Evaluation of circulation-type classifications with respect to temperature and precipitation variations in the central and eastern Tibetan Plateau

摘要

Circulation-type classifications (CTCs) are useful tools for systematically describing atmospheric circulation and studying its variation and impact on surface climate. However, few circulation classification methods (CCMs) have been applied in the Tibetan Plateau (TP). In this study, six frequently used CCMs using 500hPa geopotential height data were assessed with respect to their abilities in explaining observed daily mean temperature and precipitation variations in the central and eastern TP for the period 1980-2014. Two statistical measures, the explained variance and the Kolmogorov-Smirnov test, were used to quantify the performance of the CTCs in describing climate variable variations. SAN (simulated annealing and diversified randomization clustering) and KRZ (Kruizinga's eigenvector-based scheme) were identified as optimum CCMs for synoptic typing over the study region. The daily mean temperatures are well conditioned on SAN-CTC with 43.9-62.4% explained variance at annual scale and 11.1-25.2% at seasonal scale. For daily precipitation variations, the explained variance at the annual scale is less than 40%. The spatial characters of climate variation affected by large-scale circulation are identified. Daily mean temperature variations in the Qaidam Basin and the Qilian Mountains (90 degrees-103 degrees E and 35 degrees-40 degrees N) all year round and in the region between the Tanggula Mountains and the Himalayan Mountains (88 degrees-97 degrees E and 30 degrees-34 degrees N) in wet season (May-September) are well resolved by the CTCs. Daily precipitation variations in the eastern TP in the dry season (October-April) and a region from the Brahmaputra River basin to the Tanggula Mountains (85 degrees-92 degrees E and 28 degrees-32 degrees N) and the southern Hengduan Mountains (98 degrees-102 degrees E and 27 degrees-31 degrees N) in the wet season show a close relationship with the identified CTCs by SAN.