The United States filled a crucial gap in its weather forecasting arsenal by launching its latest geostationary satellite on March 1. The craft will allow meteorologists to track hurricanes, snowstorms and other threats as they develop. It will also transmit data that researchers can use to measure air temperature and humidity – if they can figure out how to incorporate it into their models.
Scientists currently cannot use much of the information collected by geostationary satellites, which hover above a particular location on Earth, and polar-orbiting satellites, which orbit the planet’s poles. It’s a long-standing problem caused by the type of data collected and the large uncertainties that arise when forecasters attempt to incorporate the measurements into their weather models. Today, researchers are beginning to overcome these technical challenges, with encouraging results for both short- and long-term predictions.
The Geostationary Operational Environmental Satellite-17 (GOES-17) will position itself over the equatorial Pacific Ocean. When its data is combined with that of the same GOES-16, which is already stationed over the Atlantic Ocean, they will monitor Earth from Africa to New Zealand. Meteorologists around the world use these geostationary satellites to monitor storms, and their models incorporate limited data on atmospheric humidity and wind speed and direction.
“There’s a huge treasure trove of information,” says Fuqing Zhang, a meteorologist at Pennsylvania State University at University Park. He has experimented with incorporating some of this unused data from satellites into his models, with promising results. “We can show dramatic improvements in weather forecasting, but you need a dedicated research effort.”
In a study currently being reviewed at Bulletin of the American Meteorological SocietyZhang and his colleagues show that incorporating high-resolution data from GOES-16 into an experimental weather model strengthened predictions of the early development and intensity of Hurricane Harvey, which hit Texas in August.
Without the additional data, one forecast called for the storm to become a Category 1 hurricane; in fact, it became a Category 4 monster before it hit the ground. Zhang also tried to include the additional information in a new weather model that the US National Weather Service plans to deploy as early as this year. This additional data improved predictions of precipitation amounts and storm track.
Incorporating this type of data into models has been difficult in part because geostationary data provides fewer measurements for a given vertical slice of the atmosphere than polar orbiters, which circle Earth at lower altitudes. This means researchers have less information and greater uncertainties when it comes to translating the data into measurements that models can use, such as temperature and air humidity.
“It’s not trivial,” says Dan Lindsey, a research meteorologist with the US National Oceanic and Atmospheric Administration in Fort Collins, Colorado, who works on the current series of GOES satellites. “You can’t just take a satellite image and insert it into the model.”
The science was slow to evolve on this because there was less demand for a constant stream of data when forecasting models were only run every six hours, says Jason Otkin, atmospheric scientist at The Cooperative Institute for Meteorological Satellite Studies at the University of Wisconsin at Madison. . Today, agencies are moving towards more frequent forecasting, using models that can take advantage of larger amounts of high-resolution data.
“In fact, the value of these geostationary sensors is only increasing over time,” says Otkin.
Meteorologists have also struggled to integrate measurements of cloud areas taken by polar-orbiting satellites. This is because clouds have more complex microphysics than the open sky, so even small errors in models can spill over into large uncertainties in forecasts. And that’s the fundamental problem, says Alan Geer, an atmospheric scientist at the European Center for Medium-Range Weather Forecasts (ECMWF) in Reading, UK. “These are the areas with clouds and precipitation that are associated with the most interesting weather patterns.”
The ECMWF has been leading the way in this field for more than a decade and now integrates much of the cloud region data taken by polar-orbiting satellites; most major government forecasting centers are following suit1. Zhang cites an unpublished analysis comparing the European model and the new model from the US National Weather Service. The US model performed on average as good or better than the European model when using the full suite of atmospheric data from ECMWF. But when the researchers ran the same model with the usual data from the US forecasting program, it failed.
The lesson for the United States is that satellites and models are not enough, says Zhang. “Our country has invested so much money in launching beautiful satellites, but we haven’t really put that much effort into integrating satellite information into the models.”