A Comprehensive Development Plan for a Multimodal Noise and Emissions Model (2010)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

F-1 APPENDIX F. ALTERNATIVE DESIGN CONCEPTS There were five multimodal model design concepts in the first round. The first concept, which was referred to as the Datum in the first round evaluation, is the design that was in the Wyle proposal; the initial preferred design concept. The remaining 4 are distinct alternatives for the multimodal noise and emissions model design. The design concept papers, which are presented in this appendix, are identically structured with the following sections: Description Executive Summary on the concept drawing distinctions to the other concepts Functional Specifications Description of what this design of the multimodal model will do. Justification Reasons why the design is a good idea by identifying benefits related to the evaluation criteria. Issues Addresses concerns about this design by identifying potential drawbacks related to the evaluation criteria. Design Elements Storyline on the ways and means to achieve this design end state from the current model environment. F.1 Design Alternative #0 (Datum): Step by Step Integration (Initial Preferred Design) The end state is a source (airplane, automobile, truck, marine vessel, etc.) simulation model with benefits evaluator to convert noise exposure and air quality changes into environmental costs. The model will simulate the sound propagation and air pollutant emissions for the moving sources. Rather than initiating a single, large-scale effort to design and develop the end state, the design incorporates a build sequence toward the end state in a series of steps, each step providing an improvement to some facet of the overall model. The build sequence is predicated on giving the users and agencies the tool that they need within expandable system architecture. The model will draw up ongoing model development projects sponsored by the federal government, such as, FAA’s Aviation Environmental Design Tool (AEDT) suite and DoD’s Advanced Acoustic Model (AAM). F.1.1 Description The ultimate requirement for noise would consist of a time-history of the one-third octave band spectrum produced by each vehicle operation. When combined with numbers of operations of the different vehicle types, the model would: F.1.2 Functional specifications ● Calculate any noise metric for any transportation source; ● Propagate sound over any terrain, surface, barrier, structural effects (urban canyon reverberation, etc.) and through any meteorological condition; ● Compute that propagation with a precision that is proportional to the effort spent on terrain/meteorological input (will vary by type of project); ● Include complete and validated transportation sources databases; ● Integrate background noise estimation; ● Offer the level of accuracy that meets or exceeds any regulatory requirement; and ● Provide second-by-second noise.

F-2 For emissions, the model would: ● Predict fuel consumption which would serve as a basis for energy usage (needs to take into account the different fuel types); ● Provide emissions of both criteria pollutants and Greenhouse Gases (GHG); ● Predict emissions by specific modes (e.g., acceleration, takeoff, etc.) and equipment type (e.g., light-duty vehicles, Boeing 737-200, etc.); and ● Provide second-by-second emissions. For Air Quality, the model would: ● Generate second-by-second atmospheric concentrations; ● Be able to model both transport and chemical transformations for characterized pollutants including Hazardous Air Pollutants (HAPs) and particulate matter (PM); and ● Take into account structural effects such as building wake effects, urban canyon effects, tunnels, etc. We believe a simulation model framework is the appropriate form for the end state model. The Federal Interagency Committee on Aviation Noise (FICAN), as a result of its findings and recommendations for modeling aircraft noise in national parks, concluded that “the simulation approach is considered to have the greater potential [compared to integrated models] and it is only a shortage of the comprehensive aircraft acoustic data required, and the higher demands on computing capacity, that presently limit this approach to special applications or augmentation of the more traditional integrated modeling approach.” F.1.3 Justification 1 Currently, Wyle is developing a military aircraft noise simulation model for DOD that will eventually replace the currently-used NOISEMAP integrated noise model. Hence, simulation modeling is a realistic end state. Moreover, we are examining ways of applying the existing noise database currently in NOISEMAP in order to overcome the data concerns mentioned above. Naturally, the introduction is being handled in a phased manner as a comprehensive supporting database is being developed. We envision a similarly phased strategy as we migrate towards the end state model. As an example of the limitations of air quality models that are not simulation-based, the EPA’s CAL3QHC allows for modeling of 1-hour average concentrations near roadways and intersections. However, due to the static mixing zone used in this model (and others of this type like CALINE3 and CALINE4), receptors can not be placed closer than 3 meters from each side of the roadway. As a result, health effects of pedestrians on sidewalks and crosswalks cannot be modeled. Also, microscale modeling of hotspots that take into account the relationship between traffic operations and concentrations is difficult. Wyle developed the Traffic Air Quality Simulation Model (TRAQSIM) that provides an unconstrained spatial and temporal environment that overcomes limitations of models like CAL3QHC. Using a moving source concept under a simulation framework, TRAQSIM provides a vision for a flexible, next generation highway air quality model. Feedback from the potential user communities should provide some insights on the general desirability of the proposed end state. The end state is the reference point that we will use to evaluate the current states of the art in noise and emissions modeling. We will examine the strengths and weaknesses 1 May 12, 2005 letter from the FICAN chair, Mr. Alan Zusman, to Ms. Sharon Pinkerton, Assistant Administrator for Aviation Policy, Planning, and Environment at the FAA.

F-3 of existing noise and emissions models. We will identify viable model design candidates drawing from various model development efforts both here and abroad. As such, we will recommend a model design that has undergone thorough evaluation and thoughtful consideration using structured criteria. Previous and ongoing efforts indicate that a multimodal model is certainly feasible. But, it is essential for the design and development plan for such a model to consider (a) who will use it and (b) how it will be used. F.2.4 Issues Task 1 of ACRP Project 02-09 included a preliminary market research effort to assess the viability and utility of a multimodal environmental model and help in the formulation of the model design and plan. Through widely distributed questionnaire, literature review, and personal interviews; the market research attempted to gather information from the future user communities for the multimodal noise and emissions model. A questionnaire was used to contact consultants involved in transportation planning, state and federal agencies that provide the oversight for these modes, and office staff of regional transportation administrations that organize/fund specific projects. The results were reported to the ACRP panel in the form of a Wyle Technical Note (TN09-01, Preliminary Findings on Future Utility of a Multimodal Noise and Emissions Model). The technical note discusses respondents’ reactions to this design concept. The respondents generally like the all encompassing design of the proposed end state from the aviation perspective. It indicates the potential for a comprehensive evaluation of both noise and air quality parameters, looking at the individual contribution of different modes to a location as well as the total contributions they make. The respondents also raised the following concerns about the design: ● End state as an admirable goal, but not practically achievable as evidenced by failure of other complex model development projects, such as, FHWA’s TRANSIM and EPA’s MOVES. ● Size and complexity of the end state suggest that it will be difficult to use and extremely data hungry. ● The apparent complexity of this model renders it expensive to use and thus prohibitive to all but the largest of airport operators. ● Requires considerable more specialized expertise to use than is done for current studies. ● Challenging to ensure that the model and all the data can reside on a desktop computer offering reasonable runtimes. ● The tool could be used to unfairly penalize modes of transportation if the disparity among noise metrics and impact criteria are not resolved. ● The proposed build sequence just assumes each preceding build worked successfully and does not provide a clear roadmap for full testing of each build. Clearly, achieving the envisioned end state would require a major expenditure of funds and would take many years to complete. Rather than initiating a single, large-scale effort to design and develop the end state, a more realistic approach, consistent with feasible funding streams and practical stakeholder needs, would be to approach the end state in a series of steps, each step providing an improvement to some facet of the overall model. It is important for the architecture of the model to be sufficiently flexible so as to allow for a scalable roadmap towards a future end state. For instance, when Wyle first developed the MAGENTA software back in 1997, little was known about the type and detail of the operations and noise data that would be available as inputs, or of the required outputs of the model. As a result, the architecture of MAGENTA was designed with maximum flexibility to incorporate any input or output, and any noise engine. This has proved to be invaluable to the subsequent development of the model and F.1.5 Design Elements

F-4 its integration with SAGE into AEDT Global. This is the mindset that we propose to adopt in the design of the MDP. Our current thinking on the preferred model build sequence is guided by the phrase: “Think big, start small, and act now.” This model form meets the end state objective and the ground rules that we set forth in the overall approach. The build sequence is predicated on giving the users and agencies the tool that they need within expandable system architecture. The builds and associated rationale are provided in Table F-1. For this preferred model form of build sequences in Table F-1, we believe that the iterative and incremental software development process is the best fit. This approach is actively responsive to the needs of a broad community of transportation planning users. This approach takes the form of a collaborative decision process for choosing what component (GUI, systems databases, and harmonization of inputs, outputs, or modules) should be pursued based on the most pressing user needs and their associated funding levels; while also keeping true to the design goals of the end state. The third column in Table F-1 offers some of our rationale to support our current thinking on the recommended build. The rationale for our current thinking on the model build sequence can be summarized as follows: ● Due to the prevailing “stovepipe” culture, we recommend that the first build provide some basic capability at relatively low cost. A post-processor is justified because it simply takes output from the existing tools and combines it in ways that should be helpful to multimodal planning. We see the first build as an entry-level capability to get the various stakeholders familiarized with the environmental effects of other modes. ● The premise of the second build is that the stakeholders would see utility in what the first build provides and would prefer a tool that is easier to use; thus a shell program. We envision that the stakeholders of ground-based modes (road, rail, and transit) would be amenable to harmonizing noise computation modules after they have had some experience with the first build. The timing for initiating this build would be subsequent to the planned FHWA release of TNM Version 3 in the early 2009 timeframe. ● Our third build draws on the stakeholder experiences in the application of the first two builds with the introduction of capabilities that are intended to enhance the practicality of the model for their typical projects. Based on some of the discussions in the FAA AEDT DRG, we believe that they will want noise and emissions screening tools.

