文摘
To assess the drivers of weather extremes over the Canadian prairies, a data matrix of 19 predictor and 11 predictands types was created at the monthly timeframe to explore drought severity, summer precipitation, and summer temperature. Applying composite, correlation, and regression techniques, a comprehensive data analysis produced a suite of composites and regression models for providing climatic outlooks a few weeks to a few months in advance of the critical May–July growing season for spring wheat. This was done for the prairies as a whole and for four agricultural ecological (agro-eco) zones. Among the most important predictors were the Madden–Julian Oscillation (a tropical atmospheric oscillation between 20°E and 10°W longitude) and Pacific Decadal Oscillation a slow moving oscillation covering the entire Pacific Basin both linking large-scale predictors and solar parameters like the Averaged Planetary Index. An empirical approach, using accumulated monthly values of atmosphere–ocean indices, provides useful guidance for the forecasting of summer precipitation and temperature, and hence, grain yields with a lead time of a few weeks to a few months.