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.
v research also examined how environmental analysis is currently done with existing models; pointing to potential utility for a future multimodal model. What are the impediments to successful implementation? Multimodal environmental analysis and mitigation will require greater collaboration among a variety of federal, state, and regional agencies. Multi-agency acceptance is critical, which requires an understanding of the technical and policy complexities that must be overcome. A technical complexity will involve the ability to incorporate all of the knowledge on the various modes into a single, usable model. Another technical complexity is the extent or lack of investigation of particular empirical factors and the amount of additional expertise required for their resolution. The policy complexity reflects the degree of incompatibility among existing environment assessment requirements promulgated by various agencies. The recommended model design and model build sequence calls for substantive interaction with interested Federal agencies along with periodic feedback from all stakeholders. The objective is to have the parties work together to overcome technical and policy impediments. What institutions might take leadership in applying the research product? Sole ownership of the multimodal model does not seem practical; so, the strategic plan accompanying the MDP proposes an interagency forum that could decide federal ownership and identify the technical and policy infrastructures needed to support the model development. What are necessary activities for successful implementation? The MDP includes a draft strategic plan on the funding and ownership. The use of strategic planning and tactics should offer flexibility in the model development while also providing contingency (situational) planning. How to judge the progress and consequences of implementation? Success will be judged by the actions that U.S. DOT and its modal administrations take with respect to the MDP and the suggested strategic plan. In addition, the evaluation process used in the assessment of alternative model designs should be useful to the future owners and development teams in deciding what direction to take for each model build sequence. How applicable are the results to practice? While the scope of this project is not to construct a new operational multimodal noise and emissions model, the MDP offers the fundamental pieces and considerations needed to create such a model. The approach has been to gain a thorough understanding of how the multimodal noise and emissions model would fit into intermodal and multimodal transportation projects involving decision- making that spans across the authority of several federal agencies. The potential user communities and federal agencies have been engaged in the process. The recommended model design has been rigorously assessed against viable alternative design concepts. This rigorous examination of alternatives has produced a design concept that is supported by a thorough, pragmatic MDP.
1-1 CHAPTER 1. INTRODUCTION This document describes the Model Development Plan (MDP) for a multimodal transportation environmental analysis tool. The MDP only addresses transportation sources. Considerations for other sources will only be included as needed for the transportation analysis and these efforts will be considered secondary. The overall goal of the proposed tool is to provide a capability to perform an environmental analysis consisting of noise, air quality, climate and economics of all modes of transport for various geographic scales (microscale, regional, national, and global). The proposed tool will streamline transportation-related environmental analyses by minimizing the redundant input currently required to exercise modal-specific tools, e.g., traffic, performance, etc. It will also facilitate a systems approach to transportation-related environmental analyses for air quality, climate, and noise. Additional modules on fuel burn and economic analysis will allow analysis to be more comprehensive. Inclusion of environmental cost/benefit capabilities will allow transportation planners to make more informed decisions on transportation-related projects. For example, should additional roadway access be provided to a particular airport or are limited resources better spent developing a public transit link? Details included in this MDP were developed from surveys of stakeholders (Appendix A), comprehensive literature reviews, and the experience of the research team. Several rounds of stakeholder engagement were included in the overall process. While the market research was successful in providing valuable insights for the model design, there was an issue concerning the small response to the questionnaire. Chapter 5 offers suggestion on how to build upon the initial market research. Appendix D suggests an interagency forum to engage federal stakeholders and as a vehicle for federal funding. The ACRP Panel (Panel) participated in an evaluation of five possible model development approaches put forward by the research team. As summarized in Appendix B, the approaches receiving the highest overall scores by the Panel and project team were the ones that build on the development of the FAA Aviation Environmental Design Tool (AEDT). Although aviation focused, AEDT combines many of the desired components of a multimodal environmental analysis tool (see Figure 1-1). It provides an excellent foundation for the future multimodal noise and emissions model. For example; the computation core of the AEDT component, shown in Figure 1-1, is based on EPA-approved air quality models and internationally-accepted aircraft noise models. Using the EPA- approved models, AEDT is capable of modeling roadway emissions to a limited degree and includes the architectural foundation and some of the algorithms for modeling roadway noise. The impacts and cost benefits modules of Aviation Portfolio Management Tool (APMT) provide the basis for multimodal environmental cost and economic impact analysis. It is also logical and cost effective to draw upon the extensive Graphical User Interface (GUI) that brings together the various components of the AEDT tool suite. Additionally, the new EPA emission model MOVES that was recently required for motor vehicle emission estimation also has a similar data base to AEDT using SQL and as such will allow for a direct combination of the data during analysis. Due to the substantial anticipated time and resources needed to develop the proposed end-state model, the research team recommends a multiphase approach as described in detail in Chapter 2 of this document. With this approach, the development process is recommended to be divided into a series of phases, each called a âBuildâ. An important aspect of this plan is that each Build results in a usable product. 