&EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
MD-14
EPA-454/R-97-004e
July 1997
EIIP Volume V
Biogenic Sources
Preferred and Alternative
Methods
?P^APCO
-------
PREFACE
As a result of the more prominent role given to emission inventories in the 1990 Clean
Air Act Amendments (CAAA), inventories are receiving heightened priority and
resources from the U.S. Environmental Protection Agency (EPA), state/local agencies,
and industry. More than accountings of emission sources, inventory data are now
providing the prime basis for operating permit fee systems, State Implementation Plan
(SIP) development (including attainment strategy demonstrations), regional air quality
dispersion modeling assessments, and control strategy development. This new emphasis
on the use of emissions data will require significantly increased effort by state/local
agencies to provide adequate, accurate, and transferable information to meet various
agency and regional program needs.
Existing emission inventory data collection, calculation, management, and reporting
procedures are not sufficient or of high enough quality to meet all of these needs into
the next century. To address these concerns, the Emission Inventory Improvement
Program (EIIP) was created. The EIIP is a jointly sponsored endeavor of the State and
Territorial Air Pollution Program Administrators/Association of Local Air Pollution
Control Officials (STAPPA/ALAPCO) and the U.S. EPA, and is an outgrowth of
recommendations put forth by the Standing Air Emissions Work Group (SAEWG) of
STAPPA/ALAPCO. The EIIP Steering Committee and technical committees are
composed of state/local agency, EPA, industry, consultant, and academic representatives.
In general, technical committee participation is open to anyone.
The EIIP is defined as a program to develop and use standard procedures to collect,
calculate, store, and report emissions data. Its ultimate goal is to provide cost-effective,
reliable, and consistent inventories through the achievement of the following objectives:
• Produce a coordinated system of data measurement/calculation methods as
a guide for estimating current and future source emissions;
• Produce consistent quality assurance/quality control (QA/QC) procedures
applicable to all phases of all inventory programs;
• Improve the EPA/state/local agency/industry system of data collection,
reporting, and transfer; and
• Produce an integrated source data reporting procedure that consolidates
the many current reporting requirements;
-------
EIIP goals and objectives are being addressed through the production of seven guidance
and methodology volumes. These seven are:
• Volume I: Introduction and Use of EIIP Guidance for Emissions
Inventory Development
Volume II: Point Sources Preferred and Alternative Methods
Volume III: Area Sources Preferred and Alternative Methods
Volume IV: Mobile Sources Preferred and Alternative Methods
Volume V: Biogenics Sources Preferred and Alternative Methods
Volume VI: Quality Assurance Procedures
Volume VII: Data Management Procedures
The purpose of each volume is to evaluate the existing guidance on emissions estimation
techniques, and, where applicable, to identify the preferred and alternative emission
estimation procedures. Another important objective in each volume is to identify gaps in
existing methods, and to recommend activities necessary to fill the gaps. The preferred
and alternative method findings are summarized in clear, consistent procedures so that
both experienced and entry-level inventory personnel can execute them with a reasonable
amount of time and effort. Sufficiently detailed references are provided to enable the
reader to identify any supplementary information. Users should note that the number of
source categories or topics covered in any volume is constantly expanding as a function
of EIIP implementation and availability of new information.
It is anticipated that the EIIP materials will become the guidance standard for the
emission inventory community. For this reason, the production of EIIP volumes will be
a dynamic, iterative process where documents are updated over time as better data and
scientific understanding support improved estimation, QA, and data management
methods. The number of individual source categories addressed by the guidance will
grow as well over time. The EIIP welcomes input and suggestion from all groups and
individuals on how the volumes could be improved.
-------
VOLUME V
454R97004E
BIOGENIC SOURCES PREFERRED
METHODS
Final Report
May 1996
Prepared by:
Radian Corporation
Post Office Box 13000
Research Triangle Park, North Carolina
27709
Prepared for:
Area Sources Committee
Emission Inventory Improvement Program
-------
DISCLAIMER
As the Environmental Protection Agency has indicated in Emission Inventory Improvement
Program (EIIP) documents, the choice of methods to be used to estimate emissions depends on
how the estimates will be used and the degree of accuracy required. Methods using site-specific
data are preferred over other methods. These documents are non-binding guidance and not rules.
EPA, the States, and others retain the discretion to employ or to require other approaches that
meet the requirements of the applicable statutory or regulatory requirements in individual
circumstances.
-------
ACKNOWLEDGEMENT
This document was prepared by Lucy Adams of Radian Corporation for the Biogenic Sources
Committee of the Emission Inventory Improvement Program and for Thomas Pierce of
Modeling Systems Analysis Branch, U.S. Environmental Protection Agency. Members of the
Biogenic Sources Committee contributing to the preparation of this document are:
Janet Arey, State Wide Air Pollution Research Center, University of California-Riverside
Marianne Causley, Systems Applications International
Can Furiness, North Carolina State University
Chris Geron, Air Pollution Prevention and Control Division, U.S. Environmental Protection Agency
Alex Guenther, Atmospheric Chemistry Division
Gordon L. Hutchinson, U.S. Department of Agriculture-ARS
Mike Koerber, Lake Michigan Air Directors Consortium
Richard McNider, Earth System Science Laboratory, University of Alabama-Huntsville
Michael Rogers, Earth and Atmospheric Sciences, Georgia Tech
Thomas Sharkey, Department of Botany, University of Wisconsin
Ralph Valente, Atmospheric Science Department, Tennessee Valley Authority
Patricia Valesco, California Air Resources Board
Eric Williams, National Oceanic and Atmospheric Administration Aeronomy
Other contributors to this document are:
Raymond Asregado, Technical Support Division, California Air Resources Board
Arthur Winer, Director Environmental Science and Engineering Program, UCLA School of Public Health
EIIP Volume V
-------
CONTENTS
Chapter Page
1 Introduction 1-1
1.1 Background 1-1
1.2 Natural Source Emissions vs. Biogenic Source Emissions 1-1
2 Source Category Definition 2-1
2.1 Pollutants and Sources 2-1
2.1.1 Volatile Organic Compounds 2-1
2.1.2 Oxides of Nitrogen 2-1
3 Available Methods 3-1
3.1 Methods for Estimating Biogenic Emissions 3-1
3.1.1 Overview of the Biogenic Emission Inventory System 3-3
3.1.2 Alternative Methods for Estimating Biogenic Emissions 3-14
3.2 Methods for Estimating Emissions from Lightning 3-15
3.2.1 Overview of the Preferred Method 3-16
3.2.2 Overview of the Alternative Method 3-16
3.2.3 Method Uncertainty 3-17
3.3 Methods for Estimating Emissions from Oil and Gas Seeps 3-17
3.3.1 Preferred Methods 3-17
3.3.2 Alternative Methods 3-18
3.4 Uncertainty of Biogenic Emissions Estimates 3-19
3.4.1 Sensitivity Analysis 3-20
3.5 Quality Assurance/Quality Control 3-23
3.5.1 Land Use/Biomass Density 3-24
3.5.2 Meteorological Data 3-24
3.5.3 Alternative Methods 3-25
iv EIIP Volume V
-------
CONTENTS (CONTINUED)
Chapter Page
4.0 Preferred Method For Estimating Natural Source Emissions 4-1
4.1 Biogenic Sources 4-1
4.1.1 Using BEIS 4-1
4.1.2 BEIS and UAM 4-2
4.1.3 Processing Methodology 4-2
4.1.4 Input Requirements 4-3
4.2 Using PCBEIS2.2 4-6
4.2.1 Introduction 4-6
4.2.2 Processing Methodology 4-7
4.3 Estimating Emissions from Lightning 4-21
4.4 Estimating Emissions from Oil and Gas Seeps 4-23
5 Using Alternative Methods For Estimating Emissions 5-1
5.1 Alternative Methods for Estimating Biogenic Emissions 5-1
5.1.1 Yienger/Levy Method for Soil NOX 5-1
5.1.2 BIOME 5-4
5.1.3 Using Alternate Land Use Files with BEIS 5-6
5.2 Using Alternative Methods For Estimating Emissions from Lightning ... 5-13
5.3 Using Alternative Methods For Estimating Emissions from Oil and Gas
Seeps 5-14
6 References and Bibliography 6-1
EIIP Volume V
-------
FIGURES
Page
3-1 Effect of solar radiation, relative humidity and wind speed upon
predicted isoprene 3-21
3-2 Temperature relationships within forest canopies 3-22
4-1 UAM stand-alone biogenics processor. Overview of the Biogenic
Emission Inventory System (BEIS) 4-4
4-2 Example of the County Land Use Data File for PCBEIS2.2
for St. Lawrence County, New York 4-11
4-3 PCBEIS2.2 Output File Listing 4-22
vi EIIP Volume V
-------
TABLES
Page
3-1 Emission Rates and Chemical Speciation Employed by BEIS-2 for
Canopy and Noncanopy Land Use Types 3-7
3-2 Carbon Bond IV Mechanism Speciation For BEIS-2 Biogenic Species 3-14
4-1 Example of Day Selection for PCBEIS2.2 4-9
4-2 Description of the FIPS Code File FIPSCOD.DAT 4-12
4-3 Description of the Land Use Data File cc_LU.DAT 4-12
4-4 Description of the Emission Factor Lookup Table BEIS2.TAB 4-13
4-5 Optional Meteorogical Data File 4-14
4-6 Optional Site Information Data File 4-14
4-7 Example Categorization of Underwater Gas Seepage 4-25
5-1 BEIS GRBIO Variables 5-10
5-2 BEIS UCBIO Variables 5-11
5-3 User-Input Control Variables 5-12
5-4 Emission Factors for Oil and Gas Seeps 5-14
5-5 VOC Species-Oil Seeps, Volatile Fraction 5-15
5-6 VOC Species Profile-Gas Seep 5-16
EIIP Volume V vil
-------
ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
BEIS
BEIS-2
CBM-IV
EKMA
EPA
EPS
FIPS
NMHC
PC
PC-BEIS
SIP
QA/QC
UAM
UAM BEIS-2
(updated version)
voc
—Biogenic Emissions Inventory System
—Biogenic Emissions Inventory System (updated version)
—Carbon Bond IV Mechanism
—Empirical Kinetic Modeling Approach
—U. S. Environmental Protection Agency
—Emission Preprocessor System
—Federal Information Placement System
—nonmethane hydrocarbons
—personal computer
—Personal Computer version of BEIS
—State Implementation Plan
—quality assurance/quality control
—Urban Airshed Model
—Urban Airshed Model's Biogenic Emissions Inventory System
—volatile organic compounds
SYMBOLS
CH4
CO
CO2
N2O
NH3
NO
NO2
NOX
PFC
—methane
—carbon monoxide
—carbon dioxide
—nitrous oxide
—ammonia
—nitric oxide
—nitrogen dioxide
—nitrogen oxides
—perchl orofluorocarb on
Vlll
EIIP Volume V
-------
1
INTRODUCTION
1.1 BACKGROUND
Natural source emissions can make a significant contribution to total volatile organic
compound (VOC) and oxides of nitrogen (NOX) emissions. Estimating emissions of VOC
and NOX from natural sources is an essential part of preparing an inventory of ozone
precursors. This document presents a standard approach to developing biogenic emission
estimates for ozone inventories, and VOC and NOX emission estimates from the natural
sources of lightning and oil and gas seeps.
In the past, the impacts of biogenic VOC were not considered when ozone control strategies
to limit emissions of either NOX or VOC were developed. However, the importance of
biogenic VOC emissions in an ozone inventory became apparent in some regions when the
biogenic VOC emission estimates were compared to the anthropogenic VOC emission
estimates (Chameides et a/., 1988).
Biogenic emission estimates for the United States have been reported at 30,860,000 tons of
VOC per year and 346,000 tons of NOX per year (Novak et a/., 1993). This is in comparison
to estimates of 21,090,000 tons of anthropogenic VOC and 23,550,000 tons of anthropogenic
NOX, estimated for 1990 (EPA, 1994). Isoprene, one of the major constituents of biogenic
emissions, is very photoreactive, making biogenic emissions an even more important source
of VOC. Because of the interaction between NOX and VOC in terms of atmospheric ozone
levels, biogenic emissions should be included in any inventory which will be used to predict
or to monitor atmospheric ozone levels. Inclusion of biogenic emissions is essential for
photochemical air quality modeling.
Other sources of natural VOC and NOX emissions, lightning and oil and gas seeps, are
currently assumed to be less significant as potential sources, but within a particular area may
need to be considered.
1.2 NATURAL SOURCE EMISSIONS vs. BIOGENIC SOURCE
EMISSIONS
Natural source air emissions include VOC, NOX, and greenhouse gases such as methane
(CH4), nitrous oxide (N2O), ozone (O3) and carbon dioxide (CO2). Emission sources for all
of these gases are natural processes occurring in vegetation and soils, in marine ecosystems,
El IP Volume V 1-1
-------
Biogenic Sources 5/21/96
as a result of geological activity in the form of geysers or volcanoes, as a result of
meteorological activity such as lightning, and from fauna, such as ruminants and termites.
Emissions resulting from anthropogenic activities such as those associated with the
agriculture industry are also included in this definition; these activities include fertilizer use
which triggers emissions from microbial activity, and agricultural biomass burning.
Biogenic sources, a subset of natural sources, include only those sources that result from
some sort of biological activity. Biogenic emissions represent a significant portion of the
natural source emissions, and VOC, NOX, and the greenhouse gases discussed above can all
be emitted from biogenic sources.
Vegetation is the predominant biogenic source of VOC and is typically the only source that
is used to estimate biogenic VOC emissions. Microbial activity is responsible for the
emission of NOX and the greenhouse gases of CO2, CH4, and N2O. Soil microbial activity is
responsible for NOX and N2O emissions from agricultural lands and grasslands. CH4 is
emitted through microbial action in waterlogged soils or in other anaerobic
microenvironments. CO2 is released through the aerobic decay of biomass (EPA, 1993; EPA,
1990a).
Natural sources that are not biogenic sources include lightning, a source of nitric oxide (NO)
and oil and gas seeps, which are sources of VOC, CH4 and hazardous air pollutants (HAPs).
The estimated contributions of these sources may be significant when a modeling domain
extends into areas that do not have a high anthropogenic contribution.
1-2 EIIP Volume V
-------
SOURCE CATEGORY DEFINITION
The goals of this document are to define certain natural sources of ozone precursors
(biogenic sources of VOC and NOX, lightning as a source of NOX, and natural oil or gas
seeps as sources of VOC) and to provide preferred and alternative methodologies for
estimating emissions from those sources. The sources of natural ozone precursors that can
be estimated are further limited by the practical consideration of which sources have been
studied sufficiently to provide a mechanism for emission calculations. The widespread and
complex nature of natural sources means that there is considerable uncertainty in the
emission estimates from these sources.
2.1 POLLUTANTS AND SOURCES
2.1.1 VOLATILE ORGANIC COMPOUNDS
The dominant natural source of VOC as nonmethane hydrocarbons (NMHC) is vegetation.
For the United States, natural source VOC is estimated to be higher than anthropogenic
source VOC, and the Southeastern and South Central portions of the United States account
for approximately 43 percent of the national natural NMHC estimate. Forests and agriculture
contribute the largest share of biogenic VOC (Novak and Pierce, 1993).
Less significant sources of VOC are the geogenic processes of oil and gas seeps. Natural
hydrocarbon seeps occur only in certain parts of the United States and would be important to
consider as a source category when they are near by.
Biomass burning (i.e., forest fires, burning of agricultural wastes and other prescribed
burning) is sometimes included as a natural source of VOC emissions. However, many of
the biomass burning processes are anthropogenic activities and, thus, are better grouped with
area sources.
2.1.2 OXIDES OF NITROGEN
There are four source categories for natural NOX emissions: soils, lightning, stratospheric
injection, and oceans. All of these processes produce NO, which is oxidized to NO2 in the
presence of ozone or in a photochemically reactive atmosphere. Emissions from soils are the
only biogenic source of NOX.
El IP Volume V 2-1
-------
Biogenic Sources 5/21/96
Soils emit NOX through biological and abiological pathways, and emission rates can be
categorized by land use. Most of the NOX emitted by soils is in the form of NO.
