ESRL helps develop a response system for airborne threats
ESRL’s Rich Jesuroga pulled up a map of Fort Worth, Texas on his computer, and clicked the mouse once in a suburb. “Suppose there’s a toxic release here,” he said.
Within one minute, Jesuroga, a researcher in ESRL’s Global Systems Division, learned that prevailing conditions would swirl the imaginary airborne toxin west. Within two minutes, he traced a polygon over the “hottest” part of the dispersing plume—the people there would need to be evacuated first—and he pointed to an icon that, when clicked, would send out a warning through a variety of systems, including the Federal Emergency Management Agency’s (FEMA) Integrated Public Alert and Warning System.
This is Draft One of a collaborative NOAA project called the Geo-Targeted Alerting System, or GTAS, newly funded by FEMA for up to five years. ESRL researchers and colleagues at NOAA’s Air Resources Laboratory and the National Weather Service hope to learn how scientists can best support the emergency managers who must plan for and react to airborne threats, from tornadoes and flash floods to chemical releases.
“There are a lot of questions that we need answered,” Jesuroga said. How quickly can an advanced dispersion model be run at National Weather Service Forecast Offices? Is it enough to model a toxin’s spread at 15-minute intervals, or would 5-minute intervals be more helpful? What if the threat is a severe storm? Precisely what kind of information do emergency managers need before making warning or evacuation decisions?
“We don’t know the answers yet, and that’s why we’re doing this,” Jesuroga said.
GTAS starts with HySPLIT—the Hybrid Single Particle Lagrangian Integrated Trajectory Model—a plume dispersion model developed at the Air Resources Laboratory, ARL.
“We take the meteorological fields output from forecast models, and use those to move around particles in the airfield,” said ARL’s Roland Draxler. HySPLIT accounts for the facts that some toxins—radiological ones, for example—are heavier than others, and that toxic gases will behave differently than dusts or liquids.
Jesuroga, Draxler, and their colleagues are selecting five weather forecast offices around the country as test sites—Dallas-Fort-Worth in Texas and Seattle, Wash. are the first two—and are installing HySPLIT locally, so the same people keeping an eye on local weather can turn around and quickly run a plume dispersion model.
The HySPLIT dispersion model will be run on a platform that enables real-time collaboration—emergency managers across town can watch forecasters run the dispersion model, and forecasters can watch as emergency managers draw polygonal warning areas. That kind of collaboration could strengthen helpful relationships between forecast offices and state and local emergency managers, Jesuroga said.
Draxler said that for his team, the project represents a valuable opportunity to develop a complete warning system, from front-end forecasts to final decision-making. “We are used to running models and coming up with outputs, maybe ending up with a radiation dose. But is that important? Often, I don’t know,” Draxler said. “Here, we’ll be involved from the front to the back, and that’s useful to us.”
Jesuroga led a similar pilot warning project in Boulder a few years ago, to warn about dangerous storm conditions. His team tested a system that would send out warnings to people within high-risk polygons, and that project, too, relied on a vendor to distribute the reverse-911-type warning calls.
“Homeland Security learned about what we were doing, and their idea was, could we use this kind of system in the case of a toxic plume?” Jesuroga said.