F-5 TABLE F-1 Current Thinking on the Model Build Sequence to the End State Design Description Rationale Schematic 1 Develop a post-processor to combine outputs of the executions of each of the standard noise and/or air quality models used for each transportation mode. Elements of this build:  Produces common outputs, such as, DNL and combined emission inventories.  Includes ability to produce the standard output (metric) of each of the models.  Incorporates feedback loops for iterative assessments, such as, integrated analysis of highway sound barrier design.  The user is responsible for keeping current in the versions of the standard tools. Agencies’ acceptance expected because:  Draws from agencies’ ongoing model development projects.  No invasive changes to existing models  Produces output required by current agencies’ regulations and policies  Cost effective because:  Draws from agencies’ ongoing model development projects.  Allows users to perform integrated analyses using existing noise and emissions engines.  Existing GIS tools can easily facilitate this effort. 2 1. Create a harmonized ground (including marine) noise computation module from the existing tools, which was recommended in the original ACRP problem statement. This build would take advantage of the existing TNM infrastructure, but would include rail (including horns and other warning devices) and marine sources. 2. Develop a shell program to control input preparation, execution, and output processing of each of the standard noise and/or air quality models used for each transportation mode. Agencies’ acceptance expected because:  Draws from FHWA/FRA project proposals.  Other agencies’ noise and emissions model untouched.  Produces output required by current agencies’ regulations and policies. Cost effective because:  Learn from other projects such as EUROCONTROL HARMONOISE and FAA MAGENTA.  Improves ease-of-use.

F-6 TABLE F-1 Current Thinking on the Model Build Sequence to the End State Design (continued) Description Rationale Schematic 3 Construct screening tools to allow primary agency to use the full power of its current model (like AEDT), combined with low-precision versions of the current models for other modes (like TNM). The screening tools (preprocessors) include:  Ground Noise Screener.  Aviation Noise Screener.  Airport Emissions Screener.  Non-airport Emissions Screener. Agencies’ acceptance expected because:  Lessons learned from FAAAEDT DRG discussion about AEDT complexity.  Produces output required by current agencies’ regulations and policies.  Noise and emissions screening criteria would comply with agencies’ requirements. Cost effective because the users can put appropriate level of effort to the environmental area of prime concern (like aircraft) while also assessing the other contributors (road, rail, construction, etc.). 4 1. Incorporate noise and emissions simulation into the system architecture alongside the segmented noise component. 2. Integrate output with APMT benefits valuation block (BVB) requirements for economic impact assessments. Agencies’ acceptance expected because:  Produces output required by current agencies’ regulations and policies.  DOD AAM development has already proven that simulation (NMSIM) and segmented (NOISEMAP) can work together. Cost effective because will learn from DOD AAM project and leverage on the FAA APMT effort; specifically the benefits evaluation block within APMT. This release is likely to serve as a “research model” to evaluate the simulation capability and to await interagency agreement on how to apply impact valuations in environmental assessments.

F-7 TABLE F-1 Current Thinking on the Model Build Sequence to the End State Design (concluded) Description Rationale Schematic 5+ Version 5 begins the process of integrating the ground and air components into a single multimodal model with the objective to have:  Centralized sources database (vehicle performance, noise, and emissions indices).  Common input requirements (where practical) and uniform input processes.  Harmonized computational algorithms (e.g., sound propagation).  Harmonized modules (e.g., atmospheric dispersion).  Unified output processes (e.g., metrics). The priorities for harmonization, unification, and centralization will be based on user needs, agency acceptance, and affordability. This approach depends on an almost interactive design review group (DRG) to understand how the model is being used and what is most needed next. Supporting technical and policy infrastructures are needed to achieve interagency agreement on policies and procedures. This approach also calls for the delivery of workable software in shorter intervals so that new elements can be tested across the user community before field implementation to ensure that the new version meets the needs of the broadest audience. In collaboration with the user communities and agencies, the developers could use evaluation criteria from Task 3 to reach agreement on build priorities End State Dynamic source (airplane, automobile, truck, vessel, etc.) simulation model with benefits evaluator to convert noise exposure and air quality changes into environmental costs.  Meet the emissions and noise assessment requirements (regulatory and policy) of every agency involved in an integrated regional planning process.  Highly modular design so that the model can be universally coupled to a wide range of other transportation planning tools, such as traffic simulation models (e.g. SIMMOD, TRANSIMS, etc.).

F-8 F.2 Design Alternative #1: Build on AEDT The Federal Aviation Administration (FAA) has developed a design tool named the Aviation Environmental Design Tool (AEDT). This tool is actually a suite of programs working together to perform not only environmental impact estimations, but also to allow policy decisions to be made in an informed way. This alternative explores continued development of the AEDT into a true multimodal noise and emissions model for all modes of transportation. This alternative would also include the construction of an environmental study clearinghouse where federal agencies could make available inputs and outputs of past modal studies for assistance in multimodal environmental assessments. F.2.1 Description AEDT is a tool suite that has incorporated both global and local noise and air quality modeling for aviation sources. As shown in Figure F-1, the tool suite has various modules including an economics model and cost and benefit module in addition to the environmental impacts estimation module. The environmental impacts estimation module is based on four proven (nationally and internationally) noise and air quality models: the Integrated Noise Model (INM), Magenta, Emission and Dispersion Modeling System (EDMS), and SAGE. The noise models, based on the integrated approach, allow for a wide range of outputs including a range of metrics for A- and C-Weighted, and tone perceived levels, and an approximation for time above outputs. For air quality, all criteria pollutants plus carbon dioxide and speciated hydrocarbon outputs are available. F.2.2 Functional specifications Figure F-1. AEDT tool suite overview. Source: Fleming, G.G., Aviation Environmental Design Tool (AEDT), Presentation at the 22nd Annual UC Symposium on Aviation Noise and Air Quality, Prepared by U.S. DOT/Volpe Center, March 5, 2007. The models used are very detailed including spherical spreading, atmospheric absorption, terrain shielding, lateral attenuation, and ground effects for noise propagation calculations. This would provide a solid platform for modeling the noise from other modes of transportation. The database would have to be expanded to include reference emission levels for other modes of transportation. For air quality, a detailed

F-9 emission inventory process is included and local dispersion based on the accepted air quality dispersion model, AERMOD. AERMOD has been used for many sources, is now being considered by FHWA for motor vehicles, and as such could allow for a single air dispersion model to be used. Movements of aircraft are well documented and included in AEDT. This includes simulation of vertical profiles by aircraft type and allocation of aircraft per runway, taxiway, and gate. Movement of other vehicle types would have to be included to account for all modes. The model has both local and global capabilities in predictions. Subsets of the global capability for emission inventories can be used in regional analysis. Extensive resources have been expended by FAA to build the AEDT for aviation sources. This initial expenditure has built a strong base for the aviation sources which would significantly reduce the cost in comparison to other options since implementation time would be greatly reduced and only an expansion of the model would be required. The model has also overcome a large hurdle in that air quality, noise and economic considerations have been considered and integrated into this single model. While the data base sharing would need to be improved and would need to be expanded for other modes of transportation other than aviation, AEDT would provide a platform for inclusion of the other modes of transportation. Additionally, the models used in AEDT have been promulgated and accepted by EPA for air quality and the noise model is accepted on an international basis. Implementation for the other modes of transportation could be done with the accepted modeling processes as well, again reducing the time requirements since these models have been previously accepted by other agencies. F.2.3 Justification Local modeling is accomplished with established, completely developed models (Emission & Dispersion Modeling System (EDMS) using AERMOD for dispersion and the Integrated Noise Model (INM)) so that no extensive validation is needed for implementation of other sources than aviation. Again, use of established models for the other transportation sources would significantly reduce implementation time which would result in a substantial cost savings. The model is designed to allow quick changes to input allowing mitigation and future projects to be analyzed efficiently. Global modeling is accomplished with the already existing models for aviation which include SAGE for air quality and MAGENTA for noise. While this may not be needed at this time, it could be expanded for other sources allowing climate change impacts to be evaluated. Additionally, regional modeling could be based on the same process by selection of a subset of the global database. The modular design would allow for easy inclusion of other models so that other modes of transportation could be included without extensive redesign of the basic model platform. Rail, water and highway would have to be included as modes, but again, could be done in the modular design. The database design would allow other reference levels for noise and emission levels for air quality to be included in the same way, again without extensive changes to the system architecture. The advantages are considerable and include: ● Use of established models so that development is not needed nor is the long validation process; ● The database design allow inclusion of modes of transportation in the same way so that model components can be reused and similar; ● Inputs can be shared so that repetition is not required; ● The local and global modeling will work together; ● Algorithms can be repeated, leading to more streamline design; ● Global and regional modeling could be done by model expansion for other modes; ● Future expansion is enhanced because of the modularity of the system; and

