FINAL FEASIBILITY ANALYSIS OF THE IMPLEMENTATION OF WRAP S DUST DEFINITION. Prepared for

Transcription

1 FINAL FEASIBILITY ANALYSIS OF THE IMPLEMENTATION OF WRAP S DUST DEFINITION Prepared for Western Governors Association 1515 Cleveland Place, Suite 200 Denver, CO Prepared by Alison K. Pollack Jason Conder Julia Lester Mary Sorensen Gerard Mansell ENVIRON International Corporation 707 Wilshire Boulevard, Suite 4950 Los Angeles, CA Andrew Comrie, University of Arizona Draft: May 9, 2005 Final: January 2007 P:\W\WRAP\Final Reports\Dust Defn Feasibility Report FINAL.doc [ A]

2 CONTENTS 1. INTRODUCTION 1 Report Organization 4 2. RELATED EFFORTS AND APPROACHES 6 Dust definitions at different spatial scales 6 USGS Conference 6 Local and Air Basin 6 Regional / Interstate 8 Global FEASIBILITY ASSESSMENT APPROACH 17 Purpose of the WRAP Dust definition 17 Draft Definition 19 Anthropogenic and Natural Dust 19 Technical Approach to the Implementation of the Dust Definition 20 Anthropogenic and Natural Dust Estimation and Partitioning of Category 3 Emissions DATA RESOURCE AND METHODOLOGY ASSESSMENT 34 Information Needs 34 Availability and Suitability of Information 36 Example Dust Emission Estimation FEASIBILITY ASSESSMENT PROTOCOL 51 Introduction 51 Feasibility Assessment Protocol 51 Step 5: Identify Available data Resources and Methods for Major Category 3 Sources CONCLUSIONS AND RECOMMENDATIONS 56 Related Efforts 56 Feasibility of Implementing the Draft Dust Definition 56 Data Resource and Methodology Assessment 58 Feasibility Assessment Protocol and Case Studies 59 REFERENCES 60 APPENDICES Appendix A. Data Resources Table Appendix B: Saguaro West Case Study (Separate Report) Appendix C: Salt Creek Wilderness Case Study (Separate Report) Page - i - ENVIRON

3 CONTENTS (continued) TABLES Page Table 2-1: Anthropogenic Contribution to Global Dust Loadings (from Kohfeld, et al., October 2004 USGS conference presentation) 16 Table 3-1. Potential uses of a feasible dust definition (WRAP identified). 26 Table 3-2. Current conceptual method of partitioning natural and anthropogenic dust emissions. 27 Table 3-3. Dust sources in the draft definitions of dust and preliminary identification of the natural and anthropogenic portion of each source. 28 Table 3-4. Dust sources in the draft definitions of dust and category identification. 30 FIGURES Figure 2-1. WRAP Regional Haze Rule Conceptual Glide Path 9 Figure 2-2. Class I Areas in the Western Continental US with 50-km analysis areas around each. ENVIRON has characterized the emissions within and near each of these analysis areas for the WRAP Sources In and Near Forum. 14 Figure 3-1. WRAP Draft Dust Definition. 30 Figure 3-2. Amended WRAP Dust Definition, focusing on three categories of dust sources. 31 Figure 3-3. Procedure for estimating dust emission for a Category 3 dust source. 32 Figure 3-4. Figure 4-1. Dustfall mass (heavy line) and dustfall particle mean diameter (light line) versus depth in sediment core from the bottom of Fourth of July Lake in eastern Washington. The dotted line approximates the long-term trend in dustfall mass. Arrows indicate approximate year when dustfall occurred along the length of the sediment core. Methods used to investigate pre-industrial dustfall may be useful in understanding natural dust emissions for natural lands, including the variability with climate and vegetation changes. In turn, this may facilitate partitioning of current dust emission estimates into natural and anthropogenic portions. Reproduced from Busacca et al. (1998). 33 Location of example study point in southern California (Resource: USGS National Map Viewer) ii - ENVIRON

4 CONTENTS (continued) Figure 4-2. Figure 4-3. Figure 4-4. Figure 4-5. Figure 4-6. Figure 4-7. Figure 4-8. Figure 4-9. Page Results of the query for a list of mammals that could be found at the hypothetical site (Resource: Smithsonian National Museum North American Mammals Database). 43 Life history data potentially useful in constructing a dust emission mode for mule deer (Resource: Cumulative Index for the Mammalian Species). 44 Distribution of desert scrub in California (Resource: California Wildlife Habitats Relationships software). 45 Habitat preference data (habitat suitability values) for mule deer in desert scrub habitat (Resource: California Wildlife Habitats Relationships software). 46 Hyperspectral map of predicted bare ground cover; bare soil areas subject to erosion are shown in yellow (Resource: Eolian Mapping Index). Example map only (location shown in northern Arizona). 47 GIS file depicting soil types present at hypothetical southern California site (Resource: Soil Data Mart, Soil Survey Geographic database). 48 Urban land uses near hypothetical southern California site (highlighted yellow), obtained via land use GIS data (Resource: California GAP Analysis Project). 49 Locations of roads at hypothetical southern California site (Resource: USGS National Map) iii - ENVIRON

5 PREFACE Regulatory Framework for Tribal Visibility Implementation Plans The Regional Haze Rule explicitly recognizes the authority of tribes to implement the provisions of the Rule, in accordance with principles of Federal Indian law, and as provided by the Clean Air Act 301(d) and the Tribal Authority Rule (TAR) (40 CFR ). Those provisions create the following framework: 1. Absent special circumstances, reservation lands are not subject to state jurisdiction. 2. Federally recognized tribes may apply for and receive delegation of federal authority to implement CAA programs, including visibility regulation, or "reasonably severable" elements of such programs (40 CFR 49.3, 49.7). The mechanism for this delegation is a Tribal Implementation Plan (TIP). A reasonably severable element is one that is not integrally related to program elements that are not included in the plan submittal, and is consistent with applicable statutory and regulatory requirements. 3. The Regional Haze Rule expressly provides that tribal visibility programs are not dependent on the strategies selected by the state or states in which the tribe is located (64. Fed. Reg ), and that the authority to implement 309 TIPs extends to all tribes within the GCVTC region (40 CFR (d)(12). 4. The EPA has indicated that under the TAR tribes are not required to submit 309 TIPs by the end of 2003; rather they may choose to opt-in to 309 programs at a later date (67 Fed. Reg ). 5. Where a tribe does not seek delegation through a TIP, EPA, as necessary and appropriate, will promulgate a Federal Implementation Plan (FIP) within reasonable timeframes to protect air quality in Indian country (40 CFR 49.11). EPA is committed to consulting with tribes on a government to government basis in developing tribe-specific or generally applicable TIPs where necessary (See, e.g., 63 Fed. Reg ). It is our hope that the findings and recommendations of this report will prove useful to tribes, whether they choose to submit full or partial 308 or 309 TIPs, or work with EPA to develop FIPs. The amount of modification necessary will vary considerably from tribe to tribe. The authors have striven to ensure that all references to tribes in the document are consistent with principles of tribal sovereignty and autonomy as reflected in the above framework. Any inconsistency with this framework is strictly inadvertent and not an attempt to impose requirements on tribes which are not present under existing law. -I- E N V I R O N

6 Tribes, along with states and federal agencies, are full partners in the WRAP, having equal representation on the WRAP Board as states. Whether Board members or not, it must be remembered that all tribes are governments, as distinguished from the stakeholders (private interest) which participate on Forums and Committees but are not eligible for the Board. Despite this equality of representation on the Board, tribes are very differently situated than states. There are over four hundred federally recognized tribes in the WRAP region, including Alaska. The sheer number of tribes makes full participation impossible. Moreover, many tribes are faced with pressing environmental, economic, and social issues, and do not have the resources to participate in an effort such as the WRAP, however important its goals may be. These factors necessarily limit the level of tribal input into and endorsement of WRAP products. The tribal participants in the WRAP, including Board members Forum and Committee members and co-chairs, make their best effort to ensure that WRAP products are in the best interest of the tribes, the environment, and the public. One interest is to ensure that WRAP policies, as implemented by states and tribes, will not constrain the future options of tribes who are not involved in the WRAP. With these considerations and limitations in mind, the tribal participants have joined the state, federal, and private stakeholder interests in approving this report as a consensus document. -II- E N V I R O N

7 1. INTRODUCTION Historically, particulate matter (PM) issues (and related visibility issues) have focused on anthropogenic sources, since it is these sources that are emphasized in the Clean Air Act and are subject to control to achieve or maintain air quality standards. Similarly, most modeling assessments in State Implementation Plans (SIPs) focused on anthropogenic sources and used source-receptor models and rollback analyses. In certain areas, such as the South Coast Air Basin, the presence of both secondary and primary PM 10 led to the use of increasingly sophisticated threedimensional, photochemical models, such as those used in ozone modeling. The same trend is now occurring in visibility assessments, based on the multiple PM components that affect visibility. Unlike ozone modeling, where biogenic (e.g. natural) sources of hydrocarbons are estimated and included in the modeling inventories (but not the reporting inventories), most PM (and visibility) modeling does not explicitly include natural sources of PM. Nor do standard reporting PM inventories discriminate between natural and anthropogenic sources of PM; indeed, most SIP inventories do not include natural sources of PM. In previous modeling exercises, the impact of natural PM was often described as background, and excluded from rollback analyses. As planners and policy makers formulate and communicate their plans to improve visibility and PM air quality, there is an increasing need to discriminate between uncontrollable and controllable sources to better assess PM control and visibility improvement strategies. The need exists for PM emission inventories that report both the natural and anthropogenic components of dust sources. The WRAP draft dust definition is a first step in creating such inventories. The goal of this project is to assess if the draft definition can be feasibly applied, on a source type by source type basis, to identify the natural and anthropogenic contributions of certain dust sources. To support the latest modeling tools used in PM and visibility analyses, a feasible application of the dust definition would also need to produce spatially and temporally resolved inventories of natural and anthropogenic contributions. To be feasible for these types of applications, implementation of the draft definition would have to be capable of producing these spatially and temporally resolved inventories of natural and anthropogenic emissions. Section 169 of the Clean Air Act declares as a national goal the prevention of any future, and the remedying of any existing impairment of visibility in mandatory Class I federal areas which impairment results from man-made air pollution. In 1999, the U.S. Environmental Protection Agency (EPA) promulgated the Regional Haze Rule (RHR) to meet this national goal. The Western Regional Air Partnership (WRAP) is a collaborative effort of tribal governments, state governments, and various federal agencies to implement the recommendations of the Grand Canyon Visibility Transport Commission (GCVTC) and to develop the technical (e.g. information and data gathering, data analysis, emission inventory development, and modeling) and policy tools needed by western states and tribes to comply with the RHR. WRAP also seeks to provide technical Dust Definition Feasibility Analysis -1- E N V I R O N

