星载SAR在海气边界层气象学中的应用研究

SAR(Synthetic Aperture Radar,合成孔径雷达)作为一种现代高空间分辨率成像侧视雷达,对地球表面海洋所成的图像中蕴含了极为丰富的中尺度及亚中尺度海洋大气边界层的信息,因此对边界层气象学研究有着非常重要的意义。但是,使用SAR研究海气边界层这一涵盖微波遥感、气象学及海洋学等学科的科学前沿课题在国内却少有文献报道。在此背景下,首先介绍了SAR反演海洋大气边界层的研究概况,回顾了SAR反演海气边界层参数的原理和方法。然后以2002年5月7日当地时间10时53分ERS-2卫星获取的香港地区(22.097°N,E 114.300°E)SAR海洋图像为例,进行了反演风向风速的初步试验,最终获得了较高精度的风矢量。具体过程如下:先对SAR图像进行预处理,包括ADC(Analog Digital Converter,模数转换器)补偿、精确校准及斑点滤波等过程;然后利用经典的谱分析方法求得具有180°模糊度的风向,再用香港天文台气象浮标实测资料消除这一不确定性得到了真实的相对风向;紧接着利用CMOD4地球物理模式函数计算得到了海面上10 m高的风速。与气象浮标站所记录的平均风速和风向比较,两个20 km×20 km大小的试验区域求得的风向误差分别为23.71°和7.00°,平均风速误差分别为0.18 m/s和-0.12 m/s。结果表明,如果对SAR预先进行严格的预处理,结合经典的谱分析方法和CMOD4模型,即可获取高精度的风矢量。这一结果为今后海洋大气边界层的研究奠定了良好的基础。

The Preliminary Study on Marine Atmospheric Boundary Layer Using Synthetic Aperture Radar

Synthetic Aperture Radar(SAR) is a high resolution, modern side-look imaging radar. SAR images of ocean surface contain a wealth of information of mesoscale and sub-mesoscale marine atmospheric boundary layer (MABL) phenomenon. Thus it is a very important data source for MABL meteorology study. However this hot topic, which covers microwave remote sensing, meteorology and oceanology, etc. have got reported by few Chinese open literature. Based on such situation, firstly we summarize the recent MABL studies, and introduce the principle and methods of studying MABL with SAR images. As an example,ERS-2 SAR image covering Hong Kong region acquired on May 7, 2002 is used for the preliminary test of wind vector retrieval. The processing include:①pre-processing the SAR images, including Analog Digital Converter (ADC) compensation, accurate calibration and speckle removal; and②wind direction and wind speed retrieval with the classic SAR Wind Direction Algorithm spectral method. The wind direction thus estimated is however with 180°ambiguity. Buoy data collected by the Hong Kong Observatory are then used to resolve the uncertainty. Finally, the GMF(Geophysical Model Function)-CMOD4 is adopted to estimate the wind speed at the height of 10m above sea level.
Compared with the wind direction and wind speed data recorded by Hong Kong Observatory, the error of the retrieved wind directions of two selected regions, (20 km×20 km each) are 23.71°and 7.00°,respectively, while that of the retrieved mean wind speed are 0.18 m/s and -0.12 m/s, respectively. The results show that high quality wind vector can be acquired if strict pre-processing, classical spectral analysis algorithm and CMOD4 model are adopted. The results are very encouraging for the future MABL research.