F-10 ● The model is easily adapted for other cases allowing mitigation and future enhancements to be analyzed using the same model structure reducing cost and time of implementation. Expansion of AEDT would require establishment of a common data base, inclusion of new transportation modes, and should be initially expanded in some areas such as counting time above events. Even so, it could be implemented faster than the datum alternative. However, while this end state reduces work effort and includes all in a single output, it does not result in increased fidelity, it does not lead to advancement in the understanding of the phenomenon it describes, nor does it lead to increased flexibility in use. It does not represent the most advanced noise modeling possible, nor does it allow for easy adaptation of algorithms to calculate additional supplemental metrics. Accuracy and usability remain static. F.2.4 Issues To overcome these problems, expansion to the system design, further advancement of the models being used, and development of new reference levels/emission levels would be required. These are not trivial changes but in addition to the inclusion of other modes of transportation listed above would result in a more advanced model. Unfortunately, this would not result in the flexibility offered by a true simulation model. During this expansion of AEDT, other problems would need to be overcome such as a common noise metric that could be expanded for the various metrics now required by various agencies. This assumes agencies would be resistant to change to a single metric making the base metric that and post- processing necessary. Again, this can be done in the platform as provided but requires significant programming. Another issue would be the advancement of the way various vehicle movements are modeled. Changes are needed since AEDT only works for aviation sources at the current time and in the case of emissions has a simplistic approach to motor vehicles. Inputs for trains and water craft would have to be included. Algorithms that could be shared with all modes would need to be implemented to make a streamline, flexible model. Overall model efficiency needs to be increased by the using the same algorithms, ideally for both the noise and air estimations, whenever possible. For example, the same sound propagation algorithms could be used regardless of source and the same air quality dispersion algorithms could be used. Instead of simple call statements to models in various programming languages, this streamline approach would require all models to be in the same language and porting of models not in the chosen computer language. Again, this is not trivial but should result in increased speed for the model. Finally, additional improvements in model flexibility must be considered. This requires addition of sources not currently in the models and they must be included in a consistent manner. Also, the overall interaction of the outputs from the various modes must be combined. For example, contours produced by INM would be changed when other modes were present and distances from the highway and rail where impacts existing would be increased requiring addition of noise levels from the same sources, again in the same metric. The same is true for air pollutants. In sum, a simple inclusion of exiting models, or the model data bases, into AEDT would result in time and cost savings to produce an overall transportation model, but no major advancement to the analyst’s tools. For this to occur, many non-trivial changes would be needed as described. The AEDT architecture is based on the “common thread” approach. This allows a streamlined model and modular inclusions. This would work well for the implementation of other modes of transportation. The implementation could be done in a near-term process to get to a useable model quickly, then advancements made in the mid-term, and finally major advancements made in the longer F.2.5 Design Elements

F-11 term. Table F-2 includes the major tasks and research efforts that would be required. Of course many alternatives are possible and Table F-2 is only one possible path. Getting to the final end state in the near-term would be straight-forward, but not trivial. Models for other modes of transportation would need to be included. The data bases would need to be included in a consistent manner as AEDT now operates. Noise propagation and air quality dispersion models would most likely be left out of the model import, using the common thread approach in AEDT. Sharing of algorithms would be crucial for flexibility and streamlined design. In the mid-term, advancements based on research during the near term could be implemented. Sharing of databases is crucial to the model and better ways to do this could be explored. Updates to the databases could also include advancements as they occur. For example, particulate matter estimation is advancing at a rapid pace and could be updated. Also in the mid-term, improvements in single metric analysis, post processing advancements, and improvement to the movement data base could be implemented. Care in implementing the movement algorithm advancement could be done in such a way that it could lead to true simulation modeling in the longer-term. User flexibility and improvements to user friendliness could also be accomplished in this time frame. Noise and air quality estimation is still very dynamic field. In the longer-term, advancements in modeling and data are sure to occur and could be implemented since the modular platform would be easily adapted. Additionally, Federal agencies needs and desires could change over this time period. For example, climate change modeling most likely will become more important in the longer-term. The flexibility of the modeling process would allow these changes to be implemented as well. Flexibility also extends to the manner in which multimodal noise and emissions assessments are conducted. Preliminary reactions from potential future users provide some insight on additional actions that the federal agencies could take to promote effective environmental assessments. Task 1 of ACRP Project 02-09 included a preliminary market research effort to assess the viability and utility of a multimodal environmental model and help in the formulation of the model design and plan. A questionnaire was used to contact consultants involved in transportation planning, state and federal agencies that provide the oversight for these modes, and office staff of regional transportation administrations that organize/fund specific projects. The results were reported to the ACRP panel in the form of a Wyle Technical Note (TN 09-01, Preliminary Findings on Future Utility of a Multimodal Noise and Emissions Model). Respondents are concerned that the new model would significantly increased study costs; making it prohibitively expensive for anything smaller than a regional study. They suggested that federal agency approval is important and specific guidance on when the model is to be used in the environmental process is needed. One of the respondents’ suggestion to improve study efficiency is the “ability to automatically grab via the internet project required available databases – census data, current fleet mix and operations by airport, state highway traffic data, radar data for identified time period, etc. Current methods require considerable user time and effort to collect the necessary available input data that are not project specific.” This design incorporates and expands upon this suggestion with the inclusion of action to create a federally-sponsored clearinghouse of transportation environmental study data (inputs and outs) accessible to the public. There is already federal precedent for making transportation environmental study data available to the public. For example, the Vision 100-Century of Aviation Reauthorization Act (Public Law 108-176) required the FAA to “make noise exposure and land use information from noise exposure maps [prepared under 14 CFR part 150] available to the public via the Internet on its website in an appropriate format.” FAA met the requirement with the creation of a website with links to airport noise and land use information pages and copies of the noise exposure maps. The site is:

F-12 http://www.faa.gov/airports_airtraffic/airports/environmental/airport_noise/noise_exposu re_maps This design proposal extends the precedent to the other modal agencies and suggests centralizing the study information at a single internet clearinghouse under the auspices of the appropriate federal agency, such as the Office of the Secretary of Transportation (OST). Since the intent is to provide data that would be useful in multimodal studies, the agencies need to establish standards for the type of data to be placed in the clearinghouse. For example, they need to agree on the geographic information system (GIS) for the management of the various input and out data including roads, railways, waterways, runways, flight tracks, meteorological data, computed noise contours, computed noise grids, pollutant concentrations (monitored and computed), etc. The agencies would also need to provide guidance on how the GIS data is to be used; including reaching a meeting of the minds on the common metrics to use in multimodal noise and air emissions studies. The endeavor to create the data clearinghouse is a major activity to occur in parallel with the AEDT expansion. In addition to the technical tasks in establishing an internet clearinghouse; regulatory and policymaking activities would need to be completed to establish the requirement to gather study data from mandated studies, such as Environmental Assessments (EAs) and FAA Part 150 studies, and standardize the data format and collection method. This new aspect also places the need to build into the AEDT expansion a previous study data integration module to extract necessary input and output information from the data clearinghouse. The objective is that a user would have automatic access to any previous transportation environmental study, such as, FHWA highway construction EA or FAA Part 150, to incorporate into a multimodal study covering the same geographic region. TABLE F-2 Key Milestones in the Development of the AEDT Expansion End State Implementation and Study Data Clearinghouse Track R&D Track Near-term (3 year period) • Establishment of interagency exploratory team for the development of an environmental study clearinghouse. • Implementation of post-processing modules to permit output as required by the various agencies. • Expansion of GUI to allow additional data to be included • Agreement on the architecture of the data clearinghouse • Implementation of initial model version based on existing models without use of similar algorithms. • Notice(s) of Proposed Rulemaking (NPRM) by the modal agencies to require input and outputs from federally mandated environmental studies. • Research task force and work program to develop listing of key parameters and databases for model advancement. • Development of expanded GUI and database format to include additional modes of transportation in a more exacting manner with error checking. • Development of movement data base by mode for other forms of transportation. • Development of single metric output of models to permit change to any form needed (noise). • Development of a plan for expansion of other sources than aircraft for global/regional modeling. Mid-term (3 to 6 year period) • Revision to changes in architecture, components, and data elements of the study data clearinghouse. • Implementation of generic algorithms for all sources to allow single metric analysis with post processing to permit any metric needed to meet agency requirements for noise and same dispersion algorithms for air. • Implementation of improved GUI • Initial prototype of streamlined model for testing on the Internet. • Initiate final rulemaking process(es). • Advanced movement data base development (all moving sources controlled by single algorithm). • Advanced data base control techniques. • Improved GUI control. • Updates to noise emission levels and air quality emission factors. • Development of global/regional modeling data for sources other than aviation. • Integration of cost and economics model for all sources. • Initial review of including refraction effects for noise.

