Estimating the Sea State Bias of Jason-2 Altimeter From Crossover Differences by Using a Three-Dimensional Nonparametric Model

With a standard deviation as large as 2 cm, the sea state bias (SSB) has become the dominant source of error in satellite altimetry. The operational SSB correction models are two-dimensional (2-D) empirical (parametric or nonparametric) models based on the altimeter-measured wind speed (U) and significant wave height (SWH). However, these 2-D SSB models cannot entirely parameterize the range bias variability. The SSB uncertainty may be lowered through improved SSB models including additional measurable or predictable correlatives. This paper presents a method to estimate the SSB from crossover differences by using a three-dimensional (3-D) nonparametric model. The model is based on U, SWH from Jason-2 altimeter ocean observations, and the mean wave period from the European Centre for Medium-Range Weather Forecasts reanalysis project ERA-Interim (The SSB model developed with the method presented in this paper is called “3-D SSB model” and the SSB estimated with the 3-D SSB model is called “3-D SSB estimate”). Simulations indicate that the wave period can greatly affect the SSB. Evaluated by the separate annual datasets from 2009 to 2011, the 3-D SSB estimates can increase the explained variance by 1.32 cm
²
, or 1.15-cm RMS relative to the traditional 2-D SSB estimates based on U and SWH. Spatial evaluation of improvement shows that the 3-D SSB estimates are better than the traditional 2-D SSB estimates at all latitudes. The enhancement from 2-D to 3-D SSB estimates is of great significance to improve the precision of the altimeter product. [COMP]: Please set math TYPE gin the sentence below (40) as per the authors PDF.