Submitted to: American Phytopathological Society Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: March 15, 2009
Publication Date: August 1, 2009
Citation: Coop, L., Fox, A., Mahaffee, W.F., Gent, D.H., Pfender, W.F., Thomas, C. 2009. Forecast and virtual weather driven plant disease risk modeling system. American Phytopathological Society Abstracts. p. 873. Technical Abstract: We describe a system in use and development that leverages public weather station data, several spatialized weather forecast types, leaf wetness estimation, generic plant disease models, and online statistical evaluation. Convergent technological developments in all these areas allow, with funding from CSREES (several competitive grant programs) and other sources, the creation of a broad scale decision support system that can serve the United States, particularly for plant biosecurity needs, and with forecasted weather, the western United States for some IPM needs. Significant developments of this system include: a) Public/private partnership in downscaling weather forecasts down to 2km resolution, out 6–7 days, at hourly time steps, b) Real-time data ingest of hundreds of weather networks and over 12,000 weather stations, c) Use of a near real-time data quality assurance system, d) Use of weighted distance-elevation regression of nearby weather for virtual weather estimation, e) Online statistical evaluation software, and f) A generic suite of 10+ plant disease risk models. Thus far, mean absolute error rates for virtual weather have averaged as low as 0.9°C for hourly temperature, and 1.4°C for dew point. The system shows promise for increasing adoption for IPM in western states, and can provide national decision support for plant biosecurity applications.