Different spatial cross-correlation patterns of temperature records over China: A DCCA study on different time scales

Daily mean temperature records over China during the past 50 years are studied by means of detrended cross-correlation analysis (DCCA). Taking Beijing as a center, we calculate the DCCA cross-correlation coefficient σDCCA between the temperatures in Beijing and those in other stations. After a statistical significance test, spatial cross-correlation patterns on different time scales are shown in this paper. We find the spatial cross-correlation patterns can vary with time scales. On small time scale of one week to one month, only the temperatures in nearby regions have close relations with that in Beijing, while on larger time scale of intra or inter-seasonal, temperatures in most of the regions, especially in the northeast show high level cross-correlations with that in Beijing. The southwest plateau (including the Tibetan Plateau and the YunGui Plateau) is a special region, where the temperatures take on significant anti-cross-correlations on inter-seasonal scale, but no significant correlations on inter-annual scale. By analyzing these different spatial patterns, we can better understand the influencing climatological processes of different scales. Therefore, DCCA are recommended as a reliable method in detecting the relations between two climatological variables, and further be useful for our understanding of the whole climate system.