基于商用微波链路反演降雨的空间插值方法对比研究

Comparative Study of Spatial Interpolation Methods for Retrieving Rainfall Based on Commercial Microwave Link

为探究适宜于商用微波链路降雨监测数据的空间插值方法,基于无线微波雨衰特性经验模型,分析不同线状降雨数据转换为点状降雨数据情景下,反距离权重法、克里金法和趋势面法对降雨空间插值的反演精度。研究结果表明:微波链路越密集,条数越多,其趋势的一致性越好;在微波线状数据转换为点状数据的五种情景中,将微波链路以200m为间隔概化为具体的降雨监测点时,空间插值效果较好;趋势面法的空间插值效果较好,反距离权重法次之,克里金法最差。研究结果明晰了线状降雨数据转换为格网型空间数据的原则,为商用微波链路组网构建空间二维雨量场提供了支撑,对提高线状降雨监测产品兼容性具有重要意义。

In order to explore the spatial interpolation method which is suitable for commercial microwave link rainfall monitoring data,the inverse distance weight method,Kriging method and trend surface method were analyzed based on the empirical model of wireless microwave rain decay characteristics when different linear rainfall data were converted to point-like rainfall data.The results show that the denser and more microwave links are,the better the trend consistency is.In the five scenarios of converting the microwave linear data to the point data,the spatial interpolation effect is better when the microwave link is generalized to the specific rainfall monitoring points at the interval of 200 m.The spatial interpolation effect of trend surface method is better,the inverse distance weight method is second,and the Kriging method is the worst.The results clarify the principle of converting lin⁃ear rainfall data into grid spatial data,provide support for constructing spatial two-dimensional rainfall fields in commercial micro⁃wave link networking,and have important significance for improving the compatibility of linear rainfall monitoring products.