Quantifying the bias introduced by vegetation in InSAR studies of ground deformation and surface processes

Paula Burgi, & Rowena B. Lohman

Submitted August 14, 2018, SCEC Contribution #8518, 2018 SCEC Annual Meeting Poster #140

With the advent of more frequently acquired SAR data, regions with significant vegetative cover are now easier to study. We explore the sensitivity of InSAR to vegetation and land cover variations, which can bias InSAR-based interpretations of deformation histories. We analyze ALOS-1 data from Cascadia and Sumatra, where clearcutting is prevalent. SAR interferograms are sensitive to the vegetation structure, with the radar signal interacting with the canopy at a height that depends on radar wavelength and vegetation type. Similarly, Digital Elevation Models(DEMs) used in InSAR processing do not always represent the bare-earth surface, and are affected by the vegetation structure that existed at the time of their acquisition.Our target regions have experienced ongoing clearing since the time of the SRTM DEM acquisition (2000) and throughout the timespan of ALOS-1 SAR data acquisition (2007-2011). In this study, we explore methods to address these DEM- and vegetation-related issues, and assess their impact on InSAR time series analysis. We make use of independent remote sensing datasets such as Landsat to identify cleared areas, and apply a baseline-dependent, localized phase correction to each interferogram during time periods where the vegetation characteristics of that pixel did not change. We compare our corrected time series to a time series generated using more standard approaches. Globally, many fault systems are in highly vegetated regions, making it important to understand the contribution of vegetation to InSAR deformation time series. This work constrains the bias introduced by changes in vegetation, allowing for finer scale resolution of deformation on fault systems globally.

Key Words
InSAR, methods, vegetation, time series

Citation
Burgi, P., & Lohman, R. B. (2018, 08). Quantifying the bias introduced by vegetation in InSAR studies of ground deformation and surface processes. Poster Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Tectonic Geodesy