Design Of GRACE Level-2 Data Processing Algorithm With Application In Greenland Ice Mass Monitoring

The successful launch of the GRACE satellite has provided high-precision observation data for the study of land and ocean mass changes,global water cycle and glacier and ice sheet melting,which has greatly promoted the development of environmental monitoring and earth science research.Greenland is the largest island in the world,and its ice sheet mass change directly affects the global sea level change,but also has an important impact on the global water cycle and global mass redistribution.The rapid development of satellite gravity technology provides a good opportunity to use satellite data to monitor mass changes of the Greenland ice sheet from space.GRACE data processing centers provide global users with unconstrained spherical harmonic coefficient estimates,i.e.,GRACE Level-2 products.These coefficient estimates contain significant banding errors and high frequency noise,and are often not directly useful for retrieving mass changes.Two methods can be used to solve this problem,namely,spectral domain method represented by spherical harmonic coefficient method,and spatial method represented by point mass method and Mascon method.In the spatial domain method,model parameters can be selected more flexibly and spatial constraints can be introduced more conveniently,which provides the possibility for further improving the mass inversion performance.Therefore,the spatial domain method has attracted wide attention in recent years.This thesis takes the airspace law as the research object,and the research content is as follows.Firstly,the basic theory of mass inversion using GRACE Level-2 data is introduced,with emphasis on the spatial method.The point mass method solves surface mass variation by directly establishing the functional relationship between surface mass point source and gravity field.Mascon method takes the mass change of each mass block as the unknown parameter,and establishes the functional relationship between gravity field and the unknown parameter to solve the surface mass change.A Tikhonov regularization method to solve ill-conditioned problems is introduced,including selecting the optimal regularization parameters by using generalized cross validation.Secondly,the traditional spatial method is used for GRACE Level-2 data processing.In the traditional spatial method,according to the spherical harmonic coefficient simulation provided by GRACE Level-2,the disturbance potential at the altitude of the satellite over Greenland is generated as the observation data,and the optimal range of the observation data buffer is selected.Using GRACE Level-2 data from 2003 to 2012,the mass change in Greenland region is retrieved according to the spatial method,and the annual amplitude,long-term trend and other signal distribution of mass change in Greenland island and local areas are analyzed.The results are compared with those of spectral domain method and external reference model to verify the effectiveness of the proposed method.Finally,a new spatial method is tried,which is called the direct method Mascon.In this method,the GRACE Level-2 spherical harmonic coefficient is directly used as the observation data,which no longer needs to simulate the generation of observation data in the traditional indirect method.The function model of the direct method is established,that is,the observation equation connecting the spherical harmonic coefficient and Mascon parameters.Through simulation experiment and real data experiment,the mass change of Greenland ice sheet retrieved by direct Mascon and indirect Mascon method is compared and analyzed.It is found that the two methods have a good consistency,and the accuracy is comparable to that of Mascon products calculated by other institutions,which proves the effectiveness of direct Mascon method.It provides an effective method for retrieving local mass changes and studying the melting of Greenland ice sheet.