Study On The Subsidence Monitoring Caused By Repeated Excavation With DInSAR Technology

The phenomenon of coal-mining-caused subsidence is very common in China.To solve the problem facing the mining industry, mine, miner and mining city andachieve sustainable development, people should press forward on subsidencemonitoring research. Continuous and efficient monitoring could reveal the rules ofsubsidence and the damage level it caused, thus helpful for people to arrange the coalmining plan and control ground subsidence. The traditional method for subsidencemonitoring, such as GPS, leveling, could not meet the demands of productiondevelopment due to its low coverage, labor intensive and low efficiency. The rapidlydeveloped DInSAR technology, which could work in all weather, all-day-long withhigh resolution and continuous spatial coverage, has shown great potential formining-caused subsidence monitoring. This dissertation aims to make further researchon the application of DInSAR to monitoring subsidence in the mining area. Firstly theerror propagation of some certain factors, mainly the baseline estimation and externalDEM error on DInSAR were analyzed, which showed that the theoreticallyachievable millimeter-level precision of DInSAR technology could be significantlydegraded by this two factors. A method based on high coherent points was proposedand experiment with TerraSAR-X dataset proved its efficiency for baseline estimationin mountainous area. Then three types of SAR datasets with different wavelengthwere processed by using two-pass DInSAR method with different multilook factorsand external DEMs. Results show that DInSAR technology based on TerraSAR-X andALOS-PALSAR dataset could detect a deformation rate of13.3cm/11d and40.6cm/46d respectively. For EnviSAT-ASAR, because of the limitation ofwavelength and spatial-temporal resolution, the ability to detect big deformation isway behind. However, due to the high temporal and spatial gradient of subsidence inthe research area, DInSAR could not capture the biggest subsidence value, even withL-band SAR images. And, due to the complex topography of the research area, SRTMwas proved not sufficient enough for topographic phase compensation, and therelatively coarser resolution could bring additional error during geocoding, whichcould be highly improved by using the higher resolution DEM, derived fromaero-photogrammetric materials. Furthermore, with topographic phase correctlyremoved, the well-geocoded DInSAR results could provide accurate boundary ofsubsided area, which could then be used for the angular parameters related to mining subsidence derivation.The corner reflector (CR) was integrated into subsidence monitoring in theresearch area. Hands-on experiences were accumulated from the design of CR toinstallment and identification. The precisely identified CRs based on point targetresponse analysis method were then used to evaluate and improve the precision ofhigh-resolution SAR image geocoding in mountainous area, which providesguarantees for the interpretation of DInSAR results.During DInSAR processing, the procedure to estimate atmospheric phase inPSInSAR technology was used, and the results proved this method very efficient toget rid of atmospheric phase. Results from time series DInSAR and GPS werecombined to derive mining-subsidence related parameters, such as the advancedinfluence angle, angle of subsidence, moving distance and subsidence coefficient,which could be used to evaluate the impact of repeat excavation on ground subsidenceand promote the quantitative application of DInSAR to mining subsidencemonitoring.