EMVL FY2009 Projects

• Identify new molecular biomarkers of exposure to endocrine disrupting compounds representing several modes/mechanisms of action, and
• Link those biomarkers to effects that are relevant for both diagnostic and predictive risk assessments using small fish models

This project is running large-scale modeling scenarios to quantify the effect of air-quality control measures on specific pollutants. The model runs simulate physical and chemical processes associated with the formation, transport, and interaction of such reactive pollutants as ozone and certain components of PM_2.5.

As of FY2009, this project encompasses the following subprojects:

• CLIMATE -- Model applications, data, and analyses in support of climate and climate change
• FEDRULES -- Model applications, data, and analyses in support of federal rules to help attain air quality and/or emissions standards
• MULTIPOL -- Model applications, data, and analyses in support of multi-pollutant (criteria and toxic) assessments
• NAAQS_DI -- Model applications, data, and analyses in support of implementation and nonattainment designations for National Ambient Air Quality Standards
• NAAQS_SS -- Model applications, data, and analyses in support of the review of new National Ambient Air Quality Standards
• PLATEVAL -- Model applications, data, and analyses in support of the development and evaluation of air quality modeling platforms
• SECTORS -- Model applications, data, and analyses in support of emissions sector-based assessments

The CHEMTOX project is using molecular modeling methods to predict the intermolecular interactions of environmental chemicals with protein targets. The goal is to computationally identify possible molecular toxicity as a means to prioritize chemical testing needs.

EMVL is assisting the PI with this project in two ways:

• Automating the computational procedures to increase the work flow for the molecular part of this project.
• Integrating these results into more complex multidimensional multiscale toxicity prediction approaches.

EMVL has, over the last three years, worked with EPA scientists to build the Remote Sensing Information Gateway (RSIG). RSIG offers a new way for users to get the multi-terabyte, environmental datasets they want via an interactive, Web browser-based application. RSIG reduces a file download and parsing process from months to minutes.

RSIG also allows users to integrate their selected datasets into a unified visualization. The user can tap into a wide range of key environmental models and data, such as NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS), the Environmental Protection Agency’s (EPA) Community Multi-scale Air Quality (CMAQ) model output, National Environmental Satellite, Data, and Information Service (NESDIS) biomass burning data, and ground station measurements from AIRNow and EPA's Air Quality System (AQS).

RSIG visualizes each dataset and then overlays them on a map of the selected region, automatically aligning information from various spatial and temporal scales into a unified structure. Currently, RSIG focuses on atmospheric data, but it can be extended to accommodate any type of remote or sensed data. The benefit to users and consumers of environmental data is fast acquisition of only the data they want to see and in a standard format they can save to their desktop PC.

EPA contracted with the Argonne National Lab to develop the VERDI visualization tool as a replacement for the Package for Analysis and Visualization of Environmental data (PAVE), which has been used to visualize and analyze Community Multi-scale Air Quality (CMAQ) input and output data.

For FY2009, Dr. Donna Schwede asked EMVL staff to continue its work in improving the performance of VERDI, especially in its ability to animate tile plots on ORD scientists’ personal computers when the VERDI application is being run on large-scale systems such amber or vortex, the Agency’s two largest High Performance Computing (HPC) systems. EMVL has achieved an approximately 100x speedup in the tile plot animation rate.

Researchers at the Large Lakes Research Station in Grosse Ile, MI, seek to apply their eutrophication model, LM3-Eutro, to existing hydrodynamic model output pertaining to the Gulf of Mexico. In particular, the modified LM3-Eutro, called Gulf of Mexico Dissolved Oxygen Model (GOMDOM), will be used to model hypoxia and nutrient loading events in the Gulf associated with discharge from the Mississippi River and other tributaries.