8 assistance to States in the development of emission control strategies to improve visibility. In the WRAP region, dust (as opposed to point sources of primary particulates, such as elemental carbon, and secondary particulates formed in the atmosphere) is a significant component in visibility degradation. As implied by the national goal definition, only visibility impairments from man-made pollution must be prevented and/or remedied. Mobile sources and stationary point sources are clearly man-made sources of pollution, but dust sources, particularly fugitive dust sources, can result from both man-made and natural conditions. As noted by the co-chair of WRAP s Dust Emission Joint Forum (DEJF) at the April 2004 Best Available Control Measures (BACM) Working Group meeting, The distinction between anthropogenic and natural dust will help us to identify and prioritize sources of dust that are most appropriate to control. To assist in the identification of anthropogenic and natural dust the DEJF has created a draft definition with broad criteria for categorizing dust emissions as natural or anthropogenic. Even at this broad level, there is a potential overlap where sources may be considered both natural and anthropogenic (c.f. the following figure, Examples of Anthropogenic and Natural Emissions Under a Draft Definition of Dust, taken from the RFP). The absence of a solid line separating natural from anthropogenic sources is one of the issues that increases the difficulty in applying the definition for operational purposes, such as inventories, Dust Definition Feasibility Analysis -2- E N V I R O N

9 modeling and control strategy development and assessment. The DEJF has identified several potential purposes of a feasible dust definition. These include: Clarify how the WRAP defines dust, its sources, and causes; Provide an operational definition for use in receptor- and emissions-based source apportionment techniques; and Identify and prioritize sources of dust which are most appropriate to control for purposes of improving visibility in Class I areas. A potential use of a definition would be to delineate anthropogenic and natural dust emission in the WRAP emissions inventory. Each emission type could then be tagged separately in the WRAP s air quality simulation model to estimate the contribution of natural and anthropogenic sources of dust to visibility impairment in Class I areas. The definition may also be useful for describing predominant source types (i.e., natural versus anthropogenic) when analyzing specific dust events and/or discussing them with a variety of WRAP stakeholders. Other common western regional air quality issues raised by the WRAP membership may also be addressed through the results of this project. As many areas near attainment or work to maintain their attainment status, better information on sources subject to control (anthropogenic sources) and those that are not (natural sources) would allow more targeted and cost-effective end game strategies, compared to more traditional across-the-board control strategies. Lastly, as the RFP notes, the dust definitions are not intended for use in refining EPA s estimates of natural visibility conditions, although they may be useful for that purpose. These definitions may also be used with regard to dust emanating from outside of the U.S. that is, dust from other countries can be either natural and/or anthropogenic. It will be noted when certain data/methods may be available (and at what cost) that would allow the feasible use of the dust definition for these purposes. The DEJF believes the draft definition is conceptually sound, but is uncertain whether it can be implemented in practice (e.g., for the applications above), and at a reasonable cost. This report describes the potential criteria and related data/methods that would allow for the delineation of natural and anthropogenic origins of dust. One test for feasibility is the reasonableness of the availability of and the cost of obtaining the needed data and applying the appropriate method(s). The report presents an initial assessment of data resources and methods. The assessment includes information on the availability, cost, extent, and applicability to end uses (e.g. emission characterization, estimation, or partitioning). This report identifies the breadth and depth of the needed data, the qualifications of necessary interpreters of the data, and, where available, its cost. The feasibility of implementing the dust definition may vary by project and purpose. To address this issue, the report contains a Feasibility Assessment Protocol, which will allow users to identify the data sources that are relevant to their projects based on the purposes of those projects (e.g., Dust Definition Feasibility Analysis -3- E N V I R O N

10 inventory development, current inventory partitioning, air or visibility modeling of current, future, or restored natural scenarios). The report will provide conclusions as to the general level of feasibility for implementing the dust definition for regional haze purposes based on the readily available data/methods for each source type. If such data exists for general applications of interest to WRAP and its forums, the dust definition can be generally considered feasible. For specialized applications, additional data/methods may be needed that are not currently available (e.g. satellite imagery, field studies, novel GIS applications) and the cost would be higher. The feasibility of the application of the dust definition in these cases would depend on the importance of the project and the resources available. Report Organization This report is organized as follows: Section 2 presents a brief review of other efforts related to characterizing, estimating, and partitioning natural and anthropogenic sources of dust. Section 3 presents the current WRAP draft dust definition and the feasibility assessment approach of this report. Section 4 presents the assessment of data resources and methods that could be used to implement the dust definition. Section 5 presents the Feasibility Assessment Protocol, a tool to determine the feasibility of implementing the dust definition for specific projects and purposes. This section also looks at the general feasibility of implementing the dust definition for regional haze inventory and modeling assessments. Section 6 presents the conclusions and recommendations of this report on the feasibility of implementing the draft dust definition. References Appendix A presents the data resources that have been identified and assessed as useful in implementing the draft dust definition. The data resources are presented in a comprehensive, tabular form. After the release of the draft Feasibility Assessment in May 2005, ENVIRON conducted to case studies to evaluate the effectiveness of the Feasibility Assessment Protocol proposed in this Report. During the development of these case studies, revisions and additions to the Feasibility Assessment Dust Definition Feasibility Analysis -4- E N V I R O N

11 Protocol were made. Those revisions and additions are discussed in the two case studies, which have been included under separate covers as Appendix B and C of this final Report. Appendix B is the first case study of the Draft Dust Definition. The area of study was the Saguaro West Class I area in Arizona, chosen to represent an area with significant dust impacts to visibility with relatively few data resources for implementing the Dust Definition. Appendix C is the second case study of the Draft Dust Definition. The area of study was the Salt Creek Wilderness Class I area in New Mexico, chosen to represent an area with significant dust impacts to visibility with relatively rich data resources for implementing the Dust Definition. It was also chosen because results from this case study were used in WRAP s New Mexico SIP Pilot Study. Dust Definition Feasibility Analysis -5- E N V I R O N

12 2. RELATED EFFORTS AND APPROACHES Dust definitions at different spatial scales USGS Conference The United States Geological Survey (USGS) hosted a workshop on October 23-24, 2004 entitled Linking the Scales of Process, Observation, and Modeling of Dust Emissions. Topics at the workshop included how to integrate dust studies of different scales (spatial and temporal), what physical and biotic conditions are associated with spatial and temporal variability of dust emissions, how to produce regional maps of dust-emission potential suitable for linking to meteorological and climate models, and identifying what portion of emissions are attributable to anthropogenic disturbances, directly (e.g. construction, agriculture) or indirect consequences resulting from climate change. Over 17 presentations and other contributions related to these topics can be found at the USGS web site ( Several of the presentations dealt with emissions from sources with both natural and anthropogenic contributions. The focus, however, is still on studies that produce inputs to meso- and macro-scale dust emissions, transport, and deposition models rather than identifying specific anthropogenic and natural contributions. For the field studies, workshop participants identified the desert regions of western North America, with the Mojave Desert as a good study area. They noted that western North America is historically significant because human intervention and land management practices have dramatically altered the landscape and related dust emissions. Also, dust storms have arisen from several different types of conditions in the Mojave Desert, the western Great Plains, and northern Mexico. They characterized the Mojave Desert as an area where extensive documentation exists of geological, ecological, and surface characteristics, as well as information concerning anthropogenic activities and climate changes. Local and Air Basin Most studies of fugitive dust at the local and air basin level have focused on anthropogenic sources and their emissions. For local areas, the issue has been identifying and mitigating nuisance dust. For air basins, State Implementation Plans (SIPs) have focused on attaining the National Ambient Air Quality Standards (NAAQS) for PM 10. SIP inventories and modeling have historically only described anthropogenic sources of PM 10 or PM 10 pre-cursors. (In contrast, ozone SIP modeling inventories will often include biogenic VOC emissions, although those emissions are not reported in annual average or planning emission inventories used for SIP tracking purposes. Some areas do identify a small fraction of background PM 10 during modeled attainment demonstrations.) Several areas have also prepared Natural Event Action Plans (NEAPs), as allowed under U.S. EPA s Natural Events Policy (NEP). The NEP was promulgated in June 1996 and covers high Dust Definition Feasibility Analysis -6- E N V I R O N

13 PM 10 levels from natural sources such as volcanic eruptions, wildfires, and windblown dust. The focus of the NEP is the education and alerting of populations that could be affected by natural events, controlling anthropogenic sources with best available controls, and studying other potential mitigation strategies. The policy does not require the quantification of emissions from natural sources, although they may need to be characterized in documenting individual natural events. The Arizona DEQ used a slightly different approach, determining "normal" and "extreme" climate conditions to subsequently determine if certain PM events were extraordinary because of an extreme natural climate event (windy and dry). More information about this approach can be found in Arizona Department of Environmental Quality (ADEQ) Technical Criteria Document for Determination of Natural Exceptional Events For Particulate Matter Equal to or Less Than Ten Microns in Aerodynamic Diameter (PM 10 ), dated May 31, 2000 and Climatological Analysis for PM 10 Natural Exceptional Events in Arizona (Comrie and Garfin, June 2001). The WRAP Sources In and Near Class I Areas Forum is currently conducting a review of PM 10 State Implementation Plans (SIPs) and Natural Events Action Plans (NEAPs) that have been prepared for locations within the WRAP region. These reviews are focused on the development of information which can ultimately be used for the evaluation, selection, and implementation of emission management strategies. A preliminary summary of past and current PM 10 nonattainment areas, status, and types of plans submitted has been completed (Roe, S WRAP PM 10 SIP Review Project Technical Memorandum #1: Initial Selection of Candidate Nonattainment Areas. E.H.Pechan & Associates, 24 March 2005, sips/tm-1.pdf). WRAP s review of emission management strategies is focused on control measures with proven effectiveness in reducing ambient PM levels. At this point, a group of 23 PM 10 nonattainment areas have been identified for further analysis. One result of this preliminary review is that most of the PM 10 nonattainment areas within the WRAP region are characterized by situations in which PM 10 violations are primarily attributable to a very limited number of source groups (commonly residential wood smoke, agricultural tilling and cultivation, or travel on unpaved roads). However, there are of course some very large and important areas (San Joaquin Valley, South Coast) with complex source mixtures and a greater variety of control measures. Of particular interest to the issue of anthropogenic vs. non-anthropogenic fugitive dust sources is the types of sources for which control measures have been included in the SIPs being reviewed by WRAP. The In and Near Class I Areas Forum s schedule calls for completion of this project by June 30, Dust Definition Feasibility Analysis -7- E N V I R O N

