Local tide and geoid corrections significantly improve coastal retracked Jason sea surface heights in the Philippines

Abstract

Retracking algorithms increase the accuracy of coastal sea surface height (SSH) measurements. However, it is still important to validate these retracking estimates with tide gauge (SSHtg) observations. We downloaded the freely available Jason altimeter SSH processed using the XTRACK-ALES algorithm, then detided the SSH using different tide models. The first model is the default tidal correction based on Finite Element Solution 2014 (SSHfes), and the second model is the T_Tide harmonic analysis of the nearest tide gauge (SSHaltimeter). SSHfes showed a very poor correlation (< 0.31) and very high root mean square error (RMSE, > 29 cm). In contrast, SSHaltimeter generally showed a very high correlation (> 0.91) and low RMSE (< 17.4 cm). A further quality check based on the average and standard deviation of the difference between the SSH readings (SSHfes – SSHtg and SSHaltimeter – SSHtg) also showed the superior performance of SSHaltimeter,which scored < 9.3 and < 16.5 cm, respectively; compared to SSHfes, which scored < 9.3 cm and > 27 cm for the same parameters. The poor performance from the SSHfes likely comes from the complex bathymetry and coastal geomorphology of the country, which is not accounted for in the FES. The Philippines generally has a narrow shelf, and the FES tide corrections may be related to deep-water tides rather than the shallow-water tides observed from tide gauges. Despite the high correlation and agreement between the SSHaltimeter and SSHtg, the rate of sea level rise from the SSHaltimeter in some sites is more than twice the rate from SSHtg, which indicates the possible influence of the vertical land movement.