Preserving Native Spatial Resolution in Long-Term Satellite Datasets Through Improved Projection Algorithms

Satellite datasets are growing larger due to extended mission durations and improved instrument resolutions, creating challenges in efficiently projecting measurements onto geographical grids. This requires the implementation of big data algorithms and specialized data management techniques, with a particular focus on optimizing interpolations and projections. These processing steps are critical as they propagate measurement errors and significantly increase computational time. This work presents a new interpolation algorithm for satellite missions where individual values for each measurement are retrieved. We conduct this study using the Sea Surface Salinity (SSS) processor of the Soil Moisture and Ocean Salinity (SMOS) mission. However, it can easily be extended to other multi-angular acquisition missions. We suggest keeping the measurements within the instrument coordinate system (antenna coordinates) until the final product is generated. This allows us to avoid multiple projection-related errors during the intermediate interpolations. Additionally, we introduce a novel algorithm to project those measurements, taking into account the actual area of the acquisitions instead of considering them as points. Therefore, measurements are weighted based on the area they cover over the Earth. This method is numerically optimized to transform 2D areas into discrete measurements, increasing its computational efficiency and favoring parallelization. The methodology was successfully tested with the SMOS mission's SSS processor at the Barcelona Expert Center (BEC). Final level 3 SSS maps maintain a high resolution close to the one native on the instrument, enabling the characterization of ocean dynamics at finer scales.