Retrieval of humidity and temperature profiles over the oceans from INSAT 3D satellite radiances

Journal of Earth system science

Deo Kumar1¹  C Krishnamoorthy1¹  C Balaji1¹ 

[1] Indian Institute of Technology Madras, Chennai 600 036, India$$

关键词: Artificial neural networks;    INSAT 3D;    humidity;    temperature;    maximum likelihood;    Bayesian.;   

DOI  :  

学科分类：天文学(综合)

来源: Indian Academy of Sciences

【 摘 要 】

In this study, retrieval of temperature and humidity profiles of atmosphere from INSAT 3D-observed radiances has been accomplished. As the first step, a fast forward radiative transfer model using an Artificial neural network has been developed and it was proven to be highly effective, giving a correlationcoefficient of 0.97. In order to develop this, a diverse set of physics-based clear sky profiles of pressure (P), temperature (T) and specific humidity (q) has been developed. The developed database was further used for geophysical retrieval experiments in two different frameworks, namely, an ANN and Bayesianestimation. The neural network retrievals were performed for three different cases, viz., temperature only retrieval, humidity only retrieval and combined retrieval. The temperature/humidity only ANN retrievals were found superior to combined retrieval using an ANN. Furthermore, Bayesian estimation showed superior results when compared with the combined ANN retrievals.

【 授权许可】

Unknown