14 Regional / Interstate NAPAP: As part of the National Acid Precipitation Assessment Program (NAPAP), a state-of-the-scienceand-technology report was prepared for visibility. The report (Trijonis, 1990) focused on the causes and effects of existing and historical conditions, related to visibility. The default annual natural levels of PM components in the EPA's Guideline for Estimating Natural Conditions Under the Regional Haze Rule (EPA, 2003) were based on values that were developed for the NAPAP by Trijonis, It was assumed that half of the fine soils were natural, resulting in 0.5 μg/m 3 of annual average concentration. The coarse mass was assumed to be all natural and to contribute about 3 μg/m 3 to the annual average concentrations. The coarse contribution was assumed to be the same in the eastern and western United States. These estimates were very simplistic and, particularly for the coarse contribution, arbitrary. The analysis is based on older measurements and is a top-down estimate, independent of dust source characterization or quantification. It does not account for new information on regional dust storm characterization and global transport (e.g., Gobi and Saharan dust storms). Western Regional Air Partnership In response to regulatory requirements for regional haze SIPs (see Chapter 1), WRAP provides technical and policy tools needed by the western states and tribes to comply with the RHR. WRAP has prepared a conceptual glide path to describe the goals and challenges of the RHR. Dust Definition Feasibility Analysis -8- E N V I R O N

15 Figure 2-1. WRAP Regional Haze Rule Conceptual Glide Path As shown in Figure 2-1, the goal of the RHR is to attain natural visibility conditions by A key concept is natural conditions. For WRAP, natural sources include natural windblown dust, biomass smoke, and other natural processes. Manmade sources include industrial activities and man-perturbed smoke and dust emissions. Distinguishing between natural and man-made sources has been a key issue for WRAP. From the WRAP Policy for Categorizing Fire Emissions: The WRAP Fire Emissions Joint Forum (FEJF) was established to develop policy and technical tools to address smoke effects caused by wildland and agricultural fire on public, tribal, and private lands. Due to the limitations of the current visibility monitoring technology to determine fire impacts, the FEJF was charged with addressing fire emissions contribution to natural background conditions. The FEJF formed the Natural Background Task Team (NBTT) to develop a methodology to categorize fire emissions as either natural or anthropogenic ; thus providing the basis for fire s inclusion in natural background condition values and ultimately, the tracking of reasonable progress. This Policy has been developed over an 18-month period by the NBTT; a group made up of state, tribal, and federal agency representatives, as well as those from industry, agriculture, academia, and environmental organizations. During this process, the NBTT solicited public input regarding both technical and policy issues. The resulting Recommended Policy for Categorizing Fire Emissions was granted Dust Definition Feasibility Analysis -9- E N V I R O N

16 consensus approval by the FEJF on August 30, The WRAP granted consensus approval for the Policy on November 15, The WRAP Dust Emissions Joint Forum (DEJF) has developed a draft dust definition to distinguish between natural and anthropogenic fugitive dust emission sources (see Chapter 3 for the DEJF s draft dust definition). The purpose of this report is to assess the feasibility of implementing this dust definition in the tools and policies used by the states and tribes to comply with the RHR. If feasible, this definition would provide a basis for natural dust to be included in the natural background conditions and future reasonable progress assessments. WRAP has commissioned the compilation of detailed information on PM composition, component contributions to visibility impairment, ambient PM trends, and other information which will be used to identify the causes of haze (CoH) in each Class I area (see CoH Assessment project web site at WRAP is using this and other information to develop a preliminary report on the Attribution of Haze (AoH) in each Class I area (see aoh/ars1/index.html). Results from the CoH Assessment and AoH projects will be useful for identifying potential high priority source categories for analysis. In addition, two recent studies funded by WRAP have been completed by ENVIRON which will provide important background information for the proposed study. Under contract to the WRAP and its forums, ENVIRON completed a windblown dust emission inventory that covered the WRAP air quality modeling domain (Mansell et al., 2005) and an extensive emission inventory compilation (Pollack et al., 2004) for all non-windblown PM sources located in and within 50 km of the boundaries of each of the 116 Federal and 5 Tribal Class I areas in the WRAP states. The results of these emission inventory development and analysis projects provide a better understanding of the benefits and limitations of the various methods and data available to quantify natural and anthropogenic dust sources in the inventories. Attribution of Haze and Causes of Haze Projects The WRAP has established the Attribution of Haze Workgroup to prepare a policy-level report describing the emissions source categories and geographic source regions presently contributing to visibility impairment at each of the tribal and mandatory federal Class I Areas within the WRAP region. The Workgroup included a broad representation of technical and policy representatives and established an open meeting format to foster additional input and coordination among the various stakeholders. The project was designed in two Phases, with the first Phase designed as a trial run for Phase II. Phase II will build upon the findings and recommendations of Phase I. Dust Definition Feasibility Analysis -10- E N V I R O N

17 The goals of the Attribution of Haze (AoH) project are: To provide state and tribal air regulators with an initial, regional assessment of the attribution of haze in their Class I Areas: To provide an initial assessment of how and to what extent natural and anthropogenic emissions from each state affect western Class I Areas; and Ultimately, to provide air regulators with the information and tools necessary to prepare stat and tribal implementation plans (SIPs and TIPs) under the Regional Haze Rule (RHR). The attribution of haze results of Phase I were designed neither to explicitly single out individual sources nor to identify the amount of reductions needed by a given source or group of sources in order to meet the RHR goals. Rather, the results of the project are intended to give a preliminary assessment of the natural and anthropogenic emissions from geographic source regions that contribute to visibility impairment at Class I Areas. In addition, conducting original research was not the intent of the project, but rather, the project was designed to assemble existing information from various analyses and utilize that information to determine the source types and regions impacting each of the Class I Areas. Specifically, three major of analyses and data used for the project include: Emission inventories Although in some cases the emission inventories were incomplete or uncertain, the inventories defined the geographic regions and provided estimates of the magnitude of emissions; Monitoring data Light extinction calculated from measures speciated fine mass and total coarse mass define the scope of visibility impacts in or near Class I Areas; Modeling results Atmospheric chemistry and transport models were used to provide the connection between emissions from geographic source regions and measured fine mass in or near Class I Areas. Source attribution was described and evaluated using a weight of evidence approach. The methodology involved reviewing emission inventories, monitoring data, and modeling results for the 2002 calendar year. One or two independent source apportionment methods were applied to each Class I Area, results compared and supporting data and information were used to corroborate or further scrutinize apportionment result. Supporting data and analyses used in the project included the results of the Causes of Haze Assessment (COHA) conducted by the Desert Research Institute under contract to WRAP. The COHA project used meteorological back trajectory analyses for all Class I Area monitoring Dust Definition Feasibility Analysis -11- E N V I R O N

18 locations with the WRAP region using NOAA s HYSPLIT model. Combining back trajectory results and monitoring data enabled a Trajectory Regression Analysis for each monitoring location. The analysis related the amount of time air spends over a source region, as determined by a compilation of many back trajectories,.to the aerosol species measured at a receptor site ad was applied to sulfate and aerosol extinction only. As previously noted, the AoH results and recommendations were based on a weight of evidence approach rather than any single data set, analysis result, or line of reasoning. As such, a significant amount of information and analysis results, including apportionment and attribution methods, meteorological back trajectory summaries, and emissions and monitoring data and summaries are available for use in the development and application of a refined dust definition and development of a dust definition feasibility protocol. The results of Phase II of the AoH project may provide additional supporting data and analyses for use in current efforts regarding dust definition refinements and implementation. Fugitive Windblown Dust Emissions From Wind Erosion Fugitive PM emissions from wind erosion were estimated by a team led by ENVIRON for WRAP. Emission estimates were made using a model developed based upon the most recent information available in the literature and implemented in a variety of models (Mansell, et al., 2005). The development and application of the estimation methodology relies on detailed knowledge concerning surface characteristics of vacant lands susceptible to wind erosion. Given the large regional scale domain to which the model was applied, certain assumptions were made due to the lack of detailed information available to characterize the vacant land surfaces. In particular, assumptions regarding the various landuse types, vegetative cover and disturbance levels of the soils were required. The results of the Windblown Dust Project indicated a strong dependence on the assumed values for surface roughness lengths, as determined by the land use types, and the level of disturbance of the soils. One of the major limitations of the modeling results is directly related to the assumed level of soil disturbance. Assumptions were necessary due to the large geographic extent of the modeling domain and the lack of data to adequately characterize surface conditions. It is anticipated that the results of the current feasibility study would be invaluable in providing additional supporting information to better resolve and characterize the surface parameters most important in the estimation of dust from wind erosion. In addition, while it is currently difficult, if not impossible, to distinguish between anthropogenic and natural sources of windblown dust, the results of the feasibility assessment presented herein can be useful in providing additional data and information to apply as a first attempt to address these concerns. Also, the results of the dust definition refinements and assessment protocol, particularly the various surface characterization data identified, may allow for improvements to the application Dust Definition Feasibility Analysis -12- E N V I R O N

