June 11, 2026

When an accident happens—such as an explosion at a factory or farm, a chemical leak at a public pool or a train derailment—local emergency responders need accurate information quickly to make public safety decisions and to efficiently deploy resources. Emergency managers also need to think about and plan ahead for public safety during large events such as this summer’s FIFA World Cup games. 

ARL’s HYSPLIT model is one of the most widely-used atmospheric transport and dispersion models in the atmospheric community. It is also a critical resource for emergency responders and planners. In the examples above, the HYSPLIT model can forecast the path and concentrations of hazardous chemicals or smoke and can even predict where materials might settle onto the ground. Emergency managers use this information to figure out where first responders should set up to be most effective, what areas should be placed under evacuation or shelter-in-place orders, and what areas need the most resources to deal with the emergency. For large public events, they can also think about different hypothetical scenarios to create plans ahead of time so they can confidently respond in the event of an actual emergency.

Emergency managers rely on HYSPLIT as part of their toolkit to support public safety at events such as the FIFA World Cup this summer. Each of the 11 U.S. host cities requested customized reports to explore hypothetical scenarios before, during, and after events in order to help them prepare. ARL is working with the National Weather Service’s Weather Forecast Offices in the host cities to respond to the needs of the emergency managers.

The data and research that feed the HYSPLIT model

There is a lot of research and measurements that go into the HYSPLIT model to produce accurate forecasts. HYSPLIT uses environmental data, including meteorological measurements such as wind speed/direction, temperature, humidity, and precipitation. It also ingests air quality information such as measurements of methane, carbon dioxide, and smoke. All of these measurements are inputs into the model to get accurate plume predictions showing where atmospheric materials are traveling and the concentrations of that material. HYSPLIT can also be used to trace backwards to find out the likely source for a particular material. This same technique was used in 1949 to trace back radioactive materials to find where nuclear tests were taking place.

ARL’s SOCCER-AQ field campaign—running from May 26 through June 21—is one example of the measurements that go into HYSPLIT and how the model is continuously improved. In this campaign, scientists at NOAA, NASA, the EPA, and state and regional agencies are working together to measure urban air quality at host cities on the east coast. The team is striving to understand how a large influx of people temporarily moving into an area affects urban air quality and mitigation strategies. To do this, they are collecting data using ARL’s Air Resources Car and NOAA’s Twin Otter aircraft, both of which are outfitted with the same instruments in order to get vertical snapshots of measurements from the ground and the air. The team will also rely on satellite data to get a more complete overall picture of the air quality.

Meteorological measurements from ground stations, such as ARL’s UrbanNet network in Washington, D.C., are another example of measurements that go into the HYSPLIT model. UrbanNet currently consists of five rooftop towers in and around the D.C. area that collect meteorological data including temperature, air pressure, wind speed/direction, and incoming and outgoing radiation. These stations allow scientists to get hyper-local data to input into HYSPLIT for increased accuracy and detail.

Testing and validating the HYSPLIT model

Another benefit of real-time, local measurements from sources like the Air Resources Car and UrbanNet, is that scientists can compare HYSPLIT outputs with actual measurements collected in the air, on the roads, and at ground stations. This allows them to validate the accuracy of the HYSPLIT model, and as scientists understand more about the processes happening in the atmosphere, they can use that knowledge to make continuous improvements to the model.

The measurements and research that go into HYSPLIT ensures it is constantly evolving and improving with test results validating its accuracy. This is one reason HYSPLIT has been in use for nearly three decades, it is tested and proven to be a valuable tool for emergency preparedness.