************************));*******************
   ECHO OF INPUT DATA:
 ********+*************************•****************
     ANIMAL FOOOCHAIN MODEL WITH    2  HASITAT(S)
 ******HABITAT    1   HAS      2  GROUP(S) OF RESIDENT SOIL ANIMALS
       MOVING THROUGH    2  SOIL  LAYER(S)
    PESTICIDE UPTAKE DATA FOR  L. CASTANEUS 1
          POPULATION DENSITY           = 0.120E+02  G/M"2
          INITIAL ORGANISM PEST.  CONC. = 0.0006+00  MG/MG
          8IOCONCENTRATION FACTOR     = 0.714E+01  MG/MG
          METABOLIC DEGRADATION  RATE   = 0.580E+00  OAYS"-1
          LO-10          -             = 0.100E-05  MG/MG
          LD-50                       = 0.100E-0*  MG/MG

          LOWER AND UPPER BOUNDS  ON LD10: 0.500E-06 0.150E-05
          LOWER AND UPPER BOUNDS  ON L050: 0.500E-05 0.15QE-04
                                                          t'
          MOVEMENT DATA:
             INITIAL LOCATION =  SOIL COMPARTMENT
             ANIMAL MOVES    4  TIME(S)/OAY

          SOIL MOVEMENT TRANSITION MATRIX:

          HORIZON:        1         2

              1      0.500E+00 0.500E-MJO

              2      0.800E+00 0.200E+00

Figure 7.3.    Example  of  a  portion  of  the APUM input echo.

                                 272

-------
                   AERIAL SPRAY TRANSLATION MODULE
               LOCATION OF HABITATS WITHIN SPRAY MODEL RECEPTOR GRID
              HABITAT COORDINATES
              HABITAT      SOUTHWEST CORNER   NORTHEAST CORNER

                  NO.        X         Y        X       Y
                         (METERS)  (METERS)   (METERS) (METERS)
1
2
0.
140.
0.
0.
HO.
280.
280.
280.
              NOM8GR OF X-AXIS VALUES IN RECEPTOR GRID =  14
              NUM9ER OF Y-AXIS VALUES IN RECEPTOR GRID =  14
                                 RECEPTOR GRID AND HABITAT  LOCATIONS
           Y  (METERS)

             260.   1 1  1  1  1  T1 2 2 2 2  2 2 2
             2*0.   11111112222222
             220.   11111112222222
             200.   11111112222222
             180.   11111112222222
             160.   11111112222222
             140.   11111112222222
             120.   1  1  1
             100.   1  1  1
             80.   111
             60.   111
             40.   Ill
1112222222
1112222222
1112222222
1112222222
1112222222
             20.   11111112222222
              0.   11111112222222

                   0  40  80 120 160 200 240
                    20 60  100 HO 180  220 260
                              X (METERS)


Figure  7.4.   Example of a  portion  of  the GRDDEF input echo,

-------
1FOREST SPRAY MODEL  *** TEEAM/FSC8G: Test 2, Oiazinon app. (1 gal/acre), wind din =270,no evaporation

          *** INPUTS USED 8Y ALL MODELS ***
     PROGRAM OPTIONS:  ISW(  1) =  1
                      ISW(  2) =  0
                      ISH(  5) =  0
                      ISH(  6) =  0
                      ISK(18) =  0
                      ISH(19) =  0
                      ISH(21) =  0
                      ISW(22) =  0
     IS LJQUID WATER OR NON-WATER, 2=NON-WATER, 1=HATER, (IFHATR) = 1
     AIRCRAFT WING SPAN (WNGSPN (METERS)) =  12.60
     HEIGHT OF AIRCRAFT (HGTCFT (METERS)) =   4.00
     DENSITY OF SPRAY LIQUID (OENLIQ (G/CM**3)) =   1.0000
     MOLECULAR WEIGHT OF AIR (AIRMOL) = 28.9600
     BAROMETRIC PRESSURE (AIRPRS (M8)) =1000.00
     RELATIVE HUMIDITY A80VE THE CANOPY (RELHHO (%)) = 92.000

          *** INPUTS USED 8Y THE WAKE SETTLING VELOCITY MODEL ***

     AIRCRAFT WEIGHT (ARCRHT (KG)) = 3190.000
     AIRCRAFT GROUND SPEED (ARCftSP (M/S)) =  57.200

          **» INPUTS USED 8Y THE EVAPORATION MODEL ***

     UPPER LIMITS OF DROP DIAMETERS (DRPUPR (MICRO-M)) =
      611.000, 578.000, 545.000, 512.000, 479.000, 447.000, 414.000,
      382.000, 351.000, 318.000, 284.000, 252.000, 219.000, 187.000,
      154.000, 122.000,  89.000,  56.000,
     LOWER LIMITS OF DROP DIAMETERS (DRPLWR (MICRO-M)) =
      579.000, 546.000, 513.000, 480.000, 448.000, 415.000, 383.000,
      352.000, 319.000, 285.000, 253.000, 220.000, 188.000, 155.000,
      123.000,  90.000,  57.000,  30.000,

          *** INPUTS USED BY THE DISPERSION MODELS ***

     NUM9ER OF LINE  SOURCES (NSOURC) =   7
     DISTANCE BETWEEN SPRAY LINES (SWATH (M)) =   20.000
     EMISSION OF SPRAY MATERIAL FOR EACH SOURCE (Q (GAL/ACRE)) =
      LOOOOOEtOO, LOOOOOEtOO, 1.0000QE+00, 1.00000E+00, 1.00000E+00,
      1.00000E+00, 1.00000E+00,
     TIME TO SPRAY CLOUD STABILIZATION (TAU (SEC)) =   2.500
     STAND. DEV. OF SPRAY MATERIAL ALONG SPRAY LINE (SIGXYZ («)) =   2.977
     DECAY COEFFICIENT (DECAY (/SEC)) * Q.OOQOOE-01
     DEPTH OF GA.S SOURCES (DELTAH (M)) =   1.000
     LAT., VERT. REFERENCE DISTANCE (XLRZ (M)) =*********
     SURFACE MIXING LAYER HEIGHT (H« (M)) =  200.000
     WIND DIRECTION (FROM) (THETA (OEG)) = 270.00
     RATIO OF LAGRANGIAN TO EULERIAN TIME SCALES (BETA1) = 0.00
     WIND-SPEED SHEAR (OELU (M/S)) = 0.0000

               Figure 7.5a.   Example of  the  FSCBG  input  echo.


                                             274

-------
FSCBG  aerial spray  5/ 5/50
    Air  teaperature (C) =  25.60
    Hind speed (e/sec)  =   3.58
    CALCULATION HEIGHT FOR GRID POINTS (I (METERS)) =   0.08
    STANOARD OEV.  OF HIND DIRECTION ANGLE (S1GAP (RAD))  =0.17453
    STANDARD OEV.  OF HINO ELEVATION .ANGLE (SIGEP (WO))  =0.10472
        *** FOREST SPRAY MODEL *** TEEAM/FSCBG: Test 2,  Diazinon app.  (1 gal/acre), wind dir =270. no evaporation
                                                        TABLE  1
                                    *-* DEPOSITION    (MICROGRAMS     /SQUARE  METER  ) *-*
                                    *-*       AT A HEIGHT OF   0.0815 METERS       *-»
                        (MAXIMUM DEPOSITION    = 9.1187616+05 AT X=   HO.000, Y=    60.000,1=
0.1)
Y AXIS
(M6T6RS)
- X AXIS (METERS) -
0.00
20.00
40.00
60.00
80.00

100.00
120.00

140.00
160.00
- D6POSITION
260.00
240.00
220.00
200.00
180.00
160.00
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
Y AXIS
(METERS)

260.00
240.00
220.00
200.00
180.00
160.00
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
O.OOOOOE-01
0.000006-01
O.OOOOOE-01
O.OOOOOE-01
O.OOOOOE-01
0.000006-01
O.OOOOOE-01
o.oooooe-fli
O.OOOOOE-01
O.OOOOOE-01
O.OOOOOE-01
O.OOOOOE-01
O.OOOOOE-01
O.OOOOOE-01
6.82653E+05
6.826536+05
6.82653E+05
6.826536+05
6.82653E+05
6.826536+05
6.82653E+05
6.82653E+05
6.82653E+05
6.82653E+05
6.82653E+05
6.826536*05
6.82653E+05
3.41326£*05
8.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
B.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
8.28733E+05
4.14367E+05
8.722836+05
8.722896+05
8.722896+05
8.722896+05
8.722896+05
8.722896+05
8.722896+05
8. 722896+05
8.72289E+05
8.72289E*05
8.72289E+05
8.72289E*05
8.72283E+05
4.361446*05
8.91148E+05
8.911776+05
8.91177E+05
8.911776+05
8.911776+05
8. 911776+05
8.91177E+05
8.91177E*05
8.91177E*05
8.911776+05
8. 911776+05
8.91177E*05
8.91148E+05
4.455896*05
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
012866+05
013656+05
01365E+05
013656+05
013656+05
01365E+05
013656+05
013656+05
013656+05
013656+05
013656+05
013656+05
012866+05
4.50682E+05
9.074906+05
9.076466+05
9.076466+05
9.076466+05
9.076466+05
9.076466+05
9.076466+05
9.076466+05
9.076466+05
9.076466+05
9.07646E+05
9.076466+05
9.07490E+05
4.538236+05
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
9.
g.
4.
116206*05
118766+05
11876E+05
118766+05
118766+05
118766+05
118766+05
118766+05
118766+05
118766+05
118766+05
118766+05
11620E+05
55938E+05
2.31885E+05
2.322536+05
2.32254E+05
2.322546+05
2.322546+05
2.322546+05
2.322546+05
2.322546+05
2.322546+05
2.322546+05
2.32254E+05
2.32253E+05
2.31885E+05
1.16127E+05
- X AXIS (METERS) -
180.00

6.79570E+04
8.84450E+04
8.84471E+04
8.84471E+04
8.84471E+04
8.84471£*04
8.844716+04
8.84471£*04
8.844716+04
8.844716*04
8.84471E+04
8.84450£*04
8.79570E*04
4.422366*04
200.00

4.60485E+04
4.66520E+04
4.66569E+04
4.66569£*04
4.665696+04
4.66569E*04
4.66569E+04
4. 665696+04
4.66569Ei04
4.665696*04
4.66569E+04
4.665206+04
4.60485E+04
2.33285E*04
220.00
- DEPOSITION
2.84701E+04
2.916996+04
2.91785E+04
2.91785E+04
2.91785E+04
2.91785E*04
2.91785E+04
2.91785£*04
2.91785E+04
2.91785E*04
2.91785E+04
2.916896*04
2.84701E+04
1. 458936+04
240.00
-
1.93634E+04
2.01250E+04
2.01412E+04
2.01414£*04
2.01414E+04
2.01414E*04
2.01414E+04
2.01414E*04
2.01414E+04
2.014146*04
2.01412E+0
-------
       • Daily,  monthly,  or annual  habitat  hydrology  output  summaries

       • Daily,  monthly,  or annual  habitat  pesticide  flux  summaries

       • End of  day,  end  of month,  or end of  year concentration  profiles for
         each 'habitat

       • Daily time series output (fluxes,  concentrations, masses, etc.) for
         plotting

Examples of the  TFAT output files are shown in Figures  7.6 through 7.9.   The
output is largely self-explanatory.  The labels above each column in the
time series output file echo the PLOTFL designation for the  plotted  value.

7.5  APUM

    The Terrestrial Animal Exposure Model produces two  output files.  The
first output file provides a detailed breakdown of dosages for each
simulated animal group, and may be written  out daily, monthly, or annually
depending on user-selected options.  The dosage breakdown  for each animal,
illustrated in Figure 7.10, includes total  cumulative dosage, dosages of
pesticide from each food  source and exposure  route, and the  current  average
concentration of pesticide in the animal group biomass. Also written to
this file is mass balance error for the entire food chain.   This term is
computed by summing the mass of pesticide taken up by all  animals and
subtracting losses due to decay and depuration.  The  value of the mass
balance error indicates the relative magnitude of numerical  and round off
errors, and serves as an  internal consistency check on  the APUM module.

    The second APUM output file is a time series file containing daily
values of various user-specified parameters,  and is formatted for use by
plotting and spreadsheet  software packages.  This file  is  similar in format
to the time series file produced by TFAT.   Each row in  the time series file
contains 1) the  date, 2)  the Julian day, and  3) NPLTS entries of user-
selected output  parameters.  Output parameters which  may be  selected include
the following:

       • Cumulative pesticide dosage (mg/mg)
       • Concentration of pesticide in animals (mg/mg)
       • Lethal  dosages (mg/mg)
       • Upper  and lower bounds on lethal dosages (mg/mg)

Figure 7.11 shows an example APUM time series file.
                                   276

-------
 « WKTHU ECOSYSTEM STATUS
 • HABITAT  I
  Mil:
          31 Mr.. 50
                      ALL HVOROL06V UNITS ARE Cf. OF MTER
                      SEOKEXT UNITS ARE VTRIC TONNES
                      1WSERS IN PARENTHESES ARf SOIL KATES CONTENTS
 CURRENT CONDITIONS
CROP NU'ffiER 1
FRACTION OF GROUND CO«S 0 29!0
ASOYE GROUND !1«
-------
CUMWT COWM1IONS
STOUCE UNITS IN ftfi/HA
FlUK WITS IN U/«AyoUIPUT TINESIEP




CM ««£(
funfS AND STOCAGfS H* THIS P«100
CA«C**
AlwSPKfNE
(NITMIN CAJIOPY)
»60vE 6WJNO
» gggoEgg e.nn i.«isc-ti
«.COOOMO ).(««.« g.iHtc-g; g.ixiE-t)
i.gg«E«g i.o»o(-« I.IIOK- i g.ig«E« g.g««E>gg I.MO«£-O«
i. tcooE.lt o.moE.oo g.ggogf.cg «.oooo!-oo
t.««g£-«g g.ggggE<« «.(««£^M g. ««««•««
g.toooE-io g.ggiK.gg l.oooo£.w i.gtooE-ot
e.gg«E*4g g.gggg£*gg g ggag£«gg i.goggE^g
i.gmE-gg g.gg«E-» g.»»n.«g t.wm.n
•CUKJtS EDCSION U9 kUMlf' IOSS.
«•«««•"
PU«T
IIPTAIE

g.!^ltE^l
O.JII1E-0!
g.»go
(.gogg£«
O.JOOOMO
g IMOE.OO
g.ggggEXH
i.ggggE-gg
g.gggg£»gg
g.gggo£
g gggg£*gg
g.g««E.»




CURRENT
STOUCE

1.111
g.igtg
g »»£-»
g )iti£-g<
g.SKS£-«
g.iiiiE-fii
g.siitf-ot
O.HOlf-10
g 11IIE-I!
g i
-------
           1PESTICJDE CONCENTRATION PROFILE - HABITAT  1

            DATE (OAY-MONTH-YEAR)    31 MAY.. 50


            PESTICIDE CONCENTRATIONS IN PLANTS
                   CONC. IN  ROOT BIOMASS (W3/M6)          0.7702E-06
                   CONC. IN  ABOVE GROUND BIOMASS  (MG/MG)  0.1371E-05
            PESTICIDE CONCENTRATIONS IN LOCAL SOIL ORGANISMS
                   L. CASTANEUS 1
                   L. TERRESTRIS
PEST. CONC.
  (MG/MG)
0.1308E-04
0.1591E-05
            PESTICIDE CONCENTRATIONS IN HABITAT-MOBILE ORGANISMS
                   T. MIGRATOR IDS
PEST. CONC.
  (MS/KG)
0.2138E-06
            SOIL PESTICIDE CONCENTRATIONS:
HORIZON
1.
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
COMPARTMENT
!
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IB
19
20
TOTAL ADSORBED DISSOLVED VAPOR PHASE
(M6/KG) (MG/KG) (MG/L) (MG/L)
10.18 10.15 0.3383 Q.1962E-04
0.1503 0.1500 0.4999E-02 0.2899E-06
0.2814E-Q2 0.2807E-02 0.9358E-04 0.5428E-08
0.4368E-04 0.4356E-04 0.1452E-05 0.8422E-10
0.6768E-06 0.6715E-06 0.6715E-07 0.3895E-11
0.2228E-07 0.2208E-07 0.2208E-08 0.1281E-12
0.6606E-09 0.6S37E-09 Q.6S37E-10 0.3791E-14
0.1816E-10 0.1792E-10 0.1792E-11 0.1040E-15
0.4798E-12 0.4737E-12 0.4731E-13 0.2748E-17
0.1248E-13 0.1233E-13 0.1233E-14 0.7149E-19
0.3204E-15 0.3163E-15 0.3163E-16 0.1835E-20
0.7005E-20 0.6915E-20 0.6915E-21 O.OOOOE+00
O.OOOOE+00 O.OOQOE+00 O.QOOOE+00 O.QOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 Q.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOQE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
O.OOOOE+00 O.OOOOE+00 O.OOOOE+00 O.OOOOE+00
Figure  7.8.   Example of the  habitat  pesticide  concentration output,
                                               279

-------
1'TIME SERIES OUTPUT FILES'
  'HABITAT  1 	 for DIAZINON anplication to peanuts, Tifton, 6A
 •HAS" YEAR/MO/DAY' 'JUL1    'TPST"        "TPST'
TFAT1
r
1'TIKE
1950
SERIES
'HABITAT 2 -
•HAS'
B *
2'
r
2'
1'
2"
r
2'
1"
21
r
2'
ig
2"
r
2'
1"
2'
1'
2"
i"
2"
1"
2'
1"
2'
r
2'
r
2'
i"
2'
1"
2'
r
2'
r
2'
1'
2'
r
2'
4
r
OUTPUT




•YEAR/WQAY'
i
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950
1950

4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
t
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
i.
4
4
4
B
r
r
r
r
3*
4'
4"
5'
5'
6'
6'
7'
7'
8'
8"
9'
9'
10"
10'
11"
11"
12'
12'
13"
13'
14"
14'
15'
15'
16'
16"
17"
17'
is-
ia'
19"
19'
20"
20"
21'
21'
91
FILES
0
«
.OOQGQE+QO

for DIAZINON appl
•JUL1
B B
91
92
92
93
93
94
94
95
95
96
96
97
97
98
98
99
99
100
100
101
101
102
102
103
103
104
104
105
105
106
106
107
107
108
108
109
1Q9
110
110
111
111


0
0
Q
0
0
0
Q
Q
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Q
0
•TPST'
• •
.OOQOOE+00
.OOOOOE+00
.OOQQOE+00
.OOOOOE+00
.OOOOOE+00
.OOOQOE+00
.QOOOQE+00
.OOOOOE+00
.OOOOOE+00
.OOOOOE+00
.OOOQQE+OQ
.OOOOOE+QO
.OOOOOE+00
.OOOQOE+00
.OOQOOE+00
.OOQOOE+OQ
.OOOOOE+00
.OQOQQE+00
.QOQOOE+QO
.OOOOOE+00
.OOQQQE+00
.OOOOOE+00
.QOOOOE+00
.OOOQOE+00
.OQOOOE+00
.OOOOOE+00
.OOOOOE+00
.OOOOQE+00
.OOQQOE+00
O.OQOOOE+00
O.OOOOOE+QO
0
Q
0
Q
0
0
0
0
0
0
.OOOOOE+00
.OOQOOE+00
.OOOOOE+OQ
.OOQQQE+QQ
.OOOOOE+00
.OOQOQE+00
.OOOOOE+00
.QOOOOE+OQ
.OOOOOE+00
.OOOOQE+00
0

.QQOOOE+00

ication to peanuts, Tifton, GA


0
0
Q
0
Q
0
0
0
Q
0
0
0
0
0
0
0
0
0
0
0
0
0
Q
0
0
0
0
0
0
0
0
0
0
0
Q
"TPST"
« a «
-OQQQOE+OQ
.OOQOOE+00
.QOQQOE+00
.OOOOOE+00
.OQQQOE+00
.OOOOOE+00
.OOQOOE+QO
.OOOOOE+00
.OQQOQE+00
.OOOOOE+CO
.OOQOOE+00
.OOOOOE+00
.OOOOOE+00
.QQQOOE+00
.QOOOOE+QO
.OOOOOE+00
.OQOOOE+00
.OOOOOE+00
.OOQOOE+00
.OOOOOE+00
.QOOOOE+00
.OOOOOE+00
.OOOOOE+QO
.OOOOOE+00
.OOOOQE+00
.OOOOOE+OQ
.OOOOOE+00
.OOQOOE+00
.QQOOOE+QO
.OOOOOE+00
.QOOQOE+00
.OOOOOE+00
.QQQOQE+00
.OOOOOE+00
.QQQQOE+QQ
O.OOOOOE+00
9.QOOQQE+QO
0
0
0
Q
.OOOOOE+00
.QOOQQE+OQ
.OOOOOE+00
.QOOOOE+OQ
                                                                   TFAT1
    Figure 7.9.   Example  of  the  habitat time series  output.
                                 280

-------
********:

  DETAILED SUMMARY OF FOODCHAIN MODEL RESULTS FOR DATE =
                   4-30-50
                      ;********«******
              -UPTAKE RESULTS FOR T. MIGRATORIUS
         INITIAL TISSUE CONCENTRATION = O.OOOE+00  MG/MG
         FINAL TISSUE CONCENTRATION   = 0.122E-06  MG/MG
         MAX.  TISSUE CONG. FOR PERIOD = 0.161E-06  MG/MG

         BIOMASS OF ANIMAL GROUP      = 0.80OE+02  G
         INITIAL PESTICIDE MASS
         FINAL PESTICIDE MASS

         INITIAL TOTAL DOSAGE
         FINAL TOTAL DOSAGE
O.OOOE+00
0.978E-05

O.OOOE+00
0.460E-05
G
G

MG/MG
MG/MG
         DOSAGE BREAKDOWN:
              SOIL DOSAGE            = 0.956E-06  MG/MG
              PLANT DOSAGE           = O.OOOE+00  MG/MG
              PELLET DOSAGE          = O.OOOE+00  MG/MG
              DOSAGE FROM PONDS      = O.OOOE+00  MG/MG
              INHALATION DOSAGE      = 0.801E-10  MG/MG
              METABOLIC DECAY LOSS   = 0.375E-05  MG/MG
              EXCRETION LOSS         = 0.729E-06  MG/MG

              PREDATION DOSAGE FROM L. CASTANEUS 1
              PREDATION DOSAGE FROM L. TERRESTRIS 2
                I

              TOTAL PREDATION DOSAGE = 0.364E-05  MG/MG
                   0.333E-05
                   0.308E-06
                   MG/MG
                   MG/MG
    ++++++TOTAL FOODCHAIN PESTICIDE MASS          = 0.295E+00  G
    ++++++CURRENT FOODCHAIN MASS BALANCE ERROR    = 0.386E-06  G
    ++++++CUMULATIVE FOODCHAIN MASS BALANCE ERROR =-0.270E-08  G


             Figure  7.10  APUM  dosage breakdown output file.
                                    231

-------
"NEW"
"DATE"
1 4- 1-50"
• 4_ 2-50"
1 4- 3-50"
• 4- 4-50"
• 4_ 5-50"
• 4- 6-50"
1 4- 7-50"
1 4- 8-50"
1 4- 9-50"
• 4-10-50"
" 4-11-50"
" 4-12-50"
" 4-13-50"
" 4-14-50"
11 4-15-50"
11 4-16-50"
" 4-17-50"
" 4-18-50"
" 4-19-50"
" 4-20-50"
" 4-21-50"
"JULIAN"
"DAY"
91
92
93
94
95
?o
37
98
99
100
101
102
103
104
105
106
107
108
109
110
111
"CORG "
"L. CAST"
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
0.327E-05
0.967E-05
0.112E-04
0.100E-04
0.112E-04
0."ll7E-04
0.119E-04
0.119E-04
0.118E-04
0.116E-04
0.978E-05
0.103E-04
0.892E-05
0.809E-05
0.907E-05
"CORG "
"T. MIGR"
O.OOOE+00
O.OOOE+00
O.OOOE+OO
O.OOOE+00
O.OOOE+00
O.OOOE+00
0.270E-07
0.912E-07
0.132E-06
0.142E-06
0.148E-06
0.155E-06
0.159E-06
0.161E-06
0.161E-06
0.160E-06
0.151E-06
0.147E-06
0.140E-06
0.131E-06
0.131E-06
"DOSE "
"T. MIGR"
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+OO
O.OOOE+00
O.OOOE+00
0.422E-07
0.196E-06
0.405E-06
0.615E-06
0.827E-06
0.105E-05
0.128E-05
0.151E-05
0.174E-05
0.197E-05
0.219E-05
0.239E-05
0.259E-05
0.277E-05
0.296E-05
"LD50 "
"T. MIGR"
O.OOOE+00
O.OOOE+00
O.OOOE+OO
O.OOOE+00
O.OOOE+00
O.OOOE+OO
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
                 Figure 7.11  APUM time series output file.
    Note that during Monte Carlo simulation the APUM output files are not
produced due to the potentially large amount of information that would
otherwise be written out.   See the Monte Carlo output description for
information on Monte Carlo output.