19 of the windblown dust model. Refinements to the methodology would be focused on those areas most susceptible to, or known to be caused by, human related activity (i.e., anthropogenic sources). These refinements will also provide WRAP with information to allow more focused control strategy development efforts. In And Near Class I Areas Emission Inventory Development Non-Windblown In Pollack et al., 2004, emission estimates were obtained for primary PM 10, PM 2.5, and precursors (VOC, NO x, SO 2, NH 3 ). This effort involved estimating area and mobile source emissions based on mapping of appropriate surrogate data within the identified 50 km analysis zones (shown in Figure 2-2 for the western conterminous US Class I areas are also present in Alaska) and spatially allocating area and mobile source emissions contained in the 1996 county-level WRAP inventories to the analysis zone around each Class I area. ENVIRON also surveyed state officials and Federal Land Managers with local knowledge of Class I areas in the WRAP region to obtain basic information on the presence and role of local emission sources and their impact on visibility in all Class I areas. This project resulted in the development of an extensive set of spreadsheets and maps that summarize the inventories for each area (or area group) by emissions source category. These maps and spreadsheets along with other work products from this project are available on the WRAP project web page ( Source categories for which emission summaries were developed include non-windblown dust area sources, fire and on- and off-road mobile sources. Although most of these sources are not directly related to the issues that have arisen for the draft dust definition, databases related to spatially and temporally resolving wild vs. prescribed fires, unpaved roads (e.g., penetration of human activity into native lands), and logging equipment may be useful in establishing new methods or as a source of data that could distinguish natural from anthropogenic dust emissions for related sources. In addition, the results of the project provides some information concerning the relative importance of various emission sources, particularly dust sources, which can be used to prioritize further efforts associated with the distinction between natural and anthropogenic sources. The emission inventory analysis and evaluation of survey results conducted for the in and near Class I areas project provided valuable information regarding the types of activities and their importance with respect to particulate matter emissions near Class I Areas throughout the Western US. Results showed that of the PM sources examined, fugitive dust from construction, agricultural tillage and harvesting, and paved and unpaved road dust, along with PM from residential wood combustion and industrial processes account for 94 percent of PM 10 and 84 percent of PM 2.5 emissions from non-windblown/non-natural area and nonroad mobile sources over all of the Class I analysis areas combined. Survey respondents contacted for this project provided information on trends in emission sources and on local regulations and control measures intended to address the largest source categories. Information collected and identified through the survey efforts of this project will be useful in evaluating the feasibility of applying the dust definition to specific Class I Dust Definition Feasibility Analysis -13- E N V I R O N

20 areas, addressing feasibility issues by targeted data acquisition, and ultimately, applying the dust definition to the relevant inventories for use in modeling or other assessments. Figure 2-2. Class I Areas in the Western Continental US with 50-km analysis areas around each. ENVIRON has characterized the emissions within and near each of these analysis areas for the WRAP Sources In and Near Forum. Vegetated Surfaces in the Southwestern United States Okin and Gillette (2004) reviewed a state of the science for modeling wind erosion and dust emission from naturally-vegetated lands in the Southwestern United States. They note that while many dust emission models can accommodate a fair degree of complexity at a regional scale, few can produce spatially-explicit estimates of dust flux. Modeling approaches incorporating instruments such as the Total Ozone Mapping Spectrophotometer (TOMS) are useful in very coarse (regional) scales. For example, previous work by Ginoux et al. (2001) suggests that the overwhelming majority of desert dust originates from ephemeral lake basins. Other researchers suggest that a majority of dust is associated with anthropogenic land uses (Mahowald et al., in press). Models such as the Spatially Explicit Wind Erosion and Dust Flux Model (SWEMO) are in development, using data from the Journada Basin in south-central New Mexico. SWEMO uses maps of soil texture and vegetation, in addition to knowledge of vegetation and size parameters, to create maps of wind erosion susceptibility and dust flux. A key weakness the authors identify in the Dust Definition Feasibility Analysis -14- E N V I R O N

21 model is that the bulk of dust emissions at a regional scale often appear to occur in very small hot spot areas that are beyond the predictive capacity of the model due to the comparatively low resolution of soil and vegetative cover maps. Mapping hot spots may be the key to producing reliable models for regional and site-specific dust emission. Okin and Gillette (2004) remark that hyperspectral analysis, geostatistical analysis of aerial photographs, and remote sensing of vegetation density using optical canopy models may be useful tools in this effort. Spatial modeling of dust emission on vegetated lands in the Southwestern United States remains in its infancy, and models such as SWEMO are only first steps in estimating dust emissions at a regional scale on these vegetated surfaces. Models such as SWEMO do not possess the ability to partition anthropogenic and natural dust emissions. Global Zender et al. (2004) reviewed terminology and constraints associated with current approaches which are being used to quantify mineral dust mass budgets on a global scale. The authors note several key concerns with current global modeling approaches: Emission estimates cover a wide range (more than a factor of 2). This variability is primarily due to the goals of the emission models and their inherent assumptions. Dust particle size varies widely among models. Zender et al. (2004) recommend standardization of particle size in future efforts (i.e., particle diameter < 10 μm). There is high uncertainty with all models. For example, substituting one set of meteorological dataset for another (collected by two different organizations) produce estimates of global dust loads that vary by four-fold or more. Few models distinguish between natural and anthropogenic dust emissions. Zender et al. (2004) propose a more complex partitioning scheme for emission sources. Natural sources are identified as regions that emitted dust in pre-industrial times (e.g., prior to 1750 in North America). Zender et al. (2004) further subdivide anthropogenic sources into 1st kind, which are due to direct modifications to the land that alter soil erodibility, and 2nd kind, which are due to anthropogenic changes to the global climate. As noted in the Kohfeld et al. presentation at the October 2004 USGS conference, there is a great deal of uncertainty as to the impact of natural sources in global dust loadings. From her presentation: Dust Definition Feasibility Analysis -15- E N V I R O N

22 Table 2-1: Anthropogenic Contribution to Global Dust Loadings (from Kohfeld, et al., October 2004 USGS conference presentation) Study Anthropogenic Contribution Comment IPCC, 2001 Up to 50% tuned to AVHRR AOT Prospero et al small Natural sources dominant Luo et al % New desert sources Tegen et al., 2004 <10% Comparing simulated and observed dust storm frequencies Mahowald et al. subm. 0-50% Comparing simulated and observed dust storm frequencies For future efforts, Zender et al. (2004) note that modeling and observation strategies should attempt to partition natural and anthropogenic sources of dust in order to be of relevance to decision makers. The authors admit that this is extremely challenging and will require a focused interdisciplinary effort, given the complexity and high natural variability associated with natural dust emissions. Dust Definition Feasibility Analysis -16- E N V I R O N

23 3. FEASIBILITY ASSESSMENT APPROACH Purpose of the WRAP Dust definition Historically, particulate matter (PM) issues (and related visibility issues) have focused on anthropogenic sources, since it is these sources that are emphasized in the Clean Air Act and are subject to control to achieve or maintain air quality standards. Similarly, most modeling assessments in State Implementation Plans (SIPs) focused on anthropogenic sources and used source-receptor models and rollback analyses. In certain areas, such as the South Coast Air Basin, the presence of both secondary and primary PM 10 led to the use of increasingly sophisticated threedimensional, photochemical models, such as those used in ozone modeling. The same trend is now occurring in visibility assessments, based on the multiple PM components that affect visibility. Unlike ozone modeling, where biogenic (e.g. natural) sources of hydrocarbons are estimated and included in the modeling inventories (but not the reporting inventories), most PM (and visibility) modeling does not explicitly include natural sources of PM. Nor do standard reporting PM inventories discriminate between natural and anthropogenic sources of PM; indeed, most SIP inventories do not include natural sources of PM. In previous modeling exercises, the impact of natural PM was often described as background, and excluded from rollback analyses. As planners and policy makers formulate and communicate their plans to improve visibility and PM air quality, there is an increasing need to discriminate between uncontrollable and controllable sources to better assess PM control and visibility improvement strategies. The need exists for PM emission inventories that report both the natural and anthropogenic components of dust sources. The WRAP draft dust definition is a first step in creating such inventories. To support the latest modeling tools used in PM and visibility analyses, a feasible application of the dust definition would also need to produce spatially and temporally resolved inventories of natural and anthropogenic contributions. To be feasible for these types of applications, implementation of the draft definition would have to be capable of producing these spatially and temporally resolved inventories of natural and anthropogenic emissions. Section 169 of the Clean Air Act declares as a national goal the prevention of any future, and the remedying of any existing impairment of visibility in mandatory Class I federal areas which impairment results from man-made air pollution. In 1999, the U.S. Environmental Protection Agency (EPA) promulgated the Regional Haze Rule (RHR) to meet this national goal. The Western Regional Air Partnership (WRAP) is a collaborative effort of tribal governments, state governments, and various federal agencies to implement the recommendations of the Grand Canyon Visibility Transport Commission (GCVTC) and to develop the technical (e.g. information and data gathering, data analysis, emission inventory development, and modeling) and policy tools needed by western states and tribes to comply with the RHR. WRAP also seeks to provide technical Dust Definition Feasibility Analysis -17- E N V I R O N

24 assistance to States in the development of emission control strategies to improve visibility. In the WRAP region, dust (as opposed to point sources of primary particulates, such as elemental carbon, and secondary particulates formed in the atmosphere) is a significant component in visibility degradation. As implied by the national goal definition, only visibility impairments from man-made pollution must be prevented and/or remedied. Mobile sources and stationary point sources are clearly man-made sources of pollution, but dust sources, particularly fugitive dust sources, can result from both man-made and natural conditions. The absence of a solid line separating natural from anthropogenic sources is one of the issues that increases the difficulty in applying the definition for operational purposes, such as inventories, modeling and control strategy development and assessment. To assist in the identification of anthropogenic and natural dust the DEJF has created a draft definition with broad criteria for categorizing dust emissions as natural or anthropogenic. The definition serves to: Clarify how the WRAP defines dust, its sources, and causes; Provide an operational definition for use in receptor- and emissions-based source apportionment techniques; and Identify and prioritize sources of dust which are most appropriate to control for purposes of improving visibility in Class I areas. A potential use of a definition would be to delineate anthropogenic and natural dust emission in the WRAP emissions inventory. Each emission type could then be tagged separately in the WRAP s air quality simulation model to estimate the contribution of natural and anthropogenic sources of dust to visibility impairment in Class I areas. The definition may also be useful for describing predominant source types (i.e., natural versus anthropogenic) when analyzing specific dust events and/or discussing them with a variety of WRAP stakeholders. Other common western regional air quality issues raised by the WRAP membership may also be addressed through the results of this project. As many areas near attainment or work to maintain their attainment status, better information on sources subject to control (anthropogenic sources) and those that are not (natural sources) would allow more targeted and cost-effective end game strategies, compared to more traditional across-the-board control strategies. Lastly, as the RFP notes, the dust definitions are not intended for use in refining EPA s estimates of natural visibility conditions, although they may be useful for that purpose. These definitions may also be used with regard to dust emanating from outside of the U.S. that is, dust from other countries can be either natural and/or anthropogenic. In ENVIRON s approach to this project, these other purposes will be kept in mind and notes made when certain data/methods may be available (and at what cost) that would allow the feasible use of the dust definition for these purposes. Dust Definition Feasibility Analysis -18- E N V I R O N

