NMSU branding

New Mexico State University

New Mexico State University

News Center




Researchers to develop flood-prediction capabilities

A team of New Mexico State University researchers has been funded by the National Geospatial-Intelligence Agency (NGA) to develop tools for rapid monitoring and prediction of water levels in the Rio Grande Basin.



Max Bleiweiss, director of NMSU's Center for Applied Remote Sensing in Agriculture, Meteorology and Environment, works with student Phillip Lujan on software for monitoring and predicting water levels. (NMSU photo by Ben La Marca)

The agency, a part of the Department of Defense, is providing $450,000 over three years for the project through its University Research Initiative program. The researchers will enhance existing software tools and develop new techniques for using remote-sensing data to monitor water levels and predict events such as flash floods with precision not currently possible.

The project will create a hydrologic model of the Rio Grande Basin from El Paso north to the river's headwaters in southern Colorado, said Bill Stein, senior imagery analyst with NMSU's Physical Science Laboratory and one of the lead scientists on the project.

Using 25 years' worth of meteorological and hydrological data, plus continuous updates from satellite imagery and other remote-sensing techniques, the researchers aim to develop tools to monitor and predict flows in real time.

"I'm not aware of any operational capability in the country right now to do flash flood forecasting with the level of detail that we're going to attempt," said Max Bleiweiss, a scientist in the NMSU College of Agriculture and Home Economics and co-principal investigator on the project. "With the data we'll have available to us, we should be able to come up with some fairly pointed forecasts."

A question still to be answered, Stein said, "is how rapidly can you do that?"

Data from the devastating flood that occurred in Hatch last year, as well as other flash flooding in the area in recent years, will provide useful test cases as the scientists develop the prediction capabilities of the software tools, Bleiweiss said.

The tools, which will be adaptable to other river basins and reservoirs, will be made available to the NGA for use in disaster monitoring and crisis response situations.

"In addition to supporting the Defense Department, the NGA has been involved in some of the environmental disasters we've had in the United States in recent years, like Hurricane Katrina," Stein said. "They provide imagery and help in recovery efforts. So to have software tools that could be used to predict flooding in the United States or in areas where troops might be would be very helpful to the NGA."

The tools should have other applications in managing water resources as well as predicting and responding to problems, the researchers said.

Stein is a former NGA research scientist who joined PSL after retiring in 2005. Bleiweiss, director of NMSU's Center for Applied Remote Sensing in Agriculture, Meteorology and Environment, was a physicist with the Army Research Laboratory from 1992 to 2005. The NMSU team also includes Thomas Schmugge, the Gerald Thomas Professor of Water Resources at NMSU and former senior researcher at the U.S. Department of Agriculture's Hydrology and Remote Sensing Laboratory.

The group plans to collaborate with the PSL's unmanned aerial vehicle (UAV) program, with Sandia National Laboratories, and with international consultants.

Stein said the project will be undertaken in three stages.

"The first stage is to get the software up and running, and perhaps to put a sensor on a UAV to take remote sensing of the Rio Grande Basin that we can put into the software," he said. "The second part is to look at using high-resolution commercial satellite imagery. Third is to work with Sandia to get radar data in the model and see how radar does."

The radar is a type known as synthetic aperture radar, or SAR. "Sandia has the capability to fly the SAR on an airborne platform," Bleiweiss said.

As important as the real-time data will be, the researchers said, exhaustive information on pre-existing conditions will play an equally crucial role in the system's predictive capabilities.

"You need to know what the basin conditions are prior to a rainfall event," Bleiweiss said. "You need to know what the states of the soils are and what the land cover is doing and so on. If you can update that on a daily basis, when the rainfall then occurs, you can input NEXRAD weather radar data and other kinds of information and forecast the flood event."