7.6  MCARLO

    The Monte Carlo module produces two output files, including a summary
file and a parameter file.  Note that for variables which vary in time the
Monte Carlo module performs statistics on maximum moving averages of
specified length.  Thus, if the user specifies a 5-day averaging period the
model will select the maximum 5-day average value from each Monte Carlo run
for statistical calculations.  The summary file contains summary statistics,
correlations, and cumulative distributions for the Monte Carlo simulation.
For each user-selected output variable the Monte Carlo module writes out the
mean, standard deviation,  skewness coefficient, minimum, and maximum over
all Monte Carlo runs.  Also written out is a correlation matrix containing
the value of the correlation for each pair of selected output variables.
Finally, the summary file  contains tables and line plots of cumulative
distributions for a selected subset of output variables.  Figure 7.12
illustrates an example Monte Carlo summary output file.
                                     282

-------
            MONTE-CARLO  OUTPUT

     'NOTE THAT VALUES ARE MAXIMUM N-OAY MOVING AVERAGES'
     '(I.E. MAX. AVERAGE DAILY VALUES OVER N-OAY PERIODS)1
SUMMARY STATISTICS OVER ALL MONTE-CARLO RUNS
FOR SELECTED VARIABLES:
VARIABLE      INDEX  AVERAGING  MEAN    STANDARD   COEFF. OF   SKEW   MINIMUM   MAXIMUM
                    PER.(DAYS)         DEVIATION  VARIATION  COEFF.
SOIL BULK DENS
SOIL BULK DENS
BCF
SOIL PESTICIDE
ORG PEST CONC/
ORG PEST CONC/
CORRELATION MATRIX
1
2
2
1
2
3
FOR
1
1
1
5
15
1
SUMMARY
1.49
1.54
5.00
7.06
0.176E-05
0.526E-06
0
0
0
0
0
0
.104
.802E-01
.QOOE+DO
.885E-03
.174E-06
.130E-06
0
0
0
0
0
0
.702E-01 0
.519E-01-0
.OOOE+00 0
.125E-03-0
.991E-01-0
.246 -0
.279
.250
.OOOE+00
.361E+04
.906
.279E-01
1
i
5
7
0.
0.
.33
.35
.00
.06
134E-05
289E-06
1.70
1.70
5.00
7.06
0.205E-05
0.733E-06
OUTPUT VARIABLES:
  SOIL BULK DENSITY

  SOIL BULK DENSITY

  BCF

  SOIL PESTICIDE

  ORG PEST  CONC/ANIM

  ORG PEST  CONC/ANIM
 SOIL    SOIL    BCF     SOIL    ORG P   ORG  P

 1.000   0.342   0.000   0.996  -0.808  -0.997

 0.342   1.000   0.000   0.383  -0.170  -0.371

 0.000   0.000   1.000   0.000   0.000   0.000

 0.996   0.383   0.000   1.000  -0.778  -0.999

-0.808  -0.170   0.000  -0.778   1.000   0.788

-0.997  -0.371   0.000  -0.999   0.788   1.000
           Figure 7.12   Monte Carlo summary  output  file.
                                        283

-------
       CUMULATIVE DISTRIBUTION FOR ORG PEST CONC/ANIM[3]










VALUE

0.289E-06

0.334E-06

0.378E-06

0.422E-06

0.467E-06

0.511E-06

0.556E-06

0.6QOE-06

0.645E-06

0.689E-06

0.733E-06
N
MEAN
STANDARD DEVIATION
COEFFICIENT OF VARIATION =
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
* OF TIME EQUALLED % OF
OR EXCEEDED
100.000

90.000

90.000

75.000

70.000

60.000

30.000

30.000

25.000

15.000

5.000
20
0.526E-06
Q.130E-06
0.246
0.289E-06
0.733E-06
0.654E-06
0.657E-06
0.713E-06
0.724E-06
TIME IN INTERVAL


10.000

0.000

15.000

5.000

10.000

30.000

0.000

5.000

10.000

10.000

Figure  7.12  Monte Carlo  summary  output  file (continued)
                               284

-------
          C 100 +	+•	+	+	+	+	+•	+	+•	+	+
          U    !                                                      **
          M    !                                                    ** !
          U    i                                             *********   i

          ^    j                                        *******          i
          y    i                                **********               i

          I    !                                                       !
          V 60 +•	+	+•	+	+•	+	+•	+	•*•	+	+
          E    !                               *                       !
               i                             ***                        i
          F    !                                                       !
          R 40 4-	+•	+•	+	+—******	+.	+	+	+	+
          E    I                       *                               !
          Q    !                    ****                                !
          U    i              ********                                   i

          £ 20 +•	+	+	+	+	+	+	+•	•»•	+	+•
          N    !           **+                                         !
          Q    i    *********                                            i
          Y    i*****                                                   i
             0 4.	4.	4.	+	4.	4-	f	+	+	+	4.
            0.289  0.334  0.378  0.422  0.467  0.511  0.556  0.600 0.645  0.689  0.733
                                    *  0.1E-05

                                    OR6 PEST CONC/ANIMp]



            Figure 7.12   Monte Carlo summary output  file (concluded)
    The  second Monte  Carlo output  file contains  the values  of variables for
each Monte Carlo run.   The user  selects which  variables are to be written to
this file in column format.  The data in this  file can be used to plot
cumulative distributions and to  examine the combinations of input variables
which  produce various  model results.   An example of this file is shown  in
Figure 7.13.
                                        285

-------
       "MODEL PARAMETERS FOR EACH MONTE-CARLO RUN:"
       "NOTE THAT VALUES ARE MAXIMUM N-DAY MOVING  AVERAGES"
       "(I.E. MAX. AVERAGE DAILY VALUES OVER N-DAY PERIODS)"
"RUN NO"   "SOIL BULK DEN"   "SOIL BULK DEN"    "ORG  PEST  CONG"
                123
       1     1.520             1.538            0.4751E-06
       2     1.466             1.567            0.5406E-O6
       3     1.328             1.505            0.7335E-06
       4     1.382             1.550            0.6569E-O6
       5     1.347             1.534            0.7133E-06
       6     1.657             1.575            0.3244E-O6
       7     1.484             1.551            0.5196E-06
       8     1.539             1.544            0.4531E-06
       9     1.517             1.696            0.4842E-06
      10     1.459             1.663            0.5485E-06
      11     1.461             1.665            0.5448E-06
      12     1.419             1.460            0.6126E-06
      13     1.584             1.516            0.3995E-06
      14     1.461             1.464            0.5474E-06
      15     1.334             1.348            0.7245E-06
      16     1.381             1.461            0.6538E-06
      17     1.486             1.577            0.5232E-06
      18     1.695             1.503            0.2891E-06
      19     1.602             1.609            0.3822E-06
      20     1.584             1.565            0.3991E-06

            Figure 7.13  Monte Carlo parameters output file.
                               236

-------
                                  SECTION 8

                             EXAMPLE APPLICATION
    This section gives the results of a comprehensive example application so
that the user can benchmark simulations on his machine.  It makes use of all
of the (deterministic) TEEAM modules and performs a multiple habitat
simulation.  It does not, however, utilize the model's Monte Carlo
simulation capabilities.  The problem is realistic in its assessment of
physical, chemical, and biological parameters.  Some license has been taken
with some aspects of the system in order to gain comprehensiveness.  For
instance, aerial application of the chemical in the region simulated may not
be a normal agricultural practice.  Obviously, the avian in the simulation
would eat food items in addition to the earthworms, as suggested by the
input sequence.

    The input files for this simulation are provided on the set of discs
containing the code and are not shown in this section.

8.1  GENERAL PROBLEM SETTING

    The simulation involves the estimation of exposure of diazinon to
passerine birds (American Robin) in a hypothetical peanut field near Tifton,
GA.  Meteorological data for Tifton, GA are used.  The field is square, 8 ha
in size, half of which is aerially sprayed with diazinon at a rate of 1 gal
acre'l.

8.2  FSCBG AND GRDDEF INPUTS

    Two aerial applications are simulated, one occurring on May 5 and one on
May 25.  The aircraft is assumed to be a Schweitzer Ag Cat.  As mentioned
above, the application rate is 1 gal acre  .  Wind is assumed to be out of
the west (270°) for both spray events.  Other meteorological inputs are in
the echoed output, shown later in this section.

    The aircraft is assumed to make seven passes in a north-south direction
on the western half of the 8 ha field (Habitat 1).  Therefore, the other
half of the field (Habitat 2) receives only drift from the spraying of the
western half.  Foliar deposition is computed using the exponential model
(FILTRA = 2.5 nr kg"*) and canopy penetration is modeled using the
attenuation coefficient approach (BETA = 0.15 cm~l).  No evaporation or

                                    237

-------
decay of the sprayed product is assumed  to occur during  the  application
event.

8.3  TFAT INPUTS

    Most of the TFAT inputs are taken directly from the  guidance  provided in
Section 6.  Two habitats are simulated,  as described above,  each  having an
area of 4 ha.  The soil  of each is assumed to be a Tifton sandy loam with a
saturated hydraulic conductivity of 4.4  cm hr  .  This  soil  is  in hydrologic
soil group 'B1  according to Carsel et al.  (1984).  The  top 100  cm of the
soil are simulated and the soil is assumed to have two  horizons.   These
horizons vary in their thickness, bulk density, and pesticide degradation
and adsorption properties.  Water holding  properties of  each horizon are the
same.  The degradation rate in the top horizon was estimated using values
found in the literature  for this compound.  The value for the lower soil was
taken as the low end of  the range given  in Section 6.4.1.3.

    The peanut crop is planted on April  1.  It is assumed to have an
interception potential of 0.15 cm and a  maximum rooting  depth of  60 cm.

8.4  PLTGRN AND PLTRNS INPUTS

    Inputs for the peanut plant growth model were taken directly  from the
guidance in Williams et  al. (1987) as provided in this  document.   The UPTKF
and RW factors (plant uptake factor in TFAT and reflection coefficient in
PLTRNS) were assumed to  be equal and in  the absence of  more appropriate
information, to have a value of unity.

    The degradation rate in the plant was  unknown and,  therefore, assumed to
be zero.  The KP for the chemical in the plant was calculated using a KOW of
1050, which yields a nearly equivalent KOC (Chiou's method,  see Section
6.4.1.3) and an organic  carbon content for the plant of 30%.

8.5  APUM INPUTS

    For the animal exposure module, Habitat 1 is assumed to have two species
of earthworms (L. Castaneaus), a small,  shallow borrowing lumbricid, and L.
Terrestris, which is larger and burrows  more deeply.  The movement
transition matrices reflect this behavior.  Both species were assumed to be
present in Habitat 1, while only L. Castaneaus was assumed to be present in
Habitat 2.  Biomasses were estimated using the guidance in Section 6.
Pesticide metabolic degradation rates in the lower animals were taken to be
the same as that for the upper soil.  Lethal dosages for earthworms were
taken to be on the order of 10 yg/individual as reported by Lee (1985).
Thirty percent of the diet of the robin  was assumed to  be made up of
                                    2S8

-------
contaminated earthworms, 10% from each species in each habitat.  Clearance
rates for the robin were assumed to be equivalent to those for dieldrin in
thrushes calculated from the data of Jeffries and Davis (1968).  This latter
assumption was made due to the lack of specific information on clearance
rates for diazinon in thrushes.  One would expect the diazinon clearance
rate to be somewhat higher given its slightly lower KQW (1050 for diazinon
versus 4900 for dieldrin) in the same species.  However, the rates reported
by Jeffries and Davis (1968) already seem high in comparison to rates for
other, similar compounds in avians (see Table 6-42).

8.6  TEEAM SIMULATION RESULTS

    Results of the example problem simulation are shown in the output
listing which follows (Figure 8.1).  Due to the stochastic method which is
used to simulate animal movement, the concentrations in lower animals and
simulated dosage to the birds may be slightly different in the user's
simulation than the results shown in Figure 8.1.  All other results should
be exact with the exceptions produced by roundoff errors on different
machines.
                                   289

-------
DETAILED SUMMARY OF FOOOCHAIN MODEL RESULTS FOR  DATE =   5-31-50
            -UPTAKE RESULTS FOR L. CASTANEUS 1
       INITIAL  TISSUE CONCENTRATION = O.OOQE+OQ   MG/MG
       F5NAL TISSUE CONCENTRATION   = 0.134E-04   MG/MG
       RAX.  TISSUE CONC. FOR PERIOD = 0.159E-04   MG/MG
       BIOMASS OF ANIMAL GROUP
= OJ80E+06  G
       INITIAL PESTICIDE MASS       = O.QQOE+00   G
       FINAL  PESTICIDE MASS         = 0.646E+01   G

       INITIAL TOTAL DOSAGE         = Q.OQOE+00   MG/MG
       FINAL  TOTAL DOSAGE           = 0.165E-03   MG/MS

       + + +  + LO-50 HAS SEEN EXCEEDED
       DOSAGE BREAKDOWN:
            SOIL DOSAGE
            PLANT  DOSAGE
            PELLET DOSAGE
            DOSAGE FROM PONDS
            INHALATION DOSAGE
            METABOLIC DECAY LOSS
            EXCRETION LOSS
 0.165E-03
 Q.OQOE+00
 O.OOOE+00
 O.OOOE+00
 O.OOOE+00
 0.152E-03
 O.OOOE+00
MG/MG
MG/KG
MG/MG
MG/MG
MG/MG
MG/MG
            PREOATION DOSAGE FROM L. TERRESTRIS
            PREDATIOH DOSAGE FROM L. CASTANEUS 2
            PREDATION DOSAGE FROM T. MIGRATORIUS

            TOTAL  PREDATION DOSAGE = O.OOOE+00  MG/MG
                   = O.OOOE+00  MG/MG
                   = O.OOOE+00  MG/MG
                   = O.OOOE+00  MG/MG
 Figure  8.1.   TEEAM  output  for the  example  application
                                290

-------
                   -UPTAKE  RESULTS FOR L. TERRESTRIS
              INITIAL TISSUE CONCENTRATION = Q.OQOE+OO  MG/MG
              FINAL TISSUE CONCENTRATION   = 0.169E-05  MG/MG
              MAX.  TISSUE CONC. FOR PERIOD = 0.169E-05  MG/MG
              8IOMASS OF ANIMAL GROUP
= (M80E+06  G
              INITIAL PESTICIDE MASS       = O.OODE+00  G
              FINAL PESTICIDE MASS         = 0.8106+00  G

              INITIAL TOTAL DOSAGE         = O.OOOE+QO  MG/MG
              FINAL TOTAL  DOSAGE           = 0.213E-05  MG/MG

              + + + + LD-10 HAS BEEN EXCEEDED
              DOSAGE BREAKDOWN:
                   SOIL  DOSAGE
                   PLANT DOSAGE
                   PELLET DOSAGE
                   DOSAGE FROM PONDS
                   INHALATION DOSAGE
                   METABOLIC DECAY LOSS
                   EXCRETION LOSS
                   MEDATJON DOSAGE FROM L.  CASTANEUS 1
                   PREOATION DOSAGE FROM L.  CASTANEUS 2
                   PREDATION DOSAGE FROM T.  MIGRATORIUS

                   TOTAL PREOATION DOSAGE =  O.OOOE+OO  MG/MG
0.213E-05
O.OOOE+00
O.QOQEtOO
O.OOOE+00
O.QOOE+00
0.436E-06
O.OOOE+00
MG/MG
MG/MG
MG/MG
MG/MS
MG/MG
M6/MG
MG/MG
                  = O.QOOE+OO  MG/MG
                  = O.OOOE+00  MG/MG
                  = O.OOOE+00  MS/M6
Figure 8.1.   TEEAM  output for  the  example  application  (continued),
                                          291

-------
                   -UPTAKE  RESULTS FOR L. CASTANEUS 2
              INITIAL TISSUE CONCENTRATION = O.OOOE+OQ  MG/MG
              FINAL TISSUE CONCENTRATION   = 0.433E-05  MG/MG
              MAX. TISSUE CONC. FOR PERIOD = OJ33E-05  KG/KG

              BIOMASS OF ANIMAL GROUP      = 0.480E+06  G

              INITIAL PESTICIDE MASS       = O.OOOE+OQ  G
              FINAL PESTICIDE MASS         = 0.208E+01  G

              INITIAL TOTAL DOSAGE         = 0.0006+00  MG/MG
              FINAL TOTAL DOSAGE           = 0.415E-04  MG/MG

              + + + t LO-50 HAS BEEN EXCEEDED
              DOSAGE BREAKDOWN:
                   SOIL  DOSAGE            = 0.415E-04  MG/MG
                   PUNT DOSAGE           = Q.OOOE+00  MG/MG
                   PELLET DOSAGE          = O.OOOE+00  MG/MG
                   DOSAGE FROM PONOS      = O.OOOE+00  MG/MG
                   INHALATION DOSAGE      = O.OOOE+00  MG/MG
                   METABOLIC DECAY LOSS   = 0.371E-04  MG/MG
                   EXCRETION LOSS         = O.OOOE+00  MG/MG

                   PREOATION DOSAGE FROM L. CASTANEUS  1       = Q.QQOE+00  MG/MG
                   PftEDATION DOSAGE FROM L. TERRESTRIS        = O.OOOE+00  MG/MS
                   PREOATION DOSAGE FROM T. MIGRATORIUS       = O.OOOE+00  MG/MG

                   TOTAL PREOATION DOSAGE = O.OOOE+00  MS/MG


Figure 8.1.   TEEAM output for  the  example  application  (continued)
                                          292

-------
                  -UPTAKE RESULTS FOR T.  MIGRATORIUS
              INITIAL TISSUE CONCENTRATION = Q.OOOE+00  MG/MG
              FINAL TISSUE CONCENTRATION   = 0.238E-06  MG/MG
              MAX. TISSUE CONC. FOR PERIOD = 0.248E-06  MG/MG

              8IOMASS OF ANIMAL GROUP      = 0.536E+04  G
              INITIAL PESTICIDE MASS       =-•  O.OOOE+00  G
              FINAL PESTICIDE MASS         =  0.128E-02  G

              INITIAL TOTAL DOSAGE         =  G.OOOE+00  MG/MG
              FINAL TOTAL DOSAGE           =  0.648E-05  MG/MG
              DOSAGE BREAKDOWN:
                  SOIL DOSAGE            =  Q.977E-06  MG/MG
                  PLANT DOSAGE           =  Q.QOGE+00  MG/MG
                  PELLET DOSAGE          =  O.OOOE+00  MG/MG
                  DOSAGE FROM PONOS      =  O.OOOEtOO  MG/MG
                  INHALATION DOSAGE      =  0.302E-05  MG/MG
                  METABOLIC DECAY LOSS   =  0.575E-05  MG/MG
                  EXCRETION LOSS -        =  0.496E-06  MG/MG

                  PREOATION DOSAGE FROM L.  CASTANEUS  1
                  PREDATION DOSAGE FROM L.  TERRESTRIS
                  PREDATION DOSAGE FROM L.  CASTANEUS  2

                  TOTAL PREOATION DOSAGE =  0.248E-05  MG/MG
= 0.221E-05  MG/MG
= 0.185E-05  MG/MG
* 0.836E-07  MG/MG
         mmTOTAL FOOOCHAIN PESTICIDE  MASS          = 0.935E+01  G
         ++++-H-CURRENT FOODCHAIN MASS BALANCE ERROR    = 0.128E-04  G
         +++t++CUMULAT!VE FOODCHAIN  MASS  BALANCE ERROR = 0.333E-05  G

Figure 8.1.   TEEAM output  for  the  example application  (concluded)
                                          :93

-------
                                  SECTION  9

                             MODEL ARCHITECTURE
    The development of a framework for the model,  both in an architectural
and an implementational  sense,  is an important consideration.   The utility
of a piece of software can be judged by its ease of application in the
opinion of the user and  its appropriateness in solving the problem at
hand.  This framework has been  conceived with both of these factors in mind.

    The proposed spatial and temporal  basis for the model is designed to
meet the resolution required by a management-level or regulatory
application.  The architecture  of the model hopefully balances the
capability to apply computational modules in various combinations (to
achieve optimal flexibility in  simulating generalized ecosystems) with
simplicity at the user interfaces (inputs and outputs).  The design is
modular so that computational modules may be added or exchanged following
initial development with relative ease.

    The model is designed for use on an IBM PC compatible running MS DOS.  A
significant feature in this regard is the size of the code.  To be
compatible with the 640K byte working memory limitations of most machines of
this type, the code makes use of code overlays.  The code also uses scratch
files (as opposed to working memory) to save information to "restart"
habitats when making multiple habitat simulations and to initialize Monte
Carlo runs.

    This section is organized in the following way.  Section 9.1 discusses
code architecture, including operation sequences and module content.
Section 9.2 describes intermodule communication and Section 9.3, coding
conventions.  Section 9.4 discusses the use of files to store and transfer
information critical to model operation.

9.1  CODE ARCHITECTURE

    The TEEAM code consists of  over 150 subroutines organized into 11
modules.  Of the 11 modules, 5  perform computations; the others initialize
data, perform I/O functions or  provide generalized services to other parts
of the code.  The modules are (in alphabetical order):

    APUM - Computes Pesticide Exposure to Terrestrial Animals

                                   294

-------
    FSCBG - Performs Pesticide Spray Application Deposition Calculations

    INPREA - Performs Program Input Functions

    INITEM - Initializes Program Variables

    MCARLO - Executes Monte Carlo Simulations

    PLTGRN - Performs Plant Growth Computations

    PLTRNS - Performs Plant Uptake and Translocation Calculations

    SPECIO - Performs Specialized Output Functions

    TEEAMAIN - Main Program Which Controls Input and Program Execution

    UTIL - Contains Program Utility Subroutine

    Table 9-1 lists each module and lists and describes the function of each
subroutine it contains.  Due to its length, this table appears at the end of
this section.  The modules are directed by an execution supervision _
(EXESUP).  EXESUP in turn receives its instructions from BATENT, the batch
input processor and ultimately from TEEAM, the main program.

    A flow diagram of the TEEAM main program is given in Figure 9.1  The
purpose of the main program is to determine, by reading the run file, the
various options selected.  TEEAM version 1.1 can currently only operate in
the batch mode through the subroutine BATENT.  The main program passes
control to BATENT for reading data and thence to the execution supervisor.
The structure of BATENT is depicted in Figure 9.2

    The execution supervisor (EXESUP) is the main controlling code for
TEEAM.  This subroutine plus those depicted in Figures 9.1 and 9.2 comprise
the module TEEAMAIN.  The structure of EXESUP is depicted in Figure 9.3.

