Assessing the atmospheric correction algorithms for improving the retrieval data accuracy in the remote sensing technique

For the purposes of cost-effective and rapid surface water quality monitoring, the utilization of the cutting-edge satellite remote sensing (RS) technique has increased over the years. Recently, several studies have revealed that the RS technique severely suffers from particles present in the atmosphere, especially from aerosol particles. This interference significantly influences the quality of the information extracted from remote sensing measurements and produces much more uncertainty in retrieving optically active water quality indicators (e.g., chlorophyll-a, coloured dissolved organic matter, total suspended matter) from optically complex water bodies. Therefore, it is required to minimize the uncertainty within the remotely sensed data by reducing the impact of atmospheric interference through the atmospheric correction (AC) process. Currently, a series of algorithms have been utilized in the literature for treating the AC in the RS technique, among which ACIX-Aqua, ACOLITE, BAC, C2RCC, FLAASH, iCOR, l2gen, LaSRC, POLYMER, GRS, Sen2Cor, and 6SV are widely used. Since the development of the AC algorithms, its applications have increased in handling of big data, like as remote sensing data. Recently, several studies have revealed that the existing algorithms have produced a considerable uncertainty in the retrieval data due to the architectural complexity of algorithms. Although, the application of cutting-edge machine learning and artificial intelligence techniques is increasing for atmospheric correction process. Therefore, the aim of the research is to develop an efficient algorithm utilizing the publicly available AC algorithms and incorporating machine learning and artificial intelligence approaches in order to reduce atmospheric interference from the RS data. The results of the research could be helpful for retrieving various optically active water quality indicators most efficiently in terms of reducing the uncertainty in monitoring water quality.

Keywords: surface water quality; remote sensing; atmospheric correction, artificial intelligence; optically active water quality indicators.