Agricultural lands and grasslands are the most significant emitters within this category. The
quantity of NOX emissions from agricultural land is dependant on the rate of fertilizer
application and the subsequent microbial nitrogen processing in the soil. Microbial nitrogen
processing occurs naturally in soil, but the rates are greater when soil has been fertilized with
chemical fertilizers. Emissions of NOX from soils are estimated to be as much as 16 percent
of the global budget of NOX in the troposphere, and as much as 8 percent of the NOX in
North America (EPA, 1993).
Lightning converts atmospheric nitrogen and oxygen to NO through extremely high
temperatures. Lightning may contribute as much as 14 percent of the global budget of NOX
in the troposphere, and 4 percent of the NOX in North America (EPA, 1993).
A smaller source of NO in the troposphere is the NO produced through the photodissociation
or oxidation of stratospheric N2O, which then subsides into the troposphere. This source is
estimated to contribute only 2 percent of the global budget of NOX in the troposphere and a
minimal proportion of the NOX in North America (EPA, 1993). Estimating emissions from
this source will not be discussed in this volume.
The photolysis of nitrate dissolved in seawater is responsible for NOX emissions from oceans.
Since NOX has a short lifetime in the atmosphere, this source is not considered to be a
significant source for continental areas such as North America. The global contribution of
NOX to the troposphere by oceans is estimated to be 2 percent (EPA, 1993). Estimating
emissions from this source will not be discussed in this volume.
An indirect source of NOX is the oxidation of ammonia (NH3) which, besides being emitted
by anthropogenic sources, is emitted by the decomposition of nitrogenous matter in natural
ecosystems and from animal wastes (particularly from dense populations of livestock). The
oxidation of NH3 is estimated to be as much as 8 percent of the global budget of NOX in the
troposphere (EPA, 1993). Estimating emissions from this source will not be discussed in this
volume.
2-2 EIIP Volume V
-------
AVAILABLE METHODS
This section will present the preferred and alternative methods for estimating biogenic VOC
and NOX emissions, NOX emissions from lightning, and emissions from natural oil and gas
seeps.
3.1 METHODS FOR ESTIMATING BIOGENIC EMISSIONS
Air quality modeling studies have highlighted the need to include biogenic emissions of
hydrocarbons in the prediction of ground-level ozone. The appropriateness of any estimation
method for biogenic emissions will depend on the ultimate use of the emission estimates.
Biogenic emission estimation models will typically output emission information in a format
specific to an air quality model. The input needs of the air quality models will determine the
choice of the biogenic emission estimation model. Other criteria for selection of a method
will depend on whether the model or system has land use and emission information that is
suitable for the inventory area. These factors will determine the most appropriate method to
use for a particular area.
There are three computer models that can be used to estimate biogenic emissions:
• Biogenic Emission Inventory System-2 (BEIS-2), which will estimate
speciated VOC and NOX emissions from vegetation and soils, respectively.
BEIS-2 is adapted to be used with the following quality models: Regional
Oxidant Model (ROM), the Regional Acid Deposition Model (RADM) and the
Urban Airshed Model (UAM).
• The Personal Computer version of the Biogenic Emission Inventory System-
2.2 (PCBEIS2.2), estimates speciated VOC from vegetation and NOX from
soils. Output from this model can be used in an inventory report, as text, or
in the Empirical Kinetic Modeling Approach (EKMA) model.
• Biogenic Model for Emissions (BIOME), which will estimate speciated VOC
from vegetation. Default output from this model can be used with Geocoded
Emissions Modeling and Projections (GEMAP) or in an inventory document.
An alternative to the above models involves collecting local information to substitute for
defaults in any of the above models. The most commonly used options are more recent or
El IP Volume V 3-1
-------
Biogenic Sources 5/21/96
more detailed land use or meteorological data and updated or additional emission factors and
leaf biomass.
In addition to the computer models, there is a method for biogenic NOX from soils developed
by Yienger and Levy (1995) which can be used. This method was developed in order to
estimate global soil NOX emissions, but no computerized version is available at this time.
Using this approach will require data collection for the necessary meteorology and land use
data, and computing resources for the calculations. If the necessary data and resources are
available, then this method should produce the most accurate results.
The BEIS-2 is the preferred method for air quality models using biogenic estimates, because
it is the most scientifically advanced model for estimating biogenic ozone precursors. It can
be used with several air quality models, and it estimates emissions of soil NOX, which can be
an important source in many rural areas. The PCBEIS2.2 is the preferred method when an
emission estimate is needed for reporting purposes only. The BIOME model, the collection
of local data for use in any of these models, and BEIS, the precursor of BEIS-2, are
alternative methods.
The BEIS-2 is available in several versions that can be used with the ROM, RADM, and
UAM models. The different BEIS-2 models are adaptations of the algorithm and land use
files to the meteorology and grid cell sizes used in their respective air quality models. The
most important issue in the use of a particular version of BEIS-2 is to match the meteorology
and gridding to the air quality model that it will be used with. Aside from these differences,
the scientific background of the BEIS-2 and PCBEIS2.2 models are the same, and will be
discussed below.
BEIS-2 calculates VOC from vegetation and NOX from soils. Emissions from vegetation are
calculated using vegetation types divided into 75 tree genera, 17 agricultural crops, and urban
grasses. BEIS-2 calculates three groups of VOC emissions: isoprene, monoterpenes, and
other NMHCs for the tree genera types, agricultural crops, and urban grasses (Pierce, 1994a).
Further carbon bond speciation can be done through air quality model speciation routines,
such as the UAM Emission Preprocessor System (EPS) (EPA, 1990b). The mainframe
versions of the BEIS-2, those used for air quality modeling, provide a file of gridded hourly
data, which can be converted to input for the particular air quality model for which it was
prepared.
Soil emissions of NOX are dependant on the crop type and fertilization rate (EPA, 1993), and
on a multitude of other factors. BEIS-2 calculates emissions of NOX as NO based on crop
type and fertilizer use. Emission factors have also been updated in BEIS-2.
PCBEIS2.2 generates biogenic emissions by county in kilograms per hour for NOX and each
major VOC species: isoprene, monoterpenes, and other NMHCs. Land use files and
3-2 El IP Volume V
-------
5/21/96 Biogenic Sources
emission factors for biogenic VOC and NOX emissions are the same as those used in BEIS-2.
Output files are at a level of detail suitable for use in State Inventory Plans (SIPs) and input
into the air quality model EKMA. In addition to these primary uses, PCBEIS2.2 can be used
for any application where biogenic VOC and NOX emission estimates at this level of detail
are needed.
The BIOME was developed for the Lake Michigan Ozone Study (LMOS) as the biogenic
component of a comprehensive emission modeling system used to estimate emissions for air
quality modeling (Mayenkar, 1993). Other approaches have been taken to estimate emissions
for Southern California (CARS, 1983 and Nowak, 1991), and the Houston/Galveston and
Beaumont/Port Arthur/Orange areas of Texas (TNRCC, 1993). The approaches taken for the
BIOME and the California and Texas estimates will be discussed in Chapter 5, "Using
Alternative Methods for Estimating Biogenic Emissions."
Methods to determine biogenic emissions and the scientific background that they are based
on may change in the future. For example, algorithms for calculating isoprene emission rates
may be based on leaf area. This will be made possible if remote sensing techniques become
available that permit a better characterization of the vegetation in a study area.
3.1.1 OVERVIEW OF THE BIOGENIC EMISSION INVENTORY SYSTEM
The following description of BEIS-2 and PCBEIS2.2 are drawn from the user's guides and
U. S. Environmental Protection Agency (EPA) guidance for these models (EPA, 1991a; EPA,
1990b; Birth, 1995), and communications with Chris Geron, Air and Energy Engineering
Research Laboratory (AEERL)TEPA (Geron, 1994) and Tom Pierce, Atmospheric Research
and Exposure Assessment Laboratory (AREAL)/EPA (Pierce, 1994).
In a collaborative effort, researchers at Washington State University, The National Center for
Atmospheric Research, and the EPA have developed a computerized system to estimate
hourly gridded biogenic emissions. Until the development of the BEIS, no tool had been
readily available for making these estimates of biogenic emissions. The original version of
BEIS, released in 1991, has been updated to BEIS-2. At this time, BEIS-2 is not being used
for regulatory purposes. However, BEIS-2 will be presented here as the preferred method,
and differences between BEIS-2 and BEIS will be noted.
This section describes the basic algorithm used in both the BEIS-2 and PCBEIS2.2. These
models are based on the same algorithm, but differ in the level of detail of the input used to
run them, the platform that they run on, and the use of their results. BIOME and BEIS also
use the same equation (EQ 3-1), but with different land use, emission factors and
environmental corrections. BIOME was developed as a Geographical Information System
El IP Volume V 3-3
-------
Biogenic Sources 5/21/96
(GlS)-based system. In general, all of the versions of BEIS-2 estimate biogenic NMHC
emissions based on various biomass, emission, and environmental factors.
The basic emission rate equation for forested areas can be expressed as:
n
ER; = Z [Aj • FFj • EF;j • F(S,T)] (3-1)
where:
ERj = Emission rate of each chemical species (i), (jig) (hour"1)
Aj = Area of vegetation for each vegetation type (j), meters2
FFj = Foliar density factor for each vegetation type (j),
(g leaf biomass) (meters'2)
EFy = Emission factor for each chemical species (i) and vegetation type (j),
(jig) (g leaf biomass"1) (hour"1)
F(S,T)= Environmental factor accounting for solar radiation (S) and leaf
temperature (T), unitless
For nonforested areas, the basic emission rate equation for a county can be expressed by the
following:
n
ER; = Z { Aj • EF;j • F;j(S,T) } (3-2)
where:
= Emission rate of each chemical species (i), (jig) (hour"1)
Aj = Area of each land use type (j), meters2
EF;j = Emission flux factor for each chemical species (i),
and each land use type (j), (jig) (meters'2) (hour"1)
Fjj(S,T)= Environmental factor account for solar radiation (S) and leaf
temperature (T), unitless
The BEIS-2 estimates NO emissions for forests, agricultural crops, urban trees, and
grasslands. Soil moisture and soil temperature are known to influence NO emissions from
soil. Nitrate loading can be used as an indicator of the nitrogen processing that is occurring
in the soil. BEIS-2 calculates a range of emission flux rates based on information on the
fertilizer use for certain crops.
3-4 El IP Volume V
-------
5/21/96 Biogenic Sources
The basis of the BEIS-2 calculation for soil NO emissions originates with the following
equation (Williams, et. al, 1992):
FNO = A • exp (0.071 T.) (3-3)
where:
FNO = NO flux, (ng nitrogen) (meters'2) (second"1)
Ts = Soil temperature, degrees Celsius
A = Experimentally derived constant for the land use types of grasslands
and pasture, forests and urban trees, and the individual agricultural
categories
The reader should note that in this equation, the rate is expressed in seconds rather than in
hours. In the BEIS-2 computer code, the calculations for soil NO and forest VOC have been
rewritten to take the form of Equation 3-2.
Each of the variables in Equations 3-1, 3-2 and 3-3 are discussed below, followed by a brief
description of the processing methodology employed by the BEIS-2.
The UAM BEIS-2 produces one output file, a binary UAM-format low-level emissions file.
This file contains hourly gridded biogenic emission rates (which have been corrected for
episode-specific environmental conditions) for monoterpenes, isoprene, NMHCs, and NO.
This file may be used directly as input to UAM or merged with the UAM low-level
anthropogenic emissions file using the UAM EPS Version 2.0. Other versions of BEIS-2,
such as those for the RADM and ROM models, produce output files appropriate for their
respective air quality modeling system.
The UAM BEIS produces a similar output file, which can also be used in UAM. PCBEIS2.2
produces an ASCII output file. In this file, hourly, county-level emissions for isoprene,
monoterpene, and other NMHCs are presented.
Emission Factors
The BEIS-2 estimates VOC and NOX emissions from canopy vegetation types (forests and
trees) and noncanopy vegetation types (primarily agricultural crops). Forests and agricultural
crops are the primary emitters of biogenic VOC. Emissions from canopy vegetation types
are based largely on Guenther's study of biogenic emission rates (Guenther et a/., 1994 and
Geron et a/., 1994). The emissions for canopy vegetation are broken down into factors for
75 major tree genera, and are listed in Table 3-1. Emissions for nonforest vegetation types
are estimated from areal coverage by land use type using the emission rates given in Table
3-1. All Table 3-1 emission rates are expressed in (ngXm"2)^"1), and have been standardized
El IP Volume V 3-5
-------
Biogenic Sources 5/21/96
to 30°C and bright sunlight. The VOC and NO emissions from corn fields as they were
originally used in BEIS were re-evaluated by Pierce and Van Meter, and were subsequently
changed in BEIS-2 (Pierce et al., 1992).
VOC emissions are calculated for three groups: isoprene, monoterpenes, and other NMHCs.
The Carbon Bond IV (CBM-IV) speciations assumed for these three species are shown in
Table 3-2.
The BEIS-2 seasonally adjusts biomass based on the frost dates for each county using a
simple step function (Geron, 1994). For each month, deciduous vegetation within a county is
assumed to have either full biomass or no biomass. Since most high-ozone episodes occur
during the summer months, this is not usually a critical assumption.
Land Use Definitions
Emission factors for forests are multiplied by a foliar density factor recorded in the land use
input file. The foliar density of forests is calculated in the BEIS-2 by applying allometric
equations, which relate diameter-at-breast-height measurements to foliar mass for groups of
tree species. The emission factors in Table 3-1 have already accounted for foliar density in
fully-stocked stands. The emission factors for agricultural crops are based on the crop
acreage and do not require a calculation of leaf biomass. Agricultural crop types modeled by
the BEIS-2 have been presented with their emission rates in Table 3-1.
The default foliar density database provided with the BEIS-2 is derived from land use data
compiled from U. S. Forest Service Forest Inventory and Analysis data, U. S. Census 1990
Urbanized Boundary data, U. S. Census of Agriculture data, and EROS Data Center (EDC) 1
km landcover satellite land classification data (Loveland et al., 1991). The land use database
contains gridded coverage for forests, agricultural crops, and other areas such as grasslands
and water (Geron et al., 1994; Pierce, 1994a; Geron, 1994).
EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 3-1
EMISSION RATES AND CHEMICAL SPECIATION EMPLOYED BY BEIS-2 FOR CANOPY
AND NONCANOPY LAND USE TYPES
Vegetation
ID
Abie
Acac
Acer
Aesc
Aila
Aleu
Alfa
Ainu
Amel
Asim
Avic
Barl
Barr
Betu
Borf
Bume
Carp
Gary
Isoprene Flux
OigXm'Xh1)
170
79
43
43
43
43
19
43
43
43
43
8
0
43
910
43
43
43
Monoterpenes
Flux
OigXm'Xh1)
5100
2380
680
43
43
43
8
43
43
43
43
19
0
85
713
43
680
680
Other VOC
Flux
OigXm'Xh1)
2775
1295
694
694
694
694
11
694
694
694
694
11
0
694
755
694
694
694
NO Flux
ftigXm'Xh1)
4.5
4.5
4.5
4.5
4.5
4.5
13
4.5
4.5
4.5
4.5
257
0
4.5
4.5
4.5
4.5
4.5
Vegetation ID
Description
Abies (fir)
Acacia
Acer (maple)
Aesculus
(buckeye)
Ailanthus
Aleurites
(rung oil)
Alfalfa
Alnus
(European alder)
Amelanchier,
serviceberry
Asimina (pawpaw)
Avicennia
(blk mangrove)
Barley
Barren
Betula (birch)
Boreal forest
(G94)
Bumelia
(gum bumelia)
Carpinus
(hornbean)
Carya (hickory)
EIIP Volume V
5-7
-------
Biogenic Sources
5/21/96
TABLE 3-1
(CONTINUED)
Vegetation
ID
Casp
Cast
Casu
Cata
Cedr
Celt
Cere
Cham
Citr
Cnif
Conf
Corn
Cora
Coti
Cott
Crat
Cswt
Isoprene Flux
OigXm'Xh1)
43
43
29750
43
79
43
43
170
43
745
1550
0.5
43
43
8
43
1050
Monoterpenes
Flux
OigXm'Xh1)
43
43
43
43
1269
85
43
340
680
1367
1564
0
680
43
19
43
660
Other VOC
Flux
OigXm'Xh1)
694
694
694
694
1295
694
694
2775
694
994
1036
0
694
694
11
694
770
NO Flux
ftigXm'Xh1)
4.5
4.5
4.5
4.5
4.5
4.5
4.5
4.5
4.5
4.5
4.5
578
4.5
4.5
257
4.5
0.2
Vegetation ID
Description
Castanopsis
(chinkapin)
Castanea
(chestnut)
Casuarina
(Austl pine)
Catalpa
Cedras
(Deodar cedar)
Celtis (hackberry)
Cercis (redbud)
Chamaecyparis (p-
o cedar)
Citrus (orange)
BEIS conifer
forest
Conifer forest
(G94)
Corn
Cornus (dogwood)
Cotinus
(smoke tree)
Cotton
Crataegus
(hawthorn)
Herbaceous Wetl
(G94)
3-8
EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 3-1
(CONTINUED)
Vegetation
ID
Desh
Dios
Euca
Fagu
Frax
Gled
Gord
Gras
Gymn
Hale
Harf
Hay
Ilex
Jugl
Juni
Lagu
Lari
Isoprene Flux
OigXm-'Xh-1)
65
43
29750
43
43
43
43
56
43
43
8730
38
43
43
79
43
43
Monoterpenes
Flux
OigXm'Xh1)
95
43
1275
255
43
43
43
141
43
43
436
95
85
1275
476
43
43
Other VOC
Flux
OigXm'Xh1)
57
694
694
694
694
694
694
84
694
694
882
57
694
694
1295
694
694
NO Flux
OigXm-'Xh-1)
58
4.5
4.5
4.5
4.5
4.5
4.5
58
4.5
4.5
4.5
13
4.5
4.5
4.5
4.5
4.5
Vegetation ID
Description
Desert shrub
(G94)
Diospyros
(persimmon)
Eucalyptus
Fagus
(american beech)
Fraxinus (ash)
Gleditsia
(honeylocust)
Gordonia (loblolly-
bay)
Grass
Gymnocaldus, KY
coftree
Halesia (silverbell)
Hardwood forest
(G94)
Hay
Ilex (holly)
Juglans
(black walnut)
Juniperus
(E red cedar)
Laguncularia, w
mangrove
Larix (larch)
EIIP Volume V
5-9
-------
Biogenic Sources
5/21/96
TABLE 3-1
(CONTINUED)
Vegetation
ID
Liqu
Liri
Macl
Magn
Malu
Meli
Mixf
Mora
Mscp
Nmxf
Nyss
Oak
Oats
Odcd
Ofor
Oksv
Ostr
Othe
Oxyd
Isoprene Flux
OigXm'Xh1)
29750
43
43
43
43
43
11450
43
8
10150
5950
3108
8
2112
56
7350
43
56
43
Monoterpenes
Flux
OigXm'Xh1)
1275
85
43
1275
43
43
1134
85
19
1100
255
256
19
369
141
100
43
141
255
Other VOC
Flux
OigXm'Xh1)
694
694
694
694
694
694
1140
694
11
850
694
894
11
872
84
600
694
84
694
NO Flux
ftigXm'Xh1)
4.5
4.5
4.5
4.5
4.5
4.5
4.5
4.5
13
4.5
4.5
4.5
257
4.5
4.5
4.5
4.5
58
4.5
Vegetation ID
Description
Liquidambar
(sweetgum)
Liriodendron
(y poplar)
Madura
(osage-orange)
Magnolia
Malus (apple)
Melia (chinaberry)
Mixed forest
(G94)
Moras (mulberry)
Misc crops
N Mixed Forest
(G94)
Nyssa (blackgum)
BEIS oak forest
Oats
BEIS deci forest
Open forest
Oak Savannah
(G94)
Ostrya
(hophornbeam)
Other
(assume grass)
Oxydendram
(sourwood)
3-10
EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 3-1
(CONTINUED)
Vegetation
ID
Pacp
Past
Paul
Pean
Pers
Pice
Pinu
Plan
Plat
Popu
Pota
Pros
Prun
Pseu
Quer
Rang
Rhiz
Rice
Robi
Rye
Isoprene Flux
OigXm-'Xh-1)
55
56
43
102
43
23800
79
43
14875
29750
10
43
43
170
29750
38
43
102
5950
8
Monoterpenes
Flux
OigXm'Xh1)
80
141
43
255
255
5100
2380
43
43
43
24
43
43
2720
85
95
43
255
85
19
Other VOC
Flux
OigXm'Xh1)
48
84
694
153
694
2775
1295
694
694
694
14
694
694
2775
694
57
694
153
694
11
NO Flux
OigXm-'Xh-1)
35
58
4.5
13
4.5
4.5
4.5
4.5
4.5
4.5
193
4.5
4.5
4.5
4.5
58
4.5
0.2
4.5
13
Vegetation ID
Description
Pasture cropland
(G94)
Pasture
Paulownia
Peanuts
Persea (redbay)
Picea (spruce)
Pinus (pine)
Planera
(water elm)
Platanus
(sycamore)
Populus (aspen)
Potato
Prosopis
(mesquite)
Prunus (cherry)
Pseudotsuga (doug
fir)
Quercus (oak)
Range
Rhizophora
(r mangrove)
Rice
Robinia
(black locust)
Rye
EIIP Volume V
3-11
-------
Biogenic Sources
5/21/96
TABLE 3-1
(CONTINUED)
Vegetation
ID
Sabl
Sail
Sapi
Sass
Sera
Scwd
Sere
Shrf
Smxf
Snow
Sorb
Sorg
Soyb
Spin
Swie
Taxo
Thuj
Till
Toba
Isoprene Flux
OigXm'Xh1)
5950
14875
43
43
38
2700
14875
10750
17000
0
43
8
22
1460
43
43
170
43
0
Monoterpenes
Flux
OigXm'Xh1)
43
43
43
43
95
349
43
530
1500
0
43
20
0
1983
43
1275
1020
43
59
Other VOC
Flux
OigXm'Xh1)
694
694
694
694
57
651
694
910
1250
0
694
12
0
1252
694
694
2775
694
235
NO Flux
ftigXm'Xh1)
4.5
4.5
4.5
4.5
58
31
4.5
4.5
4.5
0
4.5
578
13
4.5
4.5
4.5
4.5
4.5
257
Vegetation ID
Description
Sabal (cabbage
palmetto)
Salix (willow)
Sapium
(Chinese tallow)
Sassafras
Scrub
Scrub woodland
(G94)
Serenoa
(saw palmetto)
SE/W Decid
Forest (G94)
SE Mixed Forest
(G94)
Snow
Sorbus
(mountain ash)
Sorghum
Soybean
S pine (G94)
Swietenia
(mahogany)
Taxodium
(cypress)
Thuja
(W. red cedar)
Tilia (basswood)
Tobacco
3-12
EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 3-1
(CONTINUED)
Vegetation
ID
Tsug
Tund
Ufor
Ugra
Ulmu
Uoth
Urba
Utre
Vacc
Wash
Wate
Wcnf
Wdcp
Wetf
Whea
Wmxf
Wwdl
Isoprene Flux
OigXm'Xh1)
79
2412
1989
56
43
11
408.6
5140
43
5950
0
4270
2550
3820
15
5720
525
Monoterpenes
Flux
OigXm'Xh1)
159
121
664
141
43
28
161.9
1000
43
43
0
1120
663
923
6
620
250
Other VOC
Flux
OigXm'Xh1)
1295
151
920
84
694
17
200.5
959
694
694
0
1320
2053
1232
9
530
360
NO Flux
ftigXm'Xh1)
4.5
0.2
4.5
58
4.5
12
12.5
5
4.5
4.5
0
4.5
9
0.2
193
4.5
4.5
Vegetation ID
Description
Tsuga (Eastern
hemlock)
Tundra
BEIS urban forest
BEIS urban grass
Ulmus
(American elm)
BEIS2 other urban
(20% grass)
BEIS urban,
.2 grs/.2 for
Urban trees (50%
hardwood, 50%
coniferous)
Vaccinium
(blueberry)
Washingtonia
(fan palm)
Water
W Conifer Forest
(G94)
Woodland/
cropland (G94)
Wetland forest
(G94)
Wheat
W Mixed Forest
(G94)
Western
Woodlands (G94)
EIIP Volume V
3-13
-------
Biogenic Sources
5/21/96
TABLE 3-2
CARBON BOND IV MECHANISM SPECIATION FOR BEIS-2 BIOGENIC SPECIES
Chemical Species
Isoprene
Monoterpene
Unidentified
Moles CMB-IV Species per Mole of Chemical Species
Ole
-
0.5
0.5
PAR
-
6
8.5
ALD2
-
1.5
-
ISOP
1
-
-
Non-
Reactive
-
-
0.5
Environmental Correction Factors
Studies indicate that biogenic emissions from most plant species are strongly temperature-
dependent; isoprene emissions also vary with solar intensity. The emission factors used by
BEIS-2 are standardized for full sunlight and 30° C. The BEIS-2 adjusts these emission
factors to account for the effects of variations in ambient conditions.
BEIS-2 also simulates the vertical variation of leaf temperature and sunlight within the forest
canopy. The canopy model employed by BEIS-2 assumes that sunlight decreases
exponentially through the hypothetical forest canopy; the rate of attenuation depends on the
assumed leaf area index. Moisture stress is thought to have an important effect on
emissions, but has not been quantified. The effect is not included in the BEIS-2 model.
3.1.2 ALTERNATIVE METHODS FOR ESTIMATING BIOGENIC EMISSIONS
Of the five alternative methods available for estimating biogenic emissions, three are
computer models: PCBEIS2.2, BIOME and BEIS. The Yienger and Levy (1995) method
has not been made into a computer model. PCBEIS2.2 has been discussed with BEIS-2,
since they are based on the same scientific principals. BEIS uses the same algorithm as
BEIS-2, but contains older land use data, less detailed emission factors for forests, and an
older environmental correction factor; the BEIS-2 represents upgrades from BEIS for all of
those factors.
BIOME is a GIS-based system that calculates gridded, hourly, chemical species-specific
biogenic emission estimates for VOC. The emission rate equation is the same as the one
used in the BEIS-2 and BEIS, but it uses different factors. Although the system is built to
allow the users flexibility in using biomass and emission factors of their own choosing for
5-14
EIIP Volume V
-------
5/21/96 Biogenic Sources
plant types, and their own environmental adjustment factors, the model also provides
defaults. BIOME allows users to enter their own land use and land cover data as GIS
coverages. Enhancements include QA capabilities for the analysis of input and output.
BIOME is designed with a scheme for gridding biomass distribution that provides a higher
level of detail than the BEIS model. The choice of using BIOME as opposed to the BEIS-2
models will depend on the final uses of the results, the hardware and software available for
conducting the work, and access to and availability of the information needed to run the
models.
The fourth alternative method is only for computing NOX for soils, and is the approach used
by Yienger and Levy (1995) for global soil NOX emissions. This method is the most
advanced method available but its use as a regional inventory method may be limited by the
data and computational needs required by the model.
The final alternative method uses any of the previously mentioned computer models, but
substitutes locally collected and/or more detailed information for factors like emissions or
land use. In some cases, this may be the approach that will give the most reliable emission
estimates. Strategies and resources that may be useful will be discussed in Section 5, "Using
Alternative Methods for Estimating Biogenic Emissions."
3.2 METHODS FOR ESTIMATING EMISSIONS FROM LIGHTNING
Lightning forms NO through a high-temperature reaction from the energy released during a
lightning flash. Lightning can release about 105 Joules per meter (J/m), and produce
temperatures of about 30,000°K. NO is in thermodynamic equilibrium with N2 and O2 at
temperatures above 2300°K, and as the heated air rapidly cools below 2000°K, NO "freezes
out" as a stable species (Levine et a/., 1984).
Experimental work shows that NO production in air surrounding a lightning flash is a linear
function of the energy released in the flash (Branvold and Martinez, 1988). Thus,
simplifying the process to assume a single value for energy released and estimating the
frequency of lightning flashes allows for an estimate of NO formed by lightning. Global
estimates of NO formed by lightning have ranged from 2 to 90 Teragrams (Tg) of nitrogen
(N)/yr (Levine et a/., 1984)(1 Tg=1012g). The basis for these estimates is a combination of
atmospheric measurements, laboratory experiments, and theoretical calculations. When
constraints based on the global NOX budget are applied to these estimates, a narrower range
of 2 to 20 Tg N/yr for the global estimate is reached (Logan, 1983). Lightning is estimated
to be responsible for about 5 percent of the total NOX produced in the United States, and may
contribute to oxidant chemistry in NOx-limited rural areas (Novak and Pierce, 1993).
El IP Volume V 3-15
-------
Biogenic Sources 5/21/96
There are two available methods for estimating NOX emissions from lightning. The preferred
method uses a method discussed in Pierce and Novak (1991) and used in the ROM; the
alternative method uses emission factors and lightning strike data from other sources.
3.2.1 OVERVIEW OF THE PREFERRED METHOD
The preferred method estimates production of NO by assuming the frequency and type of
lightning strikes, and the amount of energy released. The method derives emission estimates
for cloud to ground (CG) flashes and intra-cloud (1C) flashes. The method relies on four
assumptions:
• Global production of NO by lightning is 6 Tg N/yr;
• Global flash rate is 100 flashes/sec;
• 1C flashes occur approximately 4 times more frequently than CG flashes, a
number which varies with latitude; and
• CG flashes are approximately 10 times more energetic than 1C flashes.
Emission factors have been developed using an equation from Borucki and Chameides (1984)
for CG flashes and the previous assumptions:
CG flashes: 2.9 * 1026 molecules NO per flash; and
1C flashes: 2.9 * 1025 molecules NO per flash.
In order to calculate emissions estimates for this source, the emission factors are applied to
activity for the inventory area, taking into account any corrections to the activity
measurements. This correction factor compensates for lightning flash detection network
efficiency, including a lack of detection of 1C flashes by the network. Other correction
factors correct for changes in air density as a function of height, and correct for seasonal
variations in activity. Values for these correction factors and a detailed methodology are
presented in Chapter 4, "Preferred Methods for Estimating Natural Source Emissions."
3.2.2 OVERVIEW OF THE ALTERNATIVE METHOD
The alternative method, the use of alternative emission factors and lightning strike data,
would be used only if more recent or better-constrained information becomes available.
5-16 El IP Volume V
-------
5/21/96 Biogenic Sources
3.2.3 METHOD UNCERTAINTY
Although there have been several studies of emissions from lightning, there are practical
limitations to the development of reliable estimates from this source. Any direct
measurement of this source is dependant on remote sensing techniques applied to the
unpredictable occurrence of lightning. Laboratory experiments, although they are well
controlled, do not simulate the large discharge associated with an actual lightning stroke. As
a result, all of the parameters used to estimate emissions from lightning carry a certain
amount of uncertainty. The preferred method uses an NO production factor that has an
uncertainty of as much as 10. The vertical distribution of emissions in the atmosphere is not
well defined. Activity in this method uses observations that are corrected, but still represent
an estimate. The estimate for cloud-to-ground lightning strike data is plus or minus
20 percent. (Personal communication, T. Pierce, 1994)
3.3 METHODS FOR ESTIMATING EMISSIONS FROM OIL AND GAS
SEEPS
Oil and gas seeps occur in many locations in the U.S., and emission rates and constituents
vary widely. These seeps have not been extensively studied as sources of VOC. Estimates
for this source depend upon a wide range of variables, as do other source categories in this
volume. A viable method must simplify this variability in order to be practical. The
variables that have primary importance are the amount of material or gas seeping or escaping
from the ground into the air (flow rate), and the volatile fraction of the material. Variables
that could cause similar seeps to have different emission rates include surface area of the
seep, local meteorology such as temperature and wind speed, and variability in the rate of
seepage. Seeps may become more active or new seeps may appear after seismic activity
(earthquakes) or during warmer weather.
