Washington State University’s automated weather station network, is testing a national forecasting model as a tool to predict weather systems, such as hail or freezes, in the state’s tree fruit and grape growing regions.

AgWeatherNet, a network of 137 automated weather stations across Washington State, provides current weather conditions updated every 15 minutes, but does not predict air temperature, dew points, wind speed or wind direction. Adding weather predictions would allow growers to make more informed decisions, says Dr. ­Gerritt Hoogenboom, network director.

The forecasts could be integrated into models relating to growing degree days, irrigation scheduling, cold hardiness of grapes, and pest and disease development that are available through the AgWeatherNet.

Hoogenboom hopes to use the state-of-the-art Weather Research and Forecasting model as a tool for predicting extreme weather in Washington. The WRF is a computer program created by the National Oceanic and Atmospheric Administration, the National Center for Atmospheric Research, and other organizations and ­universities in the United States and abroad.

The WRF model can provide hourly weather predictions at a high resolution. Hoogenboom and his colleagues are evaluating how accurately it would have predicted three actual freezes experienced in Washington in late February, early April, and late October 2011. Predictions are made on a grid basis. Because of Washington’s varied terrain, with abrupt changes in elevation, the scientists intend to use a grid as small as two miles square. This requires a large amount of computing power, and data are being processed on a high-performance computer purchased with part of a $95,000 grant from the Washington Tree Fruit Research Commission.

Preliminary results show that model performed well for low elevations in eastern Washington and for 24- to 48-hour predictions. However, it did poorly for longer-term predictions and for medium- and high-elevation sites.

Hoogenboom said results were encouraging, though coming up with accurate predictions in central Washington’s varied terrain is a challenge. “I know everyone wants to have an accurate forecast for two weeks out, but the reality is it’s not possible right now. What we’re trying to do is add value to what we currently provide in AgWeatherNet.”

Hoogenboom said the project will continue for at least two more years, as scientists develop protocols to improve the accuracy of the WRF model in Washington.