25 In the RFP, WRAP identified several potential uses of a feasible dust definition. Table 3-1 presents a list of these uses, their identified purposes, and a brief comment on how the use and purposes may affect the feasibility analysis. This may not be an exhaustive list, and ENVIRON will work with WRAP and its forums to identify other uses and the scope of what would define a feasible dust definition in the context of those uses. Any discussion of feasibility of implementing the draft dust definition must include the purpose for which it is to be used. The feasibility of using the draft dust definition to identify natural and anthropogenic contributions will be highly dependent on the ultimate purpose of the project to which it is being applied. Draft Definition The WRAP DEJF has proposed a draft dust definition as follows: Dust is particulate matter which is or can be suspended into the atmosphere as a result of mechanical, explosive, or wind-blown suspension of geologic, organic, synthetic, or dissolved solids. Dust does not include non-geologic particulate matter emitted directly by internal and external combustion processes. Fugitive dust was defined as dust which could not reasonably pass through a stack, chimney, vent, or other functionally equivalent opening. Anthropogenic and Natural Dust From the draft definition: Examples of anthropogenic and natural dust are provided below (Table 3-2). Any mitigation of dust for regional haze control would likely be focused on those anthropogenic sources which are most likely to contribute to visibility impairment in Class I areas and which are technically feasible and cost-effective to control. Sources that are already controlled or partially controlled may be technically infeasible or not cost-effective to control further. Anthropogenic emissions do not include any emissions which would occur if the surface were not disturbed or altered beyond a natural range. Such emissions should be subtracted, if practicable, from the total dust emissions to determine the precise anthropogenic emission quantity. Table 3-3 lists the dust sources in the definition and classifying them as anthropogenic, natural or potentially either. A key determinant of the feasibility of the dust definition is the ability to clearly classify a source in a particular area as natural or anthropogenic or an explicit combination. ENVIRON has also added a column of potential controls or mitigations for each source, since certain applications of the dust definition are for purposes of developing control strategies or assessing progress. This list is preliminary. ENVIRON intends to work with WRAP staff and others that WRAP recommends in its preparation of a final list. The final list will be an interim deliverable of the project. Dust Definition Feasibility Analysis -19- E N V I R O N

26 The focus of this report is whether anthropogenic and natural emission contribution for those categories where both exist can be distinguished through the technical implementation of the draft dust definition. Technical Approach to the Implementation of the Dust Definition WRAP s Draft Definition of Dust is summarized in Figure 3-1, Examples of Anthropogenic and Natural Emissions under a Draft Definition of Dust. An alternative to the five categories of emissions presented in Figure 3-1 is a three category approach presented in Figure 3-2. As very different dust sources may spatially co-occur on the same site, it may be more useful to express dust sources on the basis of activity rather than a description of spatial location. In this way, dust sources fall into three categories: Figure 3-1. See end of chapter. Category 1: Purely anthropogenic sources o Examples: Particle emission from cooling towers, wind erosion of agricultural soils, wind erosion and particle emission from unpaved and paved roads Category 2: Purely natural sources o Examples: Ash emission by volcanoes, mineral particle emission from wave action/sea spray, wind erosion of unstable soil following landslides Category 3: Natural sources, which may be anthropogenically, influenced. A portion of the dust emissions may be due to anthropogenic influences. o Examples: Wind erosion and mechanical suspension of soil due to animal movement, wind erosion of bare areas on natural lands, wind erosion of sediment from dried, ephemeral water bodies Examples of Anthropogenic and Natural Emissions Under a Draft Definition of Dust Anthropogenic Natural emissions emissions Category 1: Category 2: Category 3: Purely Purely natural Natural sources anthropogenic sources which may be sources anthropogenically influenced Emissions due to anthropogenic influence Emissions under healthy, natural conditions Total Dust Emissions Construction, mining, etc. Ash emission by volcanoes Wind erosion and Particle emission from Mineral particle emission mechanical suspension of cooling towers from wave action/sea spray soil due to animal Agricultural operations Wind erosion of unstable movement (native and nonnative) Wind erosion of agricultural soil following landslides soils Wind erosion of bare areas Emissions from unpaved on natural lands and paved roads (undisturbed vs. previously disturbed) Wind erosion of sediment from dried, ephemeral water bodies (natural or anthropogenic) Figure 3-2. See end of chapter. Table 3-4 allocates the fugitive dust sources in Table 3-3 into the three categories. Category 1 sources have been studied intensely through the SIP process, are relatively well defined, and several Dust Definition Feasibility Analysis -20- E N V I R O N

27 tested emission methodologies and models are available to estimate emissions from these sources. Category 2 sources are less well studied and generally have had little impact on SIP and related studies. They are unlikely to contribute significantly to NAAQS or haze violations except on a local and/or event basis. Category 3 sources provide the greatest technical challenge, both in general emission estimation and in identifying natural and anthropogenic emissions. Table 3-3 allocates the fugitive dust sources in Table 3-2 into these categories. Category 3 dust sources are the focus of the feasibility analysis. Category 3 sources are: 1) Animal movement, 2) Windblown dust from grass, range, and forest lands, 3) Windblown dust from undeveloped lands (previously disturbed), 4) Areas burned by fires, and 5) Exposed beds of lakes and rivers. For any these sources, the emissions can be natural or partly natural and partly anthropogenic. Anthropogenic and Natural Dust Estimation and Partitioning of Category 3 Emissions One of the overall goals of WRAP s draft dust definition is to partition Category 3 dust emissions into natural and anthropogenic portions. According to the above conceptual framework, a binary allocation of dust emissions between anthropogenic and natural categories is based on the level of anthropogenic disturbance for a site. For example, if a site is found to be disturbed or altered by humans beyond a natural range, Category 3 dust emissions would be classified as anthropogenic. Under the current conceptual framework, using disturbance to partition Category 3 dust sources may present several fundamental challenges. An approach which incorporates modeling or comparison with reference areas may allow more flexibility and precision. Disturbance as a Partitioning Tool: The term disturbance is potentially too simplistic for an accurate and complete understanding of Category 3 dust emissions necessary to partition emissions between natural and anthropogenic portions. The current approach also leads to a binary partitioning of dust emissions which would likely overestimate anthropogenic emissions. o Natural Disturbance: Disturbance is a natural, vital part of healthy, functioning natural lands (Pellant et al., 2000). While methods exist which may be useful in providing information regarding natural levels of disturbance on natural lands (Pellant et al., 2000; Mendoza et al., 2002; O Brien et al., 2003), disturbances cannot be clearly and cleanly classified as natural or anthropogenic (Grossman et al., 1998). Some anthropogenic disturbances are similar enough to natural disturbances that the resulting effects cannot be clearly distinguished via inspection of the land, while others may create novel modified communities that are unprecedented in the natural landscape (Grossman et al., 1998). Dust Definition Feasibility Analysis -21- E N V I R O N

28 o Indirect Disturbance: For many Category 3 dust sources, anthropogenic influences may not be directly observable, yet greatly affect dust emissions. Indirect influences would not be readily apparent from approaches which directly define natural ranges of disturbance or assess ecological health on rangelands and forestlands (Pellant et al., 2000; Mendoza et al., 2002; O Brien et al., 2003). For example: Anthropogenically-induced climate change is not easily observable as disturbance, yet it arguably could have a very large impact on natural dust emissions (Zender et al., 2004). Demonstrating global change at local scales is an extreme challenge, especially in the West, which has a highly variable climate. Although there may be a measured regional warming trend, the effect of anthropogenic influences on climate is a decadal to century-scale concern. Estimating the variability in climate due to this change and translating this to estimates of anthropogenic dust emissions will require substantial effort. Suppression of natural fire regimes, neglecting population control for large herbivores (e.g., deer, antelope), and the diversion of surface water are not readily-observable disturbances, yet they may have as much or more impact on dust emissions than more traditional, more easily-observed anthropogenic disturbances, such as suppressed vegetation cover on overgrazed natural rangelands. o Binary Partitioning: Even at natural sites severely affected by anthropogenic influenced beyond a natural range, it is unlikely that 100% of dust emissions could be credited to anthropogenic influences. This binary allocation would result in overestimation of anthropogenic dust emissions. Direct Comparison as a Partitioning Tool: The direct comparison approach encourages partitioning on a source-by-source and site-by-site basis. This method uses direct comparisons between site and reference conditions with the assistance of dust emission modeling to investigate anthropogenic influence and provide quantitative estimates of natural and anthropogenic emissions in cases in which anthropogenic influence is identified. The direct comparison approach is far superior to the more simplistic binary approach, which would tend to restrict the process to an examination of readily-observable disturbance. o Reference Areas: Comparison with reference areas is routine in natural resource management and ecological risk assessment and would enable a better understanding Dust Definition Feasibility Analysis -22- E N V I R O N

