A Study of an Inversion Technique for Application in Sea Ice Information Retrieval

For global monitoring, satellite remote sensing is a useful tool to cover wide area of earth surface for large scale observation. However, the correct interpretation of satellite images remains a challenging problem and the retrieval of earth related information from the processing and analysis of the satellite data requires the development of suitable artificial intelligence based technique. In this paper, an inverse model to retrieve sea ice thickness from active microwave remote sensing data is presented. The inverse model is a combination of an improved forward model using Radiative Transfer Theory that implements the Dense Medium Phase and Amplitude Correction Theory (DMPACT) and the Levenberg-Marquardt Optimization algorithm. The inverse model functions by interpreting a set of input data using the mentioned forward model and adjusting the wanted parameter accordingly using optimization. Through this adjustment, the sea ice thickness is estimated. Data from ground truth measurements carried out in Ross Island, Antarctica together with radar backscatter data extracted from purchased satellite images, are used as inputs during the simulations. The obtained sea ice thickness is then compared with the measured data to verify its accuracy. The results show promise towards the use of the inverse model to retrieve sea ice thickness from actual conditions in the polar region.