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
------- |