More farmers are putting up their own weather stations and utilizing new technologies and sources of data to develop information tailored to their specific areas.
A key goal of agricultural weather forecasting, says Joe Russo, is to bring the predictions down to local-farm scale on an hourly, daily, or longer basis, customizing the forecasts to specific applications such as pesticide and fertilizer applications, irrigation, and other decisions.
"New technologies with more portability, combined with more sources of weather data, are providing agriculture with new approaches to forecasting," he told members of the Mississippi Crop Improvement Association at their annual meeting at Tunica, Miss.
As an example, high-resolution data from satellites are being used to develop precipitation, cloud cover, and other information on a localized basis.
Russo, who is with ZedX, Inc., Bellefonte, Pa., said the E-Weather service provided by a company affiliate, SkyBit, Inc., offers a wide variety of products to assist decision-making in the field, including integrated pest management simulation and forecasting, irrigation scheduling, frost predictions for select crops, along with custom data for other commodities and modeling expertise for crop and pest development studies.
Producers, he says, are increasingly using portable stations that transmit field information that can help them better understand local conditions and better predict frost, and monitor cloud cover, wind direction and speed, crop canopy temperatures, etc. Micro-environment readings in crop fields and orchards can be used not only to aid forecasting but for better-informed production decisions.
Data collected can be used to "train" computer models for even more reliable forecasting. Among other uses, information collected can be combined with field grid samples to develop precise irrigation scheduling, as well as estimating how much water will be needed. Predictions can also be made for crop disease development and progress.
"There are a lot of pieces in today's agricultural production and management system," Russo says. A cooperative, integrated system of data collection and processing, decision-making, and record-keeping can help tie everything together, in the process refining farm computer models to make them more accurate and useful.
Even though technological advances and greater computer power have improved weather data generation and forecasting, he says forecasting remains "very challenging because it is still, to an extent, unpredictable."
Forecasters, Russo says, can now predict long-lasting weather systems, such as El Niños, hurricanes, etc., pretty well. "It's much harder to predict short-duration systems that have a significant impact on farming operations."
The big push, he says, is for monthly and seasonal forecasts, with many government, private, and university researchers working to increase the accuracy of weather predictions. Among the new approaches, Russo says, is the "meteorologic ensemble," which combines a number of weather models to get a consensus picture that can offer more accuracy than a single model.
The SkyBit system, he says, incorporates global, regional, and other models to develop forecasts at the local level. That weather data can then be put into disease models and used for other agricultural applications.