    Within EXESUP there are four major loops:

       • Monte Carlo
       • Year
       • Day
       • Habitat

Within the habitat loop TFAT, PLTGRN and PLTRNS are called each day.  FSCBG
and GRDDEP are called from within the day loop if a spray application is
                                    295

-------
indicated.  APUM is also  called outside the compartment  loop.   This is
because the concentrations  in the various media  (soil, plants, air, etc.)
need to be known for  all  habitats before the animal exposures  can be
calculated.  The daily  loop runs inside a year loop if the simulation
extends over several  years.  Currently, long-term simulations  are not
recommended because of  the  inability of the code to handle seasonal aspects
of animal behavior.   Finally, the entire deterministic model  is situated

                                     TEEAM
                                   Program Start
                                 CALL INITEM for
                                 determination of
                                 options selected,
                                 initialization
     Not currently
     implemented
        CALL INTENT
        for interactive data
        entry and model
        execution
                            N
CALL BATENT
for batch data entry
and model execution
                                  CALL CLOSIT
                                  to close all files
                                   f   STOP  J
                   Figure 9.1. TEEAM main program structure.

                                      296

-------
inside the Monte Carlo loop.   Inside  the  Monte  Carlo  loop,  random number
generators produce probabilistic realizations of  input  parameter sets.  In
this mode, the deterministic  model  is run a number  of times,  and frequency
distributions of various model  outputs can be generated.

    Figure 9.4 shows a schematic representation of  the  operation sequence
shown in Figures 9.1 through  9.3,  including the names of  the  subroutines
actually called within that sequence.

    The text that follows describes the structure of  computational  modules
and includes documentation for the  subroutines  within each  module.   Modules
that perform only reading, initialization, output,  and  general  utility
functions are not covered.

9.1.1  Pesticide Application/Deposition

    The models for describing the application and subsequent  transport and
deposition of pesticides are  incorporated into  the  FSCBG  and  GRDDEP
modules.  The pesticide spraying model is a modified  version  of FSCBG.  The
Module BATENT


CALL INPREA
to read and echo
input data


CALL EXESUP
to execute model


                               f RETURN  I
             Figure 9.2.  Batch input module (BATENT) structure.
                                   297

-------
^>

*N
^jr
*~s

' V
N S
rX

Module EXESUP










r
If spray a
v


	 	 1 Pall MOARI O
I
>Y

Call FSCBG





CallTFAT

Call PLTGRN


Call PLTRNS



Call APUM






— 	 I C/all MUAKLCJ
[ RETURN 1
Figure 9.3.  Execution supervisor (EXESUP) structure.
                        298

-------


















TEE AM 1—
























j— | IMTEM j—
















HRATIMT I—




















— j CLOSIT


GLBDAT
OPENF
OPECHO
ELPSE
CHKFIL



Hi
INPREA 1 —
I

























HEXESUP 1—
1






— j PRZMRD
— j PZECHO
— j PLTGRO
— j HABIO

_ CLOZEI
APINIT



APUMRD

— j RANSET
— INSTOR
— GRODEF
— CKAREA


— j CBGRD



- READM
- MCTRNS
- RANSET

- JNITMC
- RANDOM

- METRO

FSCDAT
- FSRED2
FSC8G


-1 GRDDEP


HABIO

THCALC
KDCALC
INITL
TFAT
PLTRNS
ANPR2M
OUTHYD
OUTPST
OUTTSR
OUTCNC
APUM
STATIS

- MCOUT
h

h
h









h








h



-

-




—


—




H
| PRZDAT

PLECHO
1 SEQIO





PRDRED



GRDRD
GRDOUT



FSECHO



FNDHOR
CHGHOR
SWITCH


Random
Numbers

RCHNUM



DEPRD
GRDAVE


SEQIO




STOUT
Figure 9.4  TEEAMAIN  program structure.
                299

-------
 GRDDEP  module  is  comprised  of  subroutines which  link  the  FSCBG  outputs to
 TEEAM.   These  modules  are called whenever a  pesticide spray  event  is
 simulated.   The user should be aware that pesticide can be applied directly
 into  habitats  at  specified  rates in the TFAT input sequence.  FSCBG/GRDDEP
 would be used  primarily  if  the user is  interested  in  quantifying drift of
 the spray to nontarget habitats.

 9.1.1.1  FSCBG.   The FSCBG  program, as  implemented for use with the TEEAM
 model,  consists of  three functional components  (each  implemented as one or
 more  subroutines):

        • Wake-Settling Velocity Model  (CBGS3)

        • Spray Dispersion Model  (CBGS6, CBGS7)
          -  dosage submodel
          -  deposition  submodel

        • Evaporation Model  (CBGS4)

    The  model is  called from EXESUP  and  is supplemented by several  input/
output routines.   The model  allows for  the simulation  of transport  of
sprayed  material  above  a  vegetative  canopy from  multiple spray lines or
swaths and can  be  selected for  use as modeling needs and data accessibility
merit.  Model components  are briefly described below.   Figure 9.5  shows the
structure of the  FSCBG  model as implemented  in TEEAM.

    The  Wake-Settling Velocity  submodel  permits  the user either  to  input the
aircraft wake-settling  velocity directly or  to input the aircraft weight,
wing span, and  ground  speed  which  the program uses  to  calculate  the aircraft
wake-settling velocity.  This subroutine models  the growth and position of
the pesticide cloud, during  the short period  immediately after release when
its shape is primarily  controlled  by the aircraft vortices.   For ground
spraying applications,  this  component  is not  used.

    There are two submodels  within the  spray  dispersion model.  The user has
the option of utilizing the  dosage and/or  deposition  submodels.   The
submodels can be  used  to  calculate dosage  and deposition  from a nearly
instantaneous elevated  line  source oriented  at an arbitrary angle with
respect to the mean wind  direction.   These submodels  will, if selected by
the user, calculate the dosage and/or  deposition at specified heights for
locations within  a rectangular grid.  Calculations by the  submodels are
restricted to below the height of elevated inversions.  In its present
configuration,  the FSCBG  computer program  can calculate the dosage, and
deposition at a maximum of  737 receptor nodes downwind from a maximum of  100
line sources.
                                     300

-------
 :  EXESUP
                     Figure 9.5  Module FSCBG  structure.
    Evaporation is a significant mechanism by which drop-size distribution
changes in time.  The evaporation model  allows the user to specify whether
or not drop evaporation is to be included in a calculation of dosage or
deposition.  If the user selects a drop  evaporation option, the program
automatically calculates the change in drop diameter with time, using a
polynomial expression fitted either to empirical  data or theoretical
calculations.

    The canopy penetration model available in FSCBG is not used for this
version of the TEEAM model.  In its place a simplified canopy penetration
model is used.

9.1.1.2  GRDDEP.  GRDDEP consists of three subroutines which are called from
EXESUP if a spray event has occurred on  a given simulation day.  Subroutine
GRDRD is called to read the FSCBG output file which contains the deposition
values at each receptor grid point.  Subroutine GRDAVE then averages the
deposition values within each habitat and subroutine AVEOUT writes the
GRDAVE output.

-------
9.1.2  Terrestrial  Fate and Transport

    The TFAT module performs terrestrial  fate and transport computations.
TFAT has been designed as a modular subprogram such that each important
computational step  is contained in a separate subroutine.  Central to the
module is a main subroutine or driver program (subroutine TFAT).  This
program calls all computational subroutines.

    Subroutines are grouped into two major categories by function.  These
functions can be generalized into those that:

       • Simulate hydrologic processes
       • Simulate pesticide fate and transport processes

    The subroutines which work together to perform each of these functions
are described briefly below.  Figure 9.6 depicts the system level
organization of these subroutines.  The subroutines are ordered logically
(as the main program would call them) within these sections.

    One-dimensional soil and surface water hydraulics are modeled by a
series of functional subroutines including:  PLGROW, HYDROL, POND, INFIL,
EVPOTR, HYDR1 or HYDR2, and EROSN.  These subroutines function as a group
and describe the behavior of water in the system, without regard to toxicant
chemistry.  Inherent in this functional demarcation is the assumption that
the chemical concentrations are low enough that any effect on the physical
properties of the solution are negligible.

    PLGROW is called first to simulate plant growth status and calculate
parameters which affect the hydrologic simulations (i.e., canopy
interception).  HYDROL then calculates surface hydrologic factors such as
runoff, plant interception, and snowmelt.  Subroutine EROSN computes the
soil loss from the habitat.  Potential and actual evaporation and
transpiration from the plant canopy, surface ponds, and the root zone are
computed in EVPOTR.  A precipitation event triggers the use of subroutines
INFIL and POND to simulate pond water hydraulics and infiltration on a
finer-than-daily time step.  INFIL estimates the infiltration capacity of
the soil through time allowing POND to simulate the evolution of ephemeral
surface ponds and the soil moisture profile and soil water velocity.  HYDR1
or HYDR2 calculate subsurface hydrologic factors such as pore velocities and
soil moisture content in each soil compartment in the absence of a
precipitation event or ponded surface water.

    The pesticide fate and transport subroutines activated by the
application of pesticide by the toxicant application/deposition model
include DEPMOD, PESTAP, GRANUL, VOLMT, CANOPY, CNCONC, PLPEST, PCHEM,
SLPEST, TRDIAG, and MASBAL.  Subroutine DEPMOD is called if a spray
                                     302

-------
application event occurs on a given simulation day.  It rearranges the
pesticide application information within the habitat,  inserts the
information for the spray event,  and then readjusts the pesticide habitat
information for the habitat.  If  a non-spray application of chemical occurs,
subroutine PESTAP is called instead.  PESTAP handles direct applications of
chemical to single habitats only, including soil  surface, soil
incorporation, and granular applications.  Subroutine  GRANUL then determines
the chemical  release rate from granules and quantity of pesticide remaining
in granules.   Subroutine VOLMT sets up terms for  simulation of vapor phase
movement and  volatilization of the pesticides from the soil in the absence
of surface ponding.  Subroutine CANOPY computes the canopy resistance to
vapor transfer out of the habitat by upwards diffusion.  PCHEM simulates
fate of chemical in ephemeral surface ponds and pesticide concentration in









TFAT


































Ol OQ^VA/
rLonUvv

uvnD(~>i
n YUnvJL

POND

EVPOTR
HYDR1 /
HYDR2
FRO9N

DEPMOD
PESTAP
GRANUL
PLPEST


VOLMi

PCHEM

SLPEST

MASBAL
CNCONC
PI TPRNI





— INFIL






f* A MODV
V/\N\Ji T



— TRDIAG



                     Figure 9.6  Module TFAT structure.

                                    303

-------
the ponded water.   PLPEST simulates  the degradation and  washoff of chemical
on plant foliage.   Pesticide transport through advection and dispersion, as
well as the various sources and  sink terms  are calculated in subroutine
SLPEST which predicts pesticide  concentration in soil  for the time step and
pesticide fluxes from the soil.   TRDIAG performs the Thomas algorithm to
solve the tridiagonal matrix set up  by SLPEST and SLTEMP.  MASBAL concludes
the module time step by calculating  the mass balance error for water and
pesticide in the system.  Finally,  CNCONC determines the average
concentration of chemical in the atmosphere within the plant canopy.

9.1.3  Plant Growth

    The alternate plant growth formulation, PLTGRN, is called as an option
from the TFAT code.  The structure,  the linkage of the subroutines, within
this module is presented in Figure  9.7.  The first routine called within
PLTGRN is PLTDEF,  which sets up  the  plant growth parameters for the current
plant type being simulated (the  parameters  for all plants in all
compartments are read into arrays before EXESUP begins the simulation).  The
next routine, PLTDAY, computes intermediate data which are computed once
each day (e.g., accumulated heat units and  photosynthyetically active
radiation).  PLTDAY calls the routine DAYLIT to interpolate the hours of
daylight for the current day.  This  daily datum is interpolated from the
TFAT monthly daylight hour data stored in the TFAT variable DT.  The actual
plant growth for the day is calculated by RK4GRO.  Subroutine DXGRO sets up
the system of differential equations.  The  last routine called by PLTGRN,
SAVLT, saves the results of this plant simulation into storage arrays for
use when this plant is to be simulated again.  Before returning control to
TFAT, PLTGRN computes the value for canopy  cover by using the complementary
error function ERFC located in the  UTIL'module.  At the end of PLTGRN
execution, the plant growth values  required by TFAT and the other modules
are passed back to the TFAT subroutine, PLGROW.
                              PLTDEF
                              PLTDAY
DAYLIT


INTERP
                   Figure 9.7  Module PLTGRN structure.

                                    304

-------
 9.1.4  Plant  Contaminant Transport

     The structure of the plant contaminant transport module, PLTRNS  (Figure
 9.8), is similar to the structure of PLTGRN.  The first routine called,
 DEFTRN,  sets  up the plant translocation parameters for the current plant
 type being  simulated.  The next routine called, TRNDAY, determines the daily
 constant data (e.g., daily transpiration rate).  The daily contaminant
 transport within the plant is calculated by TRNDIF using a fully implicit
                          PLTRNS
DEFTRN

TRNDAY

TRNDIF

SAVTRN
                 Figure  9.8   Module  PLTRNS  structure.
scheme.  The last routine called by PLTRNS,  SAVTRN,  saves the results of
this plant's simulation into storage arrays  for use  when this plant is to be
simulated again.  At the end of PLTRNS execution,  the plant translocation
values required by the other process modules within  TEEAM are stored in the
appropriate common arrays.

9.1.5  Terrestrial Animal Exposure

    Figure 9.9 illustrates the structure of  the Animal Exposure submodel.
The submodel consists of three primary components:

       • animal movement
       • intake rate regulation
       • animal feeding and toxicant assimilation

Subroutine MOVEM performs movement calculations for  each time step and calls
subroutine MARKOV to determine the distributions of  populations among
habitats and soil layers.  Subroutine APUM solves  the toxicant mass balance
equations by Runge-Kutta integration for the current concentrations of
toxicant in the tissues of each animal group.  APUM  is called after all
toxicant transport and animal  movement calculations  are completed for the
current time step.  Subroutine MDERIV is called by APUM to compute toxicant
mass derivatives with respect  to time; MDERIV also calls subroutines TOXASM,
FEED, and PREDAT to estimate assimilation, feeding,  and predation rates for
each animal group.  The pesticide intake rate for  the population biomass is
regulated using LD50, LD10 values in subroutine UPMOD.
                                     305

-------
    The animal  exposure submodel  also includes several  input, output, and
general utility subroutines which are shown in hatched  boxes in
Figure 9.9.  ANPRZM stores data needed by the submodel  from PRZM transport
calculations, including soil water contents and toxicant concentrations.
Subroutine  ANBAL performs mass balance calculations for the overall food
chain and writes organism toxicant concentrations to output files at the end
of each time step.  These subroutines are called directly by the execution
supervisor (EXESUP).

9.1.6  Monte Carlo Simulation

    Monte Carlo simulation is performed by running the  deterministic model a
number of times with randomly selected inputs.  The subroutines in module
MCARLO facilitate this simulation.  Of the subroutines  shown in Figure 9.10,
READM is called outside the Monte Carlo loop.  This subroutine determines
which TEEAM variables are specified by the user to be overwritten by
randomly simulated variables and reads in their distribution parameters.
Subroutine MCTRNS, called the first time outside the Monte Carlo loop
overrides parameter values in the input files with the  values of constants
in the Monte Carlo input file.  RANSET initializes the  random number
generators.  INITMC eliminates constants in the Monte Carlo file from the
          T-------I

          !  EXESUP  I-.
                        i  ANPRZM !
                        T------T

                        -I   ANBAL  I
                        i          i
                        L______J
                     Figure 9.9  Module APUM structure.

                                     306

-------
      EXESUP
      MCARLO "




























PC AP1M



MpTDMO



RANSET

IMITMP



RANDOM



MAXAVb

CT ATIC
b 1 A 1 Ib

IT
MCOUT

















	













pppuo


INCON
FNDHOR
CHGHOR
SWITCH



ncr^OMD
UtL/UMr

NMB
EXPRN
1 IMCDM
UNrHN
MTPV
TRANSB
ANRMRN







OTOI IT

























rnUI AB

/~\t ircon
UU 1 1 Ul i

.... rrnopi T
II HJI L 1
                   Figure  9.10  Module MCARLO structure.


list of Monte Carlo variables for the current run,  conditions the
correlation matrix and calls DECOMP to decompose the correlation matrix.
Within the Monte Carlo loop, random variable values are generated by any of
number of functions or subroutines called by RANDOM.  MCTRNS then transfers
these values to the proper variables in TEEAM.   After the TEEAM time loops
are complete, MCTRNS is called again to transfer the designated simulation
output variables to the post-processor routines.  MAXAVG .finds the maximum
values of an 'n1 day moving average of the output time series.  (The length
of the averaging period 'n1  is set by the user.)  STATIS accumulates the
statistics (sums, sums of  squares, etc.)  for the post-processors.

    The Monte Carlo loop ends here.  Once all Monte Carlo runs are complete,
MCOUT is called to write output.   STOUT computes means, moments, etc., of
the output cumulative distribution functions (CDFs) and OUTFOR produces
either tabular or printer  plot summaries  of the CDFs.
                                     307

-------
9.2  INTERMODULE COMMUNICATION

    Data transfers occur through the use of named common blocks,  subroutine
arguments and scratch files.   Common blocks normally contain data that are
associated by topic, i.e.,  meteorological  data are in a common block,
hydrology data are in a separate common block.  A list of common  blocks and
the topic of their contents is listed in Table 9-2 (see end of section).

    Arrays in common blocks are variably dimensioned.  Dimensions are given
in PARAMETER statements which are contained in INCLUDE files.  These INCLUDE
files,  their parameters and parameter default values are given in Table 9-3
(see end of section).

    Subroutine arguments are used rather sparingly.   They are primarily used
to pass system level information or miscellaneous variables that  are not in
common.  Occasionally, however, they are used to pass computational
variables.

    Scratch files are used to store major blocks of  data, primarily for
restarting simulations which have been temporarily suspended and
transferring data between program modules.  For more detailed descriptions
of the use of scratch files, the user should read Section 9.4.

9.3  COOING CONVENTIONS

    All of the modules contained in TEEAM were developed in, or upgraded to
ANSI Standard FORTRAN 77 (X3.9-1978) language.  The  major transformation
necessary in this conversion was to ensure that character data are stored in
CHARACTER type variables.  Some remnants of FORTRAN  IV code may remain in
the upgraded modules.

9.4  FILE UTILIZATION

    This section discusses the use of files for program operation.  Files
required include input, output, and scratch files.  Scratch files are useful
in program operation for two reasons.  First, by storing information in
external files versus memory, memory requirements are reduced.  This is
important due to the size of the program and the memory limitations of
PCs.  Second, scratch files are used to facilitate the simulation of
"multiple habitats."  Simulation of multiple habitats involves the execution
of the TFAT module,  and other modules, several times within the same time
loop iteration.  Obviously, if not saved, the information generated from a
TFAT execution would be overwritten when TFAT is next executed.   Instead of
saving all the information for each habitat in memory, it is written to
external sequential  files.  While this results in lower memory utilization,
it also slows program execution time.
                                     308

-------
    The files utilized for  the  program  are  defined in Table 9-3, along with
their default logical unit  numbers.   The  utilization of the input and output
files is self explanatory.  The scratch files  KHABS1 and KHABS2 are used in
the following way.  After the first  habitat (TFAT module)  is executed in
timestep 1, the results are written  to  the  KHABS1 file.  Results for
subsequent habitats are stacked sequentially in  the KHABS1 file.  At
timestep 2, the initial conditions are  read from the KHABS1 file and the
results are written to the  KHABS2 file.   At timestep 3, the initial
conditions are read from the KHABS2  file, results are written to the KHABS1
file and so forth.  Thus, these files are read from or written to every
other timestep.   The HABIO subroutine, which reads  and  writes  these files  at
each timestep, is also called after the initial  parameter  input  read  and
initialization,  to set up the unformatted initial  conditions  file  for the
first timestep.

    Another file which is utilized in a unique way  is  the  FSCBG  output file
(CBGOEP).   On days when a spray application event  occurs,  the  FSCBG program
writes an output file which contains pesticide deposition  information at
each receptor point in the FSCBG grid.  This file  is  read  by  subroutine
GRDRD called from the TFAT module inside the habitat  loop  of  the execution
supervisor.  For each habitat,   the grid/habitat  map overlay provided  by the
user is utilized to determine which grid points  are within the habitat
boundaries, and  the area-weighted average deposition  value for each habitat
(see Figure 9.11).  At the time of the next spray  application  event,  the
CBGDEP file is overwritten.
X X
X
X
X
X
X
X
X
X
X
X
X
X
X X
X
X
X
Habitat
X
X
X
X
X
X
X
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Habitat
X
X X
X
X
X
X
X
X
X X
X X
3 X
X X
X
X
X
Habitat
X
X
X
X
X
X
X
2
X
X
X
X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X X
                                                      X  = FSCBG Grid Output Point
                                                           NE, SW Habitat Delimiter
                    Figure  9.11  FSCBG linkage  to TEEAM  habitats.
                                     309

-------
Table 9-1.  LIST OF SUBROUTINES BY MODULE  AND DESCRIPTION  OF  THEIR
            FUNCTIONS
Module     Subroutine
                     Function
APUM
FSCB6
 APUM

MDERIV

TOXASM

 FEED

PREDAT

ANBAL

MOVEM
MARKOV

ANPRZM

UPMOD

INCON

CBGSO
FSCBG
CBGS3

CBGS4
CBGS6
Solves for Changes in Pesticide Concentrations
in Animals
Computes Chemical Mass Derivatives for Each
Animal Group
Computes Mass Derivatives Due to Uptake From
Fixed Media (i.e., soil, plants, air, ponded
water)
Computes Mass Derivatives Due to Feeding on
Lower Trophic Levels
Computes Mass Derivatives Due to Predation by
Higher Trophic Levels
Writes Pesticide Concentrations in Animals and
Performs Mass Balances
Computes Daily Animal Movement
Computes Population Distribution Among
Habitats or Soil Layers
Stores Data Needed by Animal Subroutine from
Results of Each Habitat Simulation
Modifies Uptake Rates to Account for
Exceedance of Lethal Dosages
Initializes Cumulative State Variables for
Each Monte Carlo Run
Driver Routine for FSCBG Subroutine
FSCBG Main Subroutine
Calculates Parameters Used in Evaporation
Model and Wake Settling Velocity
Drop Evaporation Model
Driver Routine for Dispersion Modules
                                   310

-------
Table 9-1.   LIST OF SUBROUTINES BY MODULE  AND  DESCRIPTION  OF  THEIR
            FUNCTIONS (continued)
Module
  Subroutine
     Function
INITEM
     CBGS7
     FSOUT
     IOPUT

     TRMVL

     REGRS
(Function) SIMUL

     TEST

     TESS
     DROPD
     FOFX
     LNPN1

     TITLR
     INITEM

     6LBDAT
     OPENF
     OPECHO
     ELPSE
     ECHOF
     CHKFIL
 Performs Dispersion or Dosage Calculations
 Prints Dispersion Model Calculations
 Prints Headings for Evaporation Model Input
 Data
 Calculates Terminal Velocities of Upper,
 Lower, and Average Drop Sizes Within a
 Category
 Performs a Second Order Polynomical Regression
 Solves a Set of Simultaneous Linear Equations
 Using Gauss-Jordan Elimination With Full
 Pivoting
 Prints Actual and Predicted Values of Below
 Canopy Evaporation
 Performs Variable Assignments
 Calculates Drop Diameters
 Sets Standard Deviation Values and Decay Terms
 Sets up Parameter Values for Dosage and
 Deposition Model Solutions
 Prints Table Headers
 Determines User-Defined Options, Opens Files,
 and Reads Global Data
 Initializes Global Data
 Opens Files
 Printing Utility
 Printing Utility
 Echoes the Names of Files Opened
Checks to See Which Files Have Been Opened
    311

