Validation of conventional and retracked Sentinel-3 observations along the Norwegian coast

Satellite altimetry is one of the fundamental techniques for Earth observation, which provides precise measurements with frequent sampling and global coverage. However, its performance is degraded in coastal areas due to different factors, such as land contamination, erroneous tropospheric corrections or complex tidal patterns. In order to improve performance of satellite altimetry in the coastal zones, an increasing number of dedicated coastal altimetry products have been developed and validated in specific areas in later years. Those products are based on the improved analysis of backscattered signals in order to increase accuracy of altimetry observations in the coastal zones. One such product is the Adaptive Leading Edge Subwaveform (ALES) retracker, specifically aimed at the issue of land contamination. As of yet, it has not been validated along the whole, complex Norwegian coastline, with thousands of small islands, narrow fjords, and rough topography.  Thus, this study aims to validate the ALES retracker along the Norwegian coast, comparing conventional and ALES-retracked Sentinel-3 A/B observations with tide gauge observations. Altimetry-tide gauge comparison pairs are found by considering altimetry observations within optimum radii around each tide gauge, determined by minimizing the root mean square of differences (RMSD) for a range of candidate radii. It was found that the optimum radii for tide gauges located towards the open ocean are smaller than for those located inside fjords, because the observation accuracy degrades in the latter areas. Thus, it was necessary to increase radii, i.e. to include more points on the open-sea, for tide gauges inside fjords in order to minimize the RMSD. It was concluded that the ALES dataset generally gave better results (in terms of RMSD and correlation to the tide gauges) than conventional datasets, as well as giving a larger number of valid observations. The results are promising for future optimal combination of altimetry observations with other available sea-level observations in the coastal zone, e.g., from tide gauges, ships, unmanned surface vehicles (USVs) or airborne LiDAR. A prerequisite for such a combination is a reliable error description of each data type, to which the current study serves as a contribution.