Study On3-Dimensional Sea State Bias Model Based On Multi-satellite Altimeters Dataset

Altimetry is particularly important in monitoring the mean sea level and it hasbeen applying to the domains of research for ocean topography detection and oceangravity field analysis. TOPEX/Poseidon (T/P) and Jason-1are one serious of satellitesand their altimetry measurements are considered to be reliable. It is required to addsome error corrections in the altimetry data and sea state bias (SSB) is one majorsource of error. The theoretical modeling and understanding of SSB remain a bigchallenge. Therefore, SSB correction still largely relies on empirical method. SSBis a particular correction since its determination and accuracy depend on thecharacteristics of sea surface height(SSH) itself.The altimetry error corrections in T/P altimeter are comparatively analyzed withsynchronous Jason-1altimeter. The above results are used to retrieve T/P data, whichis further used to establish the two altimeters’ convergent crossover dataset of SSH,significant wave height and wind speed. It demonstrates that SSB model based onthe convergent crossover dataset can adapt to different altimeters and the versatility ofthe model has largely increased compared with the model based on one altimeterdataset.In recent years, research on new parameters that potentially might represent goodSSB correlatives has been one of hot topics. In this article, mean wave period isadded to the SSB parametric model. The bilinear interpolation is used to adapt themean wave period data from ERA-40to the counterpart measuring points of Jason-1altimeter. The performance of the previous inversion models is assessed using globalcollocated validation dataset. It is demonstrated that the previous models cannotprovide enough accuracy with our global validation dataset. A quartic polynomialmodel is established based on the strong correlation between non-dimensional waveheight and wave age using an independent collocated dataset. Piecewise method is applied to remove the linear relationship between the mean wave period bias ofERA-40and altimeter with significant wave height. It indicates that the piecewisequartic polynomial model is of enough accuracy in the global range.A3-D (three-dimensional) SSB model which includes significant wave height,wind speed and mean wave period parameters is established based on the convergentcrossover dataset. The change from~2-D (significant wave height and wind speed) to3-D model provides more than1cm2improvement of SSB explained variancealmost in the whole latitude range for three different altimeters (T/P, Jason-1, Jason-2)validation datasets. It is more important that the3-D model can reflect the specificrelationship between SSB and mean wave period which is a significant descriptor ofsea state. Therefore, the3-D SSB model based on the convergent crossover datasetis more effective and versatile.