-------
Table 9-1.  LIST OF SUBROUTINES AND DESCRIPTION OF THEIR FUNCTIONS
            (continued)
Module
Subroutine
     Function
 INPREA
  INPREA

  PRZMRD
  PRZDAT
  PZECHO
  PLTGRD
  PLECHO
   HABIO

   SEQIO

  CLOZE1
  APINIT
  APUMRD

  PRDRED
  RANSET
  INSTOR

  GRDDEF

   GRDRD
  GRDOUT
  CKAREA

   CBGRD
Driver Subroutine for Reading and Initializing
Program Input Data
Reads and Checks TFAT (PRZM) Input Data
Initializes TFAT (PRZM) Data
Echoes TFAT (PRZM) Data to an Output File
Reads and Checks Plant Growth Input Data
Echoes Plant Growth Input Data
Reads and Writes Data to a Scratch File for
Restarting Habitats
Reads and Writes the Maximum Number of Records
Possible to an Unformatted Scratch File
Closes One File
Initializes Animal Model Variables
Reads, Checks, and Writes Animal Model Input
Data
Reads Animal Predation Data
Initializes the Random Number Generator
Stores Initial Conditions for the Animal Model
for Monte Carlo Simulation
Drives Subroutine for Reading and Printing
Habitat/Receptor Grid Geometry  .
Reads Habitat/Grid Receptor Geometry
Writes Habitat/Grid Receptor Geometry
Checks the Error Between TFAT Input Habitat
Areas and GRDDEF Calculated Habitat Areas
Driver Subroutine to Read FSCBG Input Data
(See also FSCDAT and FSRED2 described in
Module TFAT)
      312

-------
Table 9-1.   LIST OF SUBROUTINES AND DESCRIPTION  OF  THEIR  FUNCTIONS
            (continued)
Module
Subroutine
     Function
MCARLO
  FSECHO
  ANRMRN
  DECOMP

   EXPRN

  FRQPLT

  FRQTAB

  INITMC
             FNDHOR

             CHGHOR

             SWITCH

             MCECHO

             MCTRNS

              MTPV

             OUTFOR
             OUTPUT
 Echoes FSCBG Input Data
 Generates  Normally Distributed  Random Numbers
 Decomposes Correlation Matrix into Coefficient
 Matrix Required  to Generate  Correlated Random
 Numbers
 Generates  Exponentially Distributed Random
 Numbers
 Writes Plots of  Cumulative Frequency
 Distributions
 Writes Tabulated Cumulative  Frequency
 Distributions
 Initializes Statistical  Summation  Arrays,
 Reorganizes Monte Carlo Input Arrays to
 Account for Constant  Variables,  and Performs
 Other  Miscellenous Monte Carlo  Initializations
 Finds  the  Number of the First Layer in a Soil
 Horizon
 Assigns a  Parameter Value to All Layers in an
 Horizon
 Switches Monte Carlo  Parameter  Value for TEEAM
 Parameter  Value
Writes  Monte  Carlo  Input  Data to the Shell
Output  File
Transfers Values Between  Monte Carlo Generated
Arrays  and TEEAM Variables
Multiplies a Vector of Uncorrelated Variables
by a Coefficient Matrix to Form a Vector of
Correlated Variables
Writes  Frequency Tables and Plots
Writes Out Statistical Summaries of Monte
Carlo Runs to the Monte Carlo Output File
                                   313

-------
 Table 9-1.   LIST OF SUBROUTINES  AND  DESCRIPTION OF THEIR FUNCTION
             (continued)
 Module
Subroutine
      Function
PLT6RN
             RANDOM


              READM


             STATIS



              STOUT
 TRANSB


 TRANSM



  UNFRN


  MCOUT

 MAXAVG


  INCON


 PLTGRN

  DXGRO

 PLTDAY


 DAYLIT
Generates a Vector of Random Numbers From
Specified Distributions

Reads Monte Carlo Input Data From a User-
Specified Input File Unit Number

Performs Summations Required to Compute
Statistical Moments for Random Model Inputs
and Model Outputs Over All Monte Carlo Runs

Computes Statistical Moments (mean, standard
deviation, skewness, kurtosis, correlations,
minimum and maximum) From Summations Computed
by STATIS.  Statistics are then Written Out to
the Monte Carlo Output File

Transforms Normally Distributed Number to an
SB Distributed Number

Transforms Normally Distributed Numbers to
Numbers Having the Appropriate User-Specified
Distributions (i.e., log-normal, SB)

Generates Uniform Random Numbers Ranging
Between 0 and 1

Writes Monte Carlo Summary Statistics

Computes Moving Average Values of Time Series
Variables

Restores Initial Conditions for APUM Module at
the Beginning of a Monte Carlo Run

Driver Subroutine for Plant Growth Module

Calculates Plant Growth Differentials

Calculates Daily Information for Plant Growth
Simulation

Interpolates the Number of Daylight Hours on a
Daily Basis From Monthly Data
                                     314

-------
Table 9-1.  LIST OF SUBROUTINES AND DESCRIPTION OF THEIR FUNCTIONS
            (continued)
Module     Subroutine
Function
PLTDEF
INTERP
SAVPLT
RK4GRO
PLTRNS
'LTRNS DEFTRN
TRNDAY
TRNDIF
SAVTRN
FAT TFAT
PESTAP
PLGROW
PLPEST
SLPEST
TRDIAG
EROSN
Retrieves Data for Current Plant From Storage
Arrays
Linear Interpolation Routine
Writes Data for Current Plant Into Storage
Arrays
Fourth-Order Runge-Kutta Routine to Solve
Differential Plant Growth Equations
Driver Subroutine for Plant Translocation
Modu 1 e
Retrieves Translocation Data for Current Plant
from Storage Arrays
Calculates Daily Information for Plant
Translocation Simulation
Solves Differential Equations for Plant
Translocation Using a Fully Implicit Scheme
Stores Translocation Information for the
Current Plant into an Array
Driver Program for Terrestrial Fate and
Transport Calculations
Performs Chemical Application (not a spray
event)
Performs Plant Growth Calculations Per
Original PRZM Code if PLTGRN is OFF or Calls
PLTGRN
Performs Fate and Transport Computations for
Pesticides on Plant Foliage
Sets Up Arrays for Tridiagonal Matrix Solution
of Fate and Transport in Soil
Solves Tridiagonal Matrix Using the Thomas
Algorithm
Performs Soil Erosion Calculations
                                 315

-------
 Table 9-1.  LIST OF SUBROUTINES AND DESCRIPTION OF THEIR FUNCTIONS
             (continued)
 Module     Subroutine
                       Function
SPECIO
EVPOTR
 HYDR1

 HYDR2

MASBAL

HYDROL
 INFIL

 POND
 PCHEM

 VOLMT

CANOPY

CNCONC

DEPMOD

GRANUL

 METRO
RCHNUM

GRDDEP

 DEPRD
Computes Soil Evapotranspiration
Performs Hydraulic Computations for Freely
Draining Soils
Performs Hydraulic Computations for Soils With
Restricted Drainage
Performs Water and Pesticide Mass Balance
Computations
Performs Surface Hydrologic Computations
Performs Infiltration Computations Using a
Green-Ampt Model
Computes Depth of Surface Ponding
Performs Chemical Fate Calculations for
Surface Ponds
Performs Computations for Pesticide
Volatilization From Soil
Computes Vegetative Canopy Resistance to Mass
Transfer
Computes Average Pesticide Concentration
Within the Plant Canopy
Sets and Resets TFAT Pesticide Application
Information When Spray (FSCBG) Events Occur
Determines Chemical Release Rate from Granular
Pesticides
Reads and Checks Meteorological Data
Computes Richardson Number from Meteorological
Information
Driver Subroutine to Retrieve Deposition Data
from FSCBG Output Files
Reads FSCBG Output Deposition File
      316

-------
 Table 9-1.  LIST OF SUBROUTINES AND DESCRIPTION OF THEIR FUNCTIONS
             (continued)
 Module
Subroutine
       Function
UTIL
 GRDAVE

 AVEOUT
 FSCDAT
 FSRED2
 THCALC

 KDCALC

 OUTHYD
 OUTPST
 OUTTSR
 OUTCNC
  INITL
 ERRCHK

 TRCLIN
 LFTJUS
 SCREEN
 CENTER
 DONBAR

 COMRD

 COMRDZ
 Averages  the  Deposition  Values  in  FSCBG
 Receptor  Grid by Habitat and  Computes  Standard
 Deviation of  Deposition  Values
 Writes  GRDAVE Output
 Data  Initialization Routines  for FSCBG
 FSCBG Input Data Reading Routine
 Generates Water  Content  (Field
 Capacity/Wilting Point)  from  TFAT  Input Data
 Calculates Pesticide Distribution  Coefficient
 Values  from TFAT Input Data
 Writes  Habitat Hydrologic Summaries
 Writes  Habitat Pesticide Summaries
 Writes  Time Series Plotfile Data
 Writes  Habitat Pesticide Concentration Data
 Initializes TFAT Variables
 Writes  Error  Messages, Closes Files if Fatal
 Error
 Tracks  Line of Program Being  Executed
 Left  Justifies a Character String
 Screen  Manager Routine
 Centers a  Character String
 Tracks  the "Percentage Completeness" of the
 Program
Allows User to Insert Comments in Data Files
by Ignoring Comments When Reading
Comment Reading Routine Which Handles End-of-
File Read
                                   317

-------
Table 9-1.  LIST OF SUBROUTINES AND DESCRIPTION OF THEIR FUNCTIONS
            (concluded)
Module
Subroutine
Function
            FILCHK
            ADDSTR

            NAMFIX

            BMPCHR
            SUBIN
            SUBOUT
            EXPCHK
            LOGCHK
             ERFC
             ERF
            RELTST
                 Checks "Open"  Status of File
                 Adds  a Character String to Another Character
                 String
                 Left  Justifies and Capitalizes  a Character
                 String
                 Capitalizes  a  Character String
                 Routine Which  Tracks Entry to Subroutines
                 Routine Which  Tracks Exits From Subroutines
                 Test  for Exponent Underflow/Overflow
                 Checks for Logarithm of a  Negative Number
                 Complimentary  Error Function
                 Error Function
                 Converts Real  Numbers From Double to Single
                 Precision and  Checks Overflow/Underflow
                                    313

-------
Table 9-2.  COMMON BLOCK NAMES, TOPICS,  AND INCLUDE FILE NAMES
Common Name
ANIM
ANIC
CROP
ACCUM
HYDR
MET
MISC
GLOBAL
CHMISC
PEST
PON
CANOP
FIXED
DEPDAT
INITCM
MOVE
MSBAL
CMSBL
P6CMP
PGDAY
PGDEF
PLANTS
PLFLUX
RANDM
SPRAY
TRACE
FSDAT
Topic Description
Pesticide Uptake by Animals
Label for Animal Model
Crop Growth and Pesticide Deposition
Summary Output Accumulators for TFAT
Hydrology Information for TFAT
Meteorological Information
Miscellaneous Variables (i.e., flags, etc.) for TFAT
Global Simulation Control Data (dates)
TFAT Miscellaneous Titles, Output Control Data
Pesticide Fate and Transport for TFAT
Ponding and Pond Chemistry Information
Pesticide Canopy Penetration Information
Pesticide Application Data (TFAT)
Information Required to Link FSCBG/TFAT
Initial Conditions for Animal Exposure Module
Animal Movement Data
Mass Balance Information for Animal Model
Labels for Animal Movement Model
Plant Growth Data
Daily Plant Growth Parameters
Plant Growth Data
Buffer for Plant Model Restart
Plant Information to be Written Using Subroutine OUTPST
Used in Random Number Generator Initialization
Canopy Penetration Data for Spray Application Events
Subroutine Entry/Exit Data
Spray Application Data
Include
File Name
CANM
CANM
CCROP
CCUM
CHYDR
CMET
CMISC
CMISC
CMISC
CPEST
CPOND
DOSCAN
DOSGRN
GRID
INITCM
MOVE
MSBAL
MSBAL
PGCMP
PGDAY
PGDEF
PLANTS
PLFLUX
RANDM
SPRAY
TRACE
FSDAT
519

-------
Table 9-3.  PARAMETER STATEMENTS,  PARAMETER DEFINITIONS, AND  INCLUDE  FILE  NAMES
INCLUDE
File Name File Usage Parameter
APARM Animal Model NL
NH
NS
NM
NSG
NG
NTS
CMPLR.PAR Compiler REALMX
Specific
REALMN
MAX I NT
MAXREC
EXNMX
EXPMX
EXPMN
PCASCI
Default Value
5
4
5
5
3NH
NL + NSG
10
1.0E23
l.OE-23
2147483647
512
-53
53
REALMN
TRUE
Parameter Description
Maximum Number of Predator Groups
Maximum Number of Habitats
Maximum Number of Soil Horizons
Maximum Number of Fixed Media
(i.e., soil, water, plants,
granules, seeds)
Maximum Number of Soil Animal
Groups
Maximum Total Number of Animal
Groups
Maximum Number of APUM Time
Series Plots
Maximum Real Number
Minimum Real Number
Maximum Integer
Maximum Length of Unformatted
File Record
Minimum Exponent
Maximum Exponent
Maximum Result of Exponentiation
Logical Variable to Indicate if
                           CHR1WT
CPARM.PAR   TFAT Module     NCMPTS
TRUE
,CMPTS
NAPP
NLVL
NC

33
5
10
5
320
ANI.SYS is Present  (if true,
ANSI.SYS is present and machine
is IBM PC compatible)

Logical Variable to Indicate if
Compiler Writes First Character
in a String to the Screen (if
true, character is written)

Maximum Number of Soil Layers
Plus 1

Maximum Number of Pesticide
Applications

Maximum Number of Canopy
Deposition Levels

Maximum Number of Crops

-------
Table 9-3.  PARAMETER STATEMENTS,  PARAMETER DEFINITIONS AND INCLUDE FILE NAMES
            (continued)
INCLUDE
File Name File Usage
GRID. PAR FSCBG
Receptor Grid


HABIO Scratch File
Array Sizes
for Restarting
Habitats













Parameter
NH1
NGRDSZ
NGRDSQ
NRLMIS
NI2MIS
NCHMIS
NMIS
NRLCRP
NI2CRP
NCRP
NRLPST
NI2PST
NDPPST
NPST
NRLHYD
NI2HYD
NHYD
Default Value
NH+1
25
NGRDSZ2
NCMPTS+18
NCMPTS+58
478
NI2MIS+2
(NRLMIS)
10+4 (NC)
+NLVL(NAPP)
+2(NAPP)
8+9 (NC)
+ NAPP
NI2CRP+
2(NRLCRP)
32+14(NCMPTS)
+2(NAPP)
3+2 (NAPP)
5(NCMPTS)
NIZPST+
2(NRLPST)+
4(NDPPST)
22+15(NCMPTS)
+3(NC)
8+10(NC)
NI2HYD+
2(NRLHYD)
Parameter Description
Maximum Number of Habitats Plus 1
Maximum Number of Grid Rows/
Columns
Number of Grid Rows/Columns
Squared (I.e., Maximum Number of
Grid Points)
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
                                        321

-------
Table 9-3.   PARAMETER STATEMENTS,  PARAMETER  DEFINITIONS AND INCLUDE FILE NAMES
            (continued)
INCLUDE
File Name File Usage







IOUNIT.PAR I/O Unit
Defaults
Parameter
NRLCUM
NCUM
NRLPON
NI4PON
NPON
NRLPLT
NPLT
KRUN
KPGDEF
Default Value
76+20 (NCMPTS)
2(NRLCUM)
217
1
2(NRLPON
+ NI4PON)
NMPLNT
(NSVAL + NPARAL)
2(NRLPLT)
7
11
Parameter Description
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Equivalence Arrays
Unit No. for Run Parameters
Unit No. for Initial Conditions
                           KPZDEF
13
Input
Files



Output Files





KMCIN
KPMET
KAPIN
GRDEF
CBGDEP
KUOUT
KECHO
KRSLTS
KMCOUT
KMCOU2
KFLCN
15
9
17
19
21
0
8
10
12
26
14
                            KFLWT
16
and Definition Parameters (Plant
Growth)
Unit No. for Initial Conditions
and Definition Parameters (PRZM)
Unit No. for Monte Carlo Data
Unit No. for Time Series
(Meteorological) Data
Unit No. for Animal Model Input
Unit No. for Grid Definition
Unit No. for FSCBG Output
Unit No. for Output to User
Unit No. for Echo of Input Data
Unit No. for Animal Time Series
Unit No. for Monte Carlo Output
Unit No. for Monte Carlo Output
Unit No. for Habitat
Concentration Output
Unit No. for Hydrology Output

-------
Table 9-3.  PARAMETER STATEMENTS, PARAMETER DEFINITIONS AND INCLUDE FILE NAMES
            (concluded)
INCLUDE
File Name File Usage Parameter
KFLPS
KFLTS
KAPOUT
Scratch Files KHABS1
KHABS2
KMCINT
NMXFIL
PLANTS. PAR Plant Growth NMPLNT
and Trans location
NPARPG
NPARPT
NPARAL
NSVPG
NSVPT
NSVAL
PLTGRN.PAR NEQN
Default Value
18
20
24
23
25
27
20
10
10
6
NPARPG
+NPARPT
10
2
NSVPG
+NSVPT
9
Parameter Description
Unit No. for Habitat Pesticide
Output
Unit No. for Habitat Time Series
Output
Unit No. for Animal Model Output
Unit No. for Multiple Habitat I/O
Unit No. for Multiple Habitat I/O
Unit No. for Monte Carlo Initial
Conditions
Maximum Number of Files Which Can
be Open
Maximum Number of Plant Types
Number of Parameters for Plant
Growth
Number of Parameters for Plant
Trans location
Total Number of Plant Growth
Parameters
Number of State Variables for
Plant Growth
Number of State Variables for
Plant Translocation
Total No. of Plant Translocation
Parameters
Number of Differential Equations
                                                    Solved  for Plant Growth

-------
                                 SECTION 10

                        SIMPLE MODELS FOR PREDICTING
               TOXICANT ACCUMULATION IN TERRESTRIAL WILDLIFE:
                                   ATEEAM
10.1  MODEL DESCRIPTION

    The Analytical Terrestrial  Ecosystem Exposure Assessment Model  (ATEEAM)
code solves for pesticide concentrations in three components of a
terrestrial foodchain; the soil, a soil  dwelling organism and an upper
trophic level predator.  The system represented is shown in Figure  10.1.
There are three variations on the system shown in the figure.  These are
designated models 1 through 3 and are described below.  In all cases, the
soil mass, animal biomass and rate constants are assumed to be constant.
The equations are solved analytically which minimizes execution time.
Solutions to the equations are  given in  Section 10.2.  The code has an
interactive preprocessor which  allows the user to operate the model in
either deterministic or Monte Carlo mode.  This is discussed in
Section 10.3.

10.1.1  Model 1.  Single Toxicant Application with First-Order Soil Decay

    Model 1 assumes that a single toxicant is applied to an agricultural
field and undergoes first-order decay in the soil.  The chemical is absorbed
or otherwise taken up by a soil dwelling organism which can be in turn
ingested by a higher trophic level predator (bird or small mammal).

    The model for predicting pesticide concentration in the soil is

         dC
      Ms dT = ' kiCsMs - k'CsMs - keCsMs                            (1CM)

in which

    MS is the soil mass (kg)                                   .
    Cs is the concentration of  the toxicant in the soil (yg kg" )
    k^ is the first-order decay term for the toxicant in the soil (day  )
    k2 is the first-order assimilation rate of pesticide (via absorption.
       ingestion or other processes) by the soil dwelling organism (day"1)

                                     324

-------
    kg is the soil  ingestion rate  by the  predator  (day"1).
This model  assumes that the soil  mass,  MS,  does  not  change  over  time  t.   The
model can be rewritten so that
     dC
= - (k,
k6) C
                    2     6   s

by dividing both sides through  by the  soil  mass.
                                                                     (10-2)
   Decay


   -k5CbMb
   Decay


   -k3CoMo
    Decay
                                Predator Biomass
        Concentration =
                       Mass
                      = M
                                             Predation
                                             -k4CoMo
                                Soil Dwelling
                                Organism Biomass
                        Concentration
                        Mass
                       = C
                                          Soil
                                          Ingestion
                                          -k6CsMs
                                              Bioconcentation
                                              -k2CsMs
         Soil

         Concentration =
                                 Mass
                       = M;
              Figure  10.1.   Schematic  of  A-TEAAM Model  Structure
                                      325

-------
    The contaminant mass balance  equation  for  the  soil dwelling organism  is

         dCo
      Mo dT = k2CsMs - k3CoMo -  k*CoMo                               <10"3)

in which

    MQ is the biomass of the organism  living in  soil mass, MS, lg)
    CQ is the concentration of toxicant  in the organism   (yg g  )
    kg is the first-order depuration rate  of the toxicant by the  soil
       dwelling organism (day'1)  (Note that the  model assumes that  any
       toxicant returned to the soil in  this process is  unavailable for
       reingestion, etc.)
    k4 is the predation rate by the higher level organism (day  ) which  is
       accumulated to the next trophic level

This equation can also be simplified to
     dCn   k,CM
     dT = -M^ - 

The mass balance equation for the predator (bird or small mammal) is given
by

         dC.
      Mb dF = k^CoMo - ksCbMb + keCsMs                              <10-6

in which

    Mb is the mass of the  bird or  small  mammal  (kg)          _1
    Cb is the concentration of toxicant in the  tissue   (yg kg" )^
    k4 is the predation rate on the soil dwelling organism (day" ) and   ^
    k5 is the depuration rate of the  bird  or mammal of  the toxicant (day"
                                     326

-------
The model assumes that the mass of the bird or small mammal stays constant
over the simulation time 't'.  The model is simplified by dividing through
by the predator biomass giving
     dt
Again, for convenience, the mass terms on the right hand side can be written
in terms of densities, yielding

     dC.       C_pm      C_p_ AZ
in which

    p,  is the biomass density of the predator over the area (A) of concern
       (kg A~*).  To be consistent with equation (10-5) A must be in m  .

Also, for ease of parameter estimation the term k^MQ can be thought of  as a
zero-order predation rate (g day  ).

10.1.2  Model 2.  Continuous Toxicant Application with First-Order Decay

    Model 2 is the same as Model 1 which the exception of the description of
toxicant accumulation in the soil.  The equation for continuous application
with first-order decay is

         dCs
      Ms dT = L - k!CsMs - k2CsMs - keCsMs                          <10-9)

where L is the application rate (yg day" ) and all other terms are as
defined previously.

10.1.3  Model  3.   Steady-State  Concentrations Under Continuous Deposition

    The governing equations  for the  steady-state concentrations are simply
found by setting the left hand  sides  of  equations  (10-9),  (10-3),  and  (10-6)
to zero.

10.2  SOLUTION OF MODEL EQUATIONS

The solutions to the system  of  ordinary  differential equations which
comprise models 1 and 2 are  given  below:
                                     327

-------
MODEL 1

Soil
      Cs - C^ e~at                                                  (10-10)
in which   a,= kt  + k2  + k6   and
           C  = the in

Soil Dwelling Organism
C  = the initial  pesticide concentration in the soil.
in which  b,= k3 + k^ and
          C  = the initial  pesticide concentration in the soil  dwelling
                organism.

