基于区域GPS网大气降水量预测建模研究

The Modeling Research of the Atmospheric Precipitable Water Vapor Forecast Basedon Regional GPS Networks

可降水量是降雨强度预报的必要参数之一,为提高可降水量的准确性,需对水汽转换系数和天顶湿延迟精确计算。利用多元逐步回归分析法,建立了区域大气加权平均温度模型,并利用IGS产品中精确的对流层天顶总延迟推算出天顶湿延迟,并建立了区域GPS网大气降水量预测模型。用上述模型进行预测,相较于传统的Bevis模型和对流层天顶湿延迟模型,得到更精确的水汽转换系数和对流层湿延迟参数。实验结果表明,改进模型提高了可降水量的预测准确性。

Precipitable water vapor （PWV） is one of the essential parameters of rainfall intensity prediction. Accurate calculation of the conversion factor of water vapor and wet zenith delay can improve the accuracy of PWV. A modeling method of regional weighted mean atmosphere temperature model is proposed based on the multiple stepwise regression. A regional atmospheric weighted average temperature model is established based on the stepwise multiple regression, and IGS product is used to obtain accurate tropospheric zenith total delay （ZTD）, thus the zenith wet delay is calculated. Regional GPS network atmospheric PWV model is established based on this. On this method, water vapor conversion factor and the zenith wet delay （ZWD） are more precise than traditional models of Bevis and tropospheric zenith wet delay model. Experimental results show that this method can improve the accuracy of forecasting precipitation.