| The Earth’s gravity field is reflection of space distribution and redistribution of the earth mass, for which transportation and redistribution in the Earth system is important for world climate change, economy development and human society. Gravity Recovery and Climate Experiment (Gravity Recovery and Climate Experiment, GRACE) satellite mission dedicated to the observation and estimation the earth’s current mass transfer and redistribution. GRACE detects and measures gravity change, which is mainly associated with the redistribution of global water resources. With the help of GRACE satellites, humans can effectively monitor variation of global terrestrial water storage, as well as polar ice caps and glaciers change and land sea level change and their inner mechanism.Several organizations produce and release estimates of the Earth’s time-varying gravitational field with data provided by the GRACE mission. The released GRACE Level-2 data, are estimated using different model, different tuning parameters, and they are all independent estimates of the earth’s gravitational field. In order to effectively use GRACE model to estimate the secular, seasonal, and sub-seasonal variations in global water cycle, it is necessary to fully understand the accuracy and variations as well as correlation between these different fields, and to derive the best solution from these fields. Through comparison and analysis differences in frequency space and area space, along with correlation between models, this article use multiple GRACE model build composite model as the optimal estimation of earth gravity field model. The analysis verifies the validity and superiority of the combination model, of which RMS is 5～10 mm lower and stripe error phenomenon has certain improvement.This thesis compares how strongly a given signal is dampened through different filtering process, how much the filters spreads a localized signal over a larger region. And also analyzes and summarizes how different filtering methods and filter strength and changes in surrounding mass distribution affect the signal extracted from regional variations. This thesis expounds the "scale factor" method to restore signal attenuation caused by filter, and method to improve gravity field resolution with help of hydrological model. We assess the accuracy of global-gridded TWS estimates derived from temporal gravity field variations observed by the GRACE satellites. The TWS data set has been corrected for signal modification due to filtering and truncation.In this thesis, we use combination model derived from GRACE Release-2 models to estimate earth mass variations from 2003 to 2013, especially global large area of river basins. The amazon basin showed the strongest amplitude of seasonal variation, the Greenland and Antarctic ice water present certain mass loss. Most of the basin’s water storage show positive trends, the Amazon’s inter-annual change is 24.6 km3/y, for the Mississippi, Yenisey, Amur, Lena, Okavango basin, the inter-annual change is between 7 km3/y～9 km3/y. For the Nile, Volga, Ganges, Yukon, Euphrates and Brahmaputra basin, the inter-annual change are 3.5 km3/y,8.3 km3/y, 13.8km3/y,13.1 km3/y,16.9km3/y,7.7 km3/y. In these river basins precipitation changes are all in 1 mm/y (equivalent high water), loss of water storage can be thought of as groundwater changes or storage capacity, such as lakes and reservoirs. In 2003～2013 and TWS increased by 27.2 km3/y, which equal to 0.07 mm/y of sea level change. GRACE monitored changes need to be considered for the impact of the glacial isostatic adjustment(GIA), for Mackenzie and st.Lawrence river, the inter-annual change is mainly caused by GIA, for Asian basins, GIA correction also cannot be ignored. |
PDF Full Text Request