F-13 TABLE F-2 Key Milestones in the Development of the AEDT Expansion End State (concluded) Implementation and Study Data Clearinghouse Track R&D Track Longer-term (after 6 years) • Implementation of advanced movement data base which could include simulation components • Draft agencies’ policies and guidance for study data submission. • Interagency agreement on clearinghouse operation and maintenance. • Implementation of advanced data base control features (relational data base features). • Improved emission level and factor implementation including global/regional modeling capabilities for sources other than aircraft. • Implementation of cost and economic modules for all modes. • Exploration of hybrid options going toward a true simulation model. • Research of implementation of advanced dispersion analysis (puff modeling and regional modeling). • Refraction effects included for noise analysis. End state • Promulgation of final rule. • Implementation of study data clearinghouse. • Advanced version of AEDT for all modes of transportation based on changing modeling practices, new data, and agencies needs/desires. • Inclusion of more advanced climate change modeling capabilities • Research version of the end state based on advanced propagation models for noise (including refraction effects) and dispersion models for air (puff modeling). F.3 Design Alternative #2: Build on Existing Simulation Models This alternative design proposal outlines an approach to the development of a multimodal noise and emissions model centered on time-based simulation of source movements, source emissions, and propagation scenarios resulting in detailed output reports at receptor locations. The end-state of the model will functionally be the same as other design alternatives resulting in time-based simulation, such as the Datum. F.3.1 Description This proposal suggests a multimodal model development plan should be founded on existing single transportation mode simulation model implementations. Research and validation reports of outdoor noise and emissions algorithms are abundant both domestically and internationally. Fostering these efforts – which include studies of both heuristic and simulation approaches – will result in a model more scalable, accurate, and usable than one tethered to legacy approaches and limitations. This model provides detailed, time-varying propagation results at receptor points from which any standard or supplemental metric may be calculated for noise and emissions during the design process. This is accomplished by simulating ground, marine, and air traffic environments with discrete moving or stationary sources with detailed fundamental source characteristics. F.3.2 Functional specifications Propagation of noise is simulated via ray-tracing techniques and dispersion of pollutants is simulated via Gaussian puff dispersion algorithms when appropriate. These approaches are generally considered to be the current state-of-the-art with regards to trade-offs of analytical proficiency with respect to analysis scenarios, implementation feasibility, and computer processing requirements. For example, the US Department of Defense has already taken steps to develop and deploy a simulation model for military aviation noise – the Advanced Acoustics Model (AAM). This was a result of the recognition of a simulation model’s superior capability to handle routine scenarios in addition to the

F-14 identification of the need to model high performance aircraft in a more appropriate, realistic, and accurate manner. By defining source emanation characteristics in terms of first principles – such as spectral directivity for noise and modal fuel-burn rates for emissions – the model’s propagation algorithms may be varied according to an analyst’s scenario. For example, an analyst may perform a sophisticated environmental simulation analysis requiring careful input parameters or he may choose to default certain elements of the analysis to heuristic approaches or previous studies where appropriate. The foundation of the model’s output is a report at a receptor including time-dependent spectral noise and emissions metrics. Aggregation of these reports provides a means to not only calculate any integrated metric on a grid, but transient supplemental metrics such as number of events and time above ambient. Additionally, the inclusion of the time dimension in the output reports allows spatial and time dependent visualizations of the environmental simulation. The Federal Interagency Committee on Aviation Noise (FICAN) has already recognized simulation noise models as having the most potential for accuracy and precision in situations requiring sophisticated analysis. Examples of the adoption of noise simulation include the National Parks Service’s adoption of NMSim (Noise Model Simulation), the development of AAM (Advanced Acoustics Model) through SERDP (the US Department of Defense Strategic Environmental Research and Development Program), and the adoption of RNM (Rotorcraft Noise Model) by NASA and NATO as the de-facto standard for outdoor noise propagation helicopters and tilt-rotors. International credibility of this approach is bolstered by the fact that the European Commission has undergone a multinational research and developmental effort resulting in algorithms and technical guidance for using a harmonized ground and air noise source and propagation methodology known as IMAGINE (Improved Methods for the Assessment of the Generic Impact of Noise in the Environment). F.3.3 Justification Domestic examples of noise simulation models also exist for traffic and railway noise. CNM (Community Noise Model) simulates five motor vehicle types of sources (autos, medium trucks, heavy trucks, buses, and motorcycles) as well as multiple rail engines with trailing cars plus stationary sources such as compressors and rail yard activities. The Florida Department of Transportation funded development of a true noise simulation model called FRM (Florida Rail Model), which applies ray acoustics to rail and limited community sources with reference levels capable of being adjusted at each time step to account for vehicle operational mode and type. In terms of air quality, the U.S. Environmental Protection Agency has adopted CALPUFF – a CARB (California Air Resources Board) air-quality dispersion model – as its preferred model for assessing long range transport of pollutants and proposes its use on certain near-field applications involving complex meteorological conditions in its Guideline on Air Quality Models. Other air quality simulation models exist such as TRAQSIM (Traffic Air Quality Simulation Model, developed as part of a PhD dissertation at the University of Central Florida) and HYROAD (developed by Systems Applications International, Inc., under sponsorship of the National Cooperative Highway Research Program). HYROAD uses a particle in cell approach for close in receptors and Gaussian plume for farther away receptors. Both of these models use CALPUFF as their calculation engines, but provide modules to simulate movement of discrete sources with appropriate modal movements and puffs associated with each mode for each time step. Justification for incorporating these air quality and noise models into a simulation model capable of handling multiple modes of transportation lies in the fact that simulation modeling has already been proven to be more accurate and will provide a step forward in environmental modeling for analysts of all agencies. Considerable advantages include: ● The use of sophisticated algorithms to most accurately predict results at points of interest;

F-15 ● Building of current simulation models will circumvent developmental constraints caused by legacy approaches of lesser fidelity; ● Ray tracing algorithms for noise can be applied to any source regardless of transportation mode; ● Proper inclusion of meteorological effects, terrain, and other heterogeneous scenarios; ● Sufficient detail in the output will provide thorough understanding of any scenario; ● Inputs provide accurate representation of sources based on first principles rather than assumptions or calculated metrics (as is the case with AEDT); ● Knowledge and validation from existing simulation models will streamline development; ● Updates to propagation and dispersion algorithms can be independent of source definitions; ● Sufficient detail in output will allow any standard or supplemental metric to be calculated; ● Existing tools that model source movements may be used and tracks may be translated into time- varying spatial and conditional source trajectories; ● The main drivers for noise and emissions, such as acceleration and power setting, can be directly listed or inferred from a sufficiently detailed trajectory file; ● Potential exists for a harmonized source definition file to contain noise and air quality data together as well as rules for interpolation and extrapolation thereof; and ● The ability to define multiple emissions components emanating from a single source for a single mode (such as separate definitions for both the main and tail rotors of a helicopter). A desire of federal environmental agencies is the capability to simulate most accurately as possible the air and noise pollution affecting wildlife, vegetation, and humanity. The most accurate solution is not always the most easily constructed. A simulation model with advanced logic could be costly to develop, but it would pay dividends by providing a means to an end-state as accurate and scalable as present technology allows. F.3.4 Issues Sophisticated modeling requires the existence and development of an adequate source database. Simulation modeling of noise, for example, requires source directivity patterns and reference levels for each vehicle and vehicle condition. Obtaining such data may involve costly measurement protocols or an acceptable method of converting an existing database into one which may be applied to a new model. Another issue is the computing power required to simulate detailed scenarios in a time-dependent fashion. Current runtimes for an analysis may increase by an order of magnitude. Additionally, complex analyses may require more extensive input. Simulation modeling, while providing the potential for increased accuracy and fidelity of output could be a burden on an analyst and lead to a process for environmental analysis that is more costly and time consuming. Some users may find a full-blown simulation analysis environment too complex to use for standard environmental analyses; this may be more of an issue for one particular transportation mode than another. Most air and noise emissions problems have three fundamental drivers – the source’s movement and operational states, the paths and scenarios of propagation, and the receiver locations and metrics of interest. In terms of noise, for example, the lowest common denominator for an output report is time dependent spectral data at points of interest. Only by reporting results as a function of time may realistic scenarios be modeled and may all metrics be calculated. From this output, one may deduce such standard noise metrics as L F.3.5 Design Elements dn in addition to supplemental metrics such as Number of Events and Time Above

