加密自动气象站雨量计资料的质量控制及其相关关系的研究

Quality Control and Representativeness of Automatic Weather Station Rain Gauge Data

在长春一四平地区100km-100km的范围内，分布有平均间隔10km左右的147个自动气象站。结合该区域雷达回波强度资料，对2007-2011年4～10月的气象站雨量计小时降水数据进行质量控制。多步骤质量控制结果显示，有141个自动站雨量计的数据通过了检查，删除了6个错误站点的数据，对有疑问时段的数据作了标记。利用质量控制后的5年夏季半年自动站雨量计小时降水数据，进行相关关系统计分析表明：距离在10km以内的雨量计测量，平均相关系数均能达到0．6以上；雨量计距离小于5km，平均相关系数在0．7以上；而站点距离超过20km，相关系数普遍降到0．4以下；随着统计时间的增长（从分钟到月降水量），每个雨量计的测量值具有更高的空间代表性。

147 automatic weather stations （AWS） were deployed in 2006 over an area of about 100 km -100 km between Changchun and Siping in Jilin Province. We performed quality control （QC） of the hourly precipitation data from the densely distributed AWS rain gauges for April to October in 2007-2011. After four QC evaluations and a comparison with reflectivity data from C-band weather data in Changchun, data from six gauges were removed and questionable data from several sites were marked but kept for future use. Rain gauge data correlations were calculated after the QC. The results show that the averaged correlation coefficient （CC） is about 0.6 when the distance between gauges is less than 10 km and up to 0.7 or higher when the distance is less than 5 kin. CC generally decreases to less than0.4 when the distance exceeds 20 km. Furthermore, the spatial representativeness of measurements from a single rain gauge increases when the statistical time scale changes from minutes to months.