Study On Waveform Classification And Retracking Algorithm Based On Cryosat-2

The growth and melting of polar sea ice have significant influence on global climate changes.Satellite radar altimeter has become a very reliable and feasible method to study the changes of polar sea ice by its large scale,all-weather and all-day advantage.Cryosat - 2 radar altimeter adopted the new synthetic aperture technique and double antenna interferometry technology to improve the spatial resolution and accuracy of observation. How to acquire the high accuracy of sea surface height information from new type of synthetic aperture and interference measurement data is the key to the sea ice thickness inversion. So the radar altimeter echo waveform classification and retracking algorithm about Cryosat - 2 need to be an in-depth study.The echo waveform is the direct observation data of satellite radar altimeter.on the one hand,to do waveform classification on sea ice area is the foundation to select the optimal waveform retracking mothed aim at different reflecting surface.On the other hand, according to the different waveform corresponding to the reflective surface,to do waveform classification on sea ice area can achieve ground targets(such as the sea ice) classfication. Waveform tracking is the key technology for satellite radar altimeter to achieve high precision ground elevation information.But on the edge of ice sheet (not open sea),the traditional waveform retracking methods exist the shortcomings of low measuring precision and poor applicability.Therefore,it is necessary to study a new kind of waveform retracking method which applicable to the edge of ice sheet in order to do accurate monitoring of the polar ice caps.Using Cryosat-2 radar altimeter data, The key algorithm of sea ice thickness inversion in poles was studied, the thesis's research content includes:(1) The thesis studied the mechanism of Cryosat-2 radar altimeter satellite altimetry and sea ice thickness detection, and analyzed the characteristics of Cryosat - 2 echo waveform.(2) The target tracking classification was carried out by using Cryosat-2 synthetic aperture echo waveform.The thesis summarized the existing 10 types of waveform classification parameters(including MAX, PP, SSD, SK, PPL, PPR, LEW, TEW, LTPP, ETPP)and three commonly used waveform classification methods(including threshold value method.K-NN classifier, SVM classifier) to classify Cryosat-2 SAR mode waveforms in arctic sea ice area. The classification results mainly divide into Lead, first year ice(FI), multi-year ice(MY1,MY2) and ice free (IF) . To compare waveform classification results with the real-time ice chart provided by AARI (Arctic and Antarctic Research Institute) ,we got the applicability and characteristics of all the waveform parameters and classification mathods.(3) For the waveform retracking problem on the edge of ice sheet in pole regions, the characteristics of Cryosat-2 SARIn mode waveform were studied. Then a new waveform retracking method based on echo fitting and multiple sub-waveform threshold retracking--FMSTR(Fitted Multiple Sub-waveform Threshold Retracker ) was proposed. The method was implemented by waveform filtering, echo fitting, sub-waveform discriminating and classifying, adaptive retracking. It can attain the elevation measurements along track with more high accuracy. Compared to the ICESat DEM, the results derived from the FMSTR method show higher average accuracy than Cryosat-2 Level2 products, 0.54m, 0.26m, 0.59m and 1.02m respectively. The test proved that the FMSTR method is more suitable for the waveform retracking on the edge of ice sheet.