Predator


      Cb = Cb e" ^

      ,  MS ,     ^ ^            ke        •     at    -"**
      + Mb l(b - a) (ks - a)  + (k5  - a)'  Ls  (e     ' e    }


      + IkT^T M^ Co + (ks - b)  (a  - b)  MJ;  Csl  (e"bt - e  5 )       (10-12)

           i
in which  C.  is the initial pesticide concentration in the predator.


MODEL 2

Soil

      c   =   L   (1  _  e-atj                                           (10-13)
        s   a s
in which  a = kt +  k2 + k6

Soil Dwelling Organism
             k,L        ka-at    aa-bt
      c   =    2    [i  + £g	^ae—]                                (10-14)

in which  b = k3 +  k4
                                   328

-------
Predator
            k  k  I       k  I          -k t
            K2 K4 L      Kg L          K&t

      Cb = lks ab Mb + k5 a M^ ^ " e    ^


                    (\olNi.L.              1\ £ L.
                                                     _
             (a(k5 - a) (a - b) Mb - a(k5 - a) M^ (e    ' e    )
                        U ^            k4-    ~

                                      '
                   - bHa - b)

MODEL 3
Soil
      Cs -
where  a = k: + k2 + k6
Soil Dwelling Organisms
           ^IL_
       o ~ ab M
where  b = k3 + k4
      c  •
Predator

             I
      cb =
            U  J

10.2.1  Solution Behavior

The solutions to the governing equations (e.g.,  equation 10-11)  often have
terms of the form
     Q-at    -bt
     6  a"-b                                                       <10-19>
which are indeterminant in the limit as a approaches b.   The limit can be
found using L'Hopitals Rule by which

              t     bt   — fe~at   e~btl
      Hm  e"   - f   . aaj	1	[ = _t -at                  (10_20)
      a,b     a - b      |_ [a _ b]


Equation (10-20) shows that the limit exists and therefore that  the solution
in the limit is well behaved.

                                     329

-------
The practical consequence of this behavior is that the solutions provided
give the exact solutions for all  cases except when a identically equals b.
Such a condition is of no interest practically as there is no a priori
reason why the sum of the rate constants represented by a and b would be
equal.  If for some reason the user would like to examine such a case, the
value of either a or b should be  perturbed by a small amount (perhaps one
percent).

10.3  MODEL ACQUISITION, INSTALLATION, AND EXECUTION

    The ATEEAM model is available on 5-1/4-inch floppy diskettes for MS-DOS
compatible computers.  The model  can be obtained by contacting Dr. David S.
Brown at the EPA's Athens Georgia Environmental Research Laboratory (US EPA,
Environmental Research Laboratory, Assessment Branch, College Station Road,
Athens, GA 30613).

    The model diskette contains the model FORTRAN code (source and
executable) and sample input and  output files.  A listing of the files
contained on the disk is given below.
       ANIMAL.FOR                    ATEEAM FORTRAN code

       ANIMAL.EXE                    ATEEAM executable code

       DECAY.INC                     Common blocks containing pesticide
                                     decay and uptake rates.  This file is
                                     INCLUDED in ANIMAL.FOR during
                                     compilation.

       MCARLO.INC                    Common blocks containing the system
                                     descriptive data.  It is INCLUDED in
                                     ANIMAL.FOR during compilation.

       MODELLING                    Common blocks containing the model
                                     control  parameters.   It is INCLUDED in
                                     ANIMAL.FOR during compilation.

       PARAM.INC                     Parameter statements defining the size
                                     of arrays dimensioned in ANIMAL.FOR.
                                     It is INCLUDED in ANIMAL.FOR during
                                     compilation

       SAMPLE.IN                     Sample input file for batch mode
                                     operation.
                                     330

-------
       SAMPLES.OUT                   Sample output file produced when
                                     running in batch mode using SAMPLE.IN

       SAMPLE I.OUT                   Sample output file produced when
                                     running in interactive mode using
                                     default values.

10.3.1  Hardware

    To run the ATEEAM model an IBM PC, XT or AT compatible computer with one
floppy disk drive is required.  A hard drive is not required but is
advisable.  If running without a hard disk the floppy disk receiving the
model output must have enough storage remaining to contain all of the
output.  When running in Monte Carlo mode this output can be considerable,
depending upon the number of Monte Carlo runs and the number of variables
randomly generated.  At a minimum, the main output file requires 90 Kbytes
of storage.

    The executable file (ANIMAL.EXE) included with the model requires a math
coprocessor chip, either an INTEL 8087 or 80287.  If one is unavailable the
source code can be recompiled using a compiler which does not require a
coprocessor.

10.3.2  Software

    The only software required to run the ATEEAM model, besides an
executable version of the model, are:

       *  DOS

       *  ANSI.SYS

    The ANSI.SYS device driver is included with DOS.  This driver is
installed by inserting the command:

                              DEVICE = ANSI.SYS

in the CONFIG.SYS file.  If the CONFIG.SYS file does not exist it can be
created using either a text editor or word processor or the DOS copy
command.  The user should consult a DOS manual if more information is
needed.

10.3.3  Installation and Execution Instructions

    If a hard drive is available, copy the model from the diskette to the
hard drive.  It is best if the model is kept in its own directory on the

                                    331

-------
hard drive.  In order to create a directory, after booting the computer
type:

       MD ANIMAL
       CD ANIMAL

    To copy the model onto the hard disk, insert the floppy into the A:
drive and type:

       COPY A:*.*

    If a hard drive is not available, make a copy of the diskette containing
the model.  Save the original and use the copy.  To run the model, insert
the copy of the model diskette into the A: drive.

    To run the model for either case discussed above, type:

       ANIMAL

    A title screen will appear and the user will be prompted for information
to run the model.

10.4  MODEL INPUT SEQUENCE DEVELOPMENT

    This section discusses the development of batch input sequence and
interactive model operation.

    Each of the model parameters are assigned default values in the model.
Tables 10.1 and 10.2 list these values.  If the user wishes to run the model
with a different set of parameters there are two methods available to change
these values, either by reading new values from a file (batch mode) or by
changing them interactively.  A discussion of the formats required for the
batch mode and a sample run through the interactive mode follows the
description of the  input parameters.

10.4.1   Input Parameter Description

    There are two basic types of input data, Simulation Control Parameters,
and System Descriptive Parameters.

    There are 12 simulation control parameters.  These 12 variables are
listed below along  with a short description of each.

       NMCR                          Number of Monte Carlo simulations to
                                     perform.
                                    332

-------
       NCAT
       THRSHO
       THRSHB
       DOSAG1
       DOSAG2
       Starting and
       Ending Dates
Number of periods to divide the output
into.  The length of these periods are
defined by the user,  (see Section 10.7
for more details).

Lethal whole body concentration for
the lower trophic level animal.  This
value corresponds to a concentration
such as an LC50 (tissue concentra-
tion).  It is only used for comparison
to whole body concentration generated
by the model.

Lethal whole body concentration for
the higher trophic level animal.  This
value corresponds to a concentration
such as an LC50 (tissue
concentration).  It is only used for
comparison to whole body concentration
generated by the model.

Lethal dosage for the lower trophic
level animal.  It is used for
comparison with model output dosages.
It corresponds to a value such as the
LD50.

Lethal dosage for the higher trophic
level animal.  It is used for
comparison with the model output
dosages.  It corresponds to a value
such as the LD50.

These six control parameters are the
starting and ending day, month and
year of the simulation.
The simulation period can be divided into a number of intervals or periods
for output.  Dates corresponding to the end of each period except for the
last are input.  The last period ends on the last day of the simulation.
This results in NCAT - 1 periods.

    System descriptive parameters are those which appear in the model
equations.  In general, these parameters are defined by name, mean,
coefficient of variation, minimum and maximum allowed values and
distribution type.  If a distribution is defined for a parameter the value
                                 333

-------
    TABLE 10.1.  DEFAULT VALUES FOR ATEEAM MODEL CONTROL PARAMETERS
VARIABLE NAME
NMCR
NCAT
NP
THRSHO
THRSHB
DOSAG1
DOSAG2
MODEL1
DEFAULT VALUES
100
3
6
20
4000
45
65
SINGLE APPLICATION
of the parameter used by the model is randomly generated based upon the
distribution type and distribution parameters.  If the distribution type is
constant then the parameter takes the mean value.  If all of the parameters
are input as a constant then the model defaults to a deterministic mode.
More details on-the differences between Monte Carlo mode and deterministic
mode are given in Section 10.6.  Below is a list of the parameter names
which corresponds to designations used in Sections 10.1 and 10.2.
    PARAMETER NAME                     PARAMETER  DESIGNATION
                                  (corresponding to Equations in
                                           Section  10.1)
         DECAY1                                 14
         RATE 12                                 k
                                        2

RATE23                                 k
         DECAY2                                 k3
                                                 4
         DECAYS                                 k5
         RATE13                                 k6
         MASS1                                  M
         MASS2                                  M
         MASS3                                  Mu
         CONC1                                  C
         CONC2                                  C0
         CONC3                                  Cb
                                    334

-------
    When inputting data, keep the following points in mind.   Only the last
two digits of the year are necessary.  The coefficient of variation is in
percent; therefore, it is equal to the mean divided by the standard
deviation times 100.  The mean has to be between the minimum and maximum
values.   If it is outside these bounds the run stops and an  error message
is written to the output file.  The distribution type is an  integer
indicating the type of distribution required.   The choices are:
TABLE 10.2.  DEFAULT VALUES FOR SYSTEM DESCRIPTIVE PARAMETERS3
VARIABLE
NAME
DECAY1
RATE 12
DECAY2
RATE23
DECAYS
RATE13
MASS1
MASS2
MASS3
CONC1
CONC2
CONC3
MEAN
0.02
6.7E-06
0.02
1.02E-04
1.20
2.15E-10
5.63E+08
3.75E+04
80.0
7.50
0.0
0.0
COEFF. OF
VARIATION
100
50
100
50
100
100
10.0
10.0
10.0
10.0
0.0
0.0
MINIMUM
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.1
0.1
0.1
0.0
0.0
MAXIMUM
10.0
l.OE-03
0.10
5.0E-03
5.0
l.OE-05
0.27E+11
3.0E+06
8000
10.0
1.0
1.0
DIST.
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
CONSTANT
CONSTANT
CONSTANT
CONSTANT
a Units of all decay/transfer rates are day  ; mass, g; concentration
  yg g"1.
                                      335

-------
         DIST      DISTRIBUTION

          0        CONSTANT
          1        NORMAL
          2        LOGNORMAL

    When simulations are run in Monte Carlo mode the user has the option of
specifing correlations between descriptive parameters.   However, the model
can generate correlated parameters only when the correlated parameters are
defined using normal distributions.  The user must be careful when
specifing correlations between variables having lognormal distributions.
Parameters with lognormal distributions can be correlated with other
parameters having normal or lognormal distributions if the correlation
coefficient between the parameters is between values in normal space.  For
example, if DECAY3 and RATE13 were specified by the user as lognormally
distributed random numbers with a correlation coefficient of 0.9, the model
would generate two normally distributed random numbers with a correlation
coefficient of 0.9 then transform these results using a lognormal
transformation.  Of course, parameters specified as constants will have no
correlation with other parameters even though such a relationship is
specified.

    Lastly, the user must specify a seed for the random number generator.
This number must be an integer between 1 and 2147483647.  It is used to
initialize the generator.  A given seed will produce a unique sequence of
random numbers.  Different seeds produce different sequences of random
numbers with the same distribution statistics.

10.4.2  Batch Input Sequence

    The formats shown in Table 10.3 are used to enter the above data using
a batch input file.  The control data has to be input in the order
specified.  The system descriptive parameters can be input in any order.
For the descriptive data, the model associates the data with the name read
in, not with the order in which it is read.  Besides data, the data  file
can also contain comments.  These are identified by the presence of  three
stars ('***') as the first three non-blank characters on a line.

    The descriptive parameter data must be followed by a data line
containing the word END.  This can be anywhere on the line but must  be the
first three non-blank characters.  This indicates that there are no  more
descriptive parameters to read in.  For the balance of the required
parameters, the model will use default values.  The same is true for the
correlation data.  There must also be an END after the last correlation
data set.  This indicates the end of the data file.
                                     336

-------
TABLE 10.3.  FORMATS FOR BATCH INPUT FILE
FORTRAN NAME
TITLE
NMCR.NCAT
DESCRIPTION OF INPUT
Title of run
Number of simulations,
UNITS
NA
NA
DATA FORMATS
A80
215
  DAY,MONTH,YEAR


  DAY,MONTH,YEAR
  THRSHB,THRSHO,
   DOSAG1.DOSAG2

       LOAD
NAME,MEAN,COEF.
VAR., MIN, MAX,
DIST
       END
NAME1, NAME2, CORR
Number of output periods

Starting date of the               NA
simulation

Ending date of the                 NA
simulation

Concentrations and dosages        ug/g
used for output comparison         ug

Loading data if using            ug/day
continuous application
model (else omit)

Name, mean of distribution,    NA,
Coefficient of Variation,      I/day, %t
Minimum and Maximum of         I/day,
distribution and               I/day, NA
distribution type

Data line indicating end of    NA
descriptive data

Correlation coefficient        NA
data
      315


      315

     4F8.0


      F8.0
    A8, 2X,
 4G10.4, 215,
       A3
2(A8,2X), F10.0
  10.4.3  Interactive  Input

      The interactive  input is largely self  explanatory,  but  a  brief
  description of each  screen follows.

      The first  screen after the title screen  (screen  1,  Figure 10.2)  asks
  for the name of the  output file (screen  2, Figure  10.3).   It  is  assumed
  that this is a new file.   If the file already  exists,  (screen 3,
  Figure 10.3) will  appear  and ask if  it is  permissible  to  overwrite the file
  or to stop the program.   To overwrite type the letter  0.  To  stop the
  program type the letter S.  If the option  to overwrite  is chosen, output
                                       337

-------
                      WELCOME   TO    THE

                           A   -  T E E A M

                               MODEL

                    U.S. ENVIRONMENTAL PROTECTION
                                AGENCY

                              prepared by
                      WOODWARD-CLYDE CONSULTANTS
                               APRIL 1988

                               VERSION 1.1
                     — HIT RETURN KEV WHEN READV —


                  Figure 10.2. Screen 1 ATEEAM model title screen
from previous runs which used the same output file name  may  be  lost.   The
model creates output files which contain results from either the  Monte
Carlo simulations or the time series results from deterministic
simulations, (see Section 10.6 for description of output files).   If  these
files already exist, screen 4 (Figure 10.3)  appears and  the  same  procedure
as above should be followed.

    Screen 5 (Figure 10.4) gives the user an option as to which model to
run.  The response should be a 1, 2 or 3.  A description of  each  model is
given in Section 10.1.  The next screen (screen 6, Figure 10.4) offers the
choice of running the model in batch or interactive mode.  If batch mode is
desired, type the letter B.  The user will then be asked for the  name of
the input file and the random number generator seed.  If interactive  mode
is desired type the letter I.  The user will be prompted for additional
data by additional screens.

     Model control parameters are input next (screen 7, Figure 10.5).   The
default values for  the parameters will appear on the screen.  The default
values will  initially appear in reverse video but will convert to normal


                                    338

-------
 OUTPUT  FILE  NAME  —>

 Screen 2.  Request for output fi/e name
    ERROR  IN OPENING  OUTPUT FILE. CHOOSE  OPTION

                      2)verur i te file

                            program
 Screen 3.  Error indicated in opening output file. Possible responses:
         O - Overwrite existing file
         S -Stop program execution
ERROR IN OPENING MONTE CARLO  FILES.  CHOOSE OPTION

                     5Jverwri"te file

                     §top program
 Screen 4. Error indicated in opening model output files. Possible responses:
         O - Overwrite existing files
         S - Stop program execution
     Figure 10.3. Screens 2, 3 and 4 from ATEEAM model
                             339

-------
video when  a  value is input.   For each parameter  either the default  value
can be retained or a new value can be assigned.   The  value presently
assigned to each parameter will  appear in the  lower right hand corner of
the screen.   Entering a carriage return will retain this value and advance
the cursor.   If a new value  is desired, type in the number and a  carriage
return.  When a carriage return is entered after  the  last number  the cursor
returns to  the top of the screen to give the user a chance to correct any
mistakes or change more values.   When everything  on the screen is correct,
the user should input a Q and  a carriage return.   This will end input for
the control parameters.
                CHOOSE UHICH APPLICATION MODEL  TO RUN

                              Q SINGLE DOSE

                              § CONTINUOUS APPLICATION

                              §3 STEADY STATE CONDITIONS
           Screen 5. Indicate type of simulation to run. Possible responses:

                   1 - Simulation of single application of pesticide
                   2 - Simulation of continuous application of pesticide
                   3 -Simulation of steady-state conditions (continuous application)
               RUN MODEL  IN THE  BATCH OR INTERACTIVE  NODE

                                [Jnteractiue node

                                [jjatch node


            Screen 6. Indicate if input is in batch or interactive mode. Possible responses:

                   I - Interactive mode
                   B - Batch mode

                  Figure 10.4.  Screens 5 and 6 from ATEEAM model


                                      340

-------
    If more than one output period is requested (NCAT > 1) then screen 8
(Figure 10.5) will appear requesting the ending dates for each period
except the last.  (The last period is assumed to end on the last day of the
simulation.)  The ending day, month (1 - 12) and the last two digits of the
year are input.  The user will continue to be prompted for input until all
the necessary data are entered.  If after entering a carriage return the
cursor doesn't advance, there is an error in input.  The error could be
something such as a month greater than 12, too many days in a month or the
periods out of chronological order.  After the last date has been entered,
the cursor will go back to the top of the screen.  At this point the user
can either re-enter the data, if there was a mistake for example, or enter
a Q.  Entering a Q will bring up the next screen.

    Screen 9 (Figure 10.6) contains the descriptive parameters.  As was the
case for the control data, default values are assigned to these variables
and they will appear on the screen.  If the default data is acceptable,
enter a carriage return and the cursor will move to the next data point.
If a different value is desired, type it and enter a carriage return.  By
using the 'U' and 'D1 keys, the user can move up or down on the screen
entering only selected data points.  Typing 'D1 and return will move the
cursor to the beginning of the next line down.  Typing 'U' and return will
move the cursor to the beginning of the previous line.  When all the data
is satisfactorily entered the user types Q and carriage return.

    The last screen (screen 10, Figure 10.6) allows the user to input
parameter correlation information.  There are three pieces of data
necessary for each correlation.  The user has to input the names of the two
correlated parameters (DECAY1, RATE12, etc....) and the correlation
coefficient between them.  The correlation coefficients must be between -1
and +1 and follow the limitations specified in Section 10.4.1.   Hitting a
carriage return without inputting any data on a line will signify the end
of data input for this screen.

    After completing all of the above screens the last question asks the
user to input an integer used as the seed to the random number generator.
This number can be any integer between 1 and 2147483647.

10.5  PARAMETER ESTIMATION

    There are 12 parameters which must be given values in order to operate
the ATEEAM code.  These parameters break down into three groups:

       • Mass estimates
       • Rate constants
       • Initial pesticide concentrations in biomass
                                    341

-------
                             CONTROL PARAMETERS
                             NMCR  =

                             NCAT  =
     188
                     DOSAG1
                 STARTING TIMES
                      DAV    E

                      MONTH E

                      VEAR  ]B
             THRSHB

             IX)SAG2
                                                          888.8
           ENDING TIMES
              DAV    §2

              MONTH [E

              VEAR  IS
INPUT A  UL'JEfiilKMl  TO QUIT
HIT RETURN FOR
    Screen 7. Input model control data. Possible responses:
            Value - New value to assign to parameter
            "CR" - Retain present value of parameter that is indicated in
                  lower right hand corner of screen
              Q  - Finished model control parameters. Go to next screen
                    SIMULATION
             PERIODS
END OF
PERIOD
1
2
ENDING
DAY
38
38
ENDING
MONTH
3|
1
ENDING
VEAR
87
87
    Screen 8. Input ending dates of each output period. Possible responses:
            Date - Ending day, ending month and ending year of period
            Q   - Finished simulation period data. Go to next screen
        Figure 10.5. Screens 7 and 8 from ATEEAM model running
                    in interactive mode
                                  342

-------
                            U A  R I  A B  L E
                              DATA
NAME
DECAV1
RATE12
DECAV2
RATE23
DECAV3
RATE13
MASS1
MASS2
MASS3
CONC1
CONC2
CONC3
MEAN
.288E-81
.678E-85
.288E-81
. 182E-83
1.28
.715E-89
.563E+89
.375E+85
88.8
7.58
.888E+88

COEF. VAR.
188.
58.8
180.
58.8
188.
180.
18.8
18.8
18.8
18.8
.888E+88
.888E+88
MINIMUM
.888E+88
.088E+88
.088E+88
.BB8E+88
.880E+08
.888E+B8
.188
.188
.108
.888E+88
.888E+88
.888E+88
MAXIMUM
18.8
. 108E-82
.188
.580E-82
5.88
. 188E-84
.278E+11
.380E+87
.888E+84
18.8
1.08
1.88
DIST
1
1
1
1
1
1
1
1
1
8
0
8
INPUT A
WHEN FINISHED
HIT RETURN FOR
.8008E+88
               Screen 9. Input screen for system descriptive parameters. Possible responses:
                       Value - New value to assign to parameter
                       "CR" - Retain present value of parameter that is indicated
                              in lower right hand corner of screen
                       "U"  - Move the cursor to the beginning of the  previous line
                       "0"  - Move the cursor to the beginning of the  next line
                       "Q"  • Finished descriptive parameter input. Go to next screen
                CORRELATION    MATRIX    DATA
                 VARIABLE 1
                     NAHE

                   DECAV1
               VARIABLE 2
                  NAME

                 RATE 12
               CORRELATION
               COEFFICIENT
                  IfT RETURN FOR UAR1AKI.E  t  UHEN  FINISHED
              Screen 10.  Input screen for correlation data. Possible responses:
                        Parameter Name - Enter the names of the two parameters which
                                      are correlated
                       Correlation Coefficient -
                        Q - Finished

              Figure 10.6. Screens 9 and 10 from running ATEEAM model
                          in interactive mode
                                          343

-------
    Each of these categories of parameters is discussed in the following
sections.  Special considerations for estimating parameters required for
Monte Carlo analysis are covered last.

10.5.1  Mass Estimates

    There are three masses which must be estimated as input to the model:

       • Soil mass
       • Lower trophic level (soil dwelling organism) population biomass
       • Upper trophic level (predator) biomass

10.5.1.1  Predator Biomass—
    The easiest way to define the system is by beginning with an estimate
of the areal extent of the habitat for a single predator.  For example, if
the predator is the passerine bird T. migratorius, then the habitat or
territory is approximately 0.15 ha.  The biomass of an adult robin living
in this 0.15 ha is approximately 80 grams; therefore, Mjj the biomass of the
predator would be 80 grams.

10.5.1.2v Soil Dwelling Organism Biomass—
    The estimation of the biomass of the soil dwelling organism is
straightforward once the habitat size is known.  Normally, the density of
organisms is defined in terms of number or biomass per unit area (usually
m ).  To obtain the biomass, one simply takes the product of the habitat
area and the biomass density (e.g., g nf^).  If the number instead of
biomass per unit area is given, then the average weight of an individual of
the species must also be known in order to obtain the biomass.  Typical
biomass densities for a number of soil dwelling organisms was given in the
parameter estimation section (Section 6.7.1) of the TEEAM documentation.
As an example, for cropland, Lee (1985) provided estimates of 0.5 to 20 g
m   for earthworms.  If the size of the habitat is 0.15 ha and the biomass
density of 20 g m   is used, then the biomass MQ is 30,000 grams.