The preferred method for this category is to develop a local emission factor based on study
of the oil or gas seeps in the inventory area. Emission calculations for this method take into
account the volatile fraction of the seep, the volume and exposed area of the seep, and the
local meteorology. The alternative method uses an approach tentatively offered by the
California Air Resources Board (CARB), suitable for the estimation of emissions over a large
area, such as a nonattainment area.
3.3.1 PREFERRED METHODS
The preferred method for this source category is to develop a local emission factor based on
a study of the conditions in the inventory area. The level of detail in data collection will
depend on the needs of the inventory and the estimated magnitude of emissions from this
source category. Individual measurements of flow rate, volatile constituents, vapor pressure
El IP Volume V 3-17
-------
Biogenic Sources 5/21/96
of constituents, and surface area, as well as local meteorological observations would be
necessary to determine the most reliable estimate.
Locating seeps in an area, estimating their flow rates and other physical qualities are the
most labor intensive steps. Areas without any past or current oil or gas extraction activity
probably do not need to consider emissions from seeps in any air pollution inventories.
Areas with an oil or gas extraction history should have usable information available in the
form of exploration records and state or local surveys of seep occurrences. More detailed
information about individual sites may be available from local universities or oil or gas
exploration companies.
3.3.2 ALTERNATIVE METHODS
The alternative method for estimating emissions from this source is based on a number of
assumptions:
• Oil seeps and gas seeps can be characterized by one emission factor for either
oil or gas;
• Seeps have a steady flow rate year-round, with no seasonal variation; and
• All oil seeps are composed of the same material, and all gas seeps emit the
same type of gas.
These simplifying assumptions are necessary in order to provide a practical estimation
method for sources that are typically left unstudied, and that do not have a significant body
of supporting literature.
The method is summarized thus:
• Gather information about the location of oil or gas seeps;
• Estimate the number of barrels (42 gal) of oil or million cubic feet of gas
emitted by the seep in 1 year; and
• Multiply the annual flow rate of the seep in either barrels or million cubic feet
by the oil or gas emission factor.
3.4 UNCERTAINTY OF BIOGENIC EMISSIONS ESTIMATES
The major sources of uncertainty associated with biogenic emissions estimates have been
discussed in detail for a national inventory of biogenic emissions prepared with BEIS by
5-18 El IP Volume V
-------
5/21/96 Biogenic Sources
Lamb et al. (1993); other authors have also discussed some aspects of uncertainty (Gaudioso
et a/., 1993; Geron et al., 1994). The sources of uncertainty discussed by these authors are:
• The composition of hydrocarbons classified as "other" in the original data set
used to develop some emission factors;
• The possible disturbance of vegetation by the sampling process, resulting in
erroneously high monoterpene emissions;
• The use of canisters for storage may cause artifacts in the data resulting in an
under-estimate of emissions;
• The results of use of geometric mean rather than arithmetic mean;
• In the original BEIS, specification of emission rate factors for hydrocarbons
other than isoprene and the monotepenes and for agricultural crops is very
uncertain; and
• The canopy model is a highly simplified representation of a complex system,
and feedback mechanisms are neglected.
A thorough assessment of biogenic emissions uncertainties requires an evaluation of
emissions model (or factors) and the biomass data set. With respect to the latter, Gaudioso
et al. (1993) noted that estimates on coverage of forests can vary by a factor of 2 between
inventories for some countries. They cited another study where biomass estimates varied by
20 to 40 percent.
In the U. S., land use and biomass density for Idaho obtained from satellite imagery and the
Geoecology Data Base have been compared (Cheung et al., 1991). For the dominant land
use types — coniferous forests, scrublands and agricultural lands — the differences in
coverage between data sets were less than 20 percent. However, very large variations were
found for the more rare land use types. Despite these differences in coverage estimates,
emission estimates based on the two datasets were within 10 percent of each other. In the
same study, an independent estimate of biomass density for coniferous forests was obtained.
In comparison to the standard inventory biomass density factors, the alternative produced
emission estimates 14 percent lower.
Much of the uncertainty described by Lamb et al. (1993) is the result of the emissions data
used for BEIS. The addition of more new data in BEIS-2 has reduced or at least allowed
some quantification of that uncertainty. The emissions predicted by BEIS-2 have been
compared to measured values for some sites, and generally performed well (Geron et a/.,
El IP Volume V 3-19
-------
Biogenic Sources 5/21/96
1994). The uncertainty for the genus level emission rates used in BEIS-2 are estimated to be
approximately ±50 percent (Guenther et a/., 1994).
3.4.1 SENSITIVITY ANALYSIS
The emission of isoprene and monoterpenes is sensitive to light and leaf temperature
(Guenther et a/., 1993). The biogenic models use light and temperature levels to adjust the
empirical emission rate factors. Leaf temperatures are strongly dependent on light intensity
as well as other environmental factors.
The sensitivity of BEIS emission estimates to several environmental parameters was analyzed
by Lamb et al. (1993). The effect of the canopy model was found to be very important,
especially in deciduous forests. Without the canopy correction, peak hourly and daily
average emissions from deciduous forests were 50 percent higher than baseline. Similar
results were found for coniferous forests. The canopy both attenuates light reaching the
leaves and reduces mean leaf temperature. It is believed and almost certain that the BEIS
canopy model underestimated leaf temperatures in the lower portion of coniferous canopies.
However, there is a lack of rigorous datasets to test the canopy model. BEIS-2 computes
light attenuation within the canopy, but assumes ambient temperature throughout. As a
result, when sunlight attenuation is allowed to occur, but leaf temperatures are set equal to
above-canopy ambient temperatures, coniferous forest emissions increase by 28 percent.
Deciduous forest emissions increase by only 3 percent.
The canopy model is not directly under the control of BEIS users. However, if BIOME or
other alternative methods are used, the ability of the model to incorporate canopy effects
needs to be carefully evaluated.
In addition to temperature sensitivity, other user-supplied (or user-controlled) input data such
as solar radiation, relative humidity, and wind speed may affect emissions. Isoprene
emissions were relatively sensitive to changes in solar radiation and relative humidity, but
wind speed had only a small effect. The BEIS-2 canopy model does not use relative
humidity or wind speed. Wind speed had no effect on monoterpene emissions, while solar
radiation and relative humidity had small effects relative to isoprene. Figure 3-1 (redrawn
from Lamb et al., 1993) shows the results. In the figure, predicted isoprene is shown as a
solid line and total monoterpene is shown as a dashed line for situations of increasing solar
radiation, relative humidity and wind speed. Figure 3-2 (also from Lamb et al., 1993) shows
predicted leaf temperatures for an oak forest (solid line) and a coniferous forest (dashed line)
for cases with (a) 10% and (b) 50% relative humidity with 1 lymin"1 solar radiation, 4 m-s"
1 wind speed and 30°C ambient temperature (dotted line) at the canopy top. Other studies
have stressed the importance of temperature effects on biogenic VOC emissions (Gaudioso et
a/., 1993, Guenther et al., 1994), particularly isoprene, and the role of moisture stress.
5-20 El IP Volume V
-------
5/21/96
Biogenic Sources
T~ 4000
_c
04
O)
sooo
2000
.1 1000
CO
CO
"E
LII
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Solar Radiation (ly min )
1.4 1.6
^ 4000
'E
O)
_ 3000
x
c 2000
'co
.co
^ 1000
. , , , i i i i i i
0 10 20 30 40 50 60 70 80 90 100
Relative Humidity (%)
^T^ ^-UUU
1
-C
CM
'E sooo
D)
x 2000
3
LL
§ 1000
'co
CO
E n
iu o
-
^^^^
-
I I I I I
r\ n A r* o j rv ^
0.
EE
6
i
O>
r>
Wind Speed (m-s" )
FIGURE 3-1. EFFECT OF SOLAR RADIATION, RELATIVE HUMIDITY AND WIND SPEED
UPON PREDICTED ISOPRENE
EIIP Volume V
3-21
-------
Biogenic Sources
5/21/96
CD
Is
O
O
3
2
1
24 26 28 30 32
Temperature (deg C)
la
£
O
S4
3
2
b)
RH = 50%
i
O
24 26 28 30 32
Temperature (deg C)
FIGURE 3-2. TEMPERATURE RELATIONSHIPS WITHIN FOREST CANOPIES WITH
10% (A) AND 50% (B) RELATIVE HUMIDITY.
3-22
EIIP Volume V
-------
5/21/96 Biogenic Sources
Inventory emissions are also sensitive to land-use and biomass density data. With the
exception of the Idaho database comparison described in the previous section, very little has
been published on this topic. However, it is safe to say that the correct allocation of land
use/plant types can have significant impacts on emissions of specific pollutants. An extreme
example would be incorrectly identifying a forest type as coniferous when it was deciduous;
isoprene emissions would be underestimated and monoterpenes over-estimated.
3.5 QUALITY ASSURANCE/QUALITY CONTROL
As with any inventory development process, QA/QC procedures should be planned and
scheduled in advance. The level of effort spent on QA and on QC will depend on the
importance of biogenic sources in the overall inventory and the degree to which the user
plans to modify the defaults or use of an alternative land use dataset.
Details on specific QA/QC methods can be found in the QA Source Document, Volume VI
of this series.
At a minimum, the QA program should include:
• Peer review of input data and results;
• Quality control checks of data transcription (i.e., written or keyed into
computer); and
• Review (at QC or QA level) of any calculations, analyses (e.g., choice of
ozone days to model), or data conversions (e.g., conversion of units of
measure).
These minimum QA/QC elements may be adequate where biogenic emissions are not a
significant component of the inventory and the default data in BEIS-2 are used. However,
more extensive procedures may be warranted in some cases. Specific aspects of QA/QC
procedures are described for key inventory components in the following sections.
3.5.1 LAND USE/BIOMASS DENSITY
If BEIS-2 is being used with default land use and biomass data sets, the amount of additional
QA/QC of these datasets will depend on the importance of the biogenic emissions to the
overall inventory. At a minimum, the receiver should check the FIPS code to verify that the
correct county data were used. The reviewer may also want to perform a reality check on
the genus/land use proportions; if any of the data look unreasonable, other data sources or
local experts should be consulted for comparison values. A local university, for example,
El IP Volume V 3-23
-------
Biogenic Sources 5/21/96
may be able to supply independent estimates of forest biomass, land use patterns, or other
relevant information. If the user is modifying the land use data or is using an entirely
different dataset, then performing an uncertainty analysis is more critical. At a minimum, the
results using the default and the new data should be compared.
There are no hard and fast rules on how well such comparisons need to agree, or even how
to choose the correct dataset. If the source category is a significant one, and if these are
competing datasets of similar quality, the user is advised to conduct a detailed assessment.
The assessment could include an uncertainty analysis whereby the model is run with each
dataset in turn. If the emissions estimated for each pollutant are not markedly different (i.e.,
within the source order of magnitude), the results could be averaged. In this situation, the
rationale and the approach need to be clearly documented, and the uncertainty of the estimate
should be quantified using a standard statistic such as the standard error or the coefficient of
variation. If one dataset is chosen over others, the reasons for the choice need to be clearly
documented.
3.5.2 METEOROLOGICAL DATA
As the discussion of model sensitivity described, certain meteorological variables can have
significant effects on emission estimates. Therefore, the days chosen for modeling will have
an impact on the results; higher temperatures, in particular, will produce higher emissions.
Verification that the modeling days were chosen correctly should be a primary objective of
the QA program.
Another potential source of error lies in selecting the meteorological dataset. Generally, data
from the closest airport weather station is used, but terrain will influence meteorology
considerably. If the nearest meteorological station is at a much higher or lower elevation, or
if there are major differences in the characteristics of the landscape (e.g., mountains, large
bodies of water) alternative meteorological data may be needed. The discussion on using
PCBEIS2.2 in Section 4 of this volume describes sources of meteorological data.
3.5.3 ALTERNATIVE METHODS
If a model other than BEIS-2 is used, all of the above QA/QC procedures are still relevant.
Furthermore, sensitivity and uncertainty analyses should be conducted to provide a good
understanding of the model. These analyses should cover the key variables (temperature,
light) and critical model components (such as the canopy model).
If the results are to be used for any sort of regulatory purposes, the BEIS-2 results (using the
default datasets) should be provided as a benchmark. Differences in the results from the two
models should be explained and the reasons for using the alternative model results needs to
5-24 El IP Volume V
-------
5/21/96 Biogenic Sources
be justified. Thoroughly documented QA/QC procedures are especially important when
using a new or non-standard approach.
El IP Volume V 3-25
-------
Biogenic Sources 5/21/96
This page is intentionally left blank.
5-26 El IP Volume V
-------
PREFERRED METHOD FOR
ESTIMATING NATURAL SOURCE
EMISSIONS
In this volume, emissions of VOC and NOX from biogenic sources (vegetation and soils),
lightning, and geogenic sources (oil and gas seeps) are discussed. In this section, the
preferred emission estimation methods for these sources and pollutants will be described.
4.1 BIOGENIC SOURCES
4.1.1 USING BEIS
When an inventory requires air quality analysis using a photochemical grid model, biogenic
VOC and NOX need to be included with the modeled emissions. Under the Clean Air Act
Amendments of 1990 (CAAA), the use of photochemical grid models is required for areas
designated as serious, severe, and extreme ozone nonattainment areas, and for multistate
moderate ozone nonattainment areas in the preparation of their SIPs. For these instances, the
UAM should be used for photochemical pollutant modeling applications involving entire
urban areas.
The BEIS-2 is adapted to use with the UAM, RADM, and ROM air quality models, although
RADM and ROM are used for regional modeling and are not as widely used as UAM.
RADMBEIS-2 and ROMBEIS-2 use the same types of parameters as UAMBEIS-2 but they
apply different adaptations of the algorithm and land use files to the meteorology and grid
cell sizes. The primary differences between all of the different versions of BEIS-2, including
PCBEIS2.2, are in the treatment of the cloud cover and in the raw meteorology data used to
calculate the environmental correction factor.
For instance, PCBEIS2.2 uses a single value for sky cover. In the mainframe versions, more
detailed information for cloud cover is used. Cloud cover in these versions is defined by
thickness, level, and percent of load per level. This may provide a more accurate measure of
the reduction of solar radiation through clouds.
El IP Volume V 4-1
-------
Biogenic Sources 5/21/96
There are two preferred methods for estimating biogenic emissions of VOC and NOX, either
BEIS-2 or PCBEIS2.2. The following sections will describe how to use the UAMBEIS-2
and PCBEIS2.2.
4.1.2 BEIS AND UAM
The UAM is a three-dimensional photochemical grid model designed to calculate
concentrations of both inert and chemically reactive pollutants by simulating physical and
chemical processes which occur in the atmosphere. As of this writing, a stand-alone version
of BEIS-2 is planned to provide biogenic emission estimates that can be input to the UAM.
Inputs to BEIS-2 for the UAM are discussed in this section.
Readers should note that as of May 1996, a regulatory version of UAM BEIS-2 was not
available. As an alternative, modelers have been using emissions data extracted from ROM
BEIS-2. For regulatory purposes, readers are advised to look to the U.S. EPA Office of Air
Quality Planning and Standards (OAQPS) Technology Transfer Network (TTN)a for updates
on the status of BEIS2, or call Chet Wayland of the Emissions, Monitoring, and Analysis
Division at (919) 541-4603.
The UAM simulates the hour-by-hour photochemistry occurring for each grid cell in the
modeling domain; as a result, the input emission data from BEIS must contain a comparable
level of resolution. Total VOC and NOX emission estimates must be chemically speciated
into the chemical classes employed by the model. In addition to the temporal and chemical
allocation necessary to prepare emission information for the UAM, the data must be spatially
allocated by grid cell for each hour of the modeling episode. Chemical speciation and
temporal and spatial allocation are discussed in EPA guidance for UAM and procedures for
modeling inventories (EPA, 1991b; EPA, 1990b).
4.1.3 PROCESSING METHODOLOGY
The BEIS-2 calculates the emission rates of monoterpenes, isoprene, and other NMHCs by
multiplying foliar density for each forest type by the appropriate emission factors. The
calculated emission rates are then adjusted for the specific environmental conditions of the
modeling episode using data from the user-supplied meteorology file and the UAM
temperature preprocessor (TPBIN). Corrections are performed for each grid cell of the
modeling domain using the hourly gridded temperature and solar radiation data contained in
these input files.
