Research On The Theory And Method Of Bathymetry Prediction Using Satellite Altimetry Gravity Data

The global bathymetry map based on marine surveying and mapping occupies an extremely important position in scientific research,national economy and national defense.However,It is time-consuming and labor-consuming as well as the distribution of bathymetry points are uneven to use ships as carrier and the sonar echo to measure in the forms of point,line,and face.According to the isostasy theory,there is a certain correlation between gravity information and seafloor topography.The correlation makes it possible to use gravity information to predict the seafloor topography.Therefore,the paper focused on the theory and method of predicting seafloor topography based on satellite altimetry and emphasized on the waveform re-tracking on the coast.The methods of predicting seafloor topography such as gravity-geologic method,admittance function method,linear regression technique and least squares collocation were studied.The algorithms for predicting seafloor topography based on satellite altimetry gravity anomalies and vertical gravity gradient anomalies were analyzed.Solutions jointing gravity anomalies and vertical gravity gradient anomalies to predict seafloor topography were explored based on deeply analyzing the relationship between gravity anomalies/vertical gravity gradient anomalies and seafloor topography.The main tasks and innovations are as follows:1.Focused on the problem that satellite altimetry data is not accurate in the offshore area,the principle of formation of the waveform was discussed and the waveform re-tracking methods were studied including the Off-Center of Gravity,Threshold algorithm,Ice-2 algorithm and SeaIce algorithm,etc.The linear and index forms of 5-? method and Ice-2 method based on Brown model were used to re-track the ocean waveform of the Jason-2 in the sea area of the South China Sea.The experimental results showed that the range correction of the linear form of 5-? in the vast sea area was relatively large,the range correction and the standard deviation of the index form were close to Ice-2 method.Ice-2 method is relatively simple and has high success rate in waveform re-tracking,therefore,we can give priority to the Ice-2 method when handle the ocean echo waveform of Jason-2.2.Gravity-geologic method(GGM)which is one of the means of predicting seafloor topography was explored.The paper proposed a method using frequency domain continuation to obtain the key parameter(density difference constant),then the Mariana Trench was selected as experimental area to verify the algorithm and both sides of landform in "Challenger Abyss" was analyzed based on the results.The result showed that seafloor topography predicted by GGM model was superior to ETOPO1 model,which was better than the results of directly gridding ship sounding data.The relative error of model was not related to sea-depth and was obviously affected by the change of seafloor topography.Sea-depth difference between the models was large in the area where ship data is missing or too sparse.There was significant difference in both sides of landform in "Challenger abyss".In the south side of the trench,the sea-depth is less than 5000 m,gentle slope,about 2 °?5°;sea-depth greater than 5000 m,slope increased significantly,about 10 °?15°.In the trench north,slope is very large,about 10 °?15°.3.In order to solve the problem of predicting seafloor topography in frequency domain,the rules and effects of signal conversion from spatial domain to frequency domain were studied.At the same time,in order to obtain the waveband range of gravity signal which is strongly correlated with seafloor topography,the frequency domain coherence technique of the signal was discussed.Taking the southwestern Pacific Ocean as test object,vertical gravity gradient anomalies/gravity anomalies in the sea area and the seafloor topography have strong correlation in the 20km?200km band range.4.Considering many geophysical parameters were involved in the process of predicting seafloor topography,the paper studied the latest global crustal model CRUST 1.0,from which parameters such as crust thickness,mantle density and crustal density were obtained.Due to the effective elastic thickness cannot be observed,the internal relationship between "observed admittance" and "theoretical admittance" was analyzed and compared.Based on this,the best estimation of the effective elastic thickness in some parts of the Philippines was 10 km.5.Considering the fact that gravity anomalies and vertical gravity gradient anomalies have different sensitivity to different seafloor topography,the paper proposed self-adaptive weighting technology for predicting seafloor topography to combine gravity anomalies and vertical gravity gradient anomalies,then selected Philippine sea to test.It was found that the accuracy of the bathymetry model constructed by this method was the highest when the ratio of predicting result of vertical gravity gradient anomalies and gravity anomalies was 2:3.At the same time,using multi-source gravity data to predict seafloor topography could synthesize the superiority of gravity anomalies and vertical gravity gradient anomalies on the different seafloor topography,and the accuracy was better than bathymetry model that only used gravity anomalies or vertical gravity gradient anomalies.Taking the ship sounding data as checkpoints,the accuracy of predicting model was slightly lower than that of V18.1 model and improved by 27.17%and 39.02%respectively compared with the ETOPO1 model and the DTU10 model.Checkpoints that the absolute value of the predicting model's relative error was in the range of 5%were accounted for 94.25%of the total.6.For the traditional S&S method,constant term are usually neglected and scale factor of grid points have obvious error.Therefore,the data within a certain window radius around the grid point was taken as the input source,and the linear fitting method was used to obtain scale factor and constant term between gravity data and sea-depth.Then based on gravity anomalies/vertical gravity gradient anomalies,the linear regression analysis technique was used to carry out the bathymetry prediction test in the relevant sea area of the Southwest Pacific.The results showed that the scale factor obtained by different methods has a certain internal relation with seafloor topography.The scale factor in flat terrains is smaller and is larger in up and down terrain,which reflects the strong correlation between gravity data and seafloor topography.The accuracy of the bathymetry model constructed by linear regression analysis technique was the highest,which was about 46%higher than that of the traditional method.