Using Gravity Satellite Data On The Coseismic Signal Detection And Focal Mechanism Inversion Of The Great Earthquakes

As a new method to observe the earth, the invention and the development of the gravity satellite observation technology has been helping people study the global gravity field and its time varied changes continuously and in long-term. The gravity satellite can get the gravity field without sending anyone to the certain area, which means it will help us observe the gravity field in remoting area or the ocean sea. The great earthquakes, with their destructive power and involving huge amount of people, are always human’s nightmare. People are always eager to study the mechanism of the great earthquakes. However those great earthquakes are usually occurred under the ocean which means the common geodetic observation method are too difficult to observe. And gravity satellite are born to this. With the supplement of gravity field change data in the area of earthquakes, people can learn more for the focal mechanism of great earthquakes. This thesis was taking this as the aim. The applying of GRACE satellite data on coseismic signal detection of great earthquakes was studied. And the author innovatively combining the GRACE satellite data and GPS data to joint inverse the source parameters and fault slip distributions.The thesis introduce the GRACE satellite mission and the datasets. An appropriate GRACE level 2 dataset was chosen for the thesis study. For the sake of eliminating the errors-- especially the north-south striping error-- in GRACE data, the post-processing for GRACE level 2 data is needed before using the data. Different method to remove the error were discussed and compared in the thesis. And the author believed that removing the error would lost the chance of observing the seismic signal in local area until trying the best to keep the geophysical signal in the meantime. The Slepian function method was shown, and 2004 Sumatra-Andaman earthquake was taken as an example of using the Slepian function method to extract coseismic signal. The experiment demonstrated the effectiveness of the Slepian function method.The effect of different source parameters to the gravity field changes was simulating and analyzing. Under the circumstance of fixed earthquake magnitude, fault dip angle and rake angle can greatly affect the gravity field changes obviously both in magnitude and spatial distribution. Different great earthquakes with various magnitude and source parameters were simulated to test the probability of detecting the gravity changes by GRACE satellite. Due to these simulation, the conclusion was made that the great earthquakes with moment magnitude of 8.0 are difficult to detect by GRACE satellite only if the source parameter encounter a strict requirement. However for those earthquake with moment magnitude larger than 8.5 will be easier to detect by GRACE. The Wenchuan earthquake was taken as an example to verify the conclusion. The coseismic gravity changes are too small for GRACE to detect, because the coseismic signal is smaller than the hydrological signal. A bigger earthquake can provide a larger and detective coseismic gravity changes.The role of GRACE data in inversion for source parameters and slip distribution was discussed. And simulation of using GRACE data or combining the GRACE data and GPS data to inverse were make. After a Gaussian filtering with 350km radius for the simulating gravity data, the source parameters can still be inversed. This reveal that GRACE data can play an important role in inversing the source parameters considering the data spatial resolution. While dividing the subfaults in the fault interface, the larger of the subfaults amount, the bigger of the fluctuation and rigidity. The inversion of slip distribution showed that the upper fault are more accurate than the deeper fault which could because of the insensitivity of the gravity data to deeper movement. With introducing the Helmert Variance Component Estimation method in joint inversion to determine the weight of GRACE and GPS datasets, the result showed more accuracy than using the GRACE data alone. And in slip distribution inversion, joint inversion results showed less fluctuation and rigidity. It seems that the joint inversion can defend the instability of large amount of subfaults. Whatever in inversion alone or joint inversion, the result were more accurate with more ground observations. However in joint inversion, the weight of GRACE data was decreasing with the increasing of observation points. This should reflect the abundant of extra information by GPS than GRACE data as the relativeness was in the latter. Due to the results, the GRACE data and GPS data can supply unique information for each other, which would help a lot in accurately joint inversing the source parameters and slip distribution. This would be the basic of inversing the 2011 Tohoku-Oki earthquake.The great 2011 Tohoku-Oki earthquake was taken as a seismic example to verify the effectiveness of GRACE data in detecting the coseismic signal and inversion for the focal mechanism and slip distribution. The geological background was discussed first. For the big moment magnitude, destructive power and prominent movement in vertical direction, this earthquake would be able for GRACE detection. With analyzing the foreshock and aftershocks, these effect can be ignored because it was not big enough for GRACE detecting. On the contrary, the gravity effect due to the redistribution of ocean floor mass caused by 2011 Tohoku-Oki earthquake was prominent and should be taken into account during the GRACE data processing. The coseismic gravity data was extracted from GRACE data, which had shown the accuracy to the forward simulating gravity data in spatial distribution and the fault motion characteristic. The inversions for source parameters and slip distribution with GRACE data alone and joint data of GRACE and GPS were performed. The results were close to previous studies in a certain extent, and consistent to the seismological observations. The joint results show more accuracy. The above experiment can prove the effectiveness of GRACE satellite data can apply for great earthquake coseismic signal detection and focal mechanism inversions.