29 of the natural dust emissions in ecologically-healthy areas free (or with a minimum) of anthropogenic influence. For many dust sources and sites, reference area identification and the understanding of anthropogenic influences can be achieved via examining ecosystem health rather than disturbance. Ecosystem health is defined as the degree to which the integrity of the soil, vegetation, water, air, as well as the ecological processes of the ecosystem, are balanced and sustained (Pellant et al., 2000). This concept not only incorporates assessments of disturbance but other key factors related to the ability of the land to repair itself, integrity of ecosystem processes, and the normal variability of a site. Understanding ecological health (rather than just disturbance) would enable a more holistic approach to characterizing the anthropogenic influences and understanding the relationships between anthropogenic influences and dust emissions. In general, a difference in measurable environmental conditions (also known as metrics ) greater than 20% for two given areas may indicate differences outside the range of natural variability (Suter et al., 2000). o Reference Time Periods: Comparison of anthropogenically influenced sites with historical record of dust emissions during pre-industrial period (Busacca et al., 1998; Zender et al., 2004) may also be potentially powerful to understand the natural background of dust emissions in an area. o Modeling: Dust emission modeling is a necessary step in comparing reference situations with site conditions and estimating dust emission portions. Modeling would be conducted on a source-by-source basis. Category 3 dust sources are not equally vulnerable to anthropogenic influence. For example, the minimum level of anthropogenic influence necessary to negatively influence vegetative coverage resulting in increases in windblown dust emissions is not necessarily equal to the minimum level of influence necessary to increase dust emissions from animal movement. Anthropogenic influence could be assessed in many ways (e.g., presence of roads, livestock grazing, proximity of urban land uses), and it is likely that a wide variety of land attributes would be examined to assess anthropogenic influence. Modeling would be conducted on a site-by-site basis. Even for the same dust source type, it is likely that the relationship between anthropogenic influence Dust Definition Feasibility Analysis -23- E N V I R O N

30 and dust emission would vary by site. In some cases, site-specific models may need to be developed. At even the most anthropogenically-influenced sites there will be some portion of Category 3 dust emissions that is natural. Modeling could account for natural dust emissions and avoid the situation in which 100% of Category 3 dust emissions are attributed to anthropogenic influences. A revised figure Examples of Anthropogenic and Natural Emissions under a Draft Definition of Dust is presented in Figure 3-2 to illustrate a conceptual dust definition which would avoid potential limitations associated with the use of the term disturbance. This definition would enable the partitioning of Category 3 dust sources on the basis of more holistic investigations of ecological health and natural conditions, as facilitated by cause-and-effect dust source modeling and comparisons with reference conditions. Figure 3-3 illustrates a hypothetical example of a Category 3 direct comparison procedure for estimating the dust emissions on a site-by-site, source-by-source basis (Fig. 3-3): Step 1: Dust source identification o Example: In the area of interest (Fig. 3-3) three Category 3 dust sources are present: Area of interest Step 1: Dust source identification Mule deer population range Bare soil Ephemeral lake Unpaved road Step 3: Estimation of site-specific dust emission Site data + Model = Site-specific dust emission Step 2: Dust source characterization Generic dust emission model for dust source Wind erosion from ephemeral lakes Wind erosion from bare soil Mechanical suspension of soil due to mule deer movement Step 4: Dust partitioning Anthropogenic influences + Site-specific dust emission + Model = Anthropogenic and Natural Dust Emission Portions Figure 3-3. See end of chapter. Step 2 and Step 3: Estimation of the Category 3 dust source emissions (Step 3) for the site using previously-developed generic dust source models and site-specific data (Step 2) o Example: Estimating dust emissions from the on-site ephemeral lake using sitespecific data and generic models capable of predicting dust emissions from dry lakebeds Step 4: Investigation of the anthropogenic influences that may affect of the Category 3 dust source and quantitative assessment of anthropogenic influence on dust emission and estimation of anthropogenic portion, which may include: Dust Definition Feasibility Analysis -24- E N V I R O N

31 o Identification of anthropogenic influence For the on-site ephemeral lake dust source, this might include an analysis of the water management strategies associated with anthropogenic diversion of water from the lake, location of anthropogenically-influenced bare soil areas near/overlapping the lake, and/or quantification of vehicle traffic which may occur on the lakebed due to proximity to nearby roads o Estimation of anthropogenic dust emission portion Comparison of source-specific dust emissions at the site with source-specific dust emissions at reference areas - Example: Comparison of dust emission estimates of suitable reference ephemeral lakes which do not incur vehicle traffic Adjusting dust emission models by analyzing the effects of variables linked with anthropogenic influence - Example: Analyzing the magnitude of dust emissions associated with the time period during which the lakebed is dry due to anthropogenic diversions of water This could require a substantial effort involving many disciplines (geology, ecology, climatology, paleontology, engineering, statistics, etc.) to determine the impact of this partitioning effort on overall dust emissions. For example, Category 3 dust emissions may be very small in relation to Category 1 or 2. It may be possible to simply include Category 3 dust emissions with Category 2 if it is reasonable to assume that Category 3 dust emissions are very small in relation to Category 2 emission, or that the anthropogenic portions of Category 3 dust emissions are small relative to other dust emissions. Thus, it is recommended that WRAP assess the overall dust emissions of Category 3 sources relative to Category 1 and Category 2 emissions. o A simple comparison of the sum of Category 1, 2, and 3 dust emissions may be feasible using the data resources and approaches outlined in the next section. o Investigating case studies such as the Columbia River PM 10 Project (CP3) may also be useful in understanding approaches in partitioning dust emissions. This project is in the 12th year of its research. Although the main goal of the project is to from its inception and yet today is to develop conservation practices that will enable Dust Definition Feasibility Analysis -25- E N V I R O N

32 growers to control wind erosion and dust emissions in Washington s Columbia Plateau region, several key research thrusts may be directly useful for partitioning and modeling Category 3 dust emissions, including: Quantifying pre-farming (natural) dust emissions via a study of dust emissions from 620 A.D. to the present (Figure 3-4). Producing data to validate regional and global dust emissions models to better backcast dust emissions from the plateau and forecast future emissions. Figure 3-4. See end of chapter. Reconstructing dust accumulation from dynamic source areas in the Plateau region under the controlling influences of climate, plant, cover and topography to quantify dust emissions. Using satellite imagery on a large-scale to differentiate bare, smooth fields including those with inadequate residue from those with sufficient residue or green cover for protection against wind erosion. Table 3-1. Potential uses of a feasible dust definition (WRAP identified). Identified Use Clarification of how WRAP defines dust, its sources, and causes Operational definition for use in receptor- and emissions-based source apportionment techniques Identification and prioritization of sources of dust which are most appropriate to control for purposes of improving visibility in Class I areas Purpose Clearer policy directives, better outreach to agencies and the public Visibility projections, control strategy assessment Control strategy development and assessment Quantification Needs For Feasible Implementation of the Dust Definition Relies on general identification of methods and associated data to distinguish natural and anthropogenic contributions from a given source type. Relies on specific, quantifiable parameters that can be applied on spatially and temporally-resolved data for specific sources in specific areas. The need for high resolution data in specific area may require the acquisition of new data or the development of new methods to meet feasibility requirements. Feasibility of application would be project specific, particularly if there are resource constraints. Feasibility depends on being able to tag natural and anthropogenic portions of each specific source, for at least the major emissions contributors (or alternatively, the major visibility impairers). The definition of major may be case- or area-specific. Dust Definition Feasibility Analysis -26- E N V I R O N

33 Table 3-1. Potential uses of a feasible dust definition (WRAP identified). Identified Use Identifying anthropogenic and natural dust emissions in the WRAP inventory Addressing other common western regional air quality issues raised by WRAP members. Refining of EPA s estimate of natural visibility conditions Categorizing dust emissions from outside of the U.S. Purpose Clarify extent of sources subject to controls. Support modeling assessments of visibility trends with and without controls. Common issues can include the need for Natural Event documentation, NEAPs and NEAP revisions, end game control strategies to address remaining PM 10 hot spots within a non-attainment area. Establishing natural visibility conditions as baselines/goals as referenced in the Clean Air Act and RHR. Assess the limits of U.S. control strategies and, if requested, provide assistance in the development of non- U.S. control strategies. Quantification Needs For Feasible Implementation of the Dust Definition As above. As above. As above. Feasibility may be limited by the availability of relevant non-u.s. data. Feasibility analysis may identify data gaps and needs. Table 3-2. Current conceptual method of partitioning natural and anthropogenic dust emissions. Mechanical Anthropogenic Dust Mechanically- and explosively-suspended solids and dissolved solids from activities including but not limited to: Agriculture Construction, mining, and demolition Material handling, processing, and transport Vehicular movement on paved and unpaved surfaces Animal movement on surfaces which have been disturbed or altered by humans beyond a natural range Animal movement on undisturbed or unaltered surfaces by a number of animals which is greater than native populations Cooling towers Natural Dust Movement of a number of indigenous animals on surfaces which have not been disturbed or altered by humans beyond a natural range Natural landslides, rockslides, and avalanches Solids and dissolved solids emitted by volcanoes, geysers, waterfalls, rapids, and other types of splashing Extraterrestrial material and impacts Dust Definition Feasibility Analysis -27- E N V I R O N

34 Table 3-2. Current conceptual method of partitioning natural and anthropogenic dust emissions. Windblown Anthropogenic Dust Solids and dissolved solids entrained by wind passing over surfaces which have been disturbed or altered by humans beyond a natural range. Such surfaces may include, but are not limited to: Undeveloped lands Construction and mining sites Material storage piles, landfills, and vacant lots Agricultural crop, range, and forest lands Roadways and parking lots Artificially-exposed beds of natural lakes and rivers Exposed beds of artificial water bodies Areas burned by anthropogenic fires (as defined by the WRAP Policy for Categorizing Fire Emissions) which have yet to be revegetated or stabilized Wind-blown particulate matter from sources created by natural events over 12 months ago, similar to EPA s natural events policy Natural Dust Solids and dissolved solids entrained by wind passing over surfaces which have not been disturbed or altered by humans beyond a natural range. Such surfaces may include, but are not limited to: Naturally-dry river and lake beds Barren lands, sand dunes, and exposed rock Natural water bodies (e.g., sea spray) Non-agricultural grass, range, and forest lands Areas burned by natural fires (as defined by the WRAP Policy for Categorizing Fire Emissions) which have yet to be revegetated or stabilized Note: Anthropogenic emissions are only that portion of the total emissions which occur in excess of what would occur naturally. Wherever practical, natural emissions should be estimated and subtracted from total emissions to determine a more precise anthropogenic quantity. Table 3-3. Dust sources in the draft definitions of dust and preliminary identification of the natural and anthropogenic portion of each source. Dust Source Natural Portion Anthropogenic Portion Potential Controls or Mitigations Agriculture (mechanical and None All BACM/RACM windblown) Construction, mining, and None All BACM/RACM demolition (mechanical and windblown) Material handling, processing, None All BACM/RACM and transport (mechanical and windblown) Paved and unpaved roadways None All BACM/RACM (traffic and windblown) Cooling towers None All Unknown Dust Definition Feasibility Analysis -28- E N V I R O N

