Editor’s Note: This is one of a series of profiles of the Wisconsin Idea in action. See past profiles we have published.

Even before he started kindergarten, Jordan Gerth remembers cutting out weather forecasts from the local newspaper and comparing the predictions to reality. Gerth’s passion for weather remains today, although he’s replaced scissors and paper with some of the most powerful meteorological tools on the planet.

Since arriving at UW–Madison from Kenosha in 2005 as a freshman, the atmospheric and oceanic sciences major has taken a lead role in a project that is making campus weather research tools available to frontline forecasters in Wisconsin’s National Weather Service (NWS) regional offices.

With a shared goal of improved weather prediction, researchers from the Space Science and Engineering Center (SSEC) have been collaborating for years with state NWS forecast offices in Milwaukee, La Crosse and Green Bay to translate weather research into more accurate forecasts. But the fresh perspective provided by Gerth helped the team give a facelift to the ongoing collaboration. Gerth developed an intricate software interface that gives NWS forecasters direct computer access to UW–Madison’s wealth of satellite meteorology resources.

Undergraduate research assistant Jordan Gerth (in black) and Bob McMahon, head forecaster for the Milwaukee/Sullivan office of the National Weather Service (NWS), review MODIS sea surface temperature data at the NWS office in Dousman, Wis. Gerth helped create an outreach campaign between UW–Madison’s Space Science and Engineering Center and the NWS offices in Wisconsin, helping NWS integrate cutting-edge technology into its forecasts

Photo: Jeff Miller

“The project would be nothing if it were not for Jordan’s motivation, hard work and expertise,” says UW–Madison research meteorologist Scott Bachmeier, who helps coordinate the scientific aspects of the collaboration.

As a freshman two years ago, Gerth joined SSEC’s Cooperative Institute for Meteorological Satellite Studies (CIMSS) as an undergraduate researcher. Gerth used his past experience as a high school intern at the Milwaukee forecast office to reinvigorate the center’s NWS connections.

Forecasters improve their predictions by comparing different types of weather data, such as details about clouds, rain, snow, temperature and wind speed. This data comes from many sources, including information from weather stations, measurements from weather balloons and observations from satellites. To ensure that forecasters can easily access UW–Madison’s diverse array of satellite data, CIMSS researchers need to put the information directly into the National Weather Service’s unique processing software system, the Advanced Weather Interactive Processing System (AWIPS).

This system allows forecasters to simultaneously view images created from several different sources, allowing detailed comparisons that can indicate which forecast technique will give the most accurate prediction starting from the current weather conditions. Accommodating this system was a significant hurdle faced by the CIMSS team.

That’s where Gerth came in. As soon as he arrived, he began to pick apart the software system to get it working on his personal computer so the group could patch UW–Madison satellite data into the NWS system.

At the National Weather Service, several different machines handle various tasks. In addition to condensing the system onto one machine, Gerth also had to figure out the messages that tell the system how to process the data. “The data that feeds AWIPS has to be in a very special format,” Gerth says.

Not even the folks at the weather service know exactly how the computer system works, so Gerth had to guess. In June, almost two years after he started, Gerth finally cracked the last bit of code. “It’s like you have a shape that has 24 little points on it and you have a hole with 23 little points on it,” Gerth explains. “That just right fit is needed. There’s only one fit, and it has to be perfect.”

However, after less than six months of working with the software, Gerth had enough of the system working to allow scientists at CIMSS to begin testing different types of processed satellite data in a format useful to the NWS forecasters. By spring 2006, the team successfully patched into the pipeline of data available to operational forecasters and began transmitting weather information from satellites.

The forecast office near Milwaukee has served and continues to serve as the primary testing ground for transitioning UW–Madison research into operational forecasting. During the initial phase, the group at CIMSS began sending images generated from an imaging instrument aboard a satellite that circles the Earth from pole to pole.

A fast favorite of weather forecasters is an infrared image generated several times a day that depicts temperatures on the surface of Lake Michigan. This image helps forecasters predict snowfall associated with proximity to Lake Michigan as well as the intensity of the wind off the lake, the change in temperature along the shore, and the height of the waves on the lake itself. Since NWS began providing data to the Milwaukee forecast office, according to Craven, the satellite image of temperatures over Lake Michigan has become a primary tool used by forecasters to assess conditions on and around Lake Michigan.

Another source of weather information provided by UW–Madison is an experimental forecasting tool called the Meteorological Satellite Studies Regional Assimilation System, or CRAS. Forecasters in the Milwaukee office use CRAS to forecast how much cloud cover to expect over the next few days. “CRAS gives us a forecast of cloud cover we wouldn’t have normally,” Craven says.

According to Kim Licitar, the information technology officer at the Milwaukee forecast office, Gerth is the only person outside of the National Weather Service to successfully run the software system. “It was quite a jigsaw puzzle for Jordan to put together,” Licitar says. Because of Gerth’s work, CIMSS can develop versions of their research products specifically for distribution within the system at the National Weather Service. “No one else has taken this kind of approach,” Gerth says.

The utility of the satellite observations provided by CIMSS is evident in the NWS area forecast discussions, in which the forecasters describe the different information used to make a particular forecast as accurate as possible. Gerth also spends time talking to forecasters and encouraging them to tell forecasters at other offices about UW–Madison satellite data. While the area forecast discussions and Gerth’s frequent communication with the forecasters provide an indication of how forecasters use the data, there is one way to know for certain, according to Craven: “When the data is missing, the forecasters complain.”

Wisconsin has three National Weather Service forecast offices: Milwaukee, La Crosse and Green Bay. NWS currently provides data to Milwaukee and La Crosse, and the team has started to work with the Green Bay office. In addition to the Wisconsin offices, several locations around the country benefit from NWS information, including forecast offices in Indiana, Iowa, Wyoming, Montana and South Dakota.