1.1. Purpose The purpose of the MDP is to define the work needed to develop a multimodal transportation environmental analysis tool. The MDP presents the proposed work as a multi-phased effort, which each phase resulting in a usable tool, incrementally building towards the desired end state as time and resources allow. Each Build is expected to be completed in such a way to allow the next phase or build to
1-2 be a continuation of the model development rather than starting anew each time. It is difficult to determine the exact timeline for each build due to resource uncertainty. However, based on the assumption that resources will be available, the project Builds have been defined for short-, mid-, and long-term development. Figure 1-1. AEDT with linkages to other related FAA/AEE tools. Source: FAA, Office of Environment & Energy Details of each Build are presented in Section 2.3, Project Deliverables. The series of Builds (1 to 6) are what is currently envisioned. However, depending upon needs, progress, and future unforeseen events, these Builds could be combined, expanded in scope, and/or possibly deleted, as related science and information technology advances. 1.2. Scope and Background Based on a detailed literature review, and a gap analysis of existing tools and methodologies and the collective experience of the research team, the project team agreed that a multimodal transportation environmental analysis tool based on simulation architecture (most likely time-step based) would be the most desirable end state or final Build. The advantages of a time-simulation approach include: ⢠Increased flexibility in modeling approaches, e.g., the ability to take into account transportation schedules in a comprehensive way; ⢠Fewer needs for simplifying assumptions such as line sources or energy averaging; ⢠More comprehensive analysis of microscale conditions; and, ⢠The ability to have more detailed and representative emission factors and reference acoustic levels.
1-3 It was also recognized by the research team that initiating the Build sequence with a simulation model could have substantial pitfalls. Primary among these were the need to develop new source data, the lack of precise vehicle time-space-position information across the modes, and increased computer run time. To this end it was decided to review possible scenarios to advance the project to an end state while providing for each step of the development as outlined in Section 1.3. The evaluation and selection process was comprehensive and included two rounds of evaluations. Five designs were considered and included an option based on step by step integration, which was the initial thinking at the start of the project. The five possible model development approaches were: ⢠Step by Step Integration: 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. 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. This is the concept initially put forward by the research team and used in the stakeholder questionnaire (Exhibits E-1 and E-2 of Appendix E). ⢠Build on AEDT: The FAA is developing a tool named the Aviation Environmental Design Tool (AEDT). This tool is part of an integrated suite of tools used to perform comprehensive environmental and economic analyses, which allow policy decisions to be made in a more informed way. This development approach calls for the continued development of AEDT to a multimodal environmental analysis tool. Development is proposed in a phased fashion that would eventually result in a time-simulation-based model. ⢠Build on Existing Simulation Models: This development approach would be based on the expansion of existing single-mode, simulation-based, noise and air quality models used for transportation source analysis. The approach would be based on time-step based simulation of source movements, source emissions, and propagation scenarios resulting in detailed output reports at receptor locations. ⢠Federal Adoption of Commercial Software: This concept promotes a market-based approach for the development of a multimodal environmental analysis tool. Commercially available environmental analysis software would be the basis of this approach. The end state would be a vendor controlled software package including modeling capabilities for all modes of transportation. ⢠Build on EC IMAGINE Project: Drawing on research completed by the European Commission (EC), the fundamental principle of this model development approach would be to continue to build on IMAGINE (Improved Methods for the Assessment of the Generic Impact of Noise in the Environment). The end state, a true simulation model, would be 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. The five model development approaches were presented to the Panel with detailed instruction on evaluation procedures. Appendix B describes in detail the two round evaluation process. Five designs were evaluated in the first round. As a result of the first round scores, a new design was constructed taking the most desirable attributes from the 5 original concepts. The second round evaluated the winner of the first round against the newly constructed sixth design. Thus, the design and build requirements described in Chapter 2 are the products of two rounds of evaluation and incorporate the most desirable attributes of the alternative designs considered. Appendix B discusses how the final design was constructed with justification for the choices made.
