使用隐马尔科夫模型分析了河西走廊西部地区5个站点1955—2013年共59年的夏季逐日降水数据。结果表明4个隐状态能较好地描述研究区域的夏季日降水发生模式，状态1与2描述的区域环流特征区分了地理位置对日降水模式的影响，状态3与4表征了大尺度大气环流模式下5个站点同时下雨或都不下雨的情形。5个站点的日降水模式存在显著性的区域差异，日降水发生概率呈由东向西逐渐减小的趋势。隐状态的年际变化分析表明，所有站点（包括西部站点）的降水发生概率都存在增加趋势，东部站点湿润而西部站点干燥的情景呈显著性减少趋势，这可能与全球变暖导致的夏季风北移使该区域不同站点降水模式趋于一致有关，其原因有待进一步研究。%Daily rainfall data during the June—August 1955—2013 from 5 stations located in the western of Hexi Corridor was analyzed using a Hidden Markov model (HMM). The result indicates that a four-state HMM is able to capture the patterns of variations in daily rainfall probability for the 5 stations. One paired states (states 1 and 2) describe zonal circulation patterns and demonstrate the inlfuence of geographical location on the pattern of daily rainfall occurrence while a second paired of states (states 3 and 4) describe dry-versus-wet conditions at all stations that are controlled by the large-scale circulation of atmosphere. Difference in summer rainfall pattern is characterized by signiifcant east-west gradients in rainfall occurrence probability. Increase in states 1 and 3 and a significant decrease in state 2 are identiifed, which implicates that the occurrence probabilities of summer daily rainfall in the 5 stations are increased and become uniform. This may be caused by the fact that the northward-moving of Summer Monsoon triggered by the global warming results in a more uniform circulation pattern over the studied area. The responsibility for this phenomenon deserves a further study.
展开▼