F-16 Ambient. A time history of air quality indices such as NOX, CO2 Consideration at all times and frequencies of the three fundamental drivers could lead to a burden on the user for extensive input. Additionally, a true-to-life representation of every source would literally be impossible – there are too many unpredictable trends of spectra and in traffic patterns for an intersection, for instance, to be fully simulated a priori. , and HC allows an analyst to focus not only on the worst time period of a day, but also hone in on less polluted time periods or locations to which traffic patterns and mitigation efforts may be appropriately shifted. For these reasons, careful development and an analyst’s use of such a model must pay special attention to logical implementation of physics-based simulation algorithms. Rather than brute-force calculation of all ray paths and trajectory steps independently (the current simulation modeling state-of- the-art) internal calculation and database engines must know when a detailed analysis need be done given scenarios tied directly to mathematical functions to prioritize error and run time. This means both detailed and nominal source data and propagation algorithms would be leveraged by both a user and the model itself. The user may decide how accurate the results should be, and the model would document this accuracy in terms of the trade-offs and methodologies of its processes. For instance, consider noise simulation modeling of a highway. Traffic patterns can be sporadic and, although modal tracking data may be obtainable, historical data synchronized with real time are only of interest to emissions scientists. Emissions planners, of course, must look to the future. Using yesterday’s traffic data to model tomorrow’s emissions will never amount to anything more than educated guessing. Stochastic source movement algorithms such as those used in TRAQSIM would be leveraged for each source type to describe typical modes of operation. An environmental planner would want to be able to input some simple items, like traffic counts, time of day, fleet ratios, etc., and a time-based simulation of cars on a freeway would be run using semi-imperial traffic input data with programmed randomness, or jitter. Applying adjustable jitter to data sets and running simulations with multiple source movement and propagation class scenarios would provide results that deviate from the mean. It is important that an environmental analysis tool be able to transparently represent the limitations and applicability of its results by giving the user appropriate virtual solution boundaries representative of the inherent randomness of realistic variables such as traffic patterns and weather. For example, if enough point sources exist in a line, the sound wave from each point source will merge with adjacent sources’ spherically spreading waves and propagation will then cylindrically spread. As this limit is approached with a steady stream of traffic, environmental traffic noise levels tend to remain relatively constant with respect to time. As such, a heuristic approximation may consider this behavior as an ambient noise level for an area. On the other hand, fidelity may be lost in calculation of heterogeneous effects on emissions due to factors such as terrain, weather gradients, and turbulence. Rather than a user remaining ignorant of what is and is not is factored in, he may choose to model these effects by forcing point source propagation for each vehicle rather than allowing any heuristic approximations. The results of such a run would contain not only a grid with time-relative data, but also metadata documenting its accuracy and potential interpretation for use in planning for other modes of transportation. Moving straight to simulation modeling requires a lengthy developmental phase and, once ready for release, a complete change in modeling techniques from the current state of the art. The potential for a lengthy developmental phase for the new models means, rather than incremental improvements to existing analysis tools, the existing tools must be used as-is while the simulation model is being developed. Once deployed, training workshops to demonstrate how and why simulation modeling is to be used may be required to get analysts and engineers up to speed. Table F-3 outlines key milestones for the development of this multimodal design alternative.

F-17 TABLE F-3 Key Milestones in the Development of this Model’s Multimodal End State Operational Track R&D Track Near-term (1-2 years) • Agency guidelines on Environmental Assessments (EA) using time simulation • Continue using existing models for all EA’s • Train current environmental analysts on the utility of using simulation-based approach • Identify current state-of-the art modeling techniques, algorithms, and data sources • Evaluate utility of currently implemented simulation models • Establish methods for conversion of existing noise and emissions source databases • Develop numerical methods of propagation • Develop GUI for source movement and data input by transportation mode • Implement above R&D into a usable multimodal model of the highest fidelity Mid-term (3-4 years) • Train analysts on using simulation-based approaches • Use full simulation techniques to augment EA’s where data is available • Convert legacy source databases into new format • Agency guidelines on required levels of accuracy for EA’s • Identify heuristic algorithms available for implementation • Establish levels of accuracy for heuristic algorithms • Develop GUI that provides heuristic options as screening tools for auxiliary transportation and emissions modes • Merge noise and emissions source definitions of a source into one data file with interpolation and extrapolation rules for modes of operation • Implement stochastic movement and propagation algorithms for multiple-scenario modeling Longer-term (5-6 years) • Use agency guidelines, new source database, latest simulation modeling for EA’s • Utilize stochastic movement and propagation algorithms and report EA findings in terms of multiple possible scenarios • Develop more computationally efficient algorithms • Establish level of accuracy as a function of model input, assumptions, and user fidelity options End state • A simulation model for sophisticated analysis with heuristic approaches for screening tools and scenarios with lower fidelity requirements • Evaluate current propagation algorithms utility versus more advanced algorithms that may be implemented F.4 Design Alternative #3: Federal Adoption of Commercial Software The concept promotes a market-based option for the development of the multimodal noise and emissions model. Commercially designed software has been leveraged by engineers and designers of all disciplines to provide an efficient and documentable path to solutions of problems ranging from the simple to the complex. Commercial software is already available to noise and air quality engineers. This document focuses on two such software packages. One is maintained by the German company Braunstein + Berndt GmbH and is named SoundPLAN. The second is CadnaA, the product of another German company – DataKustik. F.4.1 Description This document does not determine which of CadnaA and SoundPLAN is the best commercially available software package. Rather, the purpose is to introduce elements of the idea that models sold commercially to the public domain could be adopted, regulated, validated, and provided with developmental assistance by the federal government. The SoundPLAN and DataKustik names imply an emphasis on noise modeling and mapping. However, the developers of the models have acknowledged the utility of incorporating modules for air F.4.2 Functional specifications

F-18 pollution evaluation along with their existing, highly flexible noise modeling frameworks. This incorporation was presumably a relatively straightforward task, considering the modular structure of their software architectures. The user interfaces act as hubs for calculation modules and other tools. For example, the SoundPLAN Manager copes with the organizational aspects of projects to allow for common resources such as geographical, meteorological, source definitions, etc. to be shared amongst multiple solution scenarios of varying fidelity and/or scope. The existing implementations of SoundPLAN and CadnaA have many commonalities amongst each other, and their modular approach is similar to that of the Datum. These software packages may already be considered multimodal; the breadth of modules provide access to a multitude of accepted, documented, and/or preferred methods of noise calculation for road, rail, aviation, parking lot, and industrial sources. This allows an analyst to compare different methodologies and even benchmark current states of the art versus legacy calculation methods. Each software package provides air pollution dispersion models. SoundPLAN provides both a Gaussian dispersion module as well as a more accurate non-hydrostatic flow and dispersion model. CadnaA utilizes a Lagrange particle model that processes time-dependent emissions. Exhibit F-1 outlines various modules of calculation for each software package, as compiled from the SoundPLAN and CadnaA evaluation tables assembled in Task 1 of this project. These commercial platforms are widespread, in spite of the cost burden of the end-user’s license agreement. The modular design allows competitive pricing to be maintained by offering customers customizable software packages. The SoundPLAN distribution package contains all the modules, and a customer may purchase license codes to unlock modules intended for use. These tools are already being used, and may be considered multimodal with respect to their ability to handle both air and noise emissions from multiple types of sources. Their position in the marketplace proves their utility. This is especially true in Europe, where some of the available modules satisfy certain government requirements for mapping. Companies and analysts in the United States have chosen to use commercial environmental software packages for various reasons, including the flexibility and portability of the input and output data, access to various calculation methods, and the built-in features for mapping and report making. F.4.3 Justification In the process of establishing their positions in the marketplace, competitive software platforms implement more advanced utilities as requested by the user community or as required to keep up with the competition. Modern operating systems now have application programming interfaces (API) that give access to powerful computational tools such as multiple-core processors, parallel processing on a network, and 64-bit system architecture. SoundPLAN and CadnaA take advantage of such API. Other utilities include commercial software’s ability to import and export to many standardized or proprietary file formats. With these, a user may now leverage other software packages within his workflow. For example, SoundPLAN can read in a noise directivity file written to the Common Loudspeaker Format (www.clfgroup.com) and use it in a standard noise calculation. CadnaA possesses the ability to read in projects saved from within SoundPLAN. Modules include the required geographical database and other optional components for definition of source properties of multiple modes of transportation; customizable propagation and transmission calculations founded on multiple internationally accepted standards; mapping and visualization utilities; data importing, exporting, and manipulation tools; and there even exists screening calculators for quick access to estimates and mitigation design cost optimization tools for noise barriers.

F-19 Existing Calculation Methodologies and Accepted Standards Implemented Exhibit F-1 Within SoundPLAN and CadnaA a. Road Noise Module Name Availability Further Information such as Associated Country Acceptance / Validation SoundPLAN CadnaA RLS 90, VBUS X X Germany DIN 18005 X X Germany CoRTN X X United Kingdom – “Calculation of Road Traffic Noise” Statens Planverk 48 X Nordic Road Noise Prediction FHWA X Federal Highway Model NMPB-Routes-96 X X France, EC Interim STL 86 X Switzerland Nord2000 X Nordtest method NT ACOU 107 Czech Method X Czech Republic TemaNord 1996:525 X Scandinavia RVS 04.02.11 X Austria RVS 3.02 X Austria b. Rail Noise Module Name Availability Further Information such as Associated Country Acceptance / Validation SoundPLAN CadnaA SCHALL-03, Schall Transrapid, VBUSch X X Germany ONR 305011 X Austria DIN 18005 X X Germany CRN 99 X X United Kingdom – “Calculation of Rail Noise” Ö-Norm S 5011 X Austria RMR, SRM II X Netherlands, EC-Interim RMR 2002 X Netherlands SEMIBEL X X Switzerland NMPB-Fer X France NMT 98 X Nordic Prediction Method for Train Noise Kilde Report 130 X Nordic Rail Prediction Method Japan Narrow Gauge Railways X Japan ÖAL 30 X Austria TemaNord 1996:525 X Scandinavia Nord2000 X Nordtest method NT ACOU 107 FTA/FRA X USA