1-4 1.3. Developmental Goals Section B.7 of Appendix B describes the winning model design. This section describes the goals of that design in terms of the time required beginning with the short term and looking to the future. Full implementation of the winning design has been divided into âBuildsâ or individual projects to allow the work to progress in a staged manner. It should be noted that these times are based on the best estimates of work effort. Many factors could extend the time such as collection of needed data, administrative requirements, and unforeseen problem areas. Short Term (1-3 Years from Initiation): The short term goals include two Builds. Build 1 would include a postprocessor for outputs of existing models, integrating those outputs into a single presentation. Table 1-1 presents a listing of existing candidate models for this process. Determination of specific models for all modes should be made during initial planning; coordination with stakeholder agencies will be required to select from the listing. Additionally, some of the models have already been integrated into AEDT, such as INM, HNM, and AERMOD for airports and AERMOD may also be adapted for other modes. To enable the combination of results from various models, an output combination rule base will be required to sum similar results of each model. This rule base is required due to conflicts in model outputs that might prevent a direct combination of results. For example, the FHWA requires the use of the noise metrics Leq or L10, while most others use Ldn Midterm (3-8 Years from Initiation): The midterm goals include Builds 3 and 4. Build 3 would be a large effort and would include integration of other modal analysis tools with AEDT, but leaving the original tool intact. The model structure and protocol would be maintained throughout the remainder of Builds unless overriding technical requirements demand otherwise. It is envisioned that the methodologies to be integrated would be again selected from and based primarily on the tools shown in Table 1-1. Build 3 would also include a substantial expansion of AEDTâs GUI, assisting the user in model input and output tasks. Build 4 would result in a harmonized set of databases and methods across the various modes to the extent possible. For example, propagation for noise and dispersion analysis for air quality would use the same algorithms for all modes of transportation. Builds 3 and 4 would also incorporate tools to move from emissions inventories and noise contours to a more direct assessment of the health and welfare impacts of transportation activities. . Other conflicts such as time of analysis, assumptions by operating mode and use of weather conditions will also occur. This rule base will utilize approximate procedures in the short term to allow combination of similar outputs. No advancements to input processing are expected in Build 1 and all existing required stakeholder agencies tools selected for integration would continue to be used in their existing fashion. Build 2 would be a series of screening tools that will be integrated and adapted in later Builds to provide mitigation tools for quick planning analysis and comparison between various future project options. All screening tools would be tested for accuracy against existing tools and sensitivity to input variables. Long-term (8+ Years from Initiation): The long-term goal would be to implement a first generation simulation architecture, which would begin in Build 5 with a hybrid approach. The initiation of this build would be limited based on the availability of high-resolution source data, e.g., noise hemispheres. This development will be sequential in that it will be based on the work in previous Builds, most notably the GUI structure, the use of a framework consistent with AEDT, and the output combination rule base. Build 5, envisioned as a hybrid tool, would include steady-state and simulation- based model processing. Build 6 is seen as the desirable end state, a full simulation-based, multimodal environmental analysis tool, with all supporting algorithms and databases fully harmonized as began in Build 4.
