Research to improve weather forecasting accuracy and honing prediction tools continue to assist farmers in making a variety of management decisions.

During a breakout session of the 2012 California Association of Pest Control Advisers (CAPCA) conference, attendees heard about the latest work being conducted by Walter Mahaffee, research plant pathologist at the USDA-ARS Horticultural Crops Research Lab in Corvallis, Ore.

Mahaffee and a large team of researchers are working on virtual weather data to improve disease forecasting systems and how inoculums detection technology can be used to determine fungicide application timing.

He cited the use of the Gubler/Thomas powdery mildew risk assessment model, developed in California, as an example of how a disease or pest model can be incorporated to time applications using virtual weather data.

“What we try to do is take weather patterns and scale them down to make decisions,” he said of the project which involves integrating information taken from more than 15,000 stations across the U.S. The exact number of stations is very fluid as the data streams from all of them are constantly monitored. Some need to be removed due to no longer meeting operational standards and conversely others are added as they become ready to go online. The group that works on this system periodically encounters issues that can arise from working with all the stations including data acquisition, storage, and processing information. Occasionally they have also had to limit users in the number of stations because of the server limitations ability to process the data.

Mahaffee suggested that those making management decisions go to the site he discussed during his talk to create their own virtual weather station based on their specific latitude and longitude. However, he added it is still a good idea to have a rain gauge in the field and visually monitor weather.

“There will always be issues with forecast accuracy since they are predictions and there will always be error associated with those predictions,” Mahaffee said.

As researchers continue to work on refining forecasting tools, Mahaffee said, investment in high resolution public weather monitoring would significantly benefit all aspects of society.

“The accuracy and spatial resolution of forecasting will continue to improve as the speed and cost of computing continues to decrease,” he said. “However, it will always be dependent on the availability of weather data.”

For those wary of using precision forecasting tools, Mahaffee encourages them to give it a try.  He recognizes that the site may give the appearance of being more for tech geeks, but he said they have plans to make it more user-friendly as resources become available.

“I hope folks start to develop confidence in using pest and disease forecasting to make more informed management decisions and that the information increases the economic and environmental sustainability of their businesses,” he said.

To create an account in the virtual forecasting tool used in Mahaffee’s presentation, go to: http://uspest.org/risk/models?id=usevertok.