F-20 Exhibit F-1 Existing Calculation Methodologies and Accepted Standards Implemented (continued) Within SoundPLAN and CadnaA c. Aviation Noise Module Name Availability Further Information such as Associated Country Acceptance / Validation SoundPLAN CadnaA AzB X X Germany AzB-MIL X Germany AzB 2007 Draft X Germany AzB (free) X Germany AzB-L (revision from 1997) X Germany DIN 45643 strict X Germany DIN 45643 (free) X Germany DIN 45684 X Germany LAI-Landeplatzleitlinie X Germany ÖAL 24 X Austria ECAC Doc 29 X X International, EC Interim d. Industrial Noise Module Name Availability Further Information such as Associated Country Acceptance / Validation SoundPLAN CadnaA VDI 2714 X X Germany VDI 2720 X X Germany ISO 9613 X X International ÖAL Richtlinie Nr. 28 X Austria BS 5228 X UK Ljud från vindkraftverk X Sweden General Prediction Method X X Scandinavia HARMONOISE X International - EC CONCAWE X X International - EC e. Air Pollution and Emissions Module Name Availability Further Information such as Associated Country Acceptance / Validation SoundPLAN CadnaA TA-Luft X Gaussian dispersion model from smoke stacks AUSTAL2000 X Lagrange particle model – time-dependent emissions MISKAM X Fulfills German VDI 3782 /8; tested in wind tunnel; 3-D non-hydrostatic flow & dispersion model

F-21 Exhibit F-1 Existing Calculation Methodologies and Accepted Standards Implemented (concluded) Within SoundPLAN and CadnaA f. Other Modules Module Name Availability Further Information such as Associated Country Acceptance / Validation SoundPLAN CadnaA DIN 18005 X X German Parking Lot Noise Utility RLS 90 X Parking Lot Noise Utility Bavarian Parking Lot Study X VDI 3760E X Indoor Noise The Indoor Factory Noise Module Calculation Method X Indoor Noise Air Absorption via 3 unique standards X ANSI 126, ISO 3891, ISO 9613 Part 1 Screening Tools Long Straight Road, City Noise Screening The federally accepted emissions models are behind the times because they are driven by modal policy and their code and interfaces are written by scientists and engineers rather than software developers. Professional developers have the ability to take off-the-shelf API and get them to operate effectively and efficiently. Specialized scientists and engineers do not necessarily have the training or expertise to effectively develop complex software architecture. Advantages include: ● Professionally developed software; ● Modular structure allows for alternative algorithms and independent updates; ● Already multimodal with respect to both multiple modes of transportation and the availability of both noise and air quality analysis tools; ● Integration of input and output data with other commercial software; ● Modules may be updated independently and new technology may be easily inserted; ● Many analysts are already familiar with the software; ● Competition for federal adoption will fuel development; ● A federal advantage is the burden of developmental cost is shifted away from federal agencies and towards commercial entities; and ● Specific modules may be licensed separately to maintain reasonable prices. The fundamental issues associated with the government’s adoption of commercially developed models will likely stem from the balance of a company’s desire to maintain propriety of their software with the government’s desire for transparent analysis results. A company may not want to disclose certain aspects of their product that could help position their competitors in the global marketplace. F.4.4 Issues Another issue will be the cost to the users. Depending on modules and customer service options, current prices for commercial engineering tools can be in the range of $20,000. This is a lot of money to ask a small environmental firm to pay, and it is not clear if an appropriate cost-sharing precedent has been

F-22 set based on a federal mandate to use a commercial product for an environmental analysis. However, it may be possible to glean some guidance from software licensing of models in other realms of analysis, such as SIMMOD – an airport and airspace simulation model validated by the FAA and maintained by ATAC. In the examples of SoundPLAN and CadnaA, the commercial models do not include U.S. databases of source data (noise levels, emissions factors, etc.). SoundPLAN contains a very limited source database; expecting the user to supply that data. CadnaA contains noise and emissions data obtained from various European environmental agencies. Allowing software design companies to periodically compete for adoption of their product by the government may require an analyst to purchase new software periodically and also learn how to properly use new software releases. This justifies providing the opportunity for renewable licensing and support contracts, but without careful consideration of contracts implementation difficult situations may arise, such as the proper course of action should a software company go out of business while analysts throughout the county are relying on them for support. Similar implications may result from changes to a software company’s business model such as management restructuring or a decision that continuing to develop or support federal environmental models is no longer a good business venture. While federal regulations and requirement for environmental analyses tend to remain static for long periods of time, government funded and independent research will always advance the state of the art. The current balance of regulatory requirements and emerging technology directs the development of commercial software packages in such a way that they remain marketable to various users. For the federal government to adopt a commercial software package as an accepted medium for analysis, software developers will be required to implement mandated standards for calculation of noise and air emissions. The federal agencies also determine and approve the source data to be used. F.4.5 Design Elements Several software developers may propose their packages and compete for federal adoption. Once a package is chosen by the government, the developers of this package will benefit from the federal requirement that accepted analyses must have been calculated with a licensed version of the federally accepted commercial model. Changing federal requirements along with emerging technologies will result in a periodic reevaluation of the chosen model and the opportunity for other software companies to develop competitive packages and propose their use. The period of time for which a certain commercial package is adopted by the government will depend on the package’s current utility compared to the existing state of the art in addition to the current users’ desire to switch to something more user-friendly, computationally efficient, and/or less expensive. Government appointed scientists and engineers provide the physics-based algorithms to the developers and the developers would provide programs to the scientists and engineers for testing. To assess a software package’s compliance with mandated modules for databases and calculations, benchmark scenarios will be used. At the very least, the inputs and results of these benchmark tests will be available to the public and may function as sample problems for an analyst to learn the functional aspects of a new software package. One way to satisfy a company’s need to meet such benchmarks would be for the government to provide explicit source code on the mandated modules so developers may compile them and use as program extensions or developers may port the algorithms to fit within their software’s framework. The federal agencies would retain responsibility for the construction of the source data to be used in the approved commercial model, such as, the process that the FAA uses to obtain aircraft noise and performance data. Table F-4 attempts to outline the general procedural requirements for implementation of this multimodal environmental model design alternative.

F-23 TABLE F-4 Key Milestones in the Development of the Commercially-based Multimodal Model End State Operational Track R&D Track Near-term (1-3 years) • Continued use of existing federal tools • Federal agencies specify which tools shall be implemented • Government announces a commercial tool will be required and accepts proposals • Analysts purchase existing tools to familiarize themselves with their use • Software companies prepare prototype versions for US • Software companies propose versions for US call for proposals • Emerging companies see future development opportunity, begin writing their own software packages Mid-term (4-5 years) • Government chooses software package • Chosen software developer releases software • Analysts now use adopted software packages • Government announces requirements for next generation model • Benchmark tests with commercial software vs. former models • Competing software companies develop model improvements for next generation End state • Next generation model is chosen and used • Government periodically chooses new model to adopt based off ongoing development (every 4 years) • Competing companies continue development • New modules and methods introduced F.5 Design Alternative #4: Build on EC IMAGINE Project Drawing on research completed by the European Commission (EC), the fundamental principle of the model design is the separation of description of the transportation source in terms of sound energy and exhaust emissions from the description of transmission to the receiver in terms of sound propagation and emissions dispersion. In May 2007, the EC completed its major noise modeling project, IMAGINE (Improved Methods for the Assessment of the Generic Impact of Noise in the Environment), which proved that it is technically feasible to build a noise model that can compute noise levels from a variety of sources. The results of the IMAGINE project fit in perfectly with the simulation modeling concepts, such as, DoD Advanced Acoustic Model (AAM). The end state is the same as the current preferred design (Datum). However, this end state is geared toward application on large, regional transportation projects where the environmental outcomes for more than one transportation mode are critical elements of the decision making. F.5.1 Description The goal is the same end state envisioned in the current preferred design (Datum), i.e., a source (airplane, automobile, truck, marine vessel, etc.) simulation model with benefits evaluator to convert noise exposure and air quality changes into environmental costs. The functional specifications are identical to the Datum. F.5.2 Functional specifications In May 2007, the EC completed its major noise modeling project, IMAGINE - Improved Methods for the Assessment of the Generic Impact of Noise in the Environment. IMAGINE and its predecessor, Harmonoise (Harmonized Accurate and Reliable Methods for the EU Directive on the Assessment and Management of Environmental Noise), produced guidelines for the common assessment methods to produce the strategic transportation and industrial noise maps as required by European Direction on the Assessment and Management of Environmental Noise 2002/49/EC. The IMAGINE and Harmonoise projects were necessary because the EC found that there did not exist harmonized methods of sufficient accuracy for the prediction and assessment of transportation and industrial noise. None of the F.5.3 Justification