1-5 TABLE 1-1 Noise, Air Quality and Related Models in Use or Developed by Federal Stakeholder Agencies Transportation Mode Air Ground Water Rail Noise AAM (NMSIM + RNM) AEDT HNM INM NIRS NOISEMAP CREATE HICNM Horn Model HSRNOISE RCNM TNM Spreadsheet: Guidance on Assessing Noise and Vibration Impacts Emissions AEDT EDMS AEDT EDMS EMIT MOBILE6 MOVES NONROAD2005 EPA AP-42 Emission Factors EPA AP-42 Emission Factors Atmospheric Dispersion AEDT EDMS (AERMOD) CALINE3 CALINE4 CAL3QHC CALPUFF AERMOD OCD Gaussian Puff-based models (e.g., CALPUFF) Impact Assessment and Valuation APMT CMAQ and BenMAP 1.4. Guiding Principles In developing the MDP, the research team followed a number of guiding principles so that the work plan will result in the final model(s) meeting stakeholder needs. The MDP development included input received as part of stakeholder engagement as well as the collective experience of the research team. Comments of the stakeholders were considered and key concerns/observations included: ⢠The users would be federal and state agencies and those conducting noise and air quality analyses on their behalf to perform project level analysis to meet requirements of regulations, perform research, and/or evaluate future scenarios in general planning. Other uses are possible but would be secondary to the stated uses. ⢠The MDP should consider how the projected model is likely to affect the cost, duration and risks of the environmental/design process. ⢠The MDP should estimate potential environmental impacts and evaluate mitigation measures. ⢠The MDP should allow estimation of the individual and aggregated contributions of the various modes. ⢠There is significant complexity, bringing the various modes into a single environmental analysis tool. This is especially true for the integration of inputs which are in some cases very different. Combination of outputs was also cited as a concern.
1-6 ⢠Current analyses are conducted by teams, with each team member adding specific expertise in one or two technical fields. If preparation and running of the model is significantly more time consuming and expensive than use of current models, the simpler project and less experienced user will be disadvantaged. As such, models should be easy to use, and scalable. ⢠The expanded use of Geographic Information Systems (GIS) is essential. ⢠Inclusion of sources beyond transportation such as point sources could be helpful. ⢠Use of a build sequence would allow the designers to adapt and learn as they go, should help to develop long-term support from stakeholders, and each build should expand upon the previous. ⢠Existing tools should be used as much as possible. Defined logic of MDP: Following good engineering practice, the logic was based on five key points: 1) development should be phased to allow advancement as resources become available and as science and technology allows, also providing a usable tool at the end of each Build; 2) development should be based on proven principles, using/adapting existing approaches in early Builds and with more comprehensive development occurring in later Builds; 3) the methodology must be responsive to the stakeholderâs needs, primarily the stakeholder agencies; 4) combination of model outputs must be an outcome of the first Build and continue to be developed with each Build; and, 5) the overall development should lead to improvements in the modeling process. In addition to the base logic, the models must be user friendly. This would require the use of the AEDTâs advanced GIS-based GUI to allow ease of use as well as a rule base to allow various outputs, especially related to different metrics, to be combined. This output combination rule base can be continually improved upon with each build allowing the combination of results to become more robust. Because the model must comply with the requirements of multiple stakeholder agencies, it is imperative that the rule base allow output of various metrics. Practical limitations: Based on the preceding logic, limitations were considered to allow each build to be completed within a specified schedule and with available resources. This required careful delineation of all tasks. Time. The Build process was divided into six stages, called Builds, so that each could be completed in a reasonable time frame. It is possible that Builds could be combined or other Build stages added as development progresses. The project team reasoned that no Build should take longer than 3 years or overall project development could be impeded because of a loss of flexibility in the technologies, and potential loss of stakeholder engagement. Project resources Improvements made to overall modeling process: The starting point for the MDP was to identify gaps in AEDT, as well as candidate models for integration with AEDT, with respect to a multimodal model. Appendix C presents the assessment of gaps identified