西南地区气象资料测试、预处理和加工研究报告

肖天贵

摘 要：西南地区是我国西部山区大型滑坡灾害的主要发生地，因此西南地区的气象资料所映射出的信息对大型滑坡致灾因子的识别和预报预警方法的研究具有非常重要的意义。由于获取的大量原始气象资料的存储格式、观测时次疏密、分辨率等均存在差异，因而造成应用过程中的困难和局限，往往需要对实际观测的多源资料进行测试和预处理。（1）通过四川、重庆、云南和贵州4省市逐日常规气象观测资料的统计加工处理，获得了西南地区各时间尺度的降水趋势、地理分布以及各级别降水量的气候概率值。（2）由于西南地区地形复杂，局地小尺度的气象要素特征在常规气象观测资料中可能会被平滑或显示不明显，因此利用观测网格更为细致的逐小时区域自动站的资料就非常必要，利用收集的西南地区2 469个区域自动站资料分析了局地小尺度的气象要素分布及变化特点。（3）运用收集到的风云二号（FY2）卫星云图资料和反演的降水估计产品，进一步分析了西部地区尤其是西藏西北部资料空白区域的降水分布；卫星资料反演降水产品与实况降水对比相似率在60%以上，精度达到40%左右。（4）通过四川凉山州对流性降水案例的雷达图像资料的加工处理，分析了西南地区雷达回波资料在强降水天气及滑坡和泥石流灾害发生时的环境条件和临近预报预警的可行性。（5）利用发生在四川盆地边缘山地的两次降水过程进行WRF数值模式的敏感性试验和环流特征分析及动力特征等分析。模拟再现了降水的雨带走向、形状特征、雨区位置、暴雨中心位置以及中尺度辐合线、中尺度低涡等中尺度系统活动过程，表明WRF模式的敏感性和应用性。

关键词：西南地区 降水量 降水反演 短临预报 数值模拟

Abstract： Heavy landslides mainly take place in southwest regions in western mountains of China. Therefore，meteorological data in southwest regions can map out the messages to identity factors or forecasting and warning of heavy landslides. The difference in the format of meteorological data， the density of the observation time and resolution causing difficulties and limitations in the process of application，we need to do some test and pretreatment of the multi-source data. （1）Counting and processing the daily routine meteorological data of Sichuan， Chongqing， Yunnan and Guizhou to get the trend of precipitation in each time scale， geographical distribution and the climate of precipitation probability value at all levels. （2） Due to the complexity of terrain in southwest China， so we need to make use of the hourly data in more detailed zone automatic station to compensate for the smooth and unconspicuous of routine data； Using the collected 2469 automatic station data in southwest to analysis the meteorological elements distribution and changes of small scale system. （3）By using the cloud pictures from FY2 satellite， the estimation products that inversion of precipitation， and the raw meteorological data received from radio satellite， try to filled in the blank space data of the regional where the automatic weather station is unrealized； compared the inversion of precipitation with actual precipitation， the similar rate over 60% and the accuracy is about 40%. （4）An analysis of radar image about a convectional precipitation in liangshan was did. Through the processing and handling the radar echo of heavy rain， analysis the environmental conditions of the landslide and debris flow disasters and the feasibility of geological disaster forecast warning. （5）Select two precipitation process occurs on the edge of the Sichuan basin mountain to proceed the numerical model WRF sensitivity test， and do the circulation characteristics analysis and the analysis of dynamic characteristics. The numerical model successfully reproduces the stretching direction of the precipitation， shape characteristics， precipitation area and the location of rainstorm center， and the mesoscale convergence line and mesoscale low vortex. It shows the sensitivity and applicability of WRF.

Key Words： Southwest Area； Precipitation； Precipitation inversion； The short forecast； Numerical Simulation

阅读全文链接（需实名注册）：http：//www.nstrs.cn/xiangxiBG.aspx？id=50459&flag=1