中国江淮、黄淮地区陆面微波比辐射率的变化特征

The Characteristics of Microwave Emissivity over Land of Chinese Jianghuai-Huanghuai Region

陆面微波比辐射率较高且易变,造成陆面上反演降水以及其它大气参数较为困难。对于地表特征复杂的中国,陆面微波比辐射率的研究还很有限。通过利用Tropical Rainfall Measuring Mis-sion(TRMM)卫星上同步扫描的VIRS(红外和可见光)与TMI(微波)资料以及微波辐射传输模式反演了中国江淮、黄淮地区陆面微波比辐射率。然后,结合MODIS提供的地表类型数据,分析了江淮、黄淮地区不同地表微波比辐射率的时空变化特征。结果表明该地区的农作物地表比辐射率最小,垂直与水平比辐射率极化差最大;而森林地表比辐射率最大,极化差最小。此外,不同地表的微波比辐射率昼夜变化明显,季节变化不明显。比辐射率估算误差中,地表温度、微波亮温和大气相对湿度3因子的准确计算对22 GHz和85 GHz的影响较为明显,对其它通道影响较小。对于小于85 GHz的通道,比辐射率估算精度受微波亮温的影响最为明显,地表温度其次,相对湿度最小;对于高频85 GHz,相对湿度的影响最明显,其次是微波亮温,最后是地表温度。

The larger and highly variable microwave emissivity over land results in more difficulties to retrieve the precipitation and other atmospheric parameters over land. The research about microwave emissivity over land is still weak due to the complicated terrains in China. In the paper, the synchronous observation of infrared and visible scanner （VIRS） and microwave imager （TMI） on the Tropical Rainfall Measuring Mission （TRMM） satellite and one microwave radiation transfer model to retrieve the microwave emissivity over land of China were utilized in Jianghuai-Huanghuai region. Then, combining with the land cover type data derived from MODIS observations, the variation characteristics of microwave emissivity over different types of land cover in the middle region of China was further analyzed. Results show that the microwave emissivity of crop land surface is smallest and its polarization difference is largest while forest emissivity is largest and its polarization difference is smallest among mainly 5 types of land cover in this region. In addition, the microwave emissivity over different types of land cover shows more significant variation in day and night time while less variation in seasons. For the error estimatioffof emissivity, the influences of land surface temperature （LST）, microwave brightness temperature （TB） and relative humidity （RH） on microwave 22 and 85 GHz channels are quite obvious. For those channels with frequencies less than 85GHz, the influence of TB on estimation of emissivity is most obvious, then LST and RH. For high frequency 85GHz, the influence of RH on estimation of emissivity is most obvious, then TB and LST.