4-2 EIIP Volume V
-------
5/21/96 Biogenic Sources
For forested areas, land use data for each county are multiplied by an emission flux value for
each tree genus. Emission flux represents the forest species emission factor and species
foliar density. The EPA provides a default land use data file for use with BEIS-2 which
contains county-level land use data for each of the tree genera and for the noncanopy
vegetation types listed in Table 3-1.
The calculated county-level vegetation coverages are gridded using a county allocation file,
which indicates the percentage of each grid cell occupied by a particular county. If more
current land use data are available, the user may include these data by using the optional
gridded land use data file or the optional county-level land use data file.
The next step in processing forested areas involves adjusting the emission rates for
temperature and solar energy changes within the forest canopy. The adjusted emission rates
are then grouped into categories for five CBM-IV species (olefins, paraffins, isoprene,
aldehydes, and nonreactives) and NO. Although UAM BEIS-2 calculates emission rates for
nonreactive hydrocarbons, it does not report this information because it is not used by UAM.
Nonforested areas, or noncanopy areas (including agricultural areas and urban trees), are
handled in a slightly different way than forested areas. The principal difference is that a
canopy model is not used. Noncanopy areas use either a winter or summer emission factor
table to reflect seasonal variations in biomass. Next, the data are gridded using the county
allocation file. Land use is then multiplied directly by the emission factor for that land use
type to produce emission rates. Speciation and environmental adjustments are handled the
same as for forested areas, except that the nonforest emission rates are not adjusted for
canopy effects.
4.1.4 INPUT REQUIREMENTS
BEIS-2 is a stand-alone processor that prepares biogenic emission estimates so they can be
used as input to the UAM or other models such as ROM or RADM. As indicated above, the
BEIS-2 uses three types of data files: UAM preprocessor data, user-supplied data, and data
supplied to the user by EPA. Figure 4-1 shows a flow chart of the BEIS-2 and BEIS
(discussed in Section 5) flow of information. Each of the input data files is briefly described
below. Differences between BEIS-2 and BEIS will be noted. The User's Guide for the
Urban Airshed Model, Volume IV: User's Manual for the Emissions Preprocessor System
2.0, Part A: Core FORTRAN System (EPA, 1990b) contains detailed format descriptions of
the various input files.
The information that must be supplied by the user for the BEIS-2 is information that will be
needed for running the UAM as well; this information includes scenario definition and
control flags, hourly meteorology information, and the Federal Information Placement System
EIIP Volume V 4-3
-------
Biogenic Sources
5/21/96
UAM PREPROCESSOR
DATA
USER-SUPPLIED DATA
METEOROLOGY COUNTY
ALLOCATION
DATA SUPPLIED TO
THE USER BY EPA
RAWMET CNTYALO
BEIS
Binary Low-Level UAM Emissions File
FIGURE 4-1. UAM STAND-ALONE BIOGENICS PROCESSOR. OVERVIEW OF THE
BIOGENIC EMISSION INVENTORY SYSTEM (BEIS)
4-4
EIIP Volume V
-------
5/21/96 Biogenic Sources
(FIPS) code of the counties being modeled. BEIS-2 calculates emission estimates for
speciated VOC from forests and agricultural crops, and NO from soil processes.
It should be noted that when a projected year is being modeled, biogenic emissions need to
be estimated for the projected year in addition to other projected emissions. However, unless
there are anticipated changes in land use or vegetation patterns for the modeling area, it is
appropriate to assume that biogenic emissions will remain the same between base and
projected years.
UAM Preprocessor Data
Two of the UAM preprocessor files are also used by BEIS and one for BEIS-2. The
WDBIN file is used by BEIS, is produced by the UAM winds preprocessor and contains
hourly gridded surface wind speed data. The second file is the TPBIN file, which is
produced by the UAM temperature preprocessor program. This file contains hourly gridded
temperature data and is used by BEIS and BEIS-2. These files should be available from the
photochemical modeler.
User-Supplied Data
The user must supply two types of data: meteorological and county allocation data for both
BEIS and BEIS-2. In addition, BEIS allows the user to specify land use data, either county-
level or for each grid cell of the modeling domain. Each of these files is described briefly
below.
RAWMET. The meteorology file RAWMET contains hourly surface meteorological
information on cloud cover, and cloud height for one station in the user's UAM domain.
The file is created by the user with data that may be obtained from the National Weather
Service.
CNTYALO. The county allocation file contains the percentage of each grid cell that a given
county occupies. To speed processing, this file should be sorted by FIPS code. The
CNTYALO file may be created by the user; however, the EPA has a Gridded
Surrogate/County Allocation Utility which will create the file given the UAM domain
coordinates and grid cell size. To request this data, contact the nearest EPA Regional Office.
GRBIO. If the user wishes to use more up-to-date land use data which are already gridded,
the optional gridded land use file, GRBIO, may be used. For each grid cell, the user must
input values in hectares for all of the land use types used by BEIS.
EIIP Volume V 4-5
-------
Biogenic Sources 5/21/96
UCBIO. If the user has more up-to-date land use data on a county level, the optional county
land use file, UCBIO, may be used. For each FIPS code, the user must input values in
hectares for all of the land use types used by BEIS.
In addition to the user-supplied data files listed above, the user must supply certain input
control data required by BEIS-2, consisting of domain and scenario-specific information
along with control option flags. The user must input start date and hour, end date and hour,
month of scenario, number of columns and rows, number of hours from Greenwich mean
time to local time, number of counties, and a list of counties. Three flags which control
several options in BEIS-2 must also be set by the user. The first flag, called GIFLAG, must
be set to TRUE if the user wishes to input his own gridded land use data using the optional
GRBIO file. The second flag, called CFLAG, must be set to TRUE if the optional UCBIO
file is being used to input the user's own county land use data. The third flag, called
GROWFLAG, must be set if the user inputs gridded land use data. This flag is set to TRUE
to indicate that vegetation is growing and FALSE if only coniferous vegetation is growing,
which might be the case in a cold season like late fall or winter.
EPA-Supplied Data
The EPA provides two data files for use with the BEIS-2. The first of these files contains
county land use in hectares for canopy, noncanopy and urban tree areas. The second file
contains visible solar energy (jiE m'V1) data.
4.2 USING PCBEIS2.2
4.2.1 INTRODUCTION
In addition to emission modeling and projection inventory preparation where a version of the
mainframe-based BEIS-2 would be used, other emission inventories may require biogenic
emission estimates. Where the speciated and spatially allocated output of BEIS-2 is not
necessary, or where a PC-based model is a more practical choice, PCBEIS2.2 is the
recommended option.
The model documentation, executable file, source code, and necessary data files are available
on EPA's CHIEF (Clearinghouse for Inventories and Emission Factors)
The following discussion of PCBEIS2.2 is drawn primarily from the model documentation
(Birth, 1995).
PCBEIS2.2 provides outputs for two specific uses:
4-6 EIIP Volume V
-------
5/21/96 Biogenic Sources
• An estimate of biogenic VOC and NO emissions required for baseline
emission inventories mandated by the CAAA; and
• Hourly estimates of biogenic VOC emissions needed for use with EKMA.
For the first item, PCBEIS2.2 gives an estimate in tons per day. In order to meet EKMA's
needs, biogenic emissions are output on a kilograms per square kilometer per hour basis for
each major species (isoprene, monoterpenes, NOX, and VOC). In addition to these primary
uses, PCBEIS2.2 can be used for any application where biogenic VOC emission estimates at
this level of detail are needed. A discussion of the model's technical background, computer
aspects and a model run example can be found in the model documentation (Birth, 1995).
4.2.2 PROCESSING METHODOLOGY
The following methodology is for using PCBEIS2.2. The PCBEIS2.2 interface has been
designed to resemble that of PC-BEIS, its predecessor, and the steps taken to run PC-BEIS
should be very similar to those described for PCBEIS2.2.
Modeling Day Selection
Modeling day selection will be partly determined by the intended application of the results,
or the reason for the emission estimate. A SIP base year inventory may need a total
biogenic VOC estimate for a single typical ozone day. The day selection procedure below
describes the steps in selecting such a day for use in the PCBEIS2.2. If results from
PCBEIS2.2 are intended for use with an air quality model such as EKMA, a number of days
will need to be simulated with PCBEIS2.2. These will correspond to the days which will be
modeled.
The procedure for selecting the typical ozone day for modeling by PCBEIS2.2 is fully
described in the documentation for the model, and involves selecting the day with the fourth-
highest temperature out of the ten highest ozone days from a 3-year period. This procedure
allows the user to use a day that is certain to have resulted in high ozone levels without
using meteorology that may have been uncharacteristic of the area. Note that the selection of
the typical ozone day for running this model may not be the same as the selection of typical
ozone days for other purposes in a base year inventory (e.g., modeling mobile emissions),
and those days do not need to be the same.
For a baseline inventory the following steps should be followed:
• Select the top ten days with the highest hourly ozone readings over the most
recent three years of monitoring data;
EIIP Volume V 4-7
-------
Biogenic Sources 5/21/96
• Obtain National Weather Service data for daily maximum temperature for each
of the ten days;
• Rank maximum daily temperatures from highest to lowest;
• Select fourth highest based upon maximum daily temperature; and
• Use hourly meteorological data (cloud cover, relative humidity, wind speed,
and temperature) for this day as input to PCBEIS2.2.
The following example illustrates the application of this procedure.
Example 4-1
The top ten ozone exceedances during the 3-year period, 1987-89, for a hypothetical
area are shown in Table 4-1. After the values are ranked by temperature, the fourth
highest is selected. In this case, it is June 30, 1987. Meteorological values for that
day should be input to PCBEIS2.2. In the event that two days have the same
maximum temperature which would be the fourth highest, the one with the lowest
average daily wind speed should be selected.
Data Requirements
The data requirements for PCBEIS2.2 are modest since the detailed land use and emission
factor information is supplied with the model. The user needs to obtain geographic
information on the county of interest and hourly meteorological data. To run PC-BEIS, the
following data are required:
Site Information
• County Federal Information Processing System (FIPS) code
• Latitude, longitude (decimal degrees, tenths)
4-8 EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 4-1
EXAMPLE OF DAY SELECTION FOR PCBEIS2.2
Date
Ozone Level (ppm)
Maximum Temperature
(°F)
Ranked by Date
30/06/87
10/08/87
06/06/88
15/06/88
26/06/88
17/07/88
05/08/88
12/08/88
04/09/88
03/07/89
0.147
0.155
0.138
0.162
0.159
0.144
0.145
0.144
0.152
0.157
90
87
85
93
88
87
88
84
91
92
Ranked by Temperature
15/06/88
03/07/89
04/09/88
30/06/87
10/08/87
26/06/88
05/08/88
17/07/88
06/06/88
12/08/88
0.162
0.157
0.152
0.147
0.155
0.159
0.145
0.144
0.138
0.144
93
92
91
90
87
88
88
87
85
84
EIIP Volume V
4-9
-------
Biogenic Sources 5/21/96
Time zone (5=EST, 6=CST, etc.)
• Month, day, year, hour(s)
Meteorological Data (hourly)
• Ambient air temperature (°C)
• Photosynthetically Active Radiation (PAR)--not required if you are using cloud
cover data
• Opaque sky cover (fraction)--not required if you provide PAR data
Meteorological data can be obtained from several sources. Locally, airport weather stations
record temperature and cloud cover throughout the day. The National Climatic Data Center
in Asheville, North Carolina, also is a resource for meteorological data. As noted above,
either PAR or opaque sky cover is needed. Usually, only opaque sky cover information will
be available. Either of the two can be used.
When the meteorological measurements have been recorded at a frequency of less than
1 hour, modelers will need to approximate the intermediate hours by averaging between the
2 hours that have data.
User data requirements for PCBEIS2.2 are minimal, but the model employs several data files
that are critical to its operation, which are supplied with the model. These files are stored
with the model executable in a file labeled B2DATA.EXE. B2DATA.EXE is a
self-extracting executable that contains land use files (cc_LU.DAT, where cc is the
2-character state postal code), a FIPS code lookup table (FIPSCODE.DAT), and the emission
factor lookup table (BEIS2.TAB). Descriptions of the parameters in these files are given in
Tables 4-2 through 4-4. An example of the data in a land use file for one example county
can be seen in Figure 4-2.
In Figure 4-2, the first line of the cc_LU.DAT file contains the FIPS code for the state and
county, the county latitude and longitude, the two-letter state abbreviation, and the county
name. Below that, the area of each land use or vegetation type is given; the total for a broad
category is given first, followed by specific genera. For example, the example in Figure 4-2
shows 438,701.312 hectares (ha) of forest areas, with 27 different genera represented. The
predominate genera are maples (Acer, 97,233.508 ha) and beech (Fagus, 37,127.871 ha).
Also, note that 8,901.898 ha of the county are water (shown at bottom of column). A reality
check of the genera and land use types should be performed as part of the QA program to
verify that the data look reasonable.
4-10 El IP Volume V
-------
5/21/96 Biogenic Sources
36089
Abie
Acer
Betu
Carp
Gary
Cora
Crat
Fagu
Frax
Gled
Jugl
Juni
Lari
Malu
Ofor
Ostr
Pice
Pinu
Popu
Pran
Quer
Robi
Sali
Thuj
Tili
Tsug
Ulmu
Hay
Corn
Pota
Oats
Mscp
Othe
Wate
44.50 75.07 NY St
438701.312
11377.75
97233.508
28381.52
600.761
1293.441
31.115
4696.688
37127.871
12950.65
57.673
1053.291
255.826
2462.424
910.761
108310.203
10004.05
22084.51
12330.27
17781.27
20458.48
4255.675
172.084
2432.601
15177.44
4551.643
10038.79
12670.88
0
98456.492
57555.641
13776.610
8.499
751.113
26364.631
190270.812
181368.906
8901.898
Lawrence
27
0
5
2
FIGURE 4-2. EXAMPLE OF THE COUNTY LAND USE DATA FILE FOR PCBEIS2.2
FOR ST. LAWRENCE COUNTY, NEW YORK
EIIP Volume V 4-11
-------
Biogenic Sources
5/21/96
TABLE 4-2
DESCRIPTION OF THE FIPS CODE FILE FIPSCOD.DAT
Variable Name
sfips
cfips
county
state
Format
12
13
C30
C2
Description
State FIPS Code
County FIPS Code
County Name
Abbreviated State Name
TABLE 4-3
DESCRIPTION OF THE LAND USE DATA FILE cc LLJ.DAT3
Variable Name
cfips
lat
long
staid
ctyid
forar
fnvg
forgen
crnar
uforar
ufnvg
uforgn
ucrar
agar
agnvg
Format
15
F5.2
F6.2
A2
C24
Floating Point
Integer
Character
Floating Point
Floating Point
Integer
Character
Floating Point
Floating Point
Integer
Descriptionb
State and County FIPS Code
Latitude
Longitude
State Abbreviation
County Name
Forest Area
Number of Forest data records
Forest Genera
Crown Area for the specific Genera
Urban Forest Area
Number of Urban Forest data records
Urban Forest Genera
Crown Area for the specific Urban Forest
Genera
Agriculture Area
Number of Agriculture data records
4-12
EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 4-3
(CONTINUED)
Variable Name
agid
aidar
othar
onvg
othid
oidar
Format
Character
Floating Point
Floating Point
Integer
Character
Floating Point
Description13
Agriculture type identifier
Area of the specific type of agriculture
Area of all other vegetation
Number of records of other vegetation
Other vegetation identifier
Area of specific other type of vegetation
a There is one file for each state; number of records for each county depends on number of specific vegetation
types and consequently varies from county to county.
b All areas are recorded in hectares (ha).
