基于微波遥感数据的雪情参数反演方法

Snow Parameter Estimation from Microwave Remote Sensing Data

微波遥感传感器在36.5 GHz通道会因雪深超过其穿透深度而出现信号饱和,从而导致雪深被低估。针对该问题,首先建立了18.7 GHz与36.5 GHz通道亮温差和10.7 GHz与18.7 GHz通道亮温差相结合的积雪深度分层反演新方法,然后利用GCOM-W1星上搭载的AMSR2传感器数据估算了2012年12月至2013年2月新疆每日积雪深度,结合同期的气象站点观测数据与野外实测数据对遥感反演结果进行了评价。结果表明,所建立模型能够很好识别新疆地区积雪的空间分布状况,雪深的估算结果明显优于常用的Chang模型。

The snow depth may be under estimated from the passive microwave remote sensing data at the frequency of 36. 5 GHz due to the saturation of the microwave signal detected by the remote sensor,thus,a new segmental modeling approach for snow depth estimation was created by combining the brightness temperature differences between 18. 7 GHz and 36. 5 GHz channels and between 10. 7 GHz and 18. 7 GHz channels. Afterwards,the brightness temperature data acquired by the AMSR2（ Advanced Microwave Scanning Radiometer 2） sensor onboard the GCOM- W1 satellite were used to test the model and the snow depth of Xinjiang from December 2012 to February2013 was estimated. The observations collected by the Xinjiang meteorological stations and field in-situ measurements of snow depth were employed to assess the estimation. The results show that the segmental approach can identify the spatial distribution of snow covers and accurate estimation of snow depth can be achieved,which obviously outperforms the result using Chang＇s algorithm.