F-24 available methods were sufficient to satisfy the requirements of the EC Directive 2002/49/EC. Similar to the current state of transportation noise modeling in the United States, no harmonized methodologies and data were available. The basic layout of the IMAGINE/Harmonise model is shown in Figure F-2. Figure F-2. Schematic of IMAGINE/Harmonise Model Source: Beuving, M. and B. Hemsworth, IMAGINE, Final Synthesis Report, Guidance on the IMAGINE methods, Doc. ID IMA10TR-06116-AEATNL10, Funded by EC 6th Framework Program, 11/16/2006. Figure F-2 shows how the model separates the source descriptions for road, rail, industry and aircraft sources, from propagation to the receiver. The result of the source models is a sound power level per source type for each source height relevant to that source, together with directivity. For example, the sound emission of an aircraft is defined in terms of sound power spectrum with directivity (longitudinal and lateral). The propagation model describes the transmission of sound along a set of propagation paths, linking the source positions to the receiver point. The number and type of the propagation paths depend on the complexity of the site. The P2P module estimates the effects of ground and obstacles on the propagation of the sound along these paths, under various meteorological conditions. For example, an aircraft flight is treated as a set of discrete point sources and the sound power spectrum at each point is consistent with the aircraft flight condition (engine thrust and speed). The result of the propagation model is a noise level at a specific receiver point for a given propagation class (the meteorological influence on the propagation paths is divided into 4 different propagation classes). By de-coupling the description of the source from the description of noise propagation, the IMAGINE/Harmonoise project provides the basis for a generic noise propagation model which can be coupled to almost any noise source. The authors claim that methods are pre-eminently suitable for harmonization of noise calculation because: ● Separation between emission and propagation allows flexibility; ● Source models are adaptable to local conditions; ● Propagation incorporates arbitrary meteorological conditions; ● Accounts for different ground conditions; ● Handles complex geometries, and; ● Usable at different fields of application.

F-25 The IMAGINE/Harmonoise project teams recognized that the quality and accuracy of the results depends upon the level of detail of the input data, especially data associated with sound propagation calculations. Therefore the project provided two kinds of propagation method: a reference model based on the numerical solution of the wave equations and an engineering model based on analytical formulae and heuristics. The reference model is able to deal with complex descriptions of the atmosphere whereas the engineering model relies on a simplified description. For example, the reference propagation model for aircraft noise is a hybrid model combining the parabolic equation (PE) model and two-ray model for application at appropriate elevation angles. The engineering propagation model for aircraft noise combines analytical models for basic phenomena, such as, ground reflections and diffraction by obstacles along with heuristic models to account for factors, such as, non-uniform ground impedances, non-flat terrain, multiple diffraction, etc. The project addresses the application of the reference or engineering propagation models as part of guidelines on modeling requirements in accordance with the following levels of application: ● High accuracy for highly critical situations where the outcome is a critical element in decision making, i.e. when noise levels are contested, possibly up to legal proceedings. ● Medium accuracy corresponds to the level of detail necessary to evaluate the costs and benefits of a specific intervention or action plan or to check conformance with regulations and limit values. ● Low accuracy may be sufficient for the more global assessment of existing situations, such as, for strategic noise mapping required by the EC. The IMAGINE/Harmonoise project achieved the EC objective to provide a harmonized, accepted and reliable method for the assessment of environmental noise from road, rail, airports and industrial sites. The IMAGINE project team identified subjects for further work including the following: F.5.4 Issues ● Investigation of propagation over built up areas as the model was validated over flat terrain. ● The aircraft noise mapping capability lacks appropriate source data. The logical sources of such data are the manufacturers, but the data is proprietary or expensive to purchase. A separate measurement campaign to obtain the data would extremely expensive. ● Pragmatic method for typical urban situations, such as effects of crossroads on road noise. ● Propagation of sound in complex geometrical situations, such as, “canyon effect.” ● Further simplification and optimization of the models because, depending on level of accuracy, the runs can tax the hardware and computation power typically used for noise mapping. The project produced executable files of the propagation models, but not a complete model. Complete models are then open to alternative implementations of many components by software developers bringing into question the reproducibility and reliability of these future models. The IMAGINE/Harmonoise project dealt strictly with noise mapping. It does not include emissions. On the subject of road noise mapping, the project team noted the existence of similar work done in the field of road traffic air pollution and suggested that a combined approach would be beneficial to both fields. Therefore, a major air quality modeling initiative is needed.

F-26 While the architecture of the Department of Defense (DoD) Advanced Acoustic Model (AAM) is conducive to the harmonized methods produced by IMAGINE; the other major federal initiative, FAA’s Aviation Environmental Design Tool (AEDT), is not. The IMAGINE project team recognized that a critical shortcoming to the development of a practical source-receptor noise model is the lack of appropriate aircraft source noise data, which would mean the need to rebuild source databases for both noise and air quality modeling. The preparation of this concept paper did not include verification of every algorithm and method put forth in the IMAGINE/Harmonoise project reports. Therefore some of the capabilities might fall short of expectations, which would require additional research to produce working capabilities in the development of the multimodal model. The AEDT development team has referred to their model architecture design concept as the “common thread.” The starting concept is the identification of a minimum operation that can be configured for an environmental analysis. The objective is to find a common thread across both the noise- emissions dimension and the local-global dimension. This common thread provides the context for both distinguishing modularity and removing redundancy in the module breakdown of the system architecture. The Build on EC IMAGINE F.5.5 Design Elements 2 The sources in Figure F-2 are the vehicle source representatives from the various transportation modes (airplanes, helicopters, automobiles, trucks, motorcycles, locomotives, railcars, and maritime vessels). Each vehicle source is a pollutant emitter defined in terms of sound power level and exhaust emissions indices (NO design concept also has a common thread based on the source-receptor relationship. The IMAGINE schematic in Figure F-2 conveys the design concept for the source-receptor multimodal noise and emissions model. X, CO2 The EC IMAGINE/Harmonoise project claims to have proven that it is technically feasible to build a noise model that can compute noise levels from a variety of sources, including all transportation sources, and produce noise maps on a common noise metric. The propagation algorithms and modeling definitions developed by this project serve as a basic blueprint for the core computation component for a future multimodal noise model. Of particular interest to ACRP Project 02-09, the EC project produced detailed guidance on: , HC, etc.). The vehicle traffic flow along flight paths, roadways, railways, or waterways are defined as consecutive vehicle source points. The transmission of pollution emissions from a vehicle source point to a receptor (home, school, hospital, park, etc.) is handled by a sound propagation model or dispersion model that is appropriate for that vehicle mode and operational situation. Unlike the IMAGAINE schematic, the multimodal model would be able to calculate any noise metric and any air pollutant associated with the various transportation modes. ● Railway source “traffic noise model” that interfaces with the IMAGINE Propagation Method to produce set of sound energy “source lines.” ● Road noise emission model describing the noise emission of an "average" European road vehicle in terms of a sound power level. ● Reference and engineering sound propagation models for aircraft noise emissions. ● Use of road traffic models (demand and flow management) for noise mapping and noise action planning. 2 Information and work products on the European Commission IMAGINE and Harmonoise projects can be found at http://www.imagine-project.org/.

F-27 The EC approach is to turn the algorithms and technical guidance over to commercial software developers who will create the models to be used for strategic noise mapping for the Member States. For development of the domestic multimodal noise and emissions model, federally funded research and development (R&D) seems the more pragmatic approach drawing upon the resources, capabilities, and knowledge gained over the years in the development of such legacy tools as INM, EDMS, NOISEMAP, NMSIM, TNM, etc. The results of the IMAGINE/Harmonoise project fit in perfectly with the DoD Advanced Acoustic Model (AAM). AAM has been developed with a spectral time series approach and uses simulation to calculate any of the temporal or spectral based noise metrics. AAM can produce the traditional integrated metrics such as Ldn, but with added capability to calculate supplemental metrics such as audibility, probability of detection, time above ambient noise levels, building transmission loss, and number of events. It also makes sense to draw in elements of the design of FAA’s ongoing AEDT development for the construction of an IMAGINE-based multimodal noise and emissions model in areas, such as, system architecture, data processing and user interface. The project would also benefit by bringing in members of the IMAGINE team to collaborate on the implementation of their algorithms and model guidance. The IMAGINE project team recognized that a critical shortcoming to the development of a practical source-receptor noise model is the lack of appropriate aircraft source noise data. The Federal Interagency Committee on Aviation Noise (FICAN) noted the same in finding that the simulation model is superior to integrated models but lacks the comprehensive aircraft acoustic data required. Therefore, the first major R&D initiative for the noise model, envisioned under this design, is the development of a practical method to create noise source data from available sources, such as aircraft noise certification testing. The IMAGINE project dealt strictly with noise; not exhaust emissions or air quality. However, that does not mean that the development of IMAGINE-based transportation emissions model is behind the development of its noise counterpart. Air quality modeling already separates source emissions from propagation (dispersion). Thanks to the Environmental Protection Agency (EPA) many components of emissions and dispersion modeling are harmonized and standardized, but this standardization has focused on static modeling for the most part and work would need to be done for simulation modeling. EPA's Air Quality Modeling Group (AQMG) provides leadership and direction on the full range of air quality models and other mathematical simulation techniques used in assessing control strategies and source impacts. This office publishes EPA's Guideline on Air Quality Models to provide consistency in the use of modeling. These guidelines are periodically revised to ensure that new model developments or expanded regulatory requirements are incorporated. The current preferred dispersion models are: AERMOD Modeling System - A steady-state plume model that incorporates air dispersion based on planetary boundary layer turbulence structure and scaling concepts, including treatment of both surface and elevated sources, and both simple and complex terrain. However, this would not work with the simulation approach without numerous adaptations. CALPUFF Modeling System - A non-steady-state puff dispersion model that simulates the effects of time- and space-varying meteorological conditions on pollution transport, transformation, and removal. CALPUFF can be applied for long-range transport and for complex terrain. Where R&D is needed is in the integration of simulation with the preferred dispersion models and source emissions models. For example, EPA’s CALINE and CAL3QHC-series roadway models provide static environments with time-averaged, aggregate variables. A simulation approach provides more realistic and robust modeling environment with no temporal or spatial constraints. The TRaffic Air Quality Simulation Model (TRAQSIM) demonstrates the possibility of modeling the effects of road grade. TRAQSIM has been shown to facilitate the emissions and dispersion modeling of both particulate