TABLE 4-4
DESCRIPTION OF THE EMISSION FACTOR LOOKUP TABLE BEIS2.TAB
Variable Name
vegid
factrs(l)
factrs(2)
factrs(3)
factrs(4)
lai
genus
Format
C4
F7.1
F6.1
F6.1
F5.1
11
C30
Description
Vegetation Identification Code
Isoprene Emission Factor
Monoterpene Emission Factor
Other VOCs Emission Factor
NO Emission Factor
Leaf Area Index
Genus and Common Namea
Units
Unitless
ug m"2 h"2
ug m"2 h"2
ug m"2 h"2
ug m"2 h"1
m2m"2
unitless
The data from the FIA database are aggregated to the genus level to be consistent with the emission factor
database of Guenther et al. (1994). Other entries are common names for agriculture or generalized terms
such as Boreal Forest for the AVHRR data source.
EIIP Volume V
4-13
-------
Biogenic Sources
5/21/96
Optional data files that can be prepared by the model user are the meteorological file and a
file containing the site information. The model also allows the user to enter this information
directly into the model, but preparing the files before the model run should allow for input
QA/QC and should be less cumbersome. Table 4-5 provides a description of the
meteorological data input file. The file format for the site information file is simply the
input information separated by commas. Site input parameters are listed in Table 4-6. Both
of these input files are named by the user. Entering the meteorological and site information
during the model run is described in the following section, Running the Model.
TABLE 4-5
OPTIONAL METEOROLOGICAL DATA FILE*
Variable Name
fflR
ACLDCV
ATMPSR
UPAR
Format
13
Floating Point
Floating Point
Floating Point
Description
Hour, local standard time (e.g., 12)
Opaque sky cover - optional (fraction)
Ambient temperature (°C)
PAR (uE/m2-sec)-optional (if PAR is not
supplied, it wil be calculated from sky cover)
Up to 24 records (hours) are allowed.
TABLE 4-6
OPTIONAL SITE INFORMATION DATA FILE
FIPS
Site Description
Latitude
Longitude
Day
Month
Year
Time Zone
Begin Hour
End Hour
4-14
EIIP Volume V
-------
5/21/96 Biogenic Sources
Running the Model
PCBEIS2.2 is a simple to use PC-based model that provides the user with default values for
all parameters except the location and meteorological data, which were discussed above.
PCBEIS2.2 is written to conform with the FORTRAN 77 standard and has been compiled on
the PC with Microsoft FORTRAN version 5.0. PCBEIS2.2 has been compiled to allow its
use on IBM-compatible personal computers. The executable, source code and necessary data
files needed to run PCBEIS2.2 will take up approximately 3.16 MB of memory. In order for
the menu interface to function properly, ANSI.SYS must be installed on your PC.
ANSI.SYS is available with MS-DOS. It is recommended that PCBEIS2.2 users have a math
co-processor.
Example 4-2
In this example, biogenic emissions for Wake County, North Carolina, are estimated
for a 24-hour period. There are 7 basic steps for running the model:
(1) Start PCBEIS2.2 program;
(2) Identify the location of any input files;
(3) Identify any input and output files;
(4) Review, edit and save site information;
(5) Select, edit and save meteorology file;
(6) Calculate emissions; and
(7) End program.
Meteorological data were obtained from the National Climatic Data Center for the weather
reporting station at Raleigh-Durham airport. The day chosen for simulation, August 19,
1988, had a maximum temperature of 103°F (313°K) and light to moderate winds.
Before beginning the model simulation, the user should assemble the data required. Site
information includes the county FIPS code and the latitude and longitude of Wake County
(35.8°N, 78.6°W).
Step 1
The user begins the model simulation by typing the command "PCBEIS."
Step 2
EIIP Volume V 4-15
-------
Biogenic Sources
5/21/96
The first question that the program asks is:
Please enter the drive and directory where input files are located:
Enter the appropriate drive and directory.
Step 3
The next query is:
If you have a default input file, enter the name:
Enter the file name of the input file, or press the RETURN key if not. The next question is:
Please enter the output filename for this run:
For this example, enter "TEST.OUT.1
4-16
EIIP Volume V
-------
5/21/96
Biogenic Sources
Step 4
The site Input Screen appears as follows:
PCBEIS version 2.2
1
FIPS
Site description
Latitude
Longitude
Day
Month
Year
Time Zone
Begin Hour
End Hour
Continue
Quit
88
1
24
Note that the default values listed above are for the Wake County example, so there is no
need to modify any of the values. If the user wishes to modify a value, however, they would
select the variable number (e.g., enter an "8" for the Time Zone). The program will position
the cursor in the appropriate location, whereupon the user may enter a new value. Once the
"Site Input Screen" is finalized, the user should enter "11" to continue.
When the site information is complete, the next screen asks:
If you want your input information saved, enter the file name here:
Enter the file name, or press the RETURN key.
EIIP Volume V
4-17
-------
Biogenic Sources
5/21/96
Step 5
The next screen asks:
If the user has already created a meteorological file, answer "YES." The next screen will ask
for the name of the meteorological file:
Enter the MET filename:
Enter "TEST.MET" for example, and the following screen is shown.
PCBEIS version 2.2
Hourly Met Input Screen
QUIT
Hour
PAR
Initially, only one line of meteorological data is displayed. One line at a time may be
modified with the options in the left-hand corner of the screen. Enter a "C" and RETURN to
enter a different cloud cover value, or a "T" for temperature, or a "P" for PAR. Proceed to
the next hour with an "N" for NEXT or the "D" to DUPLICATE the previous line. If no
4-18
EIIP Volume V
-------
5/21/96
Biogenic Sources
changes are desired, the "G" for GO should be used to fill in the entire meteorological file;
then enter RETURN to continue the program.
If PCBEIS2.2 is to be used to build the meteorological file from scratch, answer "no" to "Do
you wish to use a MET file for hourly met parameters?" One line of default meteorological
data will be displayed. The values may be modified using the options in the left-hand
corner. New values may be entered until the file is complete. Enter RETURN to continue
the program.
Preparation of a meteorological file is cumbersome using the procedure in PCBEIS2.2.
These files can also be built outside of PCBEIS2.2 using a text editor, word processing, or
spreadsheet program. The format consists of four columns of data which include: hour, sky
cover (fraction), temperature (°C), and PAR (ji E/m2-sec). Columns should be separated by a
space and must be in ASCII format. This file does not require any header information. If
PAR is recorded as 0.0, the value of PAR will be calculated based on the parameters
previously provided for latitude, longitude, month and hour. The format for these parameters
is illustrated below:
0.4
0.4
0.3 30
PAR
EIIP Volume V
4-19
-------
Biogenic Sources
5/21/96
Next, there is the option of saving the meteorological data file as the following is displayed:
In this example, the answer is "NO" because the data file was not modified. If the user had
answered "YES," the following command would have been displayed:
Enter the MET filename:
Step 6
The PCBEIS2.2 now begins its computations and displays a status line like the following:
Processing HOUR: 3
Step?
The last query from the program is as follows:
Answer "NO" and the final screen is displayed:
Program output is located in TEST.OUT
Program finished.
A listing of this output file is shown in Figure 4-3. This file is automatically saved using the
user file name specified at the outset. For this example, it is TEST.OUT. The viewer can
use the DOS TYPE command to view the data on the screen. The PRINT command can be
4-20
EIIP Volume V
-------
5/21/96
Biogenic Sources
TMPSRF ISOPRENE
C kg/h
0
NO FLUX
mg/m2-h
0.019
0. 019
0.018
0. 019
0.019
0.018
0.018
0.018
0.021
0.023
0.025
0.029
0.029
0. 033
0.034
0. 034
0.033
0 029
FIGURE 4-3. PCBEIS2.2 OUTPUT FILE LISTING
EIIP Volume V
4-21
-------
Biogenic Sources 5/21/96
used to print a hard copy of the results. Note that the output results exceed 80 characters in
width. The user must refer to a DOS manual to make the appropriate modifications for
either screen viewing or printing for their particular system. The output can also be viewed
or printed through use of a text editor or through a word processing or spreadsheet program.
4.3 ESTIMATING EMISSIONS FROM LIGHTNING
The preferred method for estimating emissions from lightning requires the collection of
activity level data (cloud-to-ground [CG] lightning flashes) and determination of the study
area's latitude. Activity for intra-cloud (1C) flashes are calculated from the CG activity. It
is assumed that 1C flashes occur about 4 times more frequently than CG flashes, and this
ratio varies with latitude (Pierce and Novak, 1991). Readers should check with their
regulatory representative to determine whether calculation of lightning NO is required.
Lightning NO calculations are not performed routinely with regional models because of their
relatively small contribution and the difficulty in obtaining lightning strike data.
The equation to calculate emissions from lightning is:
LNO = (NCG • ECG • EFCG) + [(NCG • ECG) • (10/(1 + (O/30)2) - 1)] • EFIC (4-1)
where:
LNO = NO emissions for lightning flashes in study area, molecules NO
NCG = Number of CG flashes recorded by detection network in study area
ECG = Efficiency of the CG network
EFCG = Emission factor of NO for each CG lightning flash
O = Latitude of the study area in degrees
EFIC = Emission factor of NO for each 1C lightning flash
Activity for this emission equation can be collected either from the East Coast lightning
detection network (Orville et a/., 1983), satellite data, or from the lightning strike data
archive from Global Atmospherics, Inc. in Tucson, AZ. Global Atmospherics is the only
known nationally available database for lightning strikes.
4-22 EIIP Volume V
-------
5/21/96 Biogenic Sources
The efficiency of the East Coast CG detection network is approximately 0.7, making the
efficiency factor 1.43. This efficiency factor may vary with the data source. The emission
factors for CG and 1C flashes are:
CG flashes: 2.9 * 1026 molecules NO per flash; and
1C flashes: 2.9 * 1025 molecules NO per flash.
Air models such as ROM or RADM require emission estimates to be separated into layers as
well as grid cells. Emissions from 1C flashes should be added to a 5-km layer above the CG
flash, which will vary in height from 7 km (north of 35 N) to 10 km (south of 25 N),
varying in a linear fashion between the two latitudes. The distribution of NO emitted is
assumed to decrease with altitude as a function of density (Novak and Pierce, 1993).
4.4 ESTIMATING EMISSIONS FROM OIL AND GAS SEEPS
The preferred method for estimating emissions from any source category is the one that
provides the most reliable emission estimate and also can be practically accomplished. In the
case of oil and gas seeps, the most reliable emission estimate is developed through site-
specific measurements of all of the sites or development of a local emission factor based on
a detailed study of a representative sample. Since this method is very resource-intensive, it
is recommended that it be used only when the source is potentially an important part of the
inventory, either as a contributor to total emissions, as a source emissions that is close to the
population, or as one that needs thorough definition for the sake of inventory completeness.
Parameters that will affect emissions from this source category are: flow rate, volatile
constituents, vapor pressure of constituents, surface area, and local meteorology. Collecting
information about flow rate, volatile constituents, or vapor pressure of constituents will allow
more detailed local activity estimates and emission factors to be developed. Additional
information about the seep's surface area, in the case of an oil seep, and local meteorology
should be enough to calculate the emissions for a specific seep, or to develop a local
emission factor for similar seeps.
Approaches to developing local emission factors would include categorizing the material in
the seeps, and defining the volatile constituents and their physical properties for each type of
categorized seep. Three categories of materials that might be emitted at an oil seep are:
• Asphaltum: A brownish-black solid or semisolid mixture of bitumens with a
density of 8.4 to 9.9 Ib/gal;
• Oil: Crude oil with a density of 7.08 to 7.3 Ib/gal; and
EIIP Volume V 4-23
-------
Biogenic Sources 5/21/96
• Tar: A dark, oily, viscid mixture, consisting mainly of hydrocarbons.
Produced by the destructive distillation of organic substances such as wood,
coal, or other organic substance with a density of 7.08 to 7.3 Ib/gal.
Individual gas seeps may also vent different mixtures of gases.
Seeps can also be categorized by flow rate. A difficulty in determining this rate will be in
how the oil or gas reaches the surface; seeps are often diffuse, and may exist over a
widespread area. Any estimate of flow rate will need to be based on several measurements
over that area and on an estimate of the extent of the area. It may be useful to classify
seepage rates by vent size. Table 4-7 is a copy of the guidelines used to do a visual
categorization of seepage rates in a study of off-shore oil and gas seeps. As an indicator of
the scale of seepage rate categories, flow rates for two gas vents characterized as being light-
moderate seeps were measured as being 1.08 and 0.89 cubic feet per hour (cfh). A small oil
seep was visually estimated as having a flow rate of 40 milliliters per hour (Nekton, 1982).
The approach for either a detailed site-specific study or the development of local emission
factors uses the following steps:
• Locate the occurance of all seeps in the inventory area;
• Determine the type of material that is emitted: gas, oil, asphalt or tar;
• Identify a suitable emission factor for the types of materials emitted, and the
pollutants being inventoried;
• Estimate the amount of material emitted at the seep during the inventory
interval (day, season or year) or the surface area of the material; and
• Collect the meteorological parameters that could affect emissions such as
temperature and average local wind speed.
If the site-specific approach is used, it may be necessary to identify how the emissions may
vary through time. If a local emission factor is developed, variability of the seep flow time
and variability of physical qualities of seeps from one site to another should be considered.
There are a number of factors that effect emissions from this source.
4-24 EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 4-7
EXAMPLE CATEGORIZATION OF UNDERWATER GAS SEEPAGE
Symbol
VL
L
LM
M
MH
H
VH
Assessment Term
Very Light
Light
Light Moderate
Moderate
Moderately Heavy
Heavy
Very Heavy
Definition
An intermittent single stream of gas bubbles
from undisturbed sediment or from a small
vent hole.
A continuous single stream of gas bubbles
from a small vent hole. Bubbles generally
have the same diameter.
A continuous gas flow of bubbles of various
diameters and forming a dispersed stream of
bubbles which emanate from a shallow pock-
mark in the sediment.
Multiple streams of gas bubbles flowing
continuously from a basin-sized pock-mark in
the sediment.
Multiple gas flows from the same conduit
which has formed a large pock or dug a small
crater.
Large steady gas flows forming a rising
column of gas, associated with a crater larger
than a bathtub in the sediment.
Pulsating heavy seepage.
Some of the factors affecting emissions temporally are meteorology and seismic activity.
Emissions from different seep sites can vary by: seep flow rate, volatile constituents in the
seep material, vapor pressure of constituents, and seep surface area. Inventory preparers may
want to consider including these factors in the data collection and in the emission estimation
calculation.
EIIP Volume V
4-25
-------
Biogenic Sources 5/21/96
This page is intentionally left blank.
4-26 EIIP Volume V
-------
USING ALTERNATIVE METHODS FOR
ESTIMATING EMISSIONS
5.1 ALTERNATIVE METHODS FOR ESTIMATING BIOGENIC
EMISSIONS
There are four alternate methods for estimating emissions of VOC or NOX from biogenic
sources. When investigating alternative methods, the Yienger and Levy method, for soil
NOX emissions should be considered first, if it is practical. The next alternatives are the
BIOME and the use of alternate land use files with the BEIS-2. The fourth method is using
a version of the BEIS, the earlier version of the BEIS-2. The last three of these methods
estimate gridded, hourly, chemical species-specific VOC emissions. At the time of this
writing, BIOME does not estimate NOX emissions from soils, but the BEIS-2 and BEIS
models do.
5.1.1 YIENGER/LEVY METHOD FOR SOIL NOX
This approach has been developed to estimate a world-wide level of soil NOX emissions.
Yienger and Levy associate the variation of soil NOX emissions with the following factors:
• Biomass burning;
• History of soil moisture (pulsing);
• Temperature;
• Soil moisture;
• Vegetation cover type (biome);
• Canopy reductions; and
• Fertilization rate.
El IP Volume V 5-1
-------
Biogenic Sources 5/21/96
Biomass burning
Biomass burning is recognized as a factor in soil NOX, but is included only for tropical
biomes in Yienger and Levy's approach. Emissions for tropical grasslands and tropical
woodlands are simply increased by a factor of 3 during the dry to wet season transition
period.
Pulsing
Pulsing is an increase of emissions that occurs when a dry soil is moistened. Pulsing is
believed to have its greatest impact in the tropics, but the range of emissions increase ranges
from 10 to 100 times more than background for any environment. Estimating the emissions
resulting from a pulsing event requires a record of the rainfall in the study area for the past
two weeks. When an area receives rain after two dry weeks, emissions will increase for a
period of time. The increase in emissions and the duration of the increase will depend on
the amount of rainfall.
