Extraction Of Secular Variation Signals And Estimation Of Transfer Function From Geomagnetic Stations

The geomagnetic field is a fundamental component of the Earth’s system,with significant scientific and national defense applications.These applications include geomagnetic navigation,earthquake monitoring,and forecasting.To support these endeavors,an increasing number of geomagnetic stations have been established,leveraging major national scientific and technological infrastructure initiatives such as China’s Geophysical Station network and the Meridian Project.This paper focuses on researching high-precision geomagnetic field models and geomagnetic transfer function,using existing geomagnetic observation data products in China as a basis.Firstly,this paper presents a method for extracting secular variation signals of the main magnetic field at Chinese geomagnetic stations using continuous threecomponent observation data from 2009 to 2022.The extracted signals are compared with the IGRF and WMM to assess the accuracy of these major magnetic field models in the Chinese mainland.This evaluation provides valuable insights for the analysis of geomagnetic monitoring data and the subsequent development of high-precision geomagnetic field models.Additionally,the paper explores the stable geomagnetic transfer function using data from fluxgate magnetometers and search-coil magnetometers.The bounded influence far reference method is employed to obtain reliable results.By comparing the observed data with the two-dimensional forward modeling results of Occam,the feasibility of underground electrical structure monitoring based on continuous geomagnetic station observations is evaluated.The findings serve as theoretical guidance for the quantitative analysis and interpretation of the geomagnetic transfer function.The main research content and conclusions of this study are summarized as follows:(1)The paper introduces a novel approach for extracting the exogenous field proxy signal by applying a constraint model and principal component analysis to the residual covariance matrix based on annual difference calculations.The obtained proxy signal exhibits a strong correlation with the annual difference Dst index,indicating the effectiveness of this method.Due to the limitations of NGDC-720 and MF7 lithospheric magnetic field models in accurately describing small-scale magnetic anomalies at the station,the difference between the station value removed from the external field and the newly published model value is considered as the lithosphere magnetic field to be removed.Finally,the main magnetic field and its secular variation signal are extracted.(2)The IGRF and WMM main magnetic field models can basically describe the secular trend of the main magnetic field,but the model prediction error will gradually increase with time.After updating the models every 5 years,the defnitive geomagnetic reference model is basically consistent with the actual observation.The 1-year error of the F component of IGRF-12 and WMM2015 models was within 13nT,and the maximum error was 153.65nT in the 5th year,which is far from the requirement of ultra-high precision navigation while drilling.In comparison,WMM2015 model has a higher prediction accuracy of the main magnetic field in mainland China,which is of more reference value for the construction of high-precision geomagnetic models.(3)Coherence analysis of the observation data from geomagnetic stations reveals that the coherence of induction magnetometer data is higher than that of flux-gate magnetometer data.Moreover,the coherence and stability improve when using a 24hour window length compared to a 12-hour window length.The effectiveness of the bounded influence estimation method in obtaining the geomagnetic transfer function is verified,and remote reference processing proves effective in enhancing result stability.(4)By comparing the calculation results of the geomagnetic transfer function and the apparent resistivity before and after the Lushan MS7.0 earthquake,notable changes in apparent resistivity are detected in the short period frequency band leading up to the earthquake,followed by recovery.Additionally,three two-dimensional electrical structure models are designed to simulate actual fluid migration processes in horizontal strata,and the theoretical response characteristics of the geomagnetic transfer function are analyzed using the Occam forward method.Based on comprehensive analysis,it is deemed feasible to monitor underground electrical structures using continuous observation data from geomagnetic stations in China.Further studies combining ground electric field observation data at the same site can be conducted to explore time-varying magnetotelluric inversion.