Study On Parametric Model Of Sea State Bias Based On Collinear And Crossover Convergent Dataset In Radar Altimeter

The radar altimeter is one of the most import microwave remote sensors.Altimeternot only can measure the global sea surface height with an accuracy of severalcentimeters but also get some sea surface information such as significant wave andwind speed.Sea state bias is one of source of error in altimeter range measurement.Although with a small magnitude, SSB has taken the place of the orbital error as thelargest source of error in altimeter range measurement because of the complicatedmechanism and the development of precision orbit determination. Many theoreticallyand experimental methods are used to the research on SSB.Because theoreticallymethods is very complex and the models got from these methods can not be used tocorrect error in altimeter range measurement, the experimental methods has becometo the mainline methods. Collinear and crossover data sets including sea surfaceinformation taken at the same geographic locations but different times are the base ofthe formulation of SSB experimental models.Distance weighted is used to extract collinear data set from the GDR data set ofJason-1satellite altimeter. New segmented fitting method of track is also adopted toextract crossover data set. The characteristics of collinear and crossover data sets areanalysed from latitudinal, time and global distribution. Conclusion got is that bothcollinear and crossover data have their own advantages and disadvantages for theformulation of SSB experimental models, so comparing collinear data with crossoverdata and analysing their own characteristic is a effective method to improve the SSBexperimental models.The collinear data and crossover data are respectively used to establish theparametric models of sea state bias. The latitude’s integral and local features of twosea state bias models are analyzed and it has shown that the model based on collinear data is more effective in the local feature. The collinear dataset is of large quantity ofdata and some redundant data reduces the calculating efficiency. The crossover data iscomparatively fewer; however, the local feature of sea state bias model based oncrossover data is worse.According to the characteristics of datasets and sea state bias models, severalimprovement methods on dataset are proposed and it demonstrates that thedistribution of data cannot be changed by using different datasets, and increasing thenumber of repetition cycles is an effective method. Finally, this article proposes a newmethod that using the crossover data as the frame of reference and adding thecollinear data with uniform spacing to the adjacent crossover data to modify thedataset. In the final dataset, the redundant data has been removed and theeffectiveness of sea state bias correction has been preserved. The new parametricmodel based on new dataset has improved not only the integral feature but the localfeature, and the aim of improving effectiveness of sea state bias parametric model bymodifying the dataset has been achieved.