ESTIMATING MOST PROBABLE PRAIRIE WHEAT PRODUCTION FROM PRECIPITATION DATA

A method was developed for periodically assessing the quantitative effect of antecedent precipitation on prairie wheat production as the growing season advanced. The method took into account the conservation of rain and snow on summerfallow, variation due to major soil types, and the seasonal and areal distribution of precipitation. Regression coefficients were calculated using 10 or 11 years of yield data from eight groups of crop districts in the prairies. Coefficients of determination for calculated provincial yields at the ends of June and July were statistically significant. The standard error of estimate for prairie yields at the end of July was 1.5 bu/acre compared with the standard deviation of 4.7 for actual average yield.