Monitoring of Mining Subsidence in a Sector of Central Chile through the Processing of a Time Series of Sentinel 1 Images Using Differential Interferometry Techniques (DInSAR)

: The objective of this work is to monitor the mining subsidence in a mountain range sector of central Chile between 2014 and 2018 through the processing of a time series of Sentinel 1 images using differential interferometry techniques (DInSAR). As a pretest, nine interferometric pairs were considered and processed using SNAP and SNAPHU software. Coherence levels obtained in complex topography sectors were low, due to both temporal and geometrical decorrelation. However, it is proposed that data from the PAZ project (X-band) and the Argentinean SAOCOM satellite (L-band) be used in the future.


Introduction
The subsidence of a mining sector is the sinking of the surface, due to the extraction of rocks or minerals from an underground mine [1]. The detection and monitoring of land subsidence make it possible to detect temporary changes in the sinking pattern and provide important information on the dynamics of this process. In addition, estimating where the future subsidence would occur would facilitate decision-making and the avoidance of damage to mining facilities and the environment [2].
The Interferometric Synthetic Aperture Radar (InSAR) is a remote sensing technique based on the process and analysis of SAR images which consists of measuring the optical path differences between the satellite and the earth, in two revisits of the satellite by the same area of study [3].
In this context, one of the techniques commonly used for this type of study in different parts of the world is differential interferometry (DInSAR), which is a technique based on synthetic aperture radar satellites that can be used to measure surface displacement in large regions with high spatial resolution. Under good conditions, displacements can be measured with centimeter to subcentimeter accuracy [4]. By monitoring the subsidence in the El Teniente mine in central Chile using the DInSAR technique during 2010 and 2015, subsidence was estimated in some sectors of up to −5 cm with TerraSAR-X and COSMO-SkyMED (X band) images [5]. When using the DInSAR technique with COSMO-SkyMED images in the same sector, an accumulated movement of −4 cm between 2010 and 2015 was estimated [6]. The objective of this research is to monitor the eventual subsidence that would occur in the mining sector in central Chile using DInSAR techniques with Sentinel-1 images.

Study Area
The study area corresponds to a mining sector of the Libertador General Bernardo O'Higgins Region located in the central zone of Chile between 34°1′18″ and 34°12′57″ south latitude and 70°28′18″ and 70° 28′29″ West longitude ( Figure 1). This sector has a Mediterranean climate (warm summers) with mountain influence [7].
The mining sector corresponds to an underground copper mine with a processing capacity of 140,000 tonnes per day [8].

Acquisition of Images
Nine interferometric pairs generated from SAR images of the Sentinel-1 satellite (Table 1), C-band (5.405 GHz) in Interferometric Wide Swath (IW) acquisition mode with Single Look Complex (SLC) processing level with 5 × 20 m spatial resolution and VV polarization were used for this study. The images were downloaded from the free Alaska Satellite Facility server.

Formation of Interferograms and Estimation of Displacement
Nine interferometric pairs were used from 2014 to 2018 to study subsidence in a mountain range sector in central Chile. The interferometric pairs were processed independently in the SNAP (Sentinel Application Platform) software developed by ESA, and the SNAPHU software [9] was used to phase unwrapping. Figure 2 shows the steps applied to interferometric pairs using the DInSAR technique.

Results
Nine interferograms generated from Sentinel-1 images with VV polarization, both descending and ascending pairs, were processed between 2014 and 2018. Due to time laps between acquisitions and geometrical decorrelation in this complex relief area, observed coherence level in the mining sector under study is less than 0.5 as shown in Figure 3. This is observed both in the descending and ascending pairs and to obtain displacement, coherence values greater than 0.5 should be considered, since these values are considered a good indicator of the interferometric phase quality [10].  Figure 3 shows the interferograms obtained from the nine interferometric pairs, with a low correlation in the mining sector, as a result of the low coherence in the sector. This low coherence is mainly due to time and geometric decorrelation. Due to these factors, it was not possible to measure displacement (subsidence) in this sector.

Discussions and Conclusions
Free and open access to the Sentinel-1 images allowed for a large amount of data in the study area, however, the levels of coherence obtained in sectors of complex topography were low (<0.5), due to time and geometric decorrelation, so it was not possible to estimate the subsidence in the mining area, even when using C-band. A study carried out in a mining sector in the Metropolitan Region of Chile, where TerraSAR-X and Sentinel-1 data used the MSBAS technique, obtained coherence values lower than 0.45 so it was not possible to completely remove the topographic phase in order to obtain the differential phase to estimate the subsidence [11].
Another study has estimated the subsidence through the DInSAR technique in the mining sector under study with COSMO-SkyMED images (X-band), where subsidence of up to −4 cm is monitored in some sectors [6]. However, these images are commercial and have a spatial resolution of up to 1 m.This work corresponds to the preliminary results of the first-year doctoral research and It is proposed that data from the PAZ project (X-band), the Argentinean satellite SAOCOM (L band), the MSBAS and PSI techniques, different software and all interferometric pairs available, be used in the future, which would improve surface deformation analyses like those of subsidence.