| The change of sea level is related to all aspects of human survival,directly affects the evolution of global weather and climate,and also has a position that cannot be ignored in the research industry such as geodesy and ocean mapping.Early observations of sea level mainly came from the data of the tide detection station,but the data of the tide test station can only reflect the relative sea level change,and the degree of deformation of the earth’s crust will also affect the change of sea level.With the high development of space technology,GPS technology has become one of the important observation methods in the field of geodesy,according to the GPS station vertical time series can obtain the vertical deformation information of the earth’s crust,so it can be combined with the tide station data to obtain the absolute change of sea level.In order to effectively improve the coordinate time series accuracy of GPS station and tide detection station and obtain accurate absolute sea level change value,this paper explores the GPS time series noise model information under tide detection station and different solution strategies,adopts an improved EEMD noise reduction method on the basis of EEMD noise reduction method,and applies this method to the time series of tide detection station and GPS station,so as to obtain high-precision absolute sea level change results.Specifically,the main research contents of the paper are as follows:1.According to the characteristics of long time series data and low loss rate of tide detection station,ARMA,ARFIMA and GGM noise models were selected,and the optimal noise model estimation was carried out by BIC optimal noise evaluation criterion,and the time series of the tide test station was mainly manifested as three low-order models: ARMA(1,0),ARFIMA(1,0)and ARFIMA(2,2).The improved EEMD method is written by Matlab software and applied to the tide detection station time series,and finally the speed estimation of the tide detection station time series before and after noise reduction.The results show that the noise reduction effect of this method is obvious,which can improve the speed estimation accuracy of the time series of the tide detection station by about 11%.Through three sets of tide detection station data,the relative sea level changes in different time periods and time spans were analyzed.It is proved that the sea level change in modern times is more drastic than in the past100 years,and with the increase of time span,the speed value of the tide test station will be smaller,and the speed calculation accuracy of 100 years of data can be improved by nearly 10 times compared with 20 years of data.2.In view of the differences between different solving software and models,this paper deeply analyzes different single-day solution strategies,explores the influence of different solving strategies on the optimal noise model of GPS station time series,and estimates the speed through the optimal noise model,so as to obtain the speed change value of GPS station under different solving strategies.The results show that the GPS time series is mainly manifested as FN+WN and PL+WN noise models,while the PL+WN noise model is more likely to be expressed in the GISPY solution strategy.Different combined noise models will produce a certain degree of change depending on the direction or time series length,and the change of the solution strategy will amplify this result to a certain extent.The speed estimation results show that the change of the solution strategy has the most significant effect in the U direction.3.Select the combination of GPS station and tide detection station with similar station distance,low data loss rate and long time series,preliminarily process GPS data through GAMIT software,and reduce the noise of GPS and tide detection station time series by the improved EEMD method,and FN+WN and ARMA(1,0)are selected for the speed estimation model,respectively.The speed estimation results show that the station velocity values are distributed in the range of-0.88±0.57～2.00±0.39 mm/a,and the average sea level change rate is 1.06 mm/a.After processing,the maximum change of site speed reached 0.1 mm/a,and the uncertainty of absolute sea level velocity decreased,up to 0.11 mm/a,and the correct time series processing method could effectively improve the accuracy of data estimation when analyzing sea level change. |
PDF Full Text Request