35 Table 3-3. Dust sources in the draft definitions of dust and preliminary identification of the natural and anthropogenic portion of each source. Dust Source Natural Portion Anthropogenic Portion Animal movement Native animals on Native animals on natural natural lands that lands that have been have not been disturbed by humans beyond disturbed by humans a natural range. beyond a natural Non-native animals range. introduced by humans (e.g., cattle ranching) on undisturbed natural lands. Non-native animal movement on disturbed range lands. Potential Controls or Mitigations TBD Windblown from grass, range, and forest lands Windblown from undeveloped lands (previously disturbed) Areas burned by fires Exposed beds of lakes and rivers Non-agricultural grass, range, and forest lands that have not been disturbed by humans beyond its natural range. If the undeveloped land has returned to its natural state, i.e., is no longer disturbed beyond its natural range. Areas burned by natural fires, as defined by the WRAP Policy for Categorizing Fire emissions, and which have yet to be revegetated or stabilized. Natural river and lake beds that may become naturally dry (may be seasonal or through a drought cycle). Agricultural grass, range, and forest lands Non-agricultural grass, range, and forest lands that have been disturbed by humans beyond its natural range. If the land remain disturbed beyond its natural range. Areas burned by anthropogenic fires, as defined by the WRAP Policy for Categorizing Fire emissions, and which have yet to be revegetated or stabilized. Artificially exposed by the action of humans (e.g. water transfer, damming, in-fill projects). The exposed beds of artificial (man-made) water bodies. Re-establishment of surfaces to within its natural range Others, TBD. Re-establishment of surfaces to within its natural range Accelerated revegetation or stabilization Mitigations considered and used in Owens Valley. Others, TBD. Natural water bodies All (sea spray) None Unknown Natural landslides, rockslides, All None Unknown and avalanches PM emitted by volcanoes, geysers, waterfalls, rapids, and other types of splashing All None Unknown Extraterrestrial material and impacts Windblown PM from sources created by natural events over 12 months previously All None Unknown None All Removal or stabilization of created sources Dust Definition Feasibility Analysis -29- E N V I R O N

36 Table 3-4. Dust sources in the draft definitions of dust and category identification. Dust Source Category Agriculture (mechanical and windblown) 1 Construction, mining, and demolition (mechanical and windblown) 1 Material handling, processing, and transport (mechanical and windblown) 1 Paved and unpaved roadways (traffic and windblown) 1 Cooling towers 1 Animal movement 3 Windblown from grass, range, and forest lands 3 Windblown from undeveloped lands (previously disturbed) 3 Areas burned by fires 3 Exposed beds of lakes and rivers 3 Natural water bodies (sea spray) 2 Natural landslides, rockslides, and avalanches 2 PM emitted by volcanoes, geysers, waterfalls, rapids, and other types of splashing 2 Extraterrestrial material and impacts 2 Windblown PM from sources created by natural events over 12 months previously 3 Figure 3-1. WRAP Draft Dust Definition. Dust Definition Feasibility Analysis -30- E N V I R O N

37 Examples of Anthropogenic and Natural Emissions under a Draft Definition of Dust Category 1: Purely anthropogenic sources Anthropogenic emissions Category 2: Purely natural sources Natural emissions Category 3: Natural sources which may be anthropogenically influenced Emissions due to anthropogenic influence Total Dust Emissions Emissions under healthy, natural conditions Construction, mining, etc. Particle emission from cooling towers Agricultural operations Wind erosion of agricultural soils Emissions from unpaved and paved roads Ash emission by volcanoes Mineral particle emission from wave action/sea spray Wind erosion of unstable soil following landslides Wind erosion and mechanical suspension of soil due to animal movement (native and nonnative) Wind erosion of bare areas on natural lands (undisturbed vs. previously disturbed) Wind erosion of sediment from dried, ephemeral water bodies (natural or anthropogenic) Figure 3-2. Amended WRAP Dust Definition, focusing on three categories of dust sources. Dust Definition Feasibility Analysis -31- E N V I R O N

38 Area of interest Step 1: Dust source identification Mule deer population range Bare soil Ephemeral lake Step 2: Dust source characterization Unpaved road Step 3: Estimation of site-specific dust emission Site data + Model Generic dust emission model for dust source = Site-specific dust emission Step 4: Dust partitioning Anthropogenic influences + Site-specific dust emission + Model = Anthropogenic and Natural Dust Emission Portions Figure 3-3. Procedure for estimating dust emission for a Category 3 dust source. Dust Definition Feasibility Analysis -32- E N V I R O N

39 Figure 3-4. Dustfall mass (heavy line) and dustfall particle mean diameter (light line) versus depth in sediment core from the bottom of Fourth of July Lake in eastern Washington. The dotted line approximates the long-term trend in dustfall mass. Arrows indicate approximate year when dustfall occurred along the length of the sediment core. Methods used to investigate pre-industrial dustfall may be useful in understanding natural dust emissions for natural lands, including the variability with climate and vegetation changes. In turn, this may facilitate partitioning of current dust emission estimates into natural and anthropogenic portions. Reproduced from Busacca et al. (1998). Dust Definition Feasibility Analysis -33- E N V I R O N

40 4. DATA RESOURCE AND METHODOLOGY ASSESSMENT Information Needs Four primary types of information are required to provide site-specific estimates of dust emission for Category 2 and 3 dust sources (Fig. 3-2, Fig. 3-3). 1. Dust source geographic distribution: This information type includes data needed to define the spatial extent of dust sources which vary geographically in the Western United States. Most spatial extent data are in map format. Once the spatial extents of dust sources are known, an inventory of dust sources can be compiled for any particular site. Examples include spatial extent (distribution) of: o ephemeral lakes in the United States o disturbed areas in the Mojave Desert o burrowing animal species in the United States o populations of large mammal species in the United States 2. Dust source characterization: This information type includes data and conceptual information required to construct new dust emission models or modify existing dust emission models. Models will be used to estimate dust emissions by a dust source using site-specific data. Although many dust emission models will be generic, in some cases it may be necessary to construct site-specific models. Both conceptual information and data are needed to construct dust emission models. Conceptual information is needed to construct the model framework, and includes knowledge from a wide variety of fields in various forms (expertise, examples, case studies, literature, etc.). Examples include: o approaches with which to estimate the population densities of large animal species using land cover information o understanding of the scale of climate data needed to predict the effects of precipitation on wind erosion from dry lake beds o case study illustrating the effects of cattle grazing on vegetative cover on rangelands o understanding of the most important meteorological factors affecting the formation of aerosols from sea spray Dust Definition Feasibility Analysis -34- E N V I R O N

41 o methods to predict the effects of habitat change and vegetative cover due to fire suppression Data are needed to provide constants or ranges of parameters for the model. Data and conceptual information can be in a variety of forms. Examples include: o ranges of soil moisture content of ephemeral lakes o average wind speed and direction for the United States o range of soil displacement rates for burrowing animals in various ecosystems o range of ash production per volcano eruption o ranges of population densities for large mammal populations 3. Spatially-explicit dust emission: This information type includes site-specific spatial data needed to produce site-specific dust emission estimates using the appropriate dust source model. This information is in a wide variety of forms, but serves as site-specific input for dust emission modeling. Examples of spatially-explicit data include: o percentage of bare ground coverage in southwestern Kern County, California o dominant vegetation types in Grand Canyon National Park o average monthly precipitation for northwestern Montana o physical soil structure (percent sand-silt-clay) at a site in central Idaho o number of active volcanoes within 100 miles of Yosemite National Park 4. Dust partitioning: This information type includes site-specific information required to identify and quantify site-specific natural and anthropogenic portions of Category 3 dust emissions. Information may include data with which to evaluate the anthropogenic influences in the area via modeling, comparison with suitable natural reference areas, and/or investigating natural background (pre-industrial) dust emissions. Examples of data include: o sediment coring data with which to estimate pre-industrial dust emissions in Rocky Mountain National Park Dust Definition Feasibility Analysis -35- E N V I R O N

42 o range of bare ground cover in healthy mountain big sagebrush compared to bare ground cover in areas of mountain big sagebrush in Bridger-Teton National Forest which may be anthropogenically influenced o area estimate of unpaved roads in Larimer County, Colorado o current and historic wildlife management strategy for population control of large mammal species in Yellowstone National Park o history of hydrological management of water bodies which dry due to anthropogenic diversion of water during droughts o current logging activity in Kaibab National Forest o information on land use to identify a reference site with minimal anthropogenic influence within 20 miles of Santa Fe, New Mexico Availability and Suitability of Information Appendix A provides an example list of potentially-applicable online data, models, and conceptual approaches capable of providing resources required to implement the WRAP Dust Definition. Whereas Appendix A is not an all-inclusive list of resources and it is highly likely that other resources may yield additional useful or different information, it provides a sample of the information resources available. Appendix A also shows a hierarchy of potential usefulness of a particular resource for WRAP (see the scale of 1 to 10, with 10 indicating resources that are most useful). The majority of the resources in Appendix A provide spatially-linked data which could be used to supply a wealth of site-specific information for defining dust source spatial extent and estimating site-specific dust emission via modeling. Many resources are data-rich and provide data in formats easily compatible with extremely powerful GIS applications. Resources providing information useful for dust source characterization are less abundant. Most information is in the form of conceptual models, case studies, and projects. These resources may be best utilized to create new dust emission models or understanding methods in which existing data and models can be combined to model Category 2 and 3 dust emissions. As the information is mainly conceptual, its use may require a high level of interpretation and understanding from a variety of experts. Information resources necessary for the partitioning of Category 3 dust emissions are also less abundant. As this information type is less well defined, resources include a wide range of conceptual approaches and examples and data. It is likely that many more secondary resources for this type of information (e.g., lists of experts capable of providing input during model creation) could be derived from example resources identified in Appendix A. Dust Definition Feasibility Analysis -36- E N V I R O N