EMVL is supporting the PIs in several ways for this project:

• Developing standalone FORTRAN code to calculate full 3D distribution of the horizontal dispersion coefficient (also called turbulent horizontal eddy diffusivity).
• Determine why the original LM3 Eutro model is producing negative concentration values. Once the cause is determined, make a recommendation for fixing it.
• Optimize the LM3-Eutro model, which will then carry over into the new GOMDOM program.

EMVL staff continues its work on building a fully integrated, realistic model of the human respiratory system. This includes a detailed model of the nasopharyngeal passages (nose and mouth), throat, and the lobes of the lung itself. EMVL staff also continue to write, edit, and submit papers documenting this work to peer-reviewed publications.

These realistic models are based on actual Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) data. The purpose of this work is to better understand human exposure to various pollutants through a more precise awareness of particle deposition within the human lung.

This project focuses on improving dispersion modeling tools to produce robust estimates of air pollutant concentrations in the near-road environment, which is a vital input to human exposure models. By using Computational Fluid Dynamics (CFD) simulations in connection with wind tunnel studies, the approach is to develop simplifying parameterizations that can be integrated into operational models and simplified research models.

EMVL staff have worked on:

• Developing idealized roadway models for CFD purpose based on test scenarios in the EPA wind tunnel;
• Developing a virtual model of the Las Vegas site where field experiments are being planned;
• Performing CFD simulations using the idealized roadway models and evaluation with wind tunnel results;
• Performing CFD simulations using the Las Vegas virtual model and evaluating against wind tunnel results.

EMVL is working with the PI to create a high quality visualization of lens morphogenesis in the mouse eye. The eye development project is a feasibility prototype to determine its applicability to the development of models for all embryogenesis processes.

EMVL’s effort toward this end is divided into four tasks for FY2009:

• Develop a Visualization Model to render tissue morphogenesis, as described in the NCCT Board of Science Counselors (BOSC) review presentation.
• Adapt a project management tool to describe normal and abnormal tissue morphogenesis processes. Use this tool output as a simulated input to the Interpreter application.
• Develop an Interpreter application to interface and translate incoming requests from the Computational Model and other simulated outputs to the Visualization Model.
• Use standardized ontologies, formal descriptions, and file formats, such as Gene Ontology (GO), Edinburgh Mouse Atlas Project (EMAP), and RDF/XML, so that implementations are shareable and knowledgebase-aware. Use portable open source tools.

The overall goal of this project is to develop a set of tools to generate a physiologically based pharmacokinetic (PBPK) model from scratch given a new chemical and a set of criteria, such as organism, age, organ, tissue or other compartment, and exposure route. When possible, the tools will be integrated within a custom Web-based graphical user interface (GUI).

The project’s output will include the following products:

• Report information including physiology, tissue specific partition functions, absorption, distribution, metabolism, and excretion (ADME) endpoints, parent chemical loss rates, target list Ligand Activity in Surface Similarity Order (LASSO) descriptors from SimBioSys Inc.’s eHiTS software, and target list docking.
• An estimate from a rapid calculation using the pre-computed Excel model and tissues from Cahill and MacKay.
• File(s) suitable for input into the EPA’s Exposure Related Dose Estimating Model (ERDEM).

EMVL is assisting researchers in designing a custom module to download and pre-process Geographic Information Systems (GIS), remote-sensing, and environmental data for a user-defined area. The project’s data needs are currently defined by the Analytical Framework for Coastal and Estuarine Studies (ACES) data model for estuaries and coastal systems and by tools and models already developed or under development.

The Virtual Estuary tool will allow the PI and other researchers to efficiently populate a master database as well as to develop decision-support tools that can be applied at multiple scales by stakeholders with various needs, such as National Estuary Program sites, states, and regions, without being constrained by a pre-defined population of systems.

In FY2009, EMVL staff members are assisting the PI in developing concepts for the Decision Support Framework (DSF) into well-defined requirements. EMVL is also developing visual products that communicate the DSF vision and functionality.

See ESRP’s Community Research Projects for background information related to this project.