This parameter could be used in US inventories, if the inventory preparers have access to
records of rainfall that extend over the inventory period, and are prepared to do the
calculations required. This parameter has been recognized as being important, but has not
been addressed in previous models.
Temperature, moisture, biome
The equation used to calculate the effect of these parameters takes into account the following
assumptions:
• Emissions from temperate biomes vary similarly in response to temperature.
If one biome has emissions one order of magnitude greater than another at a
given temperature, it will have proportionately larger emissions at a higher
temperature. Yienger and Levy refer to this as "stratified" emissions. Biome
parameters have been developed to fit emissions to a particular biome's
emission range, and are based on literature reviews of field work;
• In moist soils, emissions increase exponentially with temperature within the
temperature range of 10 to 30 degrees C;
Inundated soils suppress nitrifying bacteria, and thus, emissions.
Above the "optimal" soil temperature of 30 degrees C, temperature
dependance of emissions disappears.
5-2 El IP Volume V
-------
5/21/96 Biogenic Sources
Below the soil temperature of 10 degrees, moist soil emissions increase
linearly with temperature.
• In dry soils, rather than an exponential increase, emissions tend to increase
with temperature, in a weak linear relationship.
In temperate regions, the relationship between soil moisture and temperature may not be as
important in areas such as the tropics where there can be large fluctuations in soil moisture.
Soil moisture is determined using the same process of recording rainfall over a period of
time before the time period of emission estimation.
Temperature is calculated from air temperatures using the same empirical relationships used
by Williams et al. (1992), for wet soils, and by adding 5 degrees C to dry soils, based on
observations of Johansson et al. (1988).
Biomes used in the Yienger and Levy study were identified as those that could be matched
to the supporting data, such as a satellite-derived vegetation index and the empirically-
derived factors. In this case, biomes are vegetation and land cover types. Biomes that are
assumed to not have any emissions include water, ice, desert, and scrubland. Tundra,
grassland, woodland, deciduous forests, coniferous forests, drought-deciduous forests, rain
forests and agricultural lands are biomes that are assumed to have emissions.
Canopy Reduction
Canopy reduction is the loss of NOX from soil by diffusion of NO2 through plant stomata and
deposition of NO2 onto and through the cuticle. Although the dominant component of soil
NOX emissions is NO, and these processes absorb only NO2, the process is accentuated by
the conversion of NO to NO2 by O3 in the canopy. Canopy reduction is estimated for each
biome, from leaf area index (LAI) and stomatal area index (SAI), and is included in the
development of a biome parameter. However, although the biome canopy reduction is based
on SAI and LAI specific to each biome, other constants used in the equation to calculate the
canopy reduction are set to values measured in a Brazilian rain forest.
Fertilization Rate
Fertilizing soils with nitrogen fertilizers results in increased NOX emissions, through the
biological nitrification and denitrification of the fertilizer (depending on whether ammonium
or nitrate fertilizer has been applied). Increased emissions can be 1 to 10% of the added
nitrogen in the fertilizers.
Emissions are calculated using the equation:
El IP Volume V 5-3
-------
Biogenic Sources 5/21/96
Flux = fw/d (soil temperature, Aw/d(biome)) * P (precipitation) * CR(LAI, SAI) (5-1)
where:
fw/d = Function that is either constant, linear or exponential, depending on
whether the soil is wet or dry (w/d).
soil temperature = Function that is based on air temperature, varying according to soil
moisture.
Aw/d(biome) = Function dependant on soil moisture, vegetation type, and latitude.
Agricultural soils have a separate set of functions using the monthly
fertilizer rate.
P (precipitation) = Function based on precipitation over the past two weeks.
CR(LAI, SAI) = Canopy reduction factors specific to each biome that describe the
absorption by the plant canopy. Canopy reduction is roughly
dependant on the leaf area index (LAI) and the stomatal area to leaf
area, defined as the stomatal area index (SAI).
Data necessary to use this approach are:
• Rainfall—Yienger and Levy used six-hour resolution rainfall data. Emissions
are dependant on rainfall during the previous two weeks;
• Air temperature;
• Biome; and
• Monthly fertilizer application rate for agricultural soils.
5.1.2 BIOME
BIOME is a GIS-based system that calculates gridded, hourly, chemical species-specific
biogenic emission estimates for VOC. BIOME has been developed to be compatible with
the GEMAP system, an emissions processing system, which has been developed to estimate
emissions from biogenic sources, point, and area sources, and motor vehicles.
The emission rate equation used in BIOME is the same as that used in BEIS-2 and BEIS.
BIOME, however, is designed to be flexible in terms of the parameters used in the equation.
The model has recently been used to estimate emissions using 18 land use classes, and using
5-4 El IP Volume V
-------
5/21/96 Biogenic Sources
the environmental adjustment factors, leaf biomass factor, and similar emission factors to
those that are used in the BEIS (Mayenkar, 1993).
The greatest distinction between the BIOME and the BEIS-2 or BEIS is that BIOME allows
the user to change any of its parameters and the destination of the model's output files.
BIOME allows input of biomass and emission factors for plant types and input of
environmental adjustment factors. Land use and land cover can be entered as GIS coverages.
BIOME also provides flexibility with output format, but is set up to create input files ready
for GEMAP.
Enhancements to the model include quality assurance (QA) capabilities for the analysis of
input and output. BIOME is designed with a scheme for gridding biomass distribution,
which provides a higher level of detail than the BEIS model.
Disadvantages to using this model as opposed to the BEIS-2 are the greater amount of data
collection and handling required for the BIOME, and the potential loss of consistency
between areas, based on different choices being made while running the model. The
advantage of locally collected land use types and up-to-date specific emission factors is that
the result should be a higher quality emission estimate. Since biogenic VOC can be a
significant contributor to the overall total, the effort required for more detailed input data
may be justified.
The information that should be collected when using the BIOME without any of the model's
default values are listed below with some potential information sources:
• Biomass: derived from literature review of field sampling;
• Emission factors: existing data, literature review, researcher contact, and
assignment of surrogate factors;
• Land use: local records, aerial photos, state agricultural information, U.S.
Census Bureau agricultural information, satellite images; and
• Meteorology: National Weather Service, local airports, other weather stations.
Also, in addition to collecting emission factors and other relevant data, users of the BIOME
need to decide whether the default environmental adjustments, biomass adjustments, or leaf
canopy models should be used or if substitutions for those adjustment models should be
made. Users may consider the adjustments used in BEIS-2 as substitutes.
At this time, BIOME does not calculate emission estimates for NOX from soils, as BEIS-2
and BEIS do. However, BIOME may include a soil NOX module in the future. General
El IP Volume V 5-5
-------
Biogenic Sources 5/21/96
issues surrounding the development of alternate land use files or emission factors for soil
NOX and biogenic VOC, which are discussed below, apply to BIOME and the BEIS models.
5.1.3 USING ALTERNATE LAND USE FILES WITH BEIS
There are several advantages to using the BEIS-2 for biogenic emission estimates. Upgrades
to the model are planned to incorporate updates to the algorithm and data files, based on new
research and improved land use categorization. The model provides a nationally consistent
emission estimation tool that provides results comparable to those from any other area that
has used the same method. The most recent version of the BEIS-2 should provide a reliable
emission estimate in most cases. However, the model's land use files are based on the most
readily available national data, and more recent and detailed local information might provide
better results. In such a case, BEIS-2 allows the user to alter land use files. There will also
be cases when detailed local information gathering may be most effectively used with
another biogenics emission estimation model, such as BEIS and BIOME, or with other
methodologies that use the same algorithms as BEIS-2, but not the same data or the
automated features of BEIS-2.
In Section 3 of this volume, the sources used to develop the BEIS-2 default land use files are
listed. In many cases, this default file will be appropriate to use with BEIS-2. However, a
user will need to consider using an alternate file in certain situations. Those situations are
when:
• There has been a change in land use from that in the current default land use
file;
• The land use file needs to be updated to reflect changes for use in a projected
inventory; and
• It is necessary to provide more detailed and accurate land use data for
modeling of urban areas.
When BEIS-2 is being used to create biogenic emission estimates for a projection inventory,
the land use file may need to be changed. Changes to the land use file are necessary if
significant changes in land use are projected to take place (e.g., if a major new development
is planned, or population density is expected to increase in a previously rural area). This sort
of local information should be uncovered during the development of projections for mobile
and area sources. Use of BEIS-2 for projected biogenic emissions should rely on the same
meteorology that is used for other projected emissions in the air quality model being used.
The BEIS-2 land use file is derived from land use data compiled from U. S. Forest Service,
U. S. Census of Agriculture data, U. S. Census 1990 Urbanized Boundary data, and satellite
5-6 El IP Volume V
-------
5/21/96 Biogenic Sources
land classification data. The BEIS land use file is derived from land use data compiled from
the Oak Ridge National Laboratory's Geoecology data set (Olsen, et. al, 1980). The land use
data bases for both BEIS and BEIS-2 consist of gridded coverage for forests, agricultural
crops, and other areas such as grasslands and water, with other areas being interpreted from
satellite images. The satellite data covers areas in the west and arid ecosystems where there
are gaps in the other datasets; however, because it does not have the kind of control that the
Forest Service and the Agricultural Census data have, it is less reliable. Therefore, detailed
land use information may need to be added for those areas, as well as for urban areas, which
may also not be well covered. The default land use data is suitable for estimating biogenic
emissions for the ROM, but should be examined before use with the UAM. PCBEIS2.2
users that are modeling urban areas should also review the land use data in the default land
use data files before running the model.
Preparing Alternate Land Use Files
The data collection task for preparing alternate land use files for use in any of the biogenic
models (BEIS-2, BEIS or BIOME) should be planned and defined before beginning the task
itself. The major steps in the process are as follows:
• Assess agency resources and data needs;
• Determine the level of quality assurance/quality control (QA/QC) that will be
performed;
• Define QA/QC procedures;
• Conduct inventory of available information;
• Determine what can be reasonably collected, converted, and compiled into
land use files;
• Collect information;
• Convert into forms that can be used in the model's land use files (match
spatial qualities and land use types); and
• Create data files.
All inventory processes require oversight and review. The QA/QC measures will be
implemented more efficiently if they are planned as part of the entire process.
El IP Volume V 5-7
-------
Biogenic Sources 5/21/96
The agency should begin by assessing the time and resources available to gather, interpret,
and arrange the information needed to prepare alternate land use files. The agency should
also assess the level of detail needed. If time and money are available, the agency should
inventory the available information and the review format. State and local agencies that can
provide information are:
• Planning departments;
• Revenue departments;
• Parks and recreational services;
• Cooperative extension services;
• Engineering departments;
• Soil and water conservation services; and
• Geographic information services.
Other resources may be found at universities with geography, forestry, or agricultural
departments; they may be able to provide detailed studies of small areas that could be
extrapolated to larger areas.
While investigating potential sources of information, keep the model's data format
requirements in mind. For example, if gridded land use is required, paper maps or county-
level statistics may not be useful. However, many governments are storing information in
GIS, and these digital files may be useful. In addition to the spatial characteristics of the
data, determine if the land use types available from the local resources can be directly
transferred to match the land use types used by the model. If they are not, some method for
interpreting land use types from the available resources will need to be devised. Also, some
method of determining the quality of the data should be devised. Old or unreliable data
should be flagged and discarded. Land use files for BEIS-2 or BEIS can be created using
the formats described in the sections that follow.
Using BEIS or BEIS-2 with Alternate Land Use Files
In order to input alternate land use information to BEIS or BEIS-2, one of two optional land
use files must be generated. This section provides descriptions of these two optional land
use files and the user-input control data necessary to use them. The discussion and
descriptions in this section will use BEIS as the primary example.
5-8 El IP Volume V
-------
5/21/96 Biogenic Sources
Table 5-1 shows the variables and format of the optional gridded land use file, GRBIO for
BEIS. For each grid cell, the user must input values in hectares for the 25 land use types
used by BEIS. The BEIS-2 version of GRBIO has 94 land use types.
If alternate land use data is on a county level, the optional county land use file, UCBIO, may
be used. For each FIPS code, the user must input values in hectares of the amount of area
covered by the 25 land use types used by BEIS, or the 94 land use types used by BEIS-2.
The variables and format of the county land use file for BEIS are specified in Table 5-2.
In order to trigger the use of the alternate land use files, control option flags must be set in
the BEIS. The user must enter a number of variables to specify the domain and scenario-
specific information, and control option flags. Three flags which control several options in
BEIS must be set by the user. The first flag, called GIFLAG, must be set to TRUE if the
user wishes to input his own gridded land use GRBIO file. The second flag, called CFLAG,
must be set to TRUE if the optional UCBIO file is being used to input the user's own county
land use data. The third flag, called GROWFLAG, must be set if the user inputs gridded
land-use data; this flag is set to TRUE to indicate that vegetation is growing and FALSE to
indicate only coniferous vegetation is growing, which might be the case in a cold season like
late fall or winter. A listing of the user-input control data is found in Table 5-3.
Using PCBEIS2.2 with Alternate Land Use Files
Using an alternate land use file with PCBEIS2.2 can be done in order to provide more
accurate or recent data to the model, or to model a partial county area. PCBEIS2.2 uses the
file cc_LU.DAT to define the area and land use types of each county. By altering the
contents of the file, the land use types used by the model for the county can be changed.
The procedure is to:
• Make a copy of the original cc_LU.DAT file on a disk or some other safe
place;
El IP Volume V 5-9
-------
Biogenic Sources
5/21/96
TABLE 5-1
BEIS GRBIO VARIABLES
Record
No.
1
Variable
Name
COL
ROW
Data
Type
1*4
1*4
Description
Column
Row
Land Use (ha)
VEGA(l)
VEGA(2)
VEGA(3)
VEGA(4)
VEGA(5)
VEGA(6)
VEGA(7)
VEGA(8)
VEGA(9)
VEGA(IO)
VEGA(ll)
VEGA(12)
VEGA(13)
VEGA(14)
VEGA(15)
VEGA(16)
VEGA(17)
VEGA(18)
VEGA(19)
VEGA(20)
VEGA(21)
VEGA(22)
VEGA(23)
VEGA(24)
VEGA(25)
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
Oak forest area
Deciduous forest area
Coniferous forest area
Urban oak forest area
Urban deciduous forest area
Urban coniferous forest area
Alfalfa area
Sorghum area
Hay area
Soybean area
Corn area
Potato area
Tobacco area
Wheat area
Cotton area
Rye area
Rice area
Peanut area
Barley area
Oats area
Scrub area
Grass area
Urban grass area
Miscellaneous crops area
Water area
5-10
EIIP Volume V
-------
5/21/96
Biogenic Sources
TABLE 5-2
BEIS UCBIO VARIABLES
Record
No.
1
Variable Name
UCTID
Data
Type
1*4
Description
FIPS county code
Land Use (ha)
VEGA(l)
VEGA(2)
VEGA(3)
VEGA(4)
VEGA(5)
VEGA(6)
VEGA(7)
VEGA(8)
VEGA(9)
VEGA(IO)
VEGA(ll)
VEGA(12)
VEGA(13)
VEGA(14)
VEGA(15)
VEGA(16)
VEGA(17)
VEGA(18)
VEGA(19)
VEGA(20)
VEGA(21)
VEGA(22)
VEGA(23)
VEGA(24)
VEGA(25)
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
R*4
Oak forest area
Deciduous forest area
Coniferous forest area
Urban oak forest area
Urban deciduous forest area
Urban coniferous forest area
Alfalfa area
Sorghum area
Hay area
Soybean area
Corn area
Potato area
Tobacco area
Wheat area
Cotton area
Rye area
Rice area
Peanut area
Barley area
Oats area
Scrub area
Grass area
Urban grass area
Miscellaneous crops area
Water area
BIIP Volume V
5-11
-------
Biogenic Sources
5/21/96
TABLE 5-3
USER-INPUT CONTROL VARIABLES
Record
No.
