An Algorithm For Retrieving Atmospheric Temperature Profiles Directly From Visibility Function Of Aperture Synthetic Microwave Radiometer

The aperture synthetic microwave radiometer combines a sparse small-caliber antenna array into a large physical aperture,which can greatly reduce the volume and weight of the entire system,and can instantaneously image without scanning.Therefore,the aperture synthetic microwave radiometer has gained more and more people's research.Its application potential in the field of earth sensing and other remote sensing is huge.It is well known that passive remote sensors(microwave radiometer)in the application of microwave remote sensing directly measures natural radiation of the earth(bright temperature).And the geophysical parameters required for the application need to be obtained from the remote sensing brightness temperature product using a mathematical inversion algorithm.Unlike real-area radiometers,which measure brightness temperature directly,the interferometric-based aperture synthetic radiometer directly measures the spatial spectrum of the brightness temperature field—the visibility function.Therefore,it requires a data processing flow from the visibility function to the brightness temperature to the geophysical parameters when applied to Earth remote sensing.And the two steps of the inversion process are mathematically unsteady ill-conditioned problems,which are both cumbersome and lead to error amplification.Therefore,the accuracy of the remote sensing physical parameters obtained would be lower.In order to improve the precision of remote sensing physical parameter products for aperture synthetic microwave radiometer and simplify the data processing process,this paper proposes a concept for retrieving geophysical parameters directly from visibility function of aperture synthetic microwave radiometer.Concept,the paper describes the principle of the concept and the specific algorithm implementation process,and gives the simulation test results.Specifically,this paper implements a direct inversion algorithm from the visibility function to the atmospheric temperature profile using one-dimensional variational algorithm and neural network algorithm for one-dimensional aperture synthetic radiometer.Finally,this paper uses the simulated visibility function data based on the historical real atmospheric parameters to carry out end-to-end simulation inversion experiments,which are used for verification and performance evaluation of the new algorithm.The simulation test results show that the direct inversion algorithm can obtain high-precision atmospheric temperature profile,and demonstrates the feasibility of the direct inversion algorithm for the remote sensing application of the aperture synthetic microwave radiometer.Compared to the two-step inversion method,the direct inversion method has the advantages of fewer steps,faster,and less error.