Detecting the relationship between summer rainfall anomalies in eastern china and the SSTA in the global domain with a new significance test method

It is suggested that the multiple samples in a correlation map or a set of correlation maps should be examined with significance tests as per the Bernoulli probability model. Therefore, both the contemporaneous and lag correlations of summertime precipitation R in any one of the three regions of Northern China (NC), the Changjiang-Huaihe River Valley (CHRV), and Southern China (SC) with the SSTA in the global domain have been tested in the present article, using our significance test method and the method proposed by Livezey and Chen (1983) respectively. Our results demonstrate that the contemporaneous correlations of summer R in CHRV with the SSTA are larger than those in NC. Significant correlations of SSTA with CHRV R are found to be in some warm SST regions in the tropics, whereas those of SSTA with NC R, which are opposite in sign as compared to the SSTA-CHRV R correlations, are found to be in some regions where the mean SSTs are low. In comparison with the patterns of the contemporaneous correlations, the 1 to 12 month lag correlations between NC R and SSTA, and those between CHRV summer R and SSTA show similar patterns, including the magnitudes and signs, and the spatial distributions of the coefficients. However, the summer rainfall in SC is not well correlated with the SSTA, no matter how long the lag interval is. The results derived from the observations have set up a relationship frame connecting the precipitation anomalies in NC, CHRV, and SC with the SSTA in the global domain, which is critically useful for our understanding and predicting the climate variabilities in different parts of China. Both NC and CHRV summer R are connected with El Nino events, showing a ‘− −’pattern in an El Nino year and a ‘+ +’ pattern in the subsequent year.