Research On Earthquake Thermal Anomaly Based On Remote Sensing Cloud Computin

China is one of the countries most deeply suffered by earthquakes.Therefore,a large number of earthquake events and precursors have been recorded in history.Earthquake precursor and statistical earthquake are currently important fields of study in seismology.Seismic precursors are a discipline that studies the physical and chemical phenomena that occur before an earthquake.Seismic statistics is a discipline that applies statistical theory to earthquake simulation and prediction.The development of the two systems is mutually independent,yet their research goals are consistent.The thermal infrared anomaly is an important precursor to earthquakes,and the Epidemic Type Aftershock Sequence(ETAS)model is widely used in earthquake prediction applications.There are many methods for extracting infrared precursor anomalies,but there is no evaluation of the impact of thermal infrared anomaly extraction methods based on long-term statistical analysis.The ETAS model is widely used and has many branches,but its parameter estimation and simulation prediction based on earthquake catalogs are not comprehensive enough.In light of the above-mentioned issue,this article adopts a comparative study and conducts two relatively independent research projects to address the problems related to infrared precursors and the ETAS model,aiming to generate new discoveries.In the paper,daily continuous nighttime surface temperature(Con LST)data was obtained from the Google Earth Engine(GEE)platform,and each different anomaly detection method was used to detect Thermal infrared anomaly.The research area of this article is the Sichuan region(27°-37°N,97°107°E),and the research timeframe is on a daily scale from the year2000 to 2020.To compare the effectiveness of different anomaly extraction methods,we selected six methods: Robust satellite techniques,Interquartile range,Wavelet transform,Kalman filter,Isolation forest,and Autoencoder.The heated core model was applied to explore Thermal infrared anomalies which is to filter anomalies unrelated to earthquakes by setting time-space-intensity conditions.The 3D error diagram offers scores to assume the best parameter set using training-test-validation steps.In the comparative analysis of the ETAS model experiment,the issues of model simulation prediction and parameter estimation were analyzed.Four types of ETAS models and parameter estimation methods(time ETAS,spatiotemporal ETAS,forward prediction likelihood ETAS,and Bayesian estimation ETAS)were selected for comparative study.Through comparing the simulated predictions of the optimal parameters,the superiority and inferiority of the models and estimation methods were judged.The results indicate that the Kalman filter method detected the highest seismic anomaly frequency without considering the heating core condition.The Autoencoder and Isolation Forest methods were able to determine the optimal alert type and parameter set for identifying earthquake-related anomalies.The RST method performed optimally in the final part of the workflow when considering physical factors such as active faults,seismic zones,and stresses.During the comparative experiments,the refined problem(heat core model)was used for statistical and quantitative analysis of thermal anomalies,moving away from the previous notion that proximity equals correlation,and offering a reliable method for evaluating the efficacy of TIR anomaly extraction.In the ETAS model comparison experiment,four methods were evaluated,and the results indicated that the forward predictive likelihood ETAS model performed better in estimating the seismic background rate and seismic trigger rate in the study area.On the other hand,the Markov chain Monte Carlo algorithm was more effective in obtaining convergent parameters for parameter estimation.The comparative study of ETAS models serves as an excellent means to identify key areas of focus for model simulation and prediction,facilitating subsequent model improvement and development.Therefore,it can help clarify ideas for improving and developing the model.During the discussion session,an experiment was conducted using a coupled ETAS model with an infrared precursor,based on the two comparative studies mentioned earlier.The anomalous background of the infrared precursor was used as a reference for the initial parameter correction of the ETAS model.The spatio-temporal ETAS model was used as the main body to compare the simulation prediction results before and after the 10-week correction.Although the revised results are an improvement over the original spatio-temporal ETAS model,the coupled model is not yet perfect.