10.5.1.3  Soil Mass—
    The mass of the soil is easily estimated by taking the product of the
soil bulk density (g cm~3) with the habitat area (ha) and a representative
depth of soil (cm).  Bulk densities for various soil types were given in
Section 6 (Table 6-26).  The main problem in producing this mass estimate
is choosing a representative soil depth.  By choosing a depth that is too
large, concentrations of the chemical in soil may be underestimated, and
conversely, if the depth is too small they may be overestimated.  The depth
used should ideally be the depth to which the chemical is mixed over the
period of the simulation.  This might be a function of the depth of
activity of soil dwelling animals, or the depth of movement of the bulk of
the chemical mass (by leaching) over the period of the simulation.  The
latter can be estimated by assuming plug flow and'using

                                    344

-------
         D =  *VQ1-                                                  (10-21)

in which D is the depth of movement (cm)
         R is the chemical retardation factor (dimensionless)
         V is the average recharge rate over the period and
         At (cm At) and                                      _    .
         0 is the average water content of the soil over At  (cm  cm  )

The retardation factor is calculated by
         R = 1 +   °cp                                           (10-22)

in which

         KQC is the organic carbon partition coefficient for the chemical
         (cm3 g-1)

         OC is the soil organic carbon content  (fraction, g g   )

         p is the soil bulk density (gem  )

However, since the highest exposures are likely to occur immediately after
application, the user may wish to bias such a calculation towards  shallower
depths.  For convenience, in this discussion,  a depth of 15 cm is
utilized.  This is a typical "plow layer depth"  in agricultural  soils and a
depth to which the chemical  might be readily mixed by bioturbation in a
relatively short time.  Using 15  cm as  the  representative depth, and a bulk
density of 1.5 g cm"3, the soil mass in 0.15 ha would be 3.4 x 108 grams.

10.5.2  Rate Constants

    In this model, six rate  constants  can be utilized:

       • k^ is the degradation rate of  the  chemical  in soil
       • k2 is the transfer  rate  of the chemical from the soil  to the lower
         organism biomass
       • ko is the clearance rate of the chemical  from the  lower organism
         biomass
       • k^ is the transfer  rate  from  the lower organism to the predator
       • k5 is the clearance rate of the chemical  from the  predator and
       • kg is the transfer  rate  from  the soil  directly to  the predator.

10.5.2.1  Degradation Rate in Soil —
    The user should refer to Tables 6-20 and 6-21  provided  in Section 6.
If information for the chemical of interest is  not available,  then
registrant data or other sources  of information may be used.  The
degradation rate must be entered  in day  .

                                     345

-------
10.5.2.2  Transfer Rates-
    Three of the rate constants, k2, k^, and kg, deal with transfers,
related primarily to ingestion, between mass compartments.  Transfer from
the soil to the soil dwelling organism may be accomplished through
ingestion or by dermal absorption.  Incjestion rates for soil and plant
litter for various soil dwelling organisms may be found in Section
6.7.3.1.  These are normally given in grams of soil per gram of organism
body weight per day.  To obtain the rate constant, simply take the product
of this rate and the organism biomass, MQ, and divide by the soil mass
MS.  For example, in the case of earthworms, a typical soil ingestion rate
is 0.08 g g""1 day'-*-.  Therefore a typical value for the rate constant k2
would be

         k  =  0.08 (30,000 grams)  = 7<1 x 1Q-6 day-l
          *     3.4 x 10° grams
Similarly, the ingestion rate of earthworms by avian predators is typically
given in g g'1 day"  (or perhaps another time unit).  Or the user may have
information on total food intake and food preferences (%) in which case,
the product of the two will give the ingestion rate for the prey.  A
typical ingestion earthworm rate for T. migratorius would be 0.1 g g
day  .  Therefore the rate constant k4 is

         .   _  0.06 (80 grams)  _ , ,   1n-4  .  -1
         k4 	30,000 grams   ' l'6 x 10   day

    Grit or soil ingestion rates are often expressed as a percent of total
food intake per day.  The rate therefore has units of g soil g~* predator
day  .  The rate constant must be adjusted by multiplying by the predator
biomass and dividing by the soil mass.  Using a 10% soil ingestion factor
and a total food intake rate of 0.1 g g   day, the rate kg would be

         k  = 0.1 (0.1 9 g-1 day (80 grams) , 2<3 x 1Q-9 day-l^
          °        3.4 x 10B grams
10.5.2.3  Clearance Rates-
    Clearance rates are most often determined through feeding studies on
particular species with specific chemicals and may be difficult to
obtain.  The available guidance for estimating these rates (which have
units of day  ) is given in Section 6  (specifically Table 6-42).  For soil
dwelling organisms, if the ingestion rate and bioconcentration factor (BCF)
is known, then the expression
                                                                    (10-23)

in which
         AZ is the representative soil depth (cm) and
         p  is the organism soil density (g cm  )

-------
10.5.3  Initial Pesticide Concentrations

    For Model 1, the initial pesticide concentration in soil is easily
estimated by taking the application rate (kg ha  ) and dividing by the
representative soil mass with the1appropriate units conversions.  Initial
concentrations are input in yg g~ .  Therefore the concentration  (C^) may
be found by
         C- = _w«                                                   (10-24)

 in which A is the application rate in kg ha"1.

    For Model 2 an equivalent daily loading rate (yg day" ) to emulate
multiple chemical applications is found by simply dividing the total
application over the period by the length of the period in days.

10.5.4  Estimation of Parameters for Monte Carlo Analysis

    The above techniques would be used to estimate mean values for Monte
Carlo analysis.  In addition, for each parameter, the coefficient of
variation (CV), minimum and maximum and distribution type must be
selected.  If correlations between randomly generated parameters are
desired, correlation coefficients must also be specified.

10.5.4.1  Coefficients of Variation (CV) —
    The CV is a normalized measure of the spread of the uncertainty about
the mean value of the parameter.  It is equivalent to the standard
deviation divided by the mean.  Default CVs for each input parameter have
been provided in the interactive and batch input sets.  Table 10.2 shows
their values.

    Degradation and clearance rates—Specific guidance for degradation
rates in soils is available.  Rao and Davidson (1980) give CVs for soil
degradation rates for approximately 30 herbicides, insecticides and
fungicides.  They ranged from 16.1% to 130.8%.  A representative value is
on the order of 50 to 100%.  Specific guidance for CVs for clearance rates
is largely unavailable to the authors.  Calculated clearance rates for wood
thrushes fed earthworms contaminated with dieldrin (Jeffries and Davis
1968) had a CV of 97% (n = 5).  It is assumed that CVs for clearance rates
would be similar, possibly higher than, the CVs for soil degradation.

    Soil mass and animal biomass--CVs for biomasses can be obtained using
ranges of counts or biomass density found in the literature.  For instance,
Lee (1985) gives a range of 0.5 to 20 g m~2 for earthworm populations in
arable soils and Lofty (1974), a range of 2-40 g m~ .  Taking the range to
1 to 30 g m  , and assuming such a range represents approximately the 2.5%
                                     347

-------
and 97.5% quantile levels of a normal  distribution (these values lie 2
standard deviations on either side of  the mean, respectively), the mean is
15.5 g m~2, and the CV is
                     r = "47 or
The CV for the mass of the soil is determined by the uncertainty in the
volume of soil (i.e., habitat area (A)  and representative depth (d)) and
the uncertainty in soil bulk density,  p .   Bulk densities for several soils
have been estimated to have a CV of about  9% (Jury 1985).  Selection of a
representative depth is more uncertain,  more so for  less  strongly  adsorbed
chemicals.  Ideally, one could  evaluate the  aggregate  uncertainty  in the
calculation of a representative depth by propagating the  uncertainties  in
each of the parameters of the estimating equation.   According to Jury
(1985) estimates of the CV of solute velocity range  from  36  to 194%.
Therefore, a similar level of uncertainty  would be associated with the
depth of pesticide penetration.  The authors are not knowledgeable about
the CV of the estimate of habitat or territory.  Assuming a  50% variability
to size of territory, an upper  bound for the CV soil mass could be found by
(Taylor 1982)
              6M    6A    6d    6p     6M

         cv - W = W + W + T^r  W = °'5 + L94 + °-09 = 2'53
or ~ 250%,  In the case where the quantities A, d and  p   are independent
(one would expect them to be in this case),  a more reasonable estimate  of
the aggregate uncertainty is
                    A'     '   <

             = 2.01 or ~ 200%.
                                           - /(0.5)2  t (1.94)2  + (0.09)2
    Ingestion (transfer)  rates—CVs of ingestion rates  can most likely be
estimated with information given on the observed ranges for the species of
interest.  For instance,  Lofty (1974)  gives a range of  0.1 to 0.3 g g
day"  for ingestion of soil by earthworms.   Assuming that these numbers
represent the 2.5 and 97.5 percentiles, the CV would be 25%.

    Initial pesticide concentrations—The uncertainty in initial  soil
pesticide concentrations  (Model  1)  stems from uncertainty in the quotient
of the application rate and the mass of affected soil.   In a quotient,
these uncertainties would sum to give an upper bound.  To obtain an
aggregate uncertainty, assuming independence of the quantities in the
quotient, the root mean square of the fractional uncertainties of these
quantities is calculated.  The uncertainty in application rate is a
function of the type of application equipment and meteorological  conditions

                                     34G

-------
occurring during the spray event.  Smith (1988, personal communication)
reports that the CVs for granular application range from 40 to 70% and are
approximately 25% for spray applications from ground equipment.  Jury
(1985) documents CVs of 60 to 130% for measured pesticide concentrations in
soils.  These CVs do not take into account the high variability associated
with determining depth of pesticide penetration, however, and its impact on
estimated concentrations.

10.5.4.2  Minimum and Maximum Values--
    Minimum values for all the parameters should be zero or close to
zero.  Minimum values for masses (soil,  mass  biomass)  should be set  to
values slightly greater than zero to avoid zero divides in the solutions to
the model equations.  Default maximum values  in the batch and interactive
data sets are 100 times the mean.  This  should be sufficient to insure that
no truncation occurs on the upper end of distributions.  It should be noted
that normal distributions specified with low  means and large CVs will
result in frequent truncation of the distribution by rejection of negative
values.  If this happens, it is probably best to specify a log normal
distribution type for the parameter.

10.5.4.3  Distribution Type--
    Unless the user has specific knowledge about distribution type,  the
recommendation is to use the normal distribution.  It should be pointed out
that, when specifying a log normal distribution, the user inputs the
arithmetic mean and CV into the program.  The code internally calculates
the mean and variance of the corresponding log normal  distribution
according to relationships in Yevjevich  (1972).

10.5.4.4  Correlation Coefficients--
    It is difficult to give meaningful quantitative guidelines for
estimating correlation coefficients for  these parameters.  There is  reason
to suspect that soil degradation rates and clearances rates would be highly
positively correlated.  There may be weak positive correlations between the
biomass of predator and prey (due to prey availability considerations).
There is probably some correlation between ingestion rates of soil dwelling
organisms and ingestion of soil.  There  is no reason to believe that soil
mass or animal biomass is correlated with degradation or clearance rates.
It is suggested that the user adopt either of three values:

       • Zero for no correlation
       • (±) 0.5 for weak correlation and
       • (±) 0.9 for strong correlation

for use in simulations.  It is also suggested that a simulation be made
assuming independence of all parameters  to test the importance of the
correlations selected.
                                     349

-------
    In order for the  correlation matrix to be decomposed,  it must be
positive definite.  While  this has a precise mathematical  reasoning, to the
user it suggests that the  correlation matrix "makes  sense".   For instance,
if A is strongly positively correlated with B (i.e., when  A  is large, B is
large) and A is strongly negatively correlated with  C  (i.e.,  when A is
large C is small) then B must also be negatively correlated  with C.  If
such rules are not  followed, the user may find that  the  program terminates
with a message to the output file indicating that the  correlation matrix
could not be decomposed.

10.6  OUTPUT FILES

    Five output file  are created when the model is run,  SOIL.OUT,
TLEVL1.0UT, TLEVL2.0UT, MCARLO.OUT, and the user specified output file.
All of these files  except  the user specified output  file are formatted for
easy importation into LOTUS or a similar program.  The results are stored
in columns.  If titles are printed at the top of a column  they are
bracketed by quotation marks (").

    MCARLO.OUT contains all of the values of the variables that were
generated by the random number generators.  Figure 10.7  is an example of
this file for a case  of ten Monte Carlo simulations  with nine randomly
generated parameters.  The first column in the file  is the simulation
number.  If the run was deterministic this file will be  empty.

    TLEVL1.0UT and  TLEVL2.0UT contain the calculated whole body
concentrations and  dosages from the model for the lower  trophic level
animal and higher trophic  level animal respectively  for  each output period
of the simulation.  The files contain both the mean  concentration for each
       INPUT PARAME TERS TO MODEL FOR EACH RUN OF THE MODEL
•RJNNO.'
  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
, •
0
0
0
0
0
0
0
0
0
0
"OECAY1
.71619E-02
.39054E-01
.33652E-01
.14461E-01
.33238E-01
.21745E-01
.74350E-02
,380086-01
.10473E-01
.47201E-01
I
0
0
0
0
0
0
0
0
0
0
"RATE12
.11450E-04
.67081E-OS
.34689E-05
.14915E-05
.89781E-05
.25538E-05
.67093E-05
.523536-05
.79452E-05
.66163E-05
*
0
0
0
0
0
0
0
0
0
0
"OECAY2
.36964E-01
.34864E-01
.16503E-01
.97345E-02
.32909E-01
.10526E-01
.89383E-02
.343446-01
.53647E-01
.19883E-01
•
0
0
0
0
0
0
0
0
0
0
'RATE23
.10261E-03
.87746E-04
.13347E-03
.71499E-04
.63236E-04
.89265E-04
.10901E-03
.101726-03
.10186E-03
.17247E-03
' 'DECAYS
2.2893
1.1236
0.23010
1.2630
0.89912
0.85588
0.43908
2.0887
1.8503
1.1216
' 'RATE13
0.14119E-08
0.2S255E-09
0.64078E-09
0.10897E-08
0.70177E-09
0.84449E^09
0.45403E-09
0.29731E-09
0.15878E-08
0.24366E-09
•
0
0
0
0
0
0
0
0
0
0
•HASS1
.60934E+09
.616546*09
.50982E+09
,612?7E*09
.59242E+09
.57411Ef09
.60892E+09
.585886+09
.55969E+09
.607116*09
                                                                  "MASS2
  'MASS3
                                                                29807.
                                                                35372.
                                                                38750.
                                                                35320.
                                                                38727.
                                                                45998.
                                                                33403.
                                                                30968.
                                                                45671.
                                                                37883.
74.718
78.674
89.670
88.249
83.542
94.120
80.538
84.132
66.551
79.414
              Figure 10.7. Example of MCARLO.OUT output file.
                                      350

-------
period and the maximum concentration for the simulation.  Figure 10.8 is an
example of file TLEVL1.0UT.  File TLEVL2.0UT is identical in format.  The
first line in the file contains two values, the total length of the
simulation in days and the number of output periods.  The first column in
the file is the number of days during the simulation that the mean daily
concentration exceeded THRESHO (the toxicity threshold in soil-dwelling
organisms) or THRESHB (the toxicity threshold in the predator).  The next
2*NCAT columns contain the mean whole body concentrations and dosages for
each output period of the simulation.  The values are ordered by
concentration then dosage for each output period—i.e., concentration for
period 1, dosage for period 1; concentration for period 2, dosage for
period 2, etc....  The last column is the maximum concentration over the
simulation period.  If the simulation is run in deterministic mode file
TLEVL2.0UT will be empty.  File TLEVL1.0UT will contain the time series
results of the simulation.  The file will contain four columns of numbers
as shown in Figure 10.9; the time (days), the soil concentration, the whole
body concentration for the lower trophic level  animal, and, the whole body
concentration for the higher trophic level animal.

    The final output file is the user specified output file.  This file
contains an echo of the input data and the results of the simulations.
Statistics for both the lower and upper trophic levels are presented along
with printer plots.  The statistics and the plots correspond to the data
found in files TLEVLl.OUT and TLEVL2.0UT.  The  file contains two type of
printer plots as shown in Figures 10.10 and 10.11, one a histogram of the
results and the other a CDF of the results.  Results are presented for the
number of days the concentration exceeded THRESHO and THRESHB, the mean
concentration and the dosage for each trophic level  and each output period,
and, the maximum calculated concentration over  the entire simulation
period.
"DAY"
0
0
0
0
0
0
212
0
0
365
"CONG"
14.27
7.304
8.666
13.32
7.534
10.09
20.66
7.393
4.918
3
"DOSE"
76.
31.
22.
30.
31.
22.
50.
31.
34.
01
65
27
81
70
41
85
87
80
"CONC"
8.731
0.4796
2.4Q6
11.19
0.7188
6. 163
25.59
0.5179
1 .864
"DOSE"
32.66
0.2917
0.3928
5.480
0.5878
1.656
21.14
0.3331
10.02


0
0

0


0
0
"CONC"
3.658
.9876E-02
.3338
4.605
.2255E-01
1.920
16.07
.1144E-01
.5166
"DOSE"
13.46
0 2388E-02
0
0
0
0
6250E-02
9222
9839E-02
1139
8.425
0.3102E-02
2.750

17
12
11
16
12
12
28
12
6.

.25
.69
.58
.63
.13
.34
.74
.68
414
              Figure 10.8. Example of TLEVL1. OUT output file when model is
                        run in Monte Carlo mode.
                                     35l

-------
10.7  EXAMPLE PROBLEM

    This  section presents  an example problem to  enable the user to
benchmark a simulation on  his own machine.  The  example problem is a Monte
Carlo simulation run with  the default parameters on either the batch input
file or the interactive input sequence.  A portion of the example output is
given in  Appendix A.  The  first portion of the example output is an echo of
the inputs.  Selected tabular outputs are also included.  The first output
shows, for the soil dwelling organisms, the statistics for number of days
that the  whole body concentration exceeded the threshold concentration set
by the user.  The statistics shown are the mean, standard deviation, CV,
     "DAYS"
     If 1)
        1
        2
        3
        4
        5
        6
        7
        8
        9
       10
       11
       12
       13
       14
       15
       16
       17
       18
       19
       20
       21
       22
       23
       24
       25
       26
       27
       28
       29
       30
       31
"SOIL"
"CONC"
.3514408
.2058242
.0630920
.9231870
.7860533
.6516359
.5198810
.3907359
.2641488
.1400692
.0184474
.8992346
.7823832
.6678464
.5555783
.4455340
.3376694
.2319414
.1283077
.0267267
.9271578
.8295612
.7338978
.6401292
.5482180
.4581274
.3698213
.2832643
.1984219
.1152600
.0337454
"LOWER TROPHIC
    "LEVEL"
0.73944130
 1.4495200
 2. 1311105
 2.7850643
 3.4122099
 4.0133542
 4.5892825
 5.1407590
 5.6685278
 6.1733128
 6.6558186
 7.1167309
 7.5567171
 7.9764263
 8.3764903
 8.7575238
 9.1201247
 9.4648748
 9.7923400
 10.103071
 10.397602
 10.676456
 10.940137
 11.189140
 11.423942
 11.645009
 11.852795
 12.047739
 12.230269
 12.400803
 12.559743
""UPPER TROPHIC"
     "LEVEL"
 0.34082055E-01
 0.63996961E-01
 0.91883599E-01
 0.11839441
 0.14373889
 0.16799640
 0.19122636
 0.21347759
 0.23474922
 0.25510813
 0.27457043
 0.29314689
 0.31087541
 0.32780205
 0.34392114
 0.35926837
 0.37387263
 0.38776696
 0.40094832
 0.41345403
 0.42530490
 0.43654408
 0.44714213
 0.45715069
 0.46660754
 0.47547985
 0.48383697
 0.49165982
 0.49899449
 0.50583809
 0.51221586
           Figure 10.9. Example of TLEVL.OUT output file when model is
                    run in deterministic mode.
                                  352

-------
                                                              4> 1-1
 13
flu
O  +
CS
CO
O
fc,
a.
H



























.











•





•






•





— — — 4

C
•«d



•





•





•




•


















•





•


*



.

*



_ _^ „ ^

> C
1" f



•





•





•




•






h






h




•





•


* * # •»



•

* * * *



• — — — 4

> C
> C\



•





•





•




•


«



•






•


•w



«



•


* * * 4



.

# * * *



. — —. _ 4

> C
3 i-
1
# TH
1 CD
f 4- C-
1 CV1
1
* i-l
1
1
4- «
1 i-l
1
* 1-4
I
I 10
4- a
i
« o
i
i
4- in
1 0
i
* * # # o
i
1 0
1 O O
4-1-14-
i c- ta
1 • iH
* O •
1 0
1
1 O
4- C- «
i in
i
« # * » o
1
1 CO
4- O
1
* * # •» O
1
1
1 i-t
4- O
1 CM
1
# # # * O
1
1
1 i-l
4- in
i
* * * * o
i
i
1 O
- — 	 4- rH
O
s O -
1 O
bo
3
•^
£.
v>
^•4
52
*X
O
o
C2
<"-^
EH
O
EH

CO
KH
Q
O
I-J
EH
0
X
o
*•"




(13
00
C
-a
c
3
O
H—
S
a
03
•4_»

Q.
E .
CO _03
O) C^
.1 =
•^ •I-*
H— D
0 O
(u -o
"5..®
I?
LU «•
c>
^~
o
CD
O5
a.
                                            353

-------
•t
1
1




•


j
•
C3
a
j •





o

t— 1

c=
0.

o

CS

f-



cs

u


o

J


a

0

p^



£1
4
» 4
i


H






-



•



























•




•






•


t
t
t
t
-1- _ _ —.







•



•


•






•





•
















•





41
*
«
4f




f 	 _







•



•


-






•






















«
«


*
*

•





1
4. — . — —

1 if
1 *
1 *
1 *
1 *
1 4t
+ *
1 4t
1 *
1 *
+ 4f
1 *
1 *
+ *
1 41
1 *
1 «
1 *
1 *
1 *
+ 4»
1 *
1 *
1 *
1 *
1 4t
#
*
if
#
4f
*
if
if
if
if
1
1
+
1
1
1
1
-I"
1
1
1
1
1
1
+
1
1
1
1
1
1
f 	 + — _ _

• t— i
•
«— t


(O
w
~

co
o
r-t
^4

o>
0>
-
o


f
to
CD
•
O
o
o o
<-l +
t- ca
- r-i
o •
o
o
t- 4t
IO
•
0


t- o
1"
•
0
^
• a
CO
•
o


«H
• 1C
rH
-
0


1 0
f «-«
o
                                                                             ho
                                                                             s
                                                                             o
                                                                             a
                                                                             o
                                                                             H

                                                                             O
                                                                             o
                                                                             <
                                                                             o
                                                                             Z
                                                                             <
                                                                             O
                                                                             I— t
                                                                             X
                                                                             o
                                                                             H
       I
       _c

       c
       o
       M—
       s
       Q.
       OI
       v
       c
                                                                                                 a>
            O.
           +J
       M-  3
        O  O
        cu  -a
       "5. .32
       UJ
ta     r-
                                                                                              
-------
maximum and minimum values.   Due to differences in computational precision
of various machines, it is likely that the output generated by the user may
be slightly different.   The  statistics should be very similar however.

    The next output is  the tabular summary of statistics for the whole body
concentration in the lower (soil dwelling) organism,  for the first (90-day)
output summary period.   The  next tabular output is summary statistics for
dosage to the lower level  organism (from the soil) during the first 90-day
period.  The next output is  concentration statistics  for the second 91-day
period, followed by dosage statistics for that period.   This is followed by
concentration and dosage statistical  summaries for the  final 184-day
period.  The final  tabular output for the lower trophic level is for the
peak daily concentration during the entire simulation period.