43 Resources are often provided by local, state, or federal agencies at no cost (Appendix A identifies cost per resource). For resources providing spatial data, spatial extent varies from regional (e.g., Mojave Desert) to national levels. Most user interfaces are intuitive and very powerful, with many resources (such as the National Atlas) allowing a cursory analysis of areas via a simplified internet browser map interface similar to GIS software. Other sources, such as software and models, may require substantial expertise or development to apply to the WRAP Dust Definition. Data output formats include text, graphical displays (maps and figures), and tables. Especially for many of the federal sources, GIS files can be downloaded for more precise or custom analyses via GIS. There is no single resource of information capable of providing all the four necessary information types for any dust source in any given area, which means that significant effort could be required to compile and quantify natural dust emissions. The complexity will vary depending on the available resources for a particular geographic region, and it is clear that most applications of the WRAP Dust Definition will require information that originates in many different resources. In addition, substantial expertise may be required to merge this information. For example, whereas air modeling expertise will be required to adapt information to existing dust emission models or develop new dust emission models, successful interpretation and application of many resources may require the expertise of climatologists, ecologists, and/or geologists. For example, most resources are capable of providing information regarding the spatial extent of dust sources, but a multidisciplinary approach would be required to utilize resources to construct new dust models for dust source characterization. Also, the information varies greatly in their applicability to the WRAP Dust Definition due to spatial extent, ease of translation to dust emission modeling, availability, spatial resolution, and/or user-friendliness of user interface. Substantial effort may be required to rigorously evaluate applicability of the available information. Expertise will also be required to identify reference locations appropriate for a given site and to evaluate whether differences between a site and a reference location are anthropogenically influenced. For example, the designation of an appropriate reference location will be based on a variety of specific attributes of the ecoregion (geology, ecology, and meteorology). A conservative and simplistic criterion can be assumed for preliminary screening of potential anthropogenic influence (e.g., a 20% difference between a site and a reference location). However, more realistic estimates of anthropogenic influence may require close scrutiny of individual metrics to account for high natural variability (e.g., wildlife population fluctuations or vegetative cover following an extreme seasonal change, such as a drought). Dust Definition Feasibility Analysis -37- E N V I R O N

44 Example Dust Emission Estimation Dust emissions from animal movement represent one of the most significantly challenging dust sources to quantify. The following example provides a qualitative illustration of the steps necessary to estimate a Category 3 dust emission by wild mule deer (Odocoileus hemonus) movement at a hypothetical site in Figure 4-1. See end of chapter. southern California (Latitude: 33º 45' North, Longitude: 116º West; Fig. 4-1). This example integrates available information resources of Appendix A with the conceptual approaches detailed in Figures 3-2 and 3-3. As in Figure 3-3, there are four key steps to a hypothetical framework for estimating a Category 3 dust source such as mule deer movement: 1. Dust source identification. This step uses geographic distribution of potential dust sources to identify dust sources at a site. For this step, the Smithsonian North American Mammals database was queried using the hypothetical site s latitude and longitude to provide a list of mammalian species at the site that are capable of producing dust via movement. As seen in Figure 4-2, mule deer are likely present at this site. Thus, it can be assumed that one of the Category 3 dust sources present at this site is mechanical suspension of dust by mule deer movement. Figure 4-2. See end of chapter. 2. Dust source characterization. This step is not site-specific, and would be completed prior to site-specific investigations. Characterization of dust emission from mule deer movement would involve constructing a generic model capable of predicting dust emission due to the movements of a mule deer. Constants needed to parameterize the model might include life history data on the spatial ecology of mule deer Figure 4-3. See end of chapter. (average daily movement distance and home range data), size of hooves, and average individual animal mass. Life history data could be obtained from a variety of sources. For example, the results of a query for mule deer from Cumulative Index for the Mammalian Species yielded information about average population size, density, movement, average body weight, and other physiological, ecological, and behavioral characteristics of mule deer (Fig. 4-3). Other information useful in constructing the model, such as the responses of mule deer populations to anthropogenic influences (such as Dust Definition Feasibility Analysis -38- E N V I R O N

45 roads, land use, prohibition of deer hunting, etc.) could be provided by a study of these types of information sources in collaboration with ecological experts. 3. Spatially-explicit dust emission estimation. This step would involve obtaining site-specific data for input into the dust emission model used to estimate site-specific dust emissions associated with mule deer movement. For example, a key piece of information would be the size of the mule deer population at the site, as this would have a large influence on dust production by this dust source. Mule deer population would vary by habitat, land use, and other spatial factors among sites. At a coarse level, population size at a site could be estimated by combining natural history information regarding habitat preference and population density and site-specific habitat information and site suitability. As mentioned above, the Cumulative Index for the Mammalian Species provided information regarding habitat preferences and ranges of population sizes and densities for mule deer (Fig. 4-3). Habitat information for the hypothetical site can be obtained by querying the California Wildlife Habitat Relationship (CWHR) software program using the site s county (Fig. 4-4). Although this Figure 4-4. See end of chapter. software is specific to California, other state- or region-level approaches may also be available. By querying the CWHR software for mule deer, habitat preference for the species in the site s habitat type (desert scrub, Fig. 4-5) can be used to provide information regarding the mule deer population density at the site. Another key piece of information would be the coverage of bare soil areas for the site. Bare soil areas would be extremely susceptible for disturbance (dust emissions) due to mule deer movements. Site specific information on bare soil coverage can be obtained by approaches using hyperspectral satellite data, such as the Eolian Mapping Index database, which predicts bare soil areas susceptible to erosion (Fig. 4-6). GIS data downloaded from the Soil Survey Geographic database could be used to further estimate and quantify the site-specific susceptibility to dust emissions from mule deer movement (Fig. 4-7). Dust Definition Feasibility Analysis -39- E N V I R O N

46 Figure 4-5. See end of chapter. Figure 4-6. See end of chapter. Figure 4-7. See end of chapter. 4. Dust partitioning. Only category 3 dust sources would undergo this step, which partitions sitespecific dust emission estimates between natural and anthropogenic portions. As mule deer populations can be anthropogenically-influenced, dust emissions associated with mule deer movement would be classified as a Category 3 dust source. As with most Category 3 dust sources, a cause-and-effect approach may be the best way to partition natural and anthropogenic dust emissions from the mule deer movements. In many cases, anthropogenic influences would serve to negatively influence mule deer populations, and thus, dust emissions from mule deer movement. For example, it might be expected that mule deer populations near urban areas or major roads would be negatively impacted and that constrained use/movement of mule deer can lead to negative impacts on vegetative cover as the animals exploit more limited home ranges. Figure 4-8. See end of chapter. By examining land uses (such as urban land use) near the hypothetical site (Fig. 4-8), mule deer population estimates could be negatively adjusted in the dust emission model. This could be modeled according to a predefined quantitative relationship between mule deer population density and distance to roads and/or proximity to urban land uses. For this example, this would likely result in a refinement of the overall dust estimate rather than a partitioning between natural and anthropogenic sources, as partitioning for negative effects on dust emissions due to anthropogenic influence would be irrelevant. Partitioning of mule deer movement dust emissions would most likely focus on soil- and surface-specific cause-and-effect analysis. If increased coverage of bare soil was due to anthropogenic causes, such as overgrazing, off-road vehicle activity, or diversion of water from streambeds, it would be expected that a portion of the mule deer movement dust emissions would be anthropogenic. As an example, this can be evaluated explicitly by comparing the Dust Definition Feasibility Analysis -40- E N V I R O N

47 population and vegetative cover in the vicinity of the site to an appropriate reference location. If greater than a specific criterion of difference is observed (e.g., 20%), then anthropogenic influence might be considered relevant and necessitating emission partitioning. This example can be taken one step further such that if a 50% difference is observed, and 20% is considered related to natural variability in the metric, then the additional 30% difference might be considered the anthropogenic influence. The 30% difference could be accounted for within the dust emission model, allowing a quantification of the anthropogenic dust emission portion. It must be noted by WRAP and users of this approach that the actual criteria (20% vs. some other percentage) will vary for a variety of reasons; close scrutiny and rationale should be provided by knowledgeable practitioners to establish these criteria. In some cases, guidelines using a percentage-based approach to identify anthropogenic influences may not be relevant or appropriate. Further illustrating the steps that can be followed, the coverage of unpaved roads for the area could be determined via a site-specific query for the site (using longitude and latitude) using the USGS National Map (Fig. 4-9). Using this estimate of anthropogenically-influenced portion of the bare soils at the site, the portion of the original dust emission attributed to roads could be quantified within the dust emission model. This would yield an estimate of the anthropogenic portion of dust emissions from mule deer movement. Figure 4-9. See end of chapter. Dust Definition Feasibility Analysis -41- E N V I R O N

48 Figure 4-1. Location of example study point in southern California (Resource: USGS National Map Viewer). Dust Definition Feasibility Analysis -42- E N V I R O N

49 Figure 4-2. Results of the query for a list of mammals that could be found at the hypothetical site (Resource: Smithsonian National Museum North American Mammals Database). Dust Definition Feasibility Analysis -43- E N V I R O N

50 Figure 4-3. Life history data potentially useful in constructing a dust emission mode for mule deer (Resource: Cumulative Index for the Mammalian Species). Dust Definition Feasibility Analysis -44- E N V I R O N

51 Figure 4-4. Distribution of desert scrub in California (Resource: California Wildlife Habitats Relationships software). Dust Definition Feasibility Analysis -45- E N V I R O N

52 Figure 4-5. Habitat preference data (habitat suitability values) for mule deer in desert scrub habitat (Resource: California Wildlife Habitats Relationships software). Dust Definition Feasibility Analysis -46- E N V I R O N

53 Figure 4-6. Hyperspectral map of predicted bare ground cover; bare soil areas subject to erosion are shown in yellow (Resource: Eolian Mapping Index). Example map only (location shown in northern Arizona). Dust Definition Feasibility Analysis -47- E N V I R O N

54 Figure 4-7. GIS file depicting soil types present at hypothetical southern California site (Resource: Soil Data Mart, Soil Survey Geographic database). Dust Definition Feasibility Analysis -48- E N V I R O N