实孔径微波辐射计海面盐度反演方法研究

发布时间：2018-06-10 03:15

本文选题：海面盐度 + 反演算法　；参考：《华中科技大学》2016年硕士论文

【摘要】：海面盐度是海洋系统中一个极其重要的参数,它对研究海洋环境、全球水循环以及气候的监测和预测等有着巨大贡献。近年来,卫星微波遥感测量海面盐度的技术取得了快速的发展,使得大范围全天候的盐度观测成为现实。但是目前在轨运行的盐度微波遥感卫星SMOS和Aquarius的盐度数据产品在全球范围内尚未达到预期精度,因此深入开展盐度反演算法的研究具有良好的应用价值。本文以实孔径微波辐射计海面盐度反演算法为主线,首先介绍了被动微波遥感观测海面盐度的理论基础;在考虑了海面、大气层、电离层和太空中众多影响因素的条件下,建立了海面微波辐射正向传输模型,并仿真分析了入射角度、海面温度、风速和风向等因素对海面亮温贡献的影响;接着在正向模型的基础上对海面盐度反演的统计回归算法、神经网络算法和基于代价函数的迭代理论算法进行了研究,并分别基于三种盐度反演算法仿真分析了亮温、海面温度和风速等参数误差对盐度反演精度的影响;最后,将利用Aquarius卫星公布的盐度数据以及其他辅助数据通过正向模型计算得到海面模拟亮温值与Aquarius卫星公布的实测亮温数据进行比较,所得的亮温误差较小,从而验证了海面微波辐射正向模型的准确性;此外,将利用Aquarius卫星公布的实测亮温数据通过三种盐度反演算法计算得到的盐度值与Aquarius公布的盐度数据进行比较,结果表明,在低纬度开阔海域,单次盐度反演所得的盐度值与Aquarius公布的盐度数据之间的均方根误差约为0.36psu,从而验证了统计回归反演算法、神经网络反演算法和基于代价函数的迭代反演算法的有效性。
[Abstract]:Sea surface salinity is a very important parameter in ocean system. It has great contribution to the study of marine environment, global water cycle and climate monitoring and prediction. In recent years, the satellite microwave remote sensing technology for sea surface salinity measurement has made rapid development, making the large-scale all-weather salinity observation become a reality. However, the expected precision of salinity data products of the salinity microwave remote sensing satellite SMOS and Aquarius in orbit has not reached the expected precision in the world, so the further research on salinity inversion algorithm has good application value. Based on the inversion algorithm of sea surface salinity of real aperture microwave radiometer, this paper first introduces the theoretical basis of passive microwave remote sensing observation of sea surface salinity, taking into account many influence factors in sea surface, atmosphere, ionosphere and space. The forward transmission model of sea surface microwave radiation is established, and the effects of incident angle, sea surface temperature, wind speed and wind direction on the contribution of sea surface brightness temperature are simulated, and then the statistical regression algorithm for sea surface salinity inversion based on the forward model is presented. The neural network algorithm and iterative algorithm based on cost function are studied, and the effects of parameter errors such as brightness temperature, sea surface temperature and wind speed on the accuracy of salinity inversion are simulated based on three salinity inversion algorithms. The salinity data published by Aquarius satellite and other auxiliary data are compared with the measured brightness temperature data published by Aquarius satellite through forward model calculation, and the brightness temperature error is small. The accuracy of the forward model of sea surface microwave radiation is verified. In addition, the salinity values calculated by three salinities inversion algorithms using the measured brightness temperature data published by Aquarius satellite are compared with the salinity data published by Aquarius, and the results show that, In the open sea area of low latitude, the root mean square error between the salinity value obtained by single salinity inversion and the salinity data published by Aquarius is about 0.36 psu. thus the statistical regression inversion algorithm is verified. The effectiveness of neural network inversion algorithm and cost function based iterative inversion algorithm.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P715.7

【参考文献】