1
2
3
4
Control Variable
SDATE
SHOUR
EDATE
EHOUR
MONTH
NCOLS1
NROWS1
GMT
GIFLAG
CFLAG
GROWFLAG
NUMCNTY
COUNTY(NUMCNTY)
-9999
Description
Scenario starting date (Julian)
Scenario starting hour
Scenario ending date (Julian)
Scenario ending hour
Month of scenario
Number of columns
Number of rows
Number of hours from GMT to local
TRUE/FALSE flag to show if user-supplied gridded
land use data is being used
TRUE if using user-supplied data
FALSE if not
TRUE/FALSE flag to show if user-supplied county
land used data is being used.
TRUE if vegetation is growing
FALSE if not
TRUE/FALSE flag to show if vegetation is growing
or not.
TRUE if vegetation is growing
FALSE if only coniferous vegetation is
growing (late fall and winter)
Number of counties to be processed
FIPS codes of counties to be processed
Indicates end of list of FIPS codes
NOTE: The control variables data is unformatted; values are separated by commas.
5-12
EIIP Volume V
-------
5/21/96 Biogenic Sources
• Use an ASCII text editor to find and alter the listing for the county of interest.
Counties are sorted by FIPS code. The format of the individual records can
be found in the user's guide for PCBEIS2.2 (Birth, 1995);
• Check the accuracy of the land use types in the county. If a partial county is
being modeled, set the value for the area of the partial county, and area (all
areas are in hectares) of land use types in the partial county. Make sure the
land use types add up to the area being modeled;
• Run PCBEIS2.2. The output will be the value for the corrected county. This
version's output and copies of altered the land use file should be kept in a
separate directory or on a disk that is clearly labeled; and
• Include the print-out of the modified cc_LU.DAT file in the inventory
documentation, along with a justification of the data and methods used.
When using this method, it is important to remember these warnings in order to avoid
pitfalls:
• Back up the original land use file;
• Convert acres to hectares (1 acre = 0.405 hectares); and
• The new land use record should add up to the county area.
5.2 USING ALTERNATIVE METHODS FOR ESTIMATING EMISSIONS
FROM LIGHTNING
Any method for estimating emissions from this source carries with it a high amount of
uncertainty. Estimates of NOX emissions per lightning stroke require calculations involving
the length of the stroke, the number of strokes per flash, the estimated energy discharge, and
the amount of NOX produced per Joule, all of which are uncertain.
The alternative to the preferred method for this source is to define more reliable values for
the factors listed above, or to use a more accurate lightning activity estimate than that used
in the preferred method if one becomes available. Either of these alternatives would depend
on advances reported in journal articles and technical reports.
El IP Volume V 5-13
-------
Biogenic Sources 5/21/96
5.3 USING ALTERNATIVE METHODS FOR ESTIMATING EMISSIONS
FROM OIL AND GAS SEEPS
Very little work has been done to measure emissions from this source because it occurs
infrequently and because measuring emissions from this source can be difficult. Because the
more detailed site-specific approach is very resource intensive, an alternative method for
estimating emissions from oil and gas seeps is to collect local data about the location and
volume of the seeps, and multiply by an emission factor. This method is based on the
CARB's "Area Source Methodologies" Section 1.11, Oil and Gas Seeps. Information that
should be collected for this method includes:
• Location of oil or gas seeps: information should be available from the U. S.
Geological Survey, state or local geological surveys, universities, local
chapters of petroleum exploration trade organizations, or petroleum exploration
companies exploring or extracting in the area; and
• Flow rate from the seeps: information should be available from the same
resources as for location of seeps.
Assumptions that must be made at this point are that the seeps identified in the location step
represent the main portion of existing seeps in the area, and that measurements or estimates
of flow rate at a particular time can be scaled up to represent annual flow. Flow rate
estimates may need to be made from qualitative descriptions of seeps. Whenever possible
try to match these descriptions with descriptions that include a quantitative estimate of flow
and use that estimate. Emission factors for total organic gas (TOG) for oil and gas seeps are
presented in Table 5-4.
TABLE 5-4
EMISSION FACTORS FOR OIL AND GAS SEEPS
Type of Seep
Oil
Gas
Emission Factor TOG3
105
48,648.65
Units
pounds/barrel13
pounds/million cubic feet
a Total Organic Gas
b 1 barrel = 42 gallons
5-14 El IP Volume V
-------
5/21/96
Biogenic Sources
Seep emissions can be further speciated by using the speciation profiles in Table 5-5 and
Table 5-6.
TABLE 5-5
VOC SPECIES PROFILE-OIL SEEPS, VOLATILE FRACTION
Species
Benzene
w-Butane
Ethane
Isomers of Decane
Isomers of Dodecane
Isomers of Heptane
Isomers of Hexane
Isomers of Nonane
Isomers of Octane
Isomers of Pentane
Isomers of Undecane
Propane
Toluene
TOTAL
Weight Fraction
0.0100
0.0700
0.0100
0.1200
0.1200
0.1200
0.0800
0.1200
0.1200
0.0700
0.1200
0.0300
0.0100
1.0000
EIIP Volume V
5-15
-------
Biogenic Sources
5/21/96
TABLE 5-6
VOC SPECIES PROFILE-GAS SEEPS
Species
w-Butane
Ethane
Isomers of Heptane
Isomers of Hexane
Isomers of Pentane
Methane
Propane
TOTAL
Weight Fraction
0.0700
0.0600
0.0100
0.0100
0.0300
0.7500
0.0700
1.0000
5-16
EIIP Volume V
-------
REFERENCES AND BIBLIOGRAPHY
REFERENCES
Birth, T.L. 1995. User's Guide to the Personal Computer Version of the Biogenic
Emissions Inventory System (PCBEIS2.2). Prepared for U.S. Environmental
Protection Agency, Office of Research and Development. Washington, D.C.
Borucki, W.J., and W.L. Chameides. 1984. Lightning: Estimates of the Rates of
Energy Dissipation and Nitrogen Fixation. Reviews of Geophysics and Space Physics.,
vol. 22, no 4, pp. 363-372.
Brandvold, O.K., and P. Martinez. 1988. The NOX /N2O Fixation Ratio from
Electrical Discharges. Atmospheric Environment, vol. 22, no. 11, pp. 2477-2480.
California Air Resources Board. 1993. Emission Methodology for Oil and Gas
Seeps.
Chameides, W., R. Lindsay, J. Richardson, and C. Kiang. 1988. The Role
of Biogenic Hydrocarbons in Urban Photochemical Smog: Atlanta as a Case Study.
Science, vol. 241, pp. 1473-1475.
Cheung, I, B. Lamb and H. Westburg. 1991. Uncertainties in a Biogenic Emissions
Model: Use of Satellite Data to Derive Land Use and Biomass Density Data.
Presented at the AWMA International Specialty Conference on Emission Issues of the
1990's, Durham, North Carolina.
EPA. 1994. Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-1993.
U. S. Environmental Protection Agency/Office of Policy, Planning and Evaluation,
EPA 230-R-94-014. Washington, D.C.
EPA. 1993. Air Quality Criteria for NO'x , Volume I. U. S. Environmental
Protection Agency, EPA 600/8-9l/049aF. Research Triangle Park, North Carolina.
EPA. 199 la. Procedures for the Preparation of Emissions Inventories for CO and
Precursors of Ozone, Volume I: General Guidance for Stationary Sources.
U. S. Environmental Protection Agency/Office of Air Quality Planning and Standards,
450/4-91-016. Research Triangle Park, North Carolina.
El IP Volume V 6-1
-------
Biogenic Sources 5/21/96
EPA. 1991b. Procedures for the Preparation of Emissions Inventories for CO and
Precursors of Ozone, Volume II: Emission Inventory Requirements for Photochemical
Air Quality Simulation. U. S. Environmental Protection Agency/Office of Air Quality
Planning and Standards, 450/4-91-014. Research Triangle Park, North Carolina.
EPA. 1990a. Literature Review of Greenhouse Gas Emissions from Biogenic
Sources. Office of Research and Development/Air Energy Engineering Research
Laboratory, 600/68-90/017. Research Triangle Park, North Carolina.
EPA. 1990b. User's Guide for the Urban Airshed Model, Volume IV: User's
Manual for the Emissions Preprocessor System 2.0, Part A: Core FORTRAN System,
EPA-450/4-90-007D(R) (NTIS PB93-122380/XPB). U. S. Environmental Protection
Agency, Research Triangle Park, North Carolina.
Gaudioso, D., and M. C. Cirillo. 1993. Uncertainty of NMVOC Emission Estimates
from Vegetation. Presented at the EPA/AWMA International Specialty Conference on
Emission Inventory Issues in the 1990's, Durham, North Carolina.
Geron, C. D., U. S. Environmental Protection Agency, Air Energy Engineering
Research Laboratory with D. L. Jones and L. H. Adams, Radian Corporation. April
9, 1994. Contact Report.
Geron, C. D., A. B. Guenther, and T.E. Pierce. 1994. An Improved Model
for Estimating Emissions of Volatile Organic Compounds from Forests in the Eastern
United States. Journal of Geophysical Research (Atmospheres), vol. 99,
pp. 12,773-12,791.
Geron, C. D., T. E. Pierce, and T. L. Birth. 1992. An Alternative Method for
Estimating Biogenic VOC Emissions in EPA Region I. Presented at the EPA/AWMA
International Specialty Conference on Emission Inventory Issues in the 1990's,
Durham, North Carolina.
Guenther, A. B., P. R. Zimmerman, and M. Wildermuth. 1994. Natural Volatile
Organic Compound Emission Rate Estimates for U. S. Forest and Woodland
Landscapes. Atmospheric Environment, vol. 28, pp. 1197-1210.
Guenther, A., P. Zimmerman, P. Haley, R. Monson, and R. Fall. 1993. Isoprene and
Monoterpene Emission Rate Variability: Model Evaluations and Sensitivity Analysis.
Journal of Geophysical Research, vol. 98, no. D7, pp. 12,609-12,617.
6-2 EIIP Volume V
-------
5/21/96 Biogenic Sources
Hill, R.D., R.G. Rinker, and A. Coucouvinos. 1984. Nitrous Oxide Production by
Lightning. Journal of Geophysical Research, vol. 89, no. Dl, pp. 1411-1421.
Johansson, C., H. Rodhe, and E. Sanhueza. 1988. Emission of NO from Savannah
Soils during Rainy Season. Journal of Geophysical Research, vol. 93,
pp. 14,193-14,198.
Lamb B., D. Gay, H. Westberg, and T. Pierce. 1993. A Biogenic Hydrocarbon
Emission Inventory for the U.S.A. Using a Simple Forest Canopy Model. Atmospheric
Environment, vol. 27A, pp. 1673-1690.
Levine, J.S., T.R. Augustsson, 1C. Anderson, J.M. Hoell, and D.A. Brewer. 1984.
Tropospheric Sources ofNOx : Lighting and Biology. Atmospheric Environment,
vol. 18, no. 9, pp. 1797-1804.
Logan, J.A. 1983. Nitrogen Oxides in the Troposphere: Global and Regional
Budgets. Journal of Geophysical Research, vol. 88. no. CIS, pp. 10,785-10,807.
Loveland, T.R., J.W. Merchant, D.O. Ohlen and J.F. Brown. 1991. Development of
a Land-Cover Characteristics Database for the Conterminous U.S. Photogrammetric
Engineering & Remote Sensing, vol. 57, no. 11, pp. 1453-1463.
Mayenkar, K.K. 1993. Development of Biogenic Emissions for the Southeast
Michigan State Implementation Plan (SIP) Inventory. Radian Corporation,
Sacramento, CA, March 1993.
Nekton, Inc. 1982. A Manned Submersible Survey of Three Areas of Natural Oil and
Gas Seeps in State Coastal Waters in the Santa Barbara Channel. Prepared for the
California State Lands Commission, January 1982.
Nowak, D.J. 1991. Urban Forest Development and Structure: Analysis of Oakland,
California. PhD dissertation in Wildlife Resource Science, University of California at
Berkeley.
Novak, J.H., and T. E. Pierce. 1993. Natural Emissions of Oxidant Precursors.
Water, Air, and Soil Pollution, vol. 67, pp. 57-77.
Noxon, J.F. 1976. Atmospheric Nitrogen Fixation by Lightning. Geophysical
Research Letters, vol. 3, pp. 463-465.
EIIP Volume V 6-3
-------
Biogenic Sources 5/21/96
Olson, R., C. Emerson, and M. Nunsgesser. 1980. Geoecology: A County-Level
Environmental Data Base for the Conterminous United States, ORNL/TM-7351, Oak
Ridge National Laboratory, Oak Ridge, TN.
Orville, R., R. Henderson, and L. Bosart. 1983. An East Coast Lightning Detection
Network. Bulletin of the American Meteorological Society, vol. 64, pp. 1,024.
Pierce, I.E. U. S. Environmental Protection Agency, Air Research and Exposure
Assessment Laboratory with L. H. Adams, Radian Corporation. April 9, 1994a.
Personal Communication.
Pierce, T.E. U.S. Environmental Protection Agency, Air Research and Exposure
Assessment Laboratory with L. H. Adams, Radian Corporation. November 28, 1994b.
Personal Communication.
Pierce, T.E., and A. R. Van Meter. October 1992. Volatile Organic Compound and
Nitric Oxide Emissions from Corn in the Midwestern United States. Presented at the
EPA/AWMA International Specialty Conference on Emission Inventory Issues in the
1990's, Durham, NC.
Pierce, T.E. and J.H. Novak. 1991. Estimating Natural Emissions for EPA 's
Regional Oxidant Model. Presented at the EPA/AWMA International Specialty
Conference on Emission Inventory Issues in the 1990's, Durham, N.C.
Radian Corporation and Valley Research Corporation. 1993. Texas Natural Resource
Conservation Commission Biogenic Emission Factors Project, project report.
Prepared for the Texas Natural Resource Conservation Commission.
Williams, E.J., A. Guenther, and F.C. Fehsenfeld. 1992. An Inventory of Nitric
Oxide Emissions from Soils in the United States. Journal of Geophysical Research,
vol. 97, no. D7, pp. 7511-7519.
Yienger, J.J., and H. Levy II. 1995. Empirical Model of Global Soil-Biogenic NOX
Emissions. Journal of Geophysical Research, vol. 100, no. D6, pp. 11,447-11,464.
Zuidema, G., Van Den Born, J. Alcamo, and G.J.J. Kreileman. 1994. Simulating
Changes in Global Land Cover as Affected by Economic and Climatic Factors.
Water, Air and Soil Pollution, vol. 76, pp. 163-198.
6-4 EIIP Volume V
-------
5/21/96 Biogenic Sources
BIBLIOGRAPHY
Anderson, I.C., and J.S. Levine. 1986. Relative Rates of Nitric Oxide and Nitrous
Oxide Production by Nitrifiers, Denitrifters, and Nitrate Respirers. Applied and
Environmental Microbiology, May 1986, pp. 938-945
Winer, A.M., D.R. Fitz, and P.R. Miller. 1983. Investigation of the Role of Natural
Hydrocarbons in Photochemical Smog Formation in California, Final Report.
Statewide Air Pollution Research Center, University of California, report no. PB84-
108653. Prepared for California Air Resources Board.
Causley, M.C., and G.M. Wilson. 1991. Seasonal and Annual Average Biogenic
Emissions for the South Coast Air Basin Generated by the SCAQMD Biogenic Data
Base System. Prepared for South Coast Air Quality Management District.
Horie, Y., S. Sidawi, and R. Ellefsen. 1990. Inventory of Leaf Biomass for
Vegetation in California's South Coast Air Basin. Prepared for South Coast Air
Quality Management District.
Nowak, DJ. 1994. Urban Forest Structure: The State of Chicago's Urban Forest. In
Chicago's Urban Forest Ecosystem: Results of the Chicago Urban Forest Climate
Project, prepared by Northeastern Forest Experiment Station, General Technical
Report NE-186, U.S. Dept. of Agriculture. E.G. Mcpherson, D. J. Nowak, R. A.
Rowntree, eds.
Noxon, J.F. 1976. Atmospheric Nitrogen Fixation by Lightning. Geophysical
Research Letters, vol. 3, no.8, pp. 463-465.
EIIP Volume V 6-5
-------
Biogenic Sources 5/21/96
This page is intentionally left blank.
6-6 EIIP Volume V
------- |