    The same tabular summaries are then presented for the higher trophic
level.  In the actual model  output, relative frequency  and cumulative
relative frequency histograms are plotted for following each tabular
summary.
                                    355

-------
                                 SECTION  11

                                 REFERENCES

Akesson, N.B., and W.E. Yates.  1976.  Physical  Parameters Relating to
    Pesticide Application.  Iji:  Pesticide Spray Application,  Behavior and
    Assessment:  Workship Proceedings, Emeryville,  CA, 1-2 March 1973.
    U.S. Pacific Southwest Forest and Range Experiment Station,  Berkeley,
    CA.  U.S. Dept. Agric.  Forest Service.   Gen. Tech Rep.  PSW-15/1976.

Ambrose, R.B., S.I. Hill, and L.A. Mulkey.  1983.  User's Manual for the
    Chemical Fate and Transport Model.  TOXIWASP, Version I.   EPA-600-3/83-
    005.

Ashton, F.M., and A.G. Crafts.  1981.  Mode of Action of Herbicides. 2nd
    Ed.  New York: John Wiley and Sons.

Bache, D.H., and S. Uk.  1975.  Transport of Aerial Spray, II:  Transport
    Within a Crop Canopy.  Agricultural Meteorology 15:371-377.

Baker, P.B, E.R Hoebeke, J. Barnard, and T.D. Spittler.  1986.  Arthropods
    Associated with Pesticide Leach Lines in Selected Orchards in Central
    New York.  Env. Toxicol. and Chem. 5(1):  61-68.

Barker, R.J.  1958.  Some notes on ecological effects of DDT sprayed
    elms.  J. Wildlife Mgmt. 22:269-274.

Baver, L.D., W.H. Gardner, and W.R. Gardner.  1972.  Soil Physics.  Fourth
    ed.  John Wiley & Sons, Inc., New York.

Beard, K.V., and H.R. Pruppacher.  1971.  A Wind Tunnel Investigation of
    the Rate of Evaporation of Small Water Drops Falling at Terminal
    Velocity in Air.  Journal of Atmospheric Science 28:1455-1464.

Bent, A.C.  1949.  Life Histories of North American Thrushes, Kinglets, and
    Their Allies.  Smithsonian Institute, U.S. National Museum, 196 p.

Biederbeck, V.O., C.A. Campbell, and A.E. Smith.  1987.  Effects of Long
    Term 2,4-D Applications on Soil Biochemical Processes.  Journal of
    Environmental Quality 16(3):257-262.

Boersma, L., F.T. Lindstrom, C. McFarlane, and E.L. McCoy.  Uptake of
    Organic Chemicals by Plants:  A Theoretical Model.  In press.
                                     356

-------
Brakensiek, D.L., and Rawls, W.J.  1985.  Prediction of Soil Water
    Properties for Hydrologic Modeling.  In: Proceedings of Symposium
    Sponsored by Committee on Watershed Management, ASCE, Denver, Colorado.

Brackensiek, D.L., and W.J. Rawls.  1983.  Green-Ampt Infiltration Model
    Parameters for Hydrologic Classification of Soils.  In: Proceedings of
    Special Conference on Advances in Irrigation and Drainage, ASCE,
    Jackson, Wyoming.

Brady, N.C.  1974.  The Nature and Properties of Soils.  8th. ed.
    New York:  MacMillan Publishing Co.

Briggs, G.G., R.H. Bromilow, A.A. Evans, and M. Williams.  1983.
    Relationships between Lipophilicity and the Distribution of Non-ionized
    Chemicals in Barley Shoots following Uptake by Roots.  Pesticide
    Science 14:492-500.

Briggs, G.G., R.H. Bromilow, and A.A. Evans.  1982.  Relationship between
    Lipophilicity and Root Uptake and Translocation of Non-ionized
    Chemicals by Barley.  Pesticide Science 13:495-504.

Burns, L.A., D.M. Cline, and R.R. Lassiter.  1982.  Exposure Analysis
    Modeling System (EXAMS): User Manual and System Documentation.  U.S.
    Environmental Protection Agency, Athens, Georgia.  EPA-600/3-82-023.
    July.

Carsel, R.F., and R.S. Parrish.  1988.  Developing Joint Probability
    Distributions of Soil Water Retention Characteristics.  Water Resources
    Research 24(5):755-769.

Carsel, R.F., R.S. Parrish, R.L. Jones, J.L. Hansen, and R.L. Lamb.
    1988.  Characterizing the Uncertainty of Pesticide Leaching in
    Agricultural soils.  J. of Contam. Hydrol. 25:  111-124.

Carsel, R.F., C.N. Smith, L.A. Mulkey, J.D. Dean, and P. Jowise.  1984.
    User's Manual for the Pesticide Root Zone Model (PRZM):  Release 1.
    U.S. Environmental Protection Agency, Environmental Research Labora-
    tory, Athens, Georgia.  EPA-600/3-84-109.

Chapman, R.A., and P.C. Chapman.  1986.  Persistence of Granular and EC
    Formulations of Chlorpyrifos in a Mineral and in Organic Soil Incubated
    in Open and Closed Containers.  J. Environ. Sci. Health B21(6):
    447-456.

Collett, N., and D.L. Harrison.  1968.  Some Observations on the Effects of
    Using Organochlorine Sprays in an Orchard.  N.Z.J. Sci. 11: 371-379.

Criddle, W.D. 1958.  Methods of Computing Consumptive Use of Water.  Proc.
    ASCE 84(IR1).
                                     357

-------
Dean, J.D., A.S.  Donigian,  Jr.,  and  J.E.  Rafferty.   1984.   Development of a
    Mathematical  Model  to Evaluate  Pesticide  Exposure  to Birds  in  Agri-
    cultural Ecosystems.   Draft  Report  of Work  Assignment No.  11,  Contract
    No. 68-03-3116.  U.S. Environmental  Protection  Agency, Athens, Georgia.

Dean, J.D.  1979.  Determination of  Irrigation  Water Demand in  Humid
    Climates.  Master's Thesis,  University of Georgia,  Agricultural
    Engineering Department.

Dejonckheere, W., W. Steurbaut,  0.  Melkebeke, and R.H.  Kips.   1983.
    Leaching of Aldicarb and Thiofanox,  and Their Uptake in Soils  by
    Sugarbeet Plants.  Pesticide Sci.  14:99-107.

DeWeese, L.R., L.C. McEwan, G.L. Hensler, and B.E.  Petersen.   1986.
    Organochlorine Contaminants  in  Passeriformes  and Other Prey of the
    Peregrine Falcon in the Western United States.   J.  Env. Tox. and  Chem.
    5(7):675-694.

Donigian, A.S., Jr., C.S. Raju,  and  R.F.  Carsel.   1986.  Impact of
    Conservation Tillage on Environmental Pesticide Concentrations in Three
    Agricultural  Regions.  Prepared  for U.S.  EPA, Integrated Environmental
    Management Division, Office  of  Policy Analysis, Washington, D.C.

Donigian, A.S., Jr., and J.D. Dean.   1985.  Nonpoint Source Models for
    Chemicals.  In: Environmental Exposure from Chemicals, Vol. II.  Boca
    Raton, Florida:  LRC Press.

Dorst, J.  1974.   The Life of Birds.  Vol. 1.  New York:  Columbia Univ.
    Press.

Dumbauld, R.K., J.R. Bjorklund,  and S.F.  Saterlie.   1980.  Computer Models
    for Predicting Aircraft Spray Dispersion  and  Deposition above and with-
    in Forest Canopies:  User's  Manual  for the  FSCBG Computer Program.
    Report No. 80-11.  U.S.D.A.  Forest Service, FPM-MAG, Davis, California.

Dumbauld, R.D., C.R. Bowman, and J.R.  Rafferty.  1980.  Optimum'*Swath
    Widths and Application Rates for Selected Aircraft and Meteorological
    Regimes.  USDA Forest Service,  Davis, CA.  Rep No. TR-80-319-01.

Dumbauld, R.D., J.E. Rafferty, and  J.R. Bjorklund,  1977.  Prediction of
    Spray Behavior above and within a Forest  Canopy.  USDA Forest
    Service.  Pacific N.W. Forest and Range Experiment Station, Portland,
    Oregon Rep. No. TR-77-308-01.

Dyer, A.J.  1974.  A Review of Flux-Profile Relationships.  Boundary-Layer
    Meteorology 7:363-372.

Dyer, A.J., and B.B. Hicks.  1970.   Flux-Gradient Relationships in the
    Constant Flux Layer.  Quarterly Journal of the Royal Meteorological
    Society 96:715-721.
                                     358

-------
Edwards, C.A.  1974.  Macroarthropods.  In:  Biology of Plant Litter
    Decomposition, Vol 2, eds. C. H. Dickinson and G.J.F Pugh.  New York:
    Academic Press.

El-Bayoumi, M.A., H.H. Shugart, Jr., and R.W. Wein.  1983/84.  Modelling
    Succession of Eastern Canadian Mixedwood Forest.  Ecol. Modelling
    21:175-198.

Farner, D.S., J.R. King, and K.C. Parks.  1972.  Avian Biology.  New
    York:  Academic Press.

Fletcher, J.S., M.J. Muhitch, D.R. Vann, J.C. McFarlane, and F.E.
    Benanati.  1985.  PHYTOTOX Database Evaluation of Surrogate Plant
    Species Recommended by the U.S. Environmental Protection Agency and the
    Organization for Economic Cooperation and Development.  Environ.  TOX
    and Contam.  4(4):  523-532.

Freeze, R.A.  1975.  Stochastic - Conceptual Analysis of One-Dimensional
    Groundwater Flow in Non-uniform Homogeneous Media.  Water Res. Res.
    11(5):  725-741.

Fuchs, N.A.  1959.  Evaporation and Droplet Growth in Gaseous Media.  New
    York:  Pergamon Press.

Gibbs, G.H., J. Barnard, N.H. Peck, and M.T. Vittum.  1980.  Soil and Air
    Temperature at Geneva, N.Y.  Search Agriculture Number 5.  New York
    State Agricultural Experimental Station, Cornell University, Ithaca,
    N.Y.

Green, W. Heber, and G.A. Ampt.  1911.  Studies in Soil Physics.  I.  The
    Flow of Air and Water through Soils.  J. Agri. Sci. 4:1-24.

Haan, C.T.  1977.  Statistical Methods in Hydrology.  Ames, Iowa:  Iowa
    State University Press.

Haith, D.A., and R.C. Loehr, eds.  1979.  Effectiveness of Soil and Water
    Conservation Practices for Pollution Control.  U.S. Environmental
    Protection Agency, Athens, Georgia.  EPA-600/3-79-106.

Harding, D.J.L., and R.A. Stuttard.  1974.  Microarthropods.  _In:  Biology
    of Plant Litter Decomposition, Vol. 2, eds. C.H. Dickinson and G.J.F.
    Pugh.  New York:  Academic Press.

Haque, A., and W. Ebing.  Uptake, Accumulation, and Elimination of HCB and
    2,4-D by the Terrestrial Slug, Deroceras reticulatum (Muller).  Bull.
    Environ. Contam. Toxicol. 31, 727-733 (1983).

Hetrich, D.M., and L.M. McDowell-Boyer.  1984.  User's Manual for
    TOXSCREEN:  A Multimedia Screening-Level Program for Assessing the
    Potential Fate of Chemicals Released to the Environment.  ORNL-6041,
    EPA-560/5-84-024.
                                     359

-------
Hill, E.F., R.G.  Heath,  J.W.  Spann,  and  J.D.  Williams.   1975.   Lethal
    Dietary Toxicities of Environmental  Pollutants  to  Birds.   Special
    Sci.  Rep.  -  Wildlife 191.   U.S.  Dept.  of Interior,  Fish  and  Wildlife
    Services Washington, O.C.

Jeffries and Davis.  1968.   Dynamics  of  Dieldrin in Soil,  Earthworms and
    Song Thrushes.   J. Wildlife Management.   32: 441-456.

Johanson, R.C., J.C. Imhoff,  and H.H.  Davis,  Jr.  1980.   User's Manual  for
    Hydrological  Simulation Program—Fortran  (HSPF).   EPA-600/9-80-015.

Johnson, N.L.,  and  S. Kotz.  1970.   Distributions in Statistics:
    Continuous  Univariate Distributions.   Boston:  Houghton Mifflin
    Company.

Jorde, D.G., G.L. Krapu, and R.O. Crawford.   1983.   Feeding Ecology of
    Mallards Wintering in Nebraska.   J.  Wildlife Management 47(4):   1044-
    1053.

Jorgensen, S.E.  1984.  Parameter Estimation  in Toxic  Substance Models.
    Ecol. Modeling  22:1-11.

Jury, W.A.  1985.  Spatial  Variability of Soil Physical  Parameters  in
    Solute Migration:  A Critical Literature  Review.  Electric Power
    Research Institute Report No. EA-4228.   Palo Alto, California.

Jury, W.A., W.F.  Spencer and W.F. Farmer.  1984.  Behavior Assessment Model
    for Trace Organics in Soil:  III.   Application  of  Screening Model.
    13:573-579.

Jury, W.A., R.  Grover, W.F. Spencer,  and W.J. Farmer.   1983.   Behavior
    Assessment  Model for Trace Organics in Soil:  I. Model Description.
    J. Environ. Qual., 12:558-564.

Karickhoff, S.W., D.S. Brown, and T.A. Scott.  1979.  Sorption of
    Hydrophobic Pollutants on Natural  Sediments.

Kiekebusch, B.D., B.H. Edwards, J.N.  Pavillin, and  K.A.  Coghlan-Jordan.
    1981.  Procedures for Predicting Transfers of Organic Substances in the
    Terrestrial Food Chain.  Draft Report for Contract No. 68-03-3110.
    U.S. Environmental Protection Agency, Athens, Georgia.

Kirchner, T.B., and F.W. Whicker.  1984.  Validation of PATHWAY.   A
    Simulation  Model of the Transport of Radionuclides through
    Agroecosystems.  Ecological Modeling 22:21-44.

Lassey, K.R.  1982.  The Interception  and Retention of Aerosols by
    Vegetation  -  I.  The Formulation of  a Filtration Module.   Atmospheric
    Environment 16(1):13-24.

Lee, K. E.  1985.  Earthworms.   Their  Ecology and Relationships with Soils
    and Land Use.  New York:  Academic Press.

                                     35C

-------
 Linsley,  R.K., Jr., M.A. Kohler, and J.L.H. Paulus.  1975.  Hydrology  for
     Engineers.  New York:  McGraw Hill

 Lo,  A.K.   1977.  An Analytical-Empirical Method for Determining the  Rough-
     ness  Length and Zero-Plane Displacement.  Boundary-Layer Meteorology
     12:141-151.

 Lofty, J.  R.  1974.  Oligochaetes.  In:  Biology of Plant Litter
     Decomposition, Vol.  II, eds. C.H. Dickenson and G. J. F Pugh.  New
     York:  Academic Press.

 Lyman, W.J., W.F. Reohl, and D.H. Rosenblatt.  1982.  Handbook of  Chemical
     Property Estimation  Methods.  New York:  McGraw-Hill.

 Matthews,  G.A.  1979.  Pesticide Application Methods.  New York:   Longman,
     Inc.

 McDonald,  J.E.  1960.  An Aid to Computation of Terminal Fall Velocities of
     Spheres.  Journal of Meteorology 17:463.

 McFarlane, J.C., T. Pfleeger and John Fletcher.  1987.  Transpiration
     Effect on the Uptake and Distribution of Bromacil, Nitrobenzene, and
     Phenol in Soybean Plants.  J. Env. Qual. 16(4):  372-376.

 McGrath et al., E.J. and D.C. Irving, 1973.  Techniques for Efficient  Monte
     Carlo  Simulation, Volume II.  Random Number Generation for Selected
     Probability Distributions.  Report prepared for Office of Naval
     Research.  Project No. NR 366-076/1-5-72, Code 462.

 McKone, T.E., and D.W. Layton.  1986.  Screening the Potential Risks of
     Toxic  Substances Using a Multimedia Compartment Model.  Lawrence Liver-
     more National Laboratory Project No. UCRL-94548.

 Mehlenbacher, L.A., and  D.W.A. Whitfield.  1977.  Modeling Thermal Eddy
     Diffusivity at Canopy Height.  Boundary-Layer Meteorology 12:153-170.

 Mejia J. M., and I. Rodriguez - Iturbe and J.R. Cordova.  1974.
     Multivariate Generation of Mixture of Normal and Log Normal
     Variables.  Water Resources Research 10, (4):691-693.

Menzie,  C.M.   1980.   Metabolism  of  Pesticide  Update  III.   U.S.  Dept.  of the
    Interior,  Fish  and  Wildlife  Service  Spec. Sci.  Rep.  Wildlife  No.  32.
    Washington D.C.

Mockus,  V.  1972.   Estimation of Direct  Surface  Runoff  from  Storm
    Rainfall.   In:   National  Engineering  Handbook.  Section  IV,  Hydrology.
    U.  S.  Soil  Conservation  Report  NEH-Notice 4-102.  August.

Nash, R.G.  1980.   Dissipation of  Pesticides  from  soils.   In:   CREAMS:   A
    Field  Scale  Model  for Chemicals,  Runoff,  and  Erosion from  Agricultural
    Management Systems,  W.G.  Knisel,  ed.,  U.S. Dept  of  Agriculture
    Conservation Res.  Rep. No. 26.   Washington,  D.C.


                                     361

-------
Nash, R.G.  1974.   Plant Uptakes of Insecticides,  Fungicides and Fumigants
    from Soils.  _In:   Pesticides in Soil  and Water.   Soil  Sci.  Soc.  Am.,
    Madison, Wisconsin.

Nielson, O.R., J.W. Biggar,  and K.T. Erh.  1983.   Spatial  Variability of
    Field Measured Soil  Water Properties.  Hilgardia 42(7):  214-259.

Norstrom, R.J.,  T.P.  Clark,  D.A. Jeffrey, and H.T.  Won.   1986.   Dynamics of
    Organochlorine Compounds in Herring Gulls (Larus Argentatus):  I.
    Distribution and  Clearance of [  C]DDE in Free-Living  Herring Gulls
    (Larus Argentatus).   Environmental  and Toxicology and  Chemistry, Vol.
    5, pp. 41-48,  1986.

Odum, E.P. 1971.  Fundamentals of Ecology.  3rd.  ed.  Philadelphia:   W.B.
    Saunders Company.

Oliver, H.R.  1971.  Wind Profiles in and above a Forest Canopy.  Quarterly
    Journal of the Royal Meteorological Society 97:548-553.

Parmele, L.H., E.R. Lemon, and A.W. Taylor.  1972.   Micrometeorological
    Measurement  of Pesticide Vapor Flux from Bare Soil and Corn under Field
    Conditions.  Water,  Air, and Soil Pollution 1:433-451.

Patterson, M.R., T.J. Sworski, A.L. Sjoreen, M.G. Browman, C.C. Coutant,
    D.M. Hetrich,  B.D. Murphy, and R.J. Raridon.   1984.   A User's Manual
    for UTM-TOX, the  Unified Transport Model.  ORNL-6064.

Perkins, H.C.  1974.   Air Pollution.  New York:  McGraw-Hill, Inc.

Picot, J.J.C.  1979.   Comments on Droplet Spectrum Lower Limit and Droplet
    Evaporation.  University of New Brunswick, Department of Chemical
    Engineering  Report,  Fredericton, New Brunswick, Canada.

Prandtl, L., and O.g. Tietjens.  1934.   Fundamentals of Hydro- and
    Aeromecahnics.  New York:  Dover Publications, Inc.

Pruppacher, H.R.,  and R. Rasmussen.  1979.  A Wind  Tunnel  Investigation of
    Evaporation  of Large Water Drops Falling at Terminal Velocity in Air.
    Journal of Atmospheric Science 36:1255-1260.

Rafferty, J.E.,  and R.K. Dumbauld.  1980.  Selection of  Spray Operations
    Criteria for the  Withlacoochee State Seed Orchard Project.   USDA,
    Forest Service Methods Application  Group Rep.  No. 80-8.   Davis,  CA.

Rao, P.S.C., and R.J. Wagenet.  1985.  Spatial Variability of Pesticides in
    Field Soils:  Methods for Data Analysis and Consequences.  Weed  Sci.
    33: 1-26.

Rao, P.S.C, and  J.M.  Davidson.  1980.  Estimation of Pesticide Retention
    and Transformation Parameters Required in Nonpoint Source Pollution
    Models.  In:  Environmental Impact of Nonpoint Source Pollution, eds.
    M.R. Overcash  and J.M Davidson.  Ann Arbor, Michigan:   Ann Arbor
    Sciences Publishers, Inc.

                                     362

-------
Rawls, W.J.  1983.  Estimating Soil Bulk Density from Particle Size
    Analysis to Organic Matter Content.  Soil Science, 135(2):  123-125.

Richards, L.A.  1931.  Capillary Conduction of Liquids Through Porous
    Mediums.  Physics 1:318-333.

Roberts, B.C., and H.W. Dorough.  1985.  Hazards of Chemicals to Earth-
    worms.  Environmental Toxicology and Chemistry 4(3):307-323.

Rosenberg, N.J.  1974.  Microclimate:  The Biological Environment.  New
    York:  Wiley Interscience.

Ruppell, R.F., and C.W. Laughlin.  1976.  Toxicity of some Pesticides to
    Earthworms.  J. Kans.  Entomol.  Soc. So:  113-118.

Schnoor, J.L., C. Sato, D. McKechin and D. Sahoo.  1987.  Processes,
    Coefficients, and Models for Simulating Toxic Organics and Heavy Metals
    in Surface Waters.  WPA/600/3/3-87/015.  U.S. EPA, Environmental
    Research Laboratory, Athens, Georgia.

Smith, A.  1932.  Seasonal Subsoil Temperature Variations.  J. Agr.
    Research 44:421-428.

Smith, C.N., and R.F. Carsel.  1984.  Foliar Washoff of Pesticide (FWOP)
    Model:  Development and Evaluation.  Journal of Env. Sci. and Health.

Smith, R.L.  1966.  Ecology and Field Biology.  New York:  Harper and Row
    Publishers.  2nd ed.

Soil Conservation Service, U.S. Department of Agriculture.  1972.  SCS
    National Engineering Handbook, Section 4-Hydrology.

Spencer,  W.F., W.A. Jury, and W.J.  Farmer.   1984.   Importance of
    Volatilization as a Pathway for Pesticide Loss  from Forest Soils.   In
    Chemical and Biological  Controls In Forestry.   American  Chemistry
    Society, pp. 193-209.

Stanhill, G.  1969.  A Simple Instrument for the Field Measurement of
    Turbulent Diffusion Flux.  J.  Applied Meteorology 8:504-513.

Stewart,  B.A., D.A. Woolhiser, W.H.  Wischmeier,  J.H.  Caro,  and M.H.  Fere.
    1976.  Control  of Water Pollution from Cropland,  Volume  II:   An Over-
    view.  U.S.  Environmental Protection Agency,  Athens,  Georgia.
    EPA-600/2-75-026b.

Swanson,  G.A., M.I. Meyer, and V.A.  Adomaitis.   1985.   Foods Consumed  by
    Breeding Mallards on Wetlands of South Central  North  Dakota.  J.
    Wildlife Management. 49(1):197-203.

Taylor, J.  R.  1982.  An Introduction to Error Analysis.   Mill Valley,
    California:   University Science Books.
                                    363

-------
Thibodeaux, L.J.,  and H.D.  Scott.   1985.   Air/Soil  Exchange Coefficients.
    jji: Environmental Exposure from Chemicals,  Vol.  I,  eds. W.B. Neely and
    G.E. Blau, pp.  65-90, CRC Press, Inc.

Thibodeaux, L.J.   1979.  Chemodynamics:   Environmental  Movement of
    Chemicals in  Air, Water,  and Soil.   New York:   John Wiley & Sons.

