降水类天气现象自动与人工观测质量对比分析  

Quality analysis of automatic and manual observation data of precipitation weather phenomenon instrument

作　　者：周坤论 黄剑钊[1] 陶伟[1] 李艳萍 李强[1] Zhou Kunlun;Huang Jianzhao;Tao wei;Li Yanping;Li qiang(Guangxi Meteorological Technology Equipment Center,Nanning 530022,China;Guangxi Meteorological Information Center,Nanning 530022,China)

机构地区：[1]广西壮族自治区气象技术装备中心,南宁530022 [2]广西壮族自治区气象信息中心,南宁530022

出　　处：《气象研究与应用》2022年第1期112-117,共6页Journal of Meteorological Research and Application

基　　金：广西壮族自治区气象技术装备中心自立科研项目(降水过程雨滴谱的统计特征研究)。

摘　　要：选取2017年8月-2018年7月广西DSG1、DSG4和DSG5型降水类天气现象仪自动观测数据与人工观测数据,通过对比分析,研究降水类天气现象仪自动观测与人工观测数据质量的情况。结果表明,广西91个国家站点数据完整性好,设备年平均缺测率小于2%,DSG5型设备数据完整性最高(缺测率0.01%)。DSG1、DSG4和DSG5型设备对降雨的捕获率较高,均超过80%,DSG5型设备捕获率最高(90.18%),且吻合率较好(大于50%),错报率较低。三类设备对雪、雨夹雪、冰雹的识别均较困难,将雨误判为雪、雨夹雪和冰雹的分钟数较多;其中DSG1型设备容易误判为冰雹,DSG5型设备容易误判为雪。三类设备在雪、雨夹雪、冰雹等微降水和混合类降水方面进行更好的智能识别和质控算法优化,将会提高其业务应用效能。The automatic observation data of DSG1, DSG4 AND DSG5 precipitation weather phenomenon in Guangxi from August 2017 to July 2018 were selected for sorting, statistical analysis and comparison with the manual observation data. The results show that the data integrity of 91 national stations in Guangxi is good, the annual average missing rate of equipment is less than 2%, and the data integrity of DSG5 equipment is the highest(missing rate is 0.01%). The DSG1, DSG4 and DSG5 devices have higher capture rates of precipitation,all exceeding 80%, the DSG5 device has the highest capture rate(90.18%), the coincidence rate is good(higher than 50%), and the false rate is low. It is difficult for the three types of equipment to identify snow, sleet and hail, and it is easy to misjudge rain as snow, sleet or hail;in particular, the DSG1 equipment is easy to misjudge rain as hail, and the DSG5 equipment is easy to misjudge rain as snow. It is recommended that the three types of equipment perform better intelligent identification and quality control algorithm optimization in terms of snow, sleet, hail and other light precipitation and mixed precipitation.

关 键 词：人工观测 自动观测 降水类天气现象仪 智能识别 质控算法 

分 类 号：P412[天文地球—大气科学及气象学]