A model-based approach for transforming InSAR-derived vertical land motion from a local to a global reference frame

TL;DR

This paper introduces a novel model-based method to convert InSAR-derived vertical land motion data from a local to a global reference frame, enhancing its application in large-scale geophysical studies without requiring GNSS data.

Contribution

The paper presents a new approach that uses a global VLM model and polynomial fitting to transform InSAR data to a global reference frame without GNSS measurements.

Findings

Method improves the accuracy of global VLM transformation.

Validated with datasets from Los Angeles and New York City.

Enhances the usability of InSAR data in regional land subsidence monitoring.

Abstract

Vertical land motion (VLM) observations obtained from Interferometric Synthetic Aperture Radar (InSAR) have transformed our understanding of crustal deformation processes over the past 3 decades. However, these observations are often related to a local reference frame, posing challenges for studies that require large-scale observations within a global reference frame, such as assessments of relative sea level rise and associated hazards. Here, we present a novel approach that enables transforming InSAR-derived VLM at any location worldwide to a global (e.g., International Terrestrial Reference Frame) reference frame without a direct need for GNSS (Global Navigation Satellite System) measurements. To this end, we employ a coarse resolution model of global VLM obtained by interpolating rates of all available GNSS stations over the global land areas. Our rationale is that the high-resolution InSAR-derived VLM data do not capture the long-wavelength signals present in the global VLM model. Therefore, we employ a set of 2D polynomial models to evaluate the difference between InSAR-derived VLM and the global model and then add it back to the InSAR-derived VLM. We examined the validity of our rationale using normalized power spectrum analysis and tested the effect of polynomial order on the accuracy of transformed VLM and the overall success of our approach using two datasets from Los Angeles and New York City. This approach improves the usability of InSAR-derived VLM in geophysical applications, including monitoring regional land subsidence.