Thorn, A.S., J.B.  Stewart, H.R. Oliver,  and J.H.C.  Gash.  1975.  Comparison
    of Aerodynamic and Energy Budget Estimates  of  Fluxes over a Pine
    Forest.  Quarterly Journal of the Royal Meteorological  Society
    101:93-105.

Thompson, A.R, and C.A. Edwards.  1974.   Effects of Pesticides on Nontarget
    Invertibrates  in Freshwater and Soil.   In:   Pesticides  in Soil and
    Water, W.D. Guenzis, ed.   Soil. Sci.  Soc. Am.   Madison, Wisconsin.

Tinsley, I.J.  1979.  Chemical Concepts  in Pollutant Behavior.  New York:
    John Wiley &  Sons.

Todd, O.K.  1980.   Groundwater Hydrology.   Sec. Ed.  NY:  Wiley.

Todd, O.K.  1970.   The Water  Encyclopedia, Water Information Center, Port
    Washington, New York.

U.S. Department of Agriculture, Forest Service.  1984.   Measurement of Drop
    Size Frequency from Nozzles used for Aerial Applications of Pesticides
    in Forests.  October.

U.S. Department of Agriculture.  1980.   CREAMS:  A Field-Scale Model for
    Chemicals, Runoff and Erosion from Agricultural  Management Systems,
    W.G. Knisel,  ed., U.S.  Dept. of Agriculture Conservation Res. Rep. No.
    26. Washington, D.C.

Uk, S., and R.J.  Courshee.   1982.   Distribution and  Likely  Effectiveness of
    Spray Deposits Within a Cotton Canopy  From  Fine  Ultralow-Volume Spray
    Applied by Aircraft.  Pesticide Science 13:529-536.

van Genuchten, M.  Th.  1978.   Mass Transport in Saturated - Unsaturated
    Media:  One Dimensional Solutions.   Dept. of Civil  Eng.  Princeton
    University Res. Rep. 78-WR-ll.

Van Gestel, C.A.M., and W.  Ma.  1987.  Toxicity and  Bioaccumulation of
    Chlorinated Phenols in  Earthworms.   In:  Vulnerability  of Soil and
    Groundwater to Pollutants.  Proceedings of  the International
    Conference.  March 30-April 3, 1987.   Noordwijk  on  Zee, The
    Netherlands.

Wark, K., and C.F. Warner.   1976.   Air Pollution;   Its  Origin and
    Control.  New York:  Harper & Row Publishers.

Welty, J.C.  1962.  The Life  of Birds.   Philadelphia:  W.B. Saunders Co.
                                     364

-------
Welty, J.C.  1975.  The Life of Birds.  Philadelphia:  W.B. Saunders Co.

Willardson, L.S., and R.L. Hurst.  1965.  Sample Size Estimates in
    Permeabilty Studies.  ASCE J. of Irr. and Drainage.  91(IRI):  1-9.

Williams, J.R., C.A. Jones, and P.T. Dyke.  1988.  The EPIC Model.  U.S.
    Department of Agriculture.  Ag. Res. Serv., Temple, Texas.

Williams, J.R., P.T. Dyke, W.W. Fuchs, V.W. Bensen, O.W. Rice, and E.D.
    Taylor.  1987.  EPIC the Erosion-Productivity Impact Calculator.
    Volume II.  User Manual.  Draft document USDA, Agricultural Research
    Service, Economics Research Service, and Soil Conservation Service.

Williams, J.R., C.A. Jones, and P.T. Dyke.  1983.  A Modeling Approach to
    Determining the Relationship between Erosion and Soil Productivity.
    Trans. ASAE 27:129-144.

Williams, J.R., C.A. Jones, and P.T. Dyke.  1983.  EPIC - A Model for
    Assessing the Effects of Erosion on Soil Productivity.  J.n:  Analysis
    of Ecological Systems:  State-of-the-Art in Ecological Modeling, eds.
    W.K. Lauenroth, G.V. Skogerboe, and M. Flug.  New York:  Elsevier
    Scientific Publishing Company.

Williams, J.R.  1975.  Sediment Yield Prediction with Universe Equation
    Using Runoff Energy Factor.  Jji:  Present and Prospective Technology
    for Predicting Sediment Yields and Sources.  U.S. Dept of Agric.
    ARS-S-40.

Wright, M.A.  1977.  Effects of Benomyl and some other Systemic Fungicides
    on Earthworms.  Ann.  Appl.  Biol. 87:  520-524.

Yevjevich, V.  1972.  Probability and Statistics in Hydrology.  Water
    Resources Publication, Littleton, Colorado.
                                     355

-------
                             APPENDIX A

ATEEAM  EXAMPLE OUTPUT USING DEFAULT PARAMETER VALUES
*
*
*
*
*
*
*
*
*
*
**********************

A-TEEAM

ANALYTICAL TERRESTRIAL ECOSYSTEM
EXPOSURE ASSESSMENT MODEL

DECEMBER 1987

**********************
*****
*
*
*
*
*
*
*
*
* * * * *
            Example simulation for documentation.Single application
                        INPUT  DATA
           NMCR      (NUMBER OF MONTE CARLO RUMS)'        =  500

             NP      (NUMBER OF MODEL INPUT PARAMETERS
                     8EIN6 VARIED)                    =    9
           NDAYS      (LENGTH OF SIMULATION PERIOD (DAYS)  =  365

            NCAT      (NUMBER OF CATEGORIES TO DIVIDE
                      THE OUTPUT TIME SERIES INTO)       =    3
           IMODEL     Type of model used                     1
       SIMULATION STARTED ON  1/1/1987
       SIMULATION ENDED  ON 31/12/1987
                                 366

-------
SUMMARY OF INPUT PARAMETERS
PAR NAME
DECAY1
RATF12
OECAY2
RATE23
OECAY3
RATE13
MASS1
MASS?
mm
CONC1
CQNC2
CONC3
INPUT MEAN
0.2000E-01
0.6700F-05
0.2000E-01
0.1020E-03
1.200
0.7150E-09
O.S630E+09
0.3750E+05
80.00
7,500
O.OOOOE+00
O.OOOOE+00
INPUT C.V.
1.000
0.5000
1,000
0.5000
1.000
1.000
0.1000
0,1000
0.1000
O.OOOOEtOO
O.OOOOEfOO
O.OOOOE+00
OIST, MEAN.
0.2000E-01
0.6700T-05
0.2000E-01
0.1020E-03
1.200
0.7150E-09
0.5630E+09
0.3750E+OS,
80,00



OIST. STD
0.2000E-01
0.3350F-05
0.2000E-01
0.5100E-04
1.200
0.7150E-09
0.5630E+08
3750.
8.000



MINIMUM VALUE
O.OOOOE+00
O.OOOOE+00
O.OOOOEtOO
O.OOOOEfOO
O.OOOOE+00
O.OOOOE+00
0.1000
0.1000
0,1000
O.OOOOE+00
O.OOOOEfOO
O.OOOOE+00
MAXIMUM VALUE
10,00
0.1000E-02
0,1000
0.5000E-02
5.000
0.1000E-04
0.2700Et11
0.3000Et07
8000.
10.00
1.000
1.000
OIST TYPE
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
NORMAL
CONSTANT
CONSTANT
CONSTANT
                                           367

-------
         CORRELATION HHRIX



             DECAY1    RATE12    DECAY?    RATE23    OECAY3    RATE13    MASS1     IWSS2     MASS3






OECAY1       1,0000    0,9.500    0,0000    00000    0,0000    0.0000    0.0000    0.0000    0.0000
RATE12       0,9500    1,0000    0,0000    0.0000    0,0000    0,0000     0.0000    0.0000    O.C






DECAY2       0.0000    0.0000    1.0000    0.0000    0,0000    0.0000     0,0000    0,0000    0,0000






RATE23       0,0000    0,0000    0,0000    1.0000    0.0000    0.0000     0.0000    0,0000    0,






DECAY3       0.0000    0.0000    0,0000    0,0000    1.0000    0,0000     0,0000    0.0000    0,0






RATE13       0.0000    0,0000    0,0000    0.0000    0.0000    1,0000     0.0000    0.0000    O.C






west        0,0000    o.oooo    0,0000    o.oooo    0,0000    0,0000     1.0000    0,0000    o.oooo






wss?        o.oooo    0,0000    o.oooo    0,0000    o.oooo    o.oooo     o.oooo    1,0000    o.oooo






WSS3        0,0000    0,0000    0.0000    0.0000    0.0000    0,0000     0.0000    0,0000    1.0000








        NSW       Seed used  in random number generator          1234567






           543 KME CARLO RUNS REJECTED BECAUSE PARAMETER BOUNDS EXCEEDED
                                               368

-------
OUTPUT  FOR   LOWER   TROPHIC   LEVEL
      NUMER OF DAYS DURING SMJLATION PERIOD *9I «AN CONCBffRATION EXCEEDED  20.000    ug/g
                         Example slailatlm for doomentaficn.Single application

                                   N                   =  500
                                   MEAN                 -  17.8
                                   STANDARD DEVIATION     =  59.2
                                   COEFICIENT OF VARIATION =  3.33






VALUE

O.OOOE+00

34.5

69.0

103.

138.

172.

207,

241.

278.

310,

345,
MINIUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
% OF TMC EQUALLED
OR EXCEEDED
100.000

11.000

8.200

6.600

5.200

4.400

3,600

2,400

2.000

1.200

0.200
= O.OOOEtOO
= 345.
= O.OOOE+00
= O.OOOEtOO
= 48.0
= 139.
* OF TIME IN INTERVAL


89.000

2.800

1.600

1.400

0.800

0.800

1.200

0.400

0,800

1.000

                                       369

-------
 OUTPUT  FOR   LOWER   TROPHIC    LEVEL
MEAN DAILY CONCBITRATION FROM    0 TO 120 DAYS
                           Example simulation fop doomentation. Single application

                                     N                   =500
                                     MEAN                 =  10.2
                                     STANDARD DEVIATION      =  4,91
                                     COEFFICIENT OF VARIATION =  0,480






VALUE

1.00

4.02

7.04

10.1

13.1

16.1

19,1

22.1

25.2

28.2

31.2
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
% OF TIME EQUALLED
OR EXCEEDED
100.000

97,200

69,800

42.400

23.600

12,000

6.200

3.000

1.200

0.600

0,400
= 2.71
= 31.2
= 13.6
= 15,1
= 17.0
= 20,4
\ OF TIME IN INTERVAL


2,800

27.400

27,400

18.800

11.600

5,800

3.200

1,800

0.600

0.200

                                       370

-------
 OUTPUT   FOR   LOWER   TROPHIC    LEVEL


DOSAGE DURING THE PERIOD FROM    0 TO 120 DAYS
                          Exwple simulation for documentation,Stogie application

                                    N                   -   500
                                    MEAN                 --   33,8
                                    STANDARD DEVIATION      =   10,3
                                    COEFFICIENT OF VARIATION =  0,304






VALUE

1.00

8.86

18.7

24.6

32.4

40.3

48.1

56,0

63.8

71,7

79.6
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
\ OF TIME EQUALLED
OR EXCEEDED
100.000

100.000

99.600

83.600

46.800

21.600

11.000

3.800

1.200

0.400

0.200
= 15,7
= 79.8
= 41.4
= 44.4
= 49,2
* 54.0
* OF TIME IN INTERVAL


0,000

0.400

16.000

36.800

25.200

10.600

7.200

2.600

0.800

0.200

                                     371

-------
 OUTPUT  FOR   LOHER   TROPHIC    LEVEL
MEW DAILY CONCENTMHON FROM   120 TO  243 DAYS
                          Example simulation for documentation.Single application

                                    N                   =  500
                                    MEAN                =  5.6?
                                    STANDARD DEVIATION     =  6.83
                                    COEFFICIENT OF VARIATION =  1.22






VALUE

0.127E-01

4.55

9.08

13,6

18,2

22.7

27.2

31,8

36.3

40.8

45.4
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PEBCENTILE
95th PERCENTILE
* OF TIME EQUALLED
OR EXCEEDED
100.000

40.400

19.600

11.400

6.200

4,200

2.000

0,800

0.600

0,400

0.200
= 0.127E-01
= 45,4
= 8.84
= 11.3
= 14.5
= 19,7
% OF TIME IN INTERVAL


59,600

20.800

8.200

5.200

2.000

2.200

1.200

0,200

0.200

0.200

                                      372

-------
 OUTPUT  FOR   LOWER   TROPHIC    LEVEL


DOSAGE DURING THE PERIOD FROM   120 TO 243 DAYS
                           Example simulation for documentation.Single application

                                     N                   =500
                                     MEAN                 --   6.44
                                     STANDARD DEVIATION     =   10.4
                                     COEFFICIENT OF VARIATION =   1.6?






VALUE

0.165E-02

7.60

15.2

22,8

30.4

38.0

45.6

53.2

60.8

68.4

75.9
MINIMUM VALUE
MAXINJM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
* OF TIME EQUALLED
OR EXCEEDED
100.000

25.200

14.600

8.200

4.200

2.800

1.000

0.400

0.200

0,200

0.200
= 0.165E-02
= 75.9
= 10.6
= 14,8
= 20.0
= 28.0
* OF TIME IN INTERVAL


74.800

10.600

6.400

4.000

1.400

1.800

0,600

0.200

0.000

0.000

                                          373

-------
 OUTPUT  FOR   LOWER   TROPHIC    LEVEL
MEW DAILY CONCENTRATION FROM  243 TO  355 DAYS
                          Example simulation for documentation,Single application

                                     N                   =500
                                     MEAN                =   2,84
                                     STANOARD DEVIATION     =   5,64
                                     COEFFICIENT OF VARIATION =   1.99






VALUE

0.101E-04

5.63

11.3

16.9

22,5

28.2

33.8

39,4

45.0

50.7

56.3
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCFNTILE
* OF TIME EQUALLED
OR EXCEEDED
100.000

14.600

7,400

3,400

1.200

0.800

0.600

0.400

0.200

0.200

0,000
« 0.101E-04
= 56,3
= 3.91
= 5.50
= 7.54
= 14.3
% OF TIME IN INTERVAL


85.400

7,200

4.000

2.200

0.400

0.200

0.200

0.200

0,000

0,200

                                         374

-------
 OUTPUT  FOR   LOWER    TROPHIC    LEVEL
DOSAGE DURING THE PERIOD FROM   243 TO 365 DAYS
                          Example simulation for documentation.Single application

                                     N                   =  500
                                     MEAN                 «  2.95
                                     STANDARD DEVIATION      -  7.56
                                     COEFFICIENT OF VARIATION =  2.56






VALUE

0.107E-06

7.01

U.O

21.0

28.0

35.1

42.1

49.1

56,1

63.1

70.1
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
\ OF TIME EQUALLED
OR EXCEEDED
100.000

12.800

6,400

4.000

2.400

1.400

0,800

0.200

0.200

0.200

0,200
= 0.107E-06
= 70.1
= 2.75
= 5.07
= 9,26
= 16,3
% OF TIME IN INTERVAL


87.200

6.400

2.400

1.600

1,000

0.600

0.600

0.000

0.000

0.000

                                          375

-------
 OUTPUT  FOR    LOWER   TROPHIC    LEVEL


RESULTS FOR MAXIMUM CONCENTRATION
                           Example simulation for documentation,Single application

                                      N                    =500
                                      MEAN                 --   14,0
                                      STANDARD DEVIATION     =   5,17
                                      COEFFICIENT OF VARIATION =  0,439






VALUE

1.00

7.08

13.2

19,2

25,3

31,4

37.5

43.6

49,5

55.7

61,8
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
% OF TIME EQUALLED
OR EXCEEDED
100,000

95.000

46.400

14.600

5.400

2.200

0.800

0.400

0.200

0.200

0.000
= 4,49
= 61,8
= 17.5
= 19.0
= 21.1
* 25.6
\ OF TIME IN INTERVAL


5.000

48.600

31.800

9.200

3.200

1.400

0,400

0,200

0.000

0,200

                                           376

-------
OUTPUT  FOR   HIGHER   TROPHIC   LEVEL
      NUMBER OF DAYS DURING SIMULATION PERIOD WHEN MEAN CONCENTRATION EXCEEDED  4000.0    ug/g
                         Example simulation for doctnBntatlon. Single application










VALUE

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00

O.OOOE+00
N =
MEAN =
STANDARD DEVIATION
COEFFICIENT OF VARIATION =
MINIMUM VALUE
MAXIMUM VALUE
80th PERCESTILE
85th PEflCENTILE
90th PERCBITILE
95th P0KEHTILE
* OF TIME EQUALLED * OF
OR EXCEEDED
100,000

100.000

100.000

100.000

100.000

100.000

100.000

100,000

100.000

100.000

100.000
500
0,0006+00
O.OOOE+00
??????????
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
O.OOOE+00
TIME IN INTERVAL


0.000

0.000

0.000

0.000

0.000

0.000

0.000

0.000

0.000

0.000

                                         377

-------
OUTPUT  FOR   HIGHER   TROPHIC   LEVEL


   DAILY CONCENTRATION FROM    0 TO 120 OAVS
                          Example simulation for doonentation,Single application

                                    N                   =  SOU
                                    MEAN                 = 0.930
                                    STANDARD DEVIATION      =  3,53
                                    COEFFICIENT OF VARIATION =  3,79






VALUE

0.642E-02

5.50

11.0

16.5

22.0

27,5

33.0

38,5

44.0

49.5

55.0
MINIMUM VALUE
MAXIMJM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
% OF TIME EQUALLED
OR EXCEEDED
100,000

1.600

1,200

1,000

0,600

0.400

0.400

0.400

0.200

0.200

0.200
= 0.642E-02
= 55.0
= 0.753
= 1.04
= 1,57
= 2,44
% OF TIME IN INTERVAL


98,400

0.400

0.200

0.400

0,200

0.000

0.000

0.200

0.000

0,000

                                         378

-------
 OUTPUT  FOR   HIGHER   TROPHIC    LEVEL
DOSAGE DURING THE PERIOD FROM    0 TO  120 CAYS
                          Example simulation for documentation,Single application

                                    N                   =   500
                                    MEAN                --   80,9
                                    STANDARD DEVIATION     =   42,4
                                    COEFFICIENT OF VARIATION =  0,696






VALUE

0.743E-01

30,6

61,2

91,8

122,

153.

184.

214,

245.

275,

306.
MINIMUM VALUE
MAXIMUM VALUE
80th PERC9ITILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
% OF TIME EQUALLED
OR EXCEEDED
100.000

76,800

39,000

17,600

9,000

4,000

1,800

0.800

0,400

O.?00

0,200
= 0.743E-01
= 306.
= 86,6
= 101,
= 116,
= 147.
% OF TIME IN INTERVAL


23,200

37.800

21.400

8.600

5.000

2,200

1.000

0.400

0.200

0.000

                                          379

-------
 OUTPUT  FOR   HIGHER   TROPHIC   LEVEL
HEW DAILY CONCENTRATION FROM   120 TO  243 DAYS
                           Example simulation for documentation,Single application

                                      N                    =   500
                                      MEAN                  =  0.755
                                      STANDARD DEVIATION      =   4,80
                                      COEFFICIENT OF VARIATION  =   8.38






VALUE

0.362E-03

8.44

16.9

25.3

33.8

*2.2

50.6

59.1

67.5

76.0

84.4
MINIMUM VALUE
MAXIMIM VALUE
80th PERCENTILE
85th PERCENTILE
90th PEflCENTILE
95th PERCENTILE
1 OF TIME EQUALLED
OR EXCEEDED
100.000

1.000

0.800

0,600

0.600

0.400

0.200

0.200

0.200

0.200

0.000
= 0.362E-03
= 84,4
= 0.414
= 0,605
= 0.862
= 1,89
* OF TIME IN INTERVAL


99.000

0.200

0.200

0.000

0,200

0,200

0,000

0.000

0.000

0.200

                                             380

-------
 OUTPUT  FOR   HIGHER   TROPHIC   LEVEL


DOSAGE DURING THE PERIOD FROM   120 TO 243 DAYS
                          Example simulation for documentation.Single application

                                     N                   =500
                                     MEAN                 =  34.6
                                     STANDARD DEVIATION      -  50,0
                                     COEFFICIENT OF VARIATION =  1.45






VALUE

0.668E-01

39,8

79.5

119,

159,

199.

238.

278.

318,

358,

397,
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
% OF TIHE EQUALLED
OR EXCEEDED
100.000

27.600

10.800

6.800

3.000

2,000

1.000

0.800

0,400

0.200

0.200
= 0.666E-01
= 397,
= 53.1
= 64.2
= 81.9
= 142.
% OF TIME IN INTERVAL


72,400

16.800

4,200

3.600

1.000

1.000

0.200

0.400

0.200

0.000

                                          381

-------
 OUTPUT  FOR   HIGHER   TROPHIC   LEVEL


NEW DAILY CONCENTRAHCII FROM   243 TO 365
                           Example simulation for docuwntatlon. Single application

                                     N                   =500
                                     MEAN                 =  0,372
                                     STANDARD DEVIATION     =  2.36
                                     COEFFICIENT OF VARIATION =  6.36






VALUE

0.287E-06

4.04

8.08

12,1

16,2

20.2

24.2

28.3

32.3

36.4

40.4
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILE
95th PERCENTILE
* OF TIME EQUALLED
OR EXCEEDED
100.000

1.400

0.800

0.800

0.400

0,400

0,400

0.200

0,200

0.200

0,200
= 0.287E-06
= 40.4
= 0,167
= 0,233
= 0.428
= 1.12
% OF TIME IN INTERVAL


98.600

0,600

0,000

0.400

0.000

0.000

0.200

0,000

0.000

0.000

                                        382

-------
 OUTPUT  FOR   HIGHER   TROPHIC   LEVEL


DOSAGE DURING THE PERIOD FROM   243 TO  365 DAYS
                          Example simulation for documentation.Single application

                                     N                   =500
                                     MEAN                 -  17.4
                                     STANDARD DEVIATION      --  40.5
                                     COEFFICIENT OF VARIATION =  2,33






VALUE

0.645E-04

42.2

84,4

127.

169,

211,

253,

295,

338.

380.

422.
MINIMUM VALUE
MAXIMUM VALUE
80th PERCENTILE
85th PERCENTILE
90th PERCENTILF
95th PERCENTILE
% OF TIME EQUALLED
OR EXCEEDED
100.000

9,400

., 5.200

2,400

1.800

0.800

0.600

0.600

0.200

0.200

0.200
= 0.645E-04
= 422.
= 22,3
« 30.2
= 41.6
-- 86,6
\ OF TIME IN INTERVAL


90.600

4.200

2.800

0,600

1,000

0.200

0,000

0,400

0,000

0.000

                                        383

-------
  OUTPUT   FOR   HIGHER   TROPHIC   LEVEL


 RESULTS FOR MAXIMUM CONCENTRATION
                            Example simulation for doai»ntat:1«v.S1ng1e application

                                       N                    =500
                                       MEAN                  =   1.37
                                       STANDARD DEVIATION      =   5.88
                                       COEFFICIENT OF VARIATION =   4.28






VALUE

0.815E-02
9.78

19,6

29.3

39,1

48.9

58.7

68.4

78.2

88.0

97.7
MINIMUM VALUE
MAXIMUM VALUE
80th PERCfNTILE
85th PERCENTILE
90th PERCBITILE
95th PERCENTILE
* OF TIME EQUALLED
OR EXCEEDED
100,000
1.200

1,000

0.600

0.600

0.400

0.400

0.200

0,200

0.200

0.200
= 0.815E-02
= 97.7
= 1.06
= 1.39
= 2.11
= 3,48
* OF TIME IN INTERVAL

oo onn
So.BUU
0.200

0.400

0.000

0.200

0,000

0.200

0,000

0,000

0.000


-------