F-28 matter (PM) and chemically reactive pollutants. Therefore, the R&D effort would examine expansion of the capability demonstrated by TRAQSIM to the other transportation modes. Taking the Build on EC IMAGINE design approach means that a fully realized multimodal noise and emissions model is years away. EC faced a similar issue for strategic noise mapping and decided to implement “Interim Methods” for preparation of the current series of maps while the noise modeling research is underway. Annex II of EC Directive 2002/49 lays down four interim computation methods for the production of strategic noise maps. They are as follows: ● Road traffic noise: the French national computation method NMPB-Routes-96 referred to as “XPS31-133.” ● Railway n oise: the Netherlands national computation method published in Reken- en MeetvoorschriftRailverkeerslawaai 96, Ministerie Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer (20 November 1996), referred to as “RMR.” ● Aircraft n oise: ECAC.CEAC Doc. 29, Report on Standard Method of Computing Noise Contours around Civil Airports (1997), referred to as “ECAC doc. 29.” ● Industrial noise: ISO 9613-2: Acoustics — Abatement of sound propagation outdoors, Part 2: General method of calculation, referred to as “ISO 9613.” The federal government could decide to take a similar approach concerning the models to use for the environmental assessment of multimodal transportation projects. The first set of interim methods would provide specific guidance on how to use the existing approved noise and emissions models (AEDT, TNM, RCNM, HSRNOISE, NOISEMAP, NIRS, MOBILE6, etc) and screening tools (AEM, ATNS, etc.) for the full range of multimodal projects from airport-centric to system wide. The Environmental Working Group (EWG) under the Joint Program Development Office (JPDO) seems an appropriate forum to reach inter-/intra-agency agreement on interim methods for multimodal noise and emissions assessment. Specifically, the Policy and Analytical Tools Standing Committees under EWG appear to have the mandate and representation to do the job. This proposal assumes that the federal agencies act on interim methods and well. Flexibility also extends to the manner in which multimodal noise and emissions assessments are conducted. Preliminary reactions from potential future users provide some insight on additional actions that the federal agencies could take to promote effective environmental assessments. Task 1 of ACRP Project 02-09 included a preliminary market research effort to assess the viability and utility of a multimodal environmental model and help in the formulation of the model design and plan. A questionnaire was used to contact consultants involved in transportation planning, state and federal agencies that provide the oversight for these modes, and office staff of regional transportation administrations that organize/fund specific projects. The results were reported to the ACRP panel in the form of a Wyle Technical Note (TN 09-01, Preliminary Findings on Future Utility of a Multimodal Noise and Emissions Model). Respondents are concerned that the new model would significantly increased study costs; making it prohibitively expensive for anything smaller than a regional study. They suggested that federal agency approval is important and specific guidance on when the model is to be used in the environmental process is needed. One of the respondents’ suggestion to improve study efficiency is the “ability to automatically grab via the internet project required available databases – census data, current fleet mix and operations by airport, state highway traffic data, radar data for identified time period, etc. Current methods require considerable user time and effort to collect the necessary available input data that are not project specific.” This design incorporates and expands upon this suggestion with the inclusion of action to create a federally-sponsored clearinghouse of transportation environmental study data (inputs and outs) accessible

F-29 to the public. There is already federal precedent for making transportation environmental study data available to the public. For example, the Vision 100-Century of Aviation Reauthorization Act (Public Law 108-176) required the FAA to “make noise exposure and land use information from noise exposure maps [prepared under 14 CFR part 150] available to the public via the Internet on its website in an appropriate format.” FAA met the requirement with the creation of a website with links to airport noise and land use information pages and copies of the noise exposure maps. The site is: http://www.faa.gov/airports_airtraffic/airports/environmental/airport_noise/noise_exposure_maps This design proposal extends the precedent to the other modal agencies and suggests centralizing the study information at a single internet clearinghouse under the auspices of the appropriate federal agency, such as the Office of the Secretary of Transportation (OST). Since the intent is to provide data that would be useful in multimodal studies, the agencies need to establish standards for the type of data to be placed in the clearinghouse. For example, they need to agree on the geographic information system (GIS) for the management of the various input and out data including roads, railways, waterways, runways, flight tracks, meteorological data, computed noise contours, computed noise grids, pollutant concentrations (monitored and computed), etc. The agencies would also need to provide guidance on how the GIS data is to be used; including reaching a meeting of the minds on the common metrics to use in multimodal noise and air emissions studies. The endeavor to create the data clearinghouse is a major activity to occur in parallel with the AEDT expansion. In addition to the technical tasks in establishing an internet clearinghouse; regulatory and policymaking activities would need to be completed to establish the requirement to gather study data from mandated studies, such as Environmental Assessments (EAs) and FAA Part 150 studies, and standardize the data format and collection method. This new aspect also places the need to build into the AEDT expansion a previous study data integration module to extract necessary input and output information from the data clearinghouse. The objective is that a user would have automatic access to any previous transportation environmental study, such as, FHWA highway construction EA or FAA Part 150, to incorporate into a multimodal study covering the same geographic region. This proposal assumes that the federal agencies act on interim methods and Table F-5 lays out milestones on a dual track system leading to the end state. The operational track lists the steps that the federal government would take to update guidance on multimodal environmental modeling in coordination with the model developments achieved on the R&D track. The idea is that the developers and policymakers both learn through practical application and adjust the interim methods, accordingly. A simulation model is more computationally intensive than existing integrated models. Multimodal environmental assessments would require a greater array of data and specialized expertise than current single mode projects. Therefore, it would be reasonable for the federal agencies to require a simulation multimodal model for those applications where the environmental outcome for more than one transportation mode are critical elements of the decision making, such as, the regional transportation solutions envisioned by the NextGen Metropolitan Areas Solution Set that emphasizes innovative approaches to regional planning and multimodal systems. For other, smaller projects, such as individual airport or highway construction projects, the agencies would retain interim methods guidelines for use of existing tools.

F-30 TABLE F-5 Key Milestones in the Development of the “Build on EC IMAGINE Project” Multimodal End State Operational Track R&D Track Near-term (Years 1-3) • Inter- and intra-agency guidelines for application of existing noise and emissions models to environmental assessments (EAs) of multimodal transportation projects (airport to system level) • Inter- and intra-agency agreement on levels of accuracy (high, medium, or low) for multimodal noise modeling in EAs. • Research task force and work program to develop technical guidelines for IMAGINE-like transportation emissions modeling • Practical methods to create detailed noise source data from available sources, such as, aircraft noise certification tests. • Prototype simulation noise model based on DoD AAM and EC IMAGINE reference model algorithms. Mid-term (Years 4-6) • All new source noise data for existing model databases meet data requirements for simulation source-receptor modeling • Release of hybrid simulation (for aircraft) /integrated noise model. • Revised inter-/intra-agency guidelines on multimodal modeling to integrate new hybrid noise model. • Inter- and intra-agency agreement on levels of accuracy (high, medium, or low) for multimodal air quality and emissions modeling in EAs. • Prototype simulation noise model based on DoD AAM and EC IMAGINE engineering model algorithms integrating output with APMT benefits valuation block (BVB) requirements for economic impact assessments. • Simulation noise model Version 2 (reference model). • Simulation transportation emissions model algorithms and source data definitions. • Practical methods to create detailed exhaust emissions source data from available sources, such as, aircraft engine emissions certification tests. Longer-term (Years 7-11) • All new source emissions data for existing model databases meet data requirements for simulation modeling • Release of hybrid simulation (all sources) /integrated noise model (source data availability) • Revised inter-/intra-agency guidelines on multimodal modeling to integrate new model. • Prototype simulation emissions model (reference model) based on TRAQSIM approach and EPA preferred source and dispersion models. • Simulation noise model Version 2 (engineering) • Prototype simulation emissions model (engineering model) integrating output with APMT BVB requirements for economic impact assessment. End state • Dynamic source (airplane, automobile, truck, vessel, etc.) simulation model using “best available” engineering propagation models with benefits evaluator to convert noise exposure and air quality changes into environmental costs. • New inter-/intra-agency guidelines requiring simulation model for multimodal EISs and continued use of existing tools for smaller projects. Research version of the end state based on reference propagation models.