资源环境科技发展态势分析平台

Climate oscillations are one of the primary factors that influence precipitation. This study uses canonical correlation analysis (CCA) to examine how El Nino-Southern Oscillation (ENSO), Atlantic Multidecadal Oscillation, North Atlantic Oscillation, Pacific Decadal Oscillation, and the Pacific-North American pattern influence precipitation in Texas. This study identifies the months, regions, and time lags where the relationships between climate oscillations and precipitation are strongest. Correlation results indicate that ENSO accounts for the greatest amount of precipitation variance in Texas. However, including all five climate oscillations is important and together they account for a greater amount of the variance in precipitation than any individual climate oscillation. Precipitation in southern Texas is more strongly influenced by climate oscillations than other regions in Texas. The CCA results demonstrate that there are statistically significant relationships between the climate oscillations and precipitation at time lags longer than 6months during the summer and at time lags shorter than 6months during the winter. Based on the CCA results, a precipitation forecast model was developed for the three climate regions that we defined. In the cases of January, the Heidke Skill Score (HSS) of our model is comparable or higher to those achieved by the Climate Prediction Center (CPC) in each region. For all of the 36month/region cases (12monthsx3 regions), there are 50% cases that the HSS of our model is comparable or higher to those achieved by the CPC. The results of this study illustrate that including multiple teleconnections can increase forecast skill, and statistical methods are useful for precipitation forecasting at a 0-month lead time.