Cluster Analysis Applied to the Spatiotemporal and Climatological Variability of Precipitation over China

Rainfall is one of the important climate variables, as it directly affects agriculture, livestock, electricity generation, supplies. And, with the objective of performing a cluster analysis, define homogeneous regions of precipitation for China. For this, the TRMM time series from 1998 to 2017, monthly frequency, was used. With this, a space-time cluster analysis was carried out for China. For verification and validation of these clusters, a linear regression of the homogeneous regions was elaborated with the help of multiple linear regression and its level of significance. The results showed that 8 clusters were needed, ranging from 7.2% (cluster 5) to 18.4% (cluster 3) of China's territory. Although 70.96% of the China region obtained R² greater than 0.99, and 94.88% of its area with R² greater than 0.9, with a valid significance level for this region. Therefore, this work shows an application of cluster analysis with the aid of TRMM data, corroborating to improve the climate understanding of the region.