星载微波辐射计计算植被光学厚度研究

Vegetation opacity estimation from microwave radiometer data

大尺度的土壤水分变化监测对于建立全球的水循环模型意义重大,是实现气候变化预测和洪涝预警的基础。星载辐射计为实现大尺度土壤水分的监测提供了有效途径。由于地表植被对地表的上行微波辐射存在衰减作用,已有的反演算法过程中需要大量的地表植被信息辅助数据来计算地表植被层的影响,其中一个重要的参数就是植被光学厚度。该文基于地表微波辐射传输模型,分析微波极化差异指数(MPDI)与植被光学厚度的关系,结果表明6.9GHz通道MPDI与地表的植被覆盖度之间存在一定的负相关关系。在此基础上,该文建立了MPDI反演植被光学厚度的方法,并应用于土壤水分反演。实验中采用2004年AMSR/AMSR-E的亮温数据计算我国植被光学厚度并进行土壤水分反演,反演结果表明MPDI计算地表植被光学厚度能够取得较好的效果。

Monitoring the soil moisture variety in large-scale is very important to establishing the global water-cycle model, and to forecasting the weather and the floods. The space borne microwave radiometer is an efficient way to monitor the soil moisture variety in large-scale. During the radiometer observation from the space, the absorbance and scattering of the vegetation layer will attenuate the up-forward radiance from the soil, and this affection must be counted and reduced during the process of soil moisture retrieval. Much ancillary data are needed for computing the attenuating effects of the vegetation layer in the soil moisture retrieval model before, and the ancillary data are often unavailable. Based on the microwave radiation transfer model, we analyzed the relationship between vegetation opacity and MPDI （Multi-Polarization Difference Index）, and developed vegetation opacity estimation model. The vegetation opacity has a linearly relationship with natural logarithm of MPDI at C-band （about 6.6GHz）. The MPDI decreases with the vegetation coverage increase. Its lowest value （〈0.008） is corresponding to the forest region, and the higher value stands for the lower vegetation coverage. The highest value （〉0.3） is corresponding to the water region. For soil moisture retrieval, we eliminate the effects of vegetation directly from the brightness temperature of AMSR observed in the space using the vegetation opacity estimation model. Sets of AMSR-E brightness temperature data are used to retrieve soil dielectric constant and the land surface temperature with the two channels （V- and H-polarization） at 6.92GHz in the study. The results showed that the developed model works well for eliminating the effects of vegetation during soil moisture retrieval.