Computer modeling and homeland security

HPAC software developed at ORNL models atmospheric transport and ground deposition of hazardous releases (shown as colorful plume).

Suppose that a group of terrorists releases anthrax disease spores from a crop duster. Could these airborne spores be detected? How many people might be exposed?

What if a group of terrorists detonates a "dirty bomb" in which radioactive materials are dispersed by a conventional explosive? How many people could be exposed to hazardous radioactivity from such an explosion?

Suppose that a Boeing 767 airplane hijacked by terrorists crashes into a nuclear power plant, destroying the building in which spent nuclear fuel is being stored under water and causing the release of large amounts of radioactivity. How many people could be exposed to dangerous amounts of radiation?

These questions have been raised in the news media in the wake of the September 11, 2001, terrorist attacks on America. Research at ORNL using computational simulation and modeling has resulted in products that can help America and other countries address these questions concerning the use of weapons of mass destruction. Computer software and modeling and simulation techniques developed at ORNL can provide American citizens with the information they need to protect themselves from terrorist threats, improving homeland security.

Combining technologies

The only device in the world capable of detecting both chemical and biological threats in minutes has been built at ORNL with help from its partners. This system won an R&D 100 Award in 2000 as one of the most significant technological advances of the year. This Block II chemical-biological mass spectrometer (CBMS) is the first integrated system capable of detecting and identifying chemical warfare agents, such as mustard gas and sarin, and biological warfare agents, such as anthrax spores.

ORNL researchers in the newly formed Computational Sciences and Engineering Division (CSED), Chemical Sciences Division (CSD), and Engineering Science and Technology Division (ESTD) worked with several partners to develop the CBMS for the U.S. Army Soldier Biological Chemical Command. The goal of this five-year program, now in its fifth year, has been to develop a technology to rapidly detect and identify chemical and biological agents in the mobile battlefield. The CBMS achieves this goal by combining a high-sensitivity mass spectrometer, integrated electronics, and an expert system (using software developed by CSED).

The $45-million CBMS program, headed by CSD's Wayne Griest, is responsible for designing, building, and demonstrating prototypes and pre-production units of the Block II CBMS. Tests performed thus far at the Dugway Proving Ground in Utah and by an independent contractor at ORNL have shown that ORNL is on track for meeting program goals. "The CBMS can be easily operated by soldiers in the field," says Griest. "It also can be used for homeland security by detecting and identifying chemical and biological threats to civilian populations and determining when it is safe for people to return to evacuated areas."

CSED researchers Bob Morris and Ron Lee are working on improving the Version 4.0 software architecture for the Hazard Prediction and Assessment Capability (HPAC) Program. This program is funded by the Defense Threat Reduction Agency and Strategic Command of the U.S. Department of Defense (DoD).

HPAC is an integrated system of codes and data that provide useful information on the dispersal of hazardous nuclear, biological, and chemical materials released to the atmosphere. HPAC includes detailed, three-dimensional (3D) information on transport and dispersion of hazardous materials through the atmosphere, and their deposition to the ground. All this information is linked to worldwide population information at a one- kilometer (1-km) resolution (see next section), to predict the consequences of various releases (e.g., how many people are being exposed to potentially dangerous concentrations of these materials). The HPAC system provides estimates of acute and long-term doses of radioactive, chemical, or biological materials to the population.

HPAC is currently being used at most military command centers throughout the world. It was used during the Bosnia conflict in the 1990s; the 1996 Summer Olympic Games in Atlanta, Georgia; the inauguration of President George Bush in 2001; and studies of illnesses suffered by soldiers who participated in the Persian Gulf War of 1991. It was used at the 2002 Winter Olympic Games in Salt Lake City, Utah.

In March 2001, Jim Kulesz of CSED chaired an internal quality-assurance surveillance of the HPAC project. Familiar with the CBMS development, he believed that CBMS could be modified to measure the plume concentrations of biological and chemical warfare agents it detects. This information could then be entered into modified HPAC air-dispersion models to predict future migration rates and concentrations of hazardous materials in the plume (prognostics). The models could also use this information to locate the source of the hazardous plume (forensics).

After the incidents of September 11 and subsequent anthrax releases in the U.S. mail system, it became apparent that the combined CBMS-HPAC technologies could be used to enhance homeland security. Sponsors of ORNL work on the CBMS and HPAC projects are now discussing the possibility of funding a collaborative effort at ORNL to develop improved products for both battlefield and homeland security. ORNL's National Security Directorate is actively discussing the concept with military and homeland security agencies.

