SPATIAL VARIABILITY AND DOWNSCALLING OF PRECIPITATION

Authors: ROSSEL, FREDERICK - FORMER POST-DOC; Garbrecht, Jurgen

Submitted to: Physics and Chemistry of the Earth
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/9/2002
Publication Date: 5/9/2002
Citation: N/A

Interpretive Summary: NOAA is issuing monthly and seasonal precipitation forecasts for large regions that may cover several crop reporting districts. The forecasted precipitation represents average precipitation for the region as a whole. However, for individual crop reporting districts within the region and at individual field locations, the actual amount of precipitation may vary. To assess the reliability and effectiveness of regional precipitation forecast at the farm, ranch and field scale, it is necessary to determine the difference between precipitation at the different geographical scales. This difference has been estimated here for the central Oklahoma region. Monthly precipitation differences between a field location and a crop reporting district can reach up to 50% of the mean year-to-year variation in precipitation. As the length of the time interval over which the precipitation is measured increases (monthly, seasonal, yearly), the range of precipitation differences between field and crop reporting districts decreases. As the size of the area under consideration increases from the crop reporting district to the region, the precipitation difference also increases. The findings of this study suggest that large differences may exist between the regional forecasted precipitation and the precipitation at a location within the region. The range of precipitation differences has been used to define a measure of confidence for application of regional forecasted precipitation at a field location.

Technical Abstract: To fully take advantage of regional climate forecast information for agricultural applications, the relationship between precipitation characteristics at regional and local scales must be quantified. The spatial variability of precipitation within a region is defined as the difference between the standardized values at these two scales. The spatial variability is quantified at the monthly, seasonal and yearly time scales for the Central Oklahoma "super climate division" and the four climate divisions within the super-climate division. For the super-climate division, the standard deviation of the differences between station and regional standardized values ranged from a minimum value during the winter around 0.55 to a maximum value during the summer around 0.80. The decrease in spatial variability associated with an increase in time scale from month to season is about 35%, and about 45% as one increases the time scale from season to year. On the other hand, the decrease of spatial variability associated with the decrease in space scale from super-climate division to climate division is about 35%. An analysis of the distribution defines the confidence envelope for local precipitation around the regional value. This study demonstrates the critical influence of the spatial variability of precipitation and the additional uncertainty that is introduced in downscaling regional information to local applications.