宁德市降水量异常点剔除的影响分析

发现异常点是对逐时降水资料质量控制的有效方法,但简单的剔除未修正的异常点必然会影响其周围一定范围内降水量的空间变化。本文提出综合评分法定量的分析不同时段不同异常监测点被剔除后对降水量变化幅度的影响,探讨剔除区域异常点后降水量的时空变化分布特征。对宁德市降水资料进行实例验证,分析结果表明:在相邻站点数大于12个、相邻站点最近距离小于3km、相邻站点平均监测值大于6mm时,其降水量变化幅度相对较小,可剔除异常点。综合考虑三个相邻站点指标的综合分析可以显著表现不同时段不同监测站点降水量空间变化的地带性分布,在宁德市西北和东南区域出现高变化幅度站点的概率最高,呈现出高变化幅度站点多、分布范围广特点;在其东北区域监测站点分布不均匀,降水高变化幅度站点易受逐时降水量大小影响,呈现出小雨大范围、大雨小范围分布;在其西南和中部区域监测站点分布密集,出现高变化幅度站点的可能性最小。因此,综合分析可剔除区域异常点对降水量空间变化的影响特征,确定对降水量变化幅度影响最小的监测站点,有效提高了降水资料的质量控制水平。

Spatial Variability Analysis of Precipitation Anomaly Point Removal in Ningde City

To find anomaly point is an effective method for controlling the quality of hourly precipitation data. However, simply removingthe uncorrected anomaly point will inevitably affect the spatial variation of the precipitation within a certain range. This paperproposed a comprehensive scoring method to quantitatively analyze the influence of different anomaly monitoring points removalon the variation of precipitation during different durations, and discussed the temporal and spatial distribution of precipitation afteranomaly point removal. The case study of Ningde precipitation data shows that the variation of precipitation is relatively small whenthe number of neighbors is more than 12, the nearest neighbor distance is less than 3km, and the average monitoring value of adjacentsites is greater than 6mm. The comprehensive analysis of three adjacent stations can significantly show the spatial distributionof precipitation at different monitoring stations in different durations. The probability of occurrence of high-variation sites in thenorthwest and southeast of Ningde is the highest, showing a high variation range distributed in a wide range of areas. In thenortheast, the monitoring sites are unevenly distributed, and the site of high variation in precipitation is susceptible to the size ofthe hourly precipitation, showing a wide range of light rain, and heavy rain distribution in small areas. In the southwest and center,the monitoring sites are densely distributed, the possibility of a site showing a high degree of change is the smallest. Therefore, acomprehensive analysis can eliminate the influence characteristics of the regional anomaly points on the spatial variation of precipitation,and determine the monitoring stations that have the least impact on the variation range of precipitation, and effectively improvethe quality control of precipitation data.