Assessing populations in crisis

When a hazardous material is released into the atmosphere, emergency responders to the crisis need to know who is being exposed, where aid is needed the most, how many people are leaving the affected area, and where the "refugees" are going. To help answer these and other questions, CSED's Geographic Information Science and Technology (GIST) Group has developed a world population database that has the world's finest resolution, as well as computer models and 3D visualizations that use this unique database. For this project, the group has applied and improved geographic information systems (GIS) techniques.

The GIST LandScan Population Distribution Project, funded by DoD, developed the LandScan 1998 and LandScan 2000 databases for the entire world. The LandScan database allows the user to assess how many people are present in any given 1-km2 area. CSED researchers Eddie Bright, Phil Coleman, and Budhendra Bhaduri are currently developing a very high-resolution population distribution database (LandScan USA) for the United States. LandScan USA indicates the number of people in any 90-m cell. The unique feature of LandScan USA is that in addition to residential (nighttime) population, it predicts a daytime population at the same spatial resolution. Because natural and human-induced disasters often strike during the day, the LandScan USA daytime population distribution data is critical in emergency management and response applications.

LandScan uses population numbers provided by the U.S. Census Bureau's International Program and geographic data derived from satellite imagery. The LandScan Population Distribution Model takes into account various factors—such as land cover (e.g., buildings, trees, grassy areas), nighttime lights, the slope of the ground, and the proximity to roads—to determine the likelihood that a cell has a population.

Recent terrorist attacks on the United States have made emergency responders, legislators, and the general public more aware of their need to know where people are and where they are going. "Most natural and manmade disasters strike very unexpectedly, placing vast populations at risk," says Bhaduri, the leader of the GIST Group. "The lack of any efficient advanced warning system compels emergency responders to quickly assess how far and in what direction will a contaminant release disperse, how many people are at risk, who are they, and where are they? Geographic information can significantly aid in quickly answering these critical questions for emergency planning, response, and recovery activities."

To answer these questions more accurately, CSED researchers are combining the capabilities of the LandScan population database with those of the HPAC system. In an HPAC simulation, a plume containing a disease-causing agent travels from its source through the air over a rural area toward a large city. Using its imagery and data, the LandScan database generates 3D animations of potentially affected populations. This visual tool used in conjunction with HPAC allows investigators and planners to "see" where a chemical or biological agent will disperse and which populations will be exposed.

The recent terrorist attacks also aroused concerns about resources being at risk. CSED researchers could provide useful information on the risks to the population of contaminated food and water. Currently, Bhaduri is leading a national pesticide usage impact modeling project. He and his group are assessing which of 10,000 drinking water intakes are most likely to receive pesticides used to protect agricultural crops from attack by insects and pathogens. The project makes its analyses using National Hydrography Data, National Land Cover Data, county-based pesticide usage data, and the national agricultural census.

CSED researchers also can aid authorities in reducing possible terrorist threats to U.S. transportation systems. For the Department of Energy's National Transportation Program, Paul Johnson and Richard Michelhaugh, both of ORNL's Nuclear Science and Technology Division, have developed the Transportation Routing Analysis GIS (TRAGIS) model. This routing tool can be used to determine the highway, railroad, or waterway transportation routes that pass through areas having the least population in the unlikely event of an accident that could release hazardous materials. TRAGIS also has the capability to determine alternative routes (e.g., to avoid areas considered at risk of an imminent terrorist attack).

Information processing agent

CSED's Collaborative Technologies Group, led by Thomas Potok, has created and demonstrated a new way to integrate, manage, and discover information. Called the Virtual Information Processing Agent Research (VIPAR) system, this tool gathers and integrates information from a number of open sources on the Internet. It then rapidly searches, clusters, and analyzes the combined data and presents the results visually to information analysts. This project, sponsored by DoD's Office of Naval Research, was conducted jointly with the Advanced Technologies Program at DOE's Oak Ridge Y-12 National Security Complex.

The VIPAR system is installed at the U.S. Pacific Command Virtual Information Center, which calls it a "grand-slam home run," as well as at ORNL and the Pacific Disaster Center. VIPAR users no longer have to read through large quantities of information to find the answers they seek. VIPAR graphically displays relevant new information relationships that may suggest trends of interest to military, intelligence, and law-enforcement personnel. This technology could also be applied to other text-intensive processes, such as message routing and organizing electronic mail.

Development of a virtual information processing agent was possible because of several research breakthroughs in managing newspaper information, dynamically adding and clustering new information entering the system, and graphically representing the organized infor-mation so that it is easily understandable. Thanks to the VIPAR system, meaningful information can be rapidly extracted from the vast repository of data on the Internet.