Skip to content

Monitoring Volcanic Deformation Using InSAR: Deformation Time Series at Seasonally Snow-covered Volcanoes

Tianyuan Zhu1, Juliet Biggs1, Alison Rust1, Milan Lazecký2, Loreto Córdova3

  • Affiliations: 1School of Earth Sciences, University of Bristol, Bristol, UK, BS8 1RJ 2COMET, School of Earth and Environment, University of Leeds, Leeds LS2 9JT, UK 3Servicio Nacional de Geología y Minería (SERNAGEOMIN), Santiago, Chile                 

  • Presentation type: Talk

  • Presentation time: Friday 09:30 - 09:45, Room R380

  • Programme No: 2.3.5

  • Theme 2 > Session 3

Abstract

Satellite-based Interferometric Synthetic Aperture Radar (InSAR) is widely used for volcanic deformation monitoring, particularly since Sentinel-1 launched in 2014, providing scientists with an unprecedented volume of openly-available data. Automatic processing and analysis systems process raw satellite data into interferograms and calculate deformation time series. However, snow often leads to coherence loss in interferograms, resulting in unwrapping errors and affecting monitoring accuracy, especially for automated systems. Here, we use MODIS MOD10A2 snow cover products to identify volcanoes that experience seasonal snow cover and find that roughly half have a snow persistence of 7-90%, primarily in high-latitude and high-altitude regions. Here we focused on Laguna del Maule (LdM), Chile which has been steadily uplifting since 2007 and is characterized by seasonal snow cover (snow persistence of 73%) and steep-sided lava flows. Automated LiCSBAS products underestimate the line of sight deformation rate by 26% at GPS site MAU2. First, we test the ability of MODIS MOD10A2 snow maps to predict InSAR coherence, and found an average accuracy of 78%, peaking at 99% during early winter, confirming that low coherence is primarily caused by snow cover. We then adapt the LiCSBAS processing strategy by generating long time-span summer-to-summer interferograms and applying a re-unwrapping algorithm. This changes the observed uplift rate and improves the agreement with the GPS observation. This study shows that MODIS snow cover products can predict interferogram coherence, helping to improve InSAR auto-processing efficiency. Furthermore, we demonstrate the importance of adjusting global-averaged parameters of processing system according to regional characteristics.