in existing emissions, dispersion and noise models being considered as a part of MDP development. As each Build occurs, there should be a well-defined plan to eliminate these gaps. It must be recognized that while many of the gaps may be overcome in the Build options depending upon the concurrent advancement of the knowledge base, others will require scientific advancement and basic research. The MDP may identify . Using the AEDT development as a basis for estimates, significant resources will be needed to develop the end state. However, since the development of the multimodal model will be based on the work of AEDT, significant leveraging of resources should occur leading to reduced time and lower costs. For example, beginning with Build 1, the architecture and protocol of AEDT will be utilized for GUI development. This will help define programming efforts leading to reduced time requirements when in Build 3 the models are integrated into a single platform. Many of the development issues that were addressed during AEDT development can be recognized and avoided. The large work effort already accomplished for aviation sources will not need to be repeated.
1-7 some scientific gaps, but the scope of this document, including estimated schedule and resources, does not include the research effort required to address these scientific gaps. Single model concept advancement of a multimodal tool: The advancement of a single model concept of a multimodal tool would require not only inclusion of the existing single-mode models beginning in Build 3, but would require that these models be integrated more completely with each Build. This will require adaptation of the models. In the early stages this will be external to the models as the outputs are the foundation for âintegrationâ. With each successive build the inclusion of these models would become more comprehensive and seamless, resulting in a fully integrated tool. This principle will result in an end state in which each mode uses the same fundamental emission, dispersion and acoustic modules. Additionally, as these advancements occur, gaps in the models should be addressed to the extent possible, given that expected resources for the MDP will not permit advancement of fundamental science. It must also be noted that significant changes to any model will require extensive testing to validate and meet federal requirements. Additionally, inclusion of the regional, national, and global scale modeling will require large databases and models to be included in the overall framework. It is very likely that these options will require multiple platforms (e.g., portable computers and server-based systems) to be used. Additionally, it is recognized that global modeling for some modes may be difficult due to the extensive data requirements. For example, in the case of rail, non-regulatory models, such as the Florida Department of Transportation Rail Model, and European models such as the German Rail Model will need to be reviewed to establish what is available in terms of both method and database. This is also complicated for rail sources since the various types will need additional classification including heavy rail, light rail, high speed rail, etc. Integration into the Federal system: Federal stakeholder requirements must be observed during development. Any Build that would result in processing, calculations, outputs, or metrics that are not accepted by the federal stakeholders is unacceptable. Vehicle. Each mode of transportation has different vehicles. This results in different operational parameters, emission (noise and air quality) considerations, and different operating environments. While the early Builds will not change what is now being modeled, later Builds must find ways to define these vehicles using defined, acceptable methods available at the time. Work to derive new methods is not expected to be part of this work unless a specific need occurs. Model response. Many inputs are similar for various models now being used, e.g., meteorology. While not realistic in the early Builds, later Builds would need to incorporate a methodology to have input in a simple fashion that could be used for all modes when possible. Metrics. While air pollution computations from multiple models can be combined directly due to the use of similar metrics and the same standards, varying metrics now exist for noise. Since the output must be combined to determine the total noise exposure at any modeled location1 1 Note that in AEDT development, due to the differing use of the words âreceiverâ and âreceptorâ in legacy models, all modeled locations are now referred to as receptors. a rule based structure will be needed, beginning with Build 1, to allow prediction of the various metrics required for federal impact analysis (see Section 1.2). This rule base should be continued and strengthened with each Build.
