Volcanic Activity Change Detection Using SqueeSAR-InSAR and Backscatter Analysis

Proper dealing with and preparedness for natural hazards require an efficient system of monitoring related activities. Some catastrophic events such as volcanic eruptions, show a precursory ground deformation caused by the magma ascending to the earth's surface that begins months to years before an eruption. Radar remote sensing is one of the most reliable geodetic techniques that can provide timely information over a vast area and is not affected by cloud and ash contamination. However, there is a challenge for dealing with decorrelations in natural terrain with low to high vegetation cover.

In this research, we present our approach to detect changes due to magma ascend based on two different aspects. First we estimate line of sight deformation based on the well-known SqueeSAR method and filtering of single look complex images to process both persistent and distributed scatterers together and increase the density of our observations. Second, we analyze the changes using backscatter values and polarimetric features over the areas where there is high decorrelation due to vegetation cover and cannot be detected through interferometric SAR. We apply these methods using Sentinel-1 and TerraSAR-X data to partially vegetated volcanoes in Indonesia and Ecuador (Merapi, Rijani, Agung, Pichincha and Sierra Negra).