雷达与雨量计联合估测降水的相关性分析

Correlation Analysis on Estimating Rainfall Using Radar-rain Gauge Calibration

在对比分析质量控制前后雷达估测降水量与自动雨量计降水量之间相关性的基础上,采用雷达-雨量计联合校准方法,对14种不同密度雨量计校准雷达估测降水的效果进行分析。结果表明:在使用雷达资料和雨量计资料前有必要对资料的质量进行分析与控制。联合雨量计校准雷达能明显提高雷达对降水的估测能力;采用不同密度雨量计校准雷达,随着校准雨量计密度的加大,雷达估测降水的精度不断提高并趋于稳定。校准雷达的效果及所需雨量计密度与降水类型有关,当校准效果相同时,积云强降水过程需要的雨量计密度最大,积混对流性降水过程次之,层云稳定性降水过程需要的雨量计密度最小。不同方法的校准效果不同,卡尔曼滤波方法适合于对稳定性降水的校准,或在雨量计密度低的地区对雷达进行校准;变分校准法和最优插值法的校准效果相当,适合对积混对流性降水的校准,或在雨量计密度高的地区对雷达进行校准。

In order to give full play to the advantages of radar-gauge calibration algorithms,the correlation of radar data and rain gauge data is studied before and after quality control by analyzing the quality of radar data and rain gauge data.Based on 11 major precipitation processes during 2008-2009,impacts of 14 types of rain gauge densities on radar rainfall estimation are analyzed by using three radar-gauge calibration algorithms, which are variational calibration method,optimal interpolation method and Kalman filter method. The results show that the quality of rain gauge precipitation data of Tianjin is reliable,only 0.5%of the rain gauge precipitation data has larger error,and equipment failure or external factors is the main cause of its larger error.A reflectivity quality control（QC） procedure has been developed by Chinese Academy of Meteorological Sciences for identifying and removing non-precipitation echoes（such as ground clutter or anomalously propagated ground returns） from the radar base reflectivity fields,and quality control of radar data is implemented using the QC procedure.These non-meteorological echoes can be effectively removed, while retaining precipitation echoes,and thus the rainfall overestimation phenomenon of radar can be significantly improved.The correlation of radar reflectivity data and rain gauge data is analyzed before and after controlling their qualities by selecting different types of precipitation process,results show that quality control of the radar and rain gauge data is necessary to significantly increase the correlation between them and to improve the capacity of radar rainfall estimation.Using some radar-gauge calibration algorithms, impacts of different rain gauge densities on radar rainfall estimation are analyzed.The conclusion is that the capacity of radar rainfall estimation on rain gauge calibration can be improved significantly.The precision of radar rainfall estimation is continuously improved and then become stable with the density of rain gauge increased.The impacts of radar rainfall estimation and the calibration gauge density are related to the types of rainfall.To achieve equal calibration results,convective precipitation caused by cumulus needs the rain gauge density of about a gauge per 182 km^2,mixed cloud precipitation needs about a gauge per 211 km^2,and for stratiform precipitation a gauge per 405 km^2 is enough.The impacts of different radargauge calibration algorithms are different.It shows that Kalman filter method is suitable for the calibration of stratiform precipitation or for the low rain gauge density area,and variational method and optimal interpolation method are suitable for the calibration of convective precipitation or for the high rain gauge density area.