The application of Convolutional Neural Networks to Detect Slow, Sustained Deformation in InSAR Timeseries

TL;DR

This paper investigates the use of convolutional neural networks (CNNs) to detect slow, sustained deformation in InSAR satellite data, establishing detection thresholds and demonstrating practical application on volcanic and geothermal sites.

Contribution

It introduces a CNN-based method for detecting deformation in InSAR data, evaluates detection thresholds with synthetic data, and proposes a mean-filtering approach for improved classification.

Findings

Detection threshold of 3.9cm for deformation signals in synthetic data

Over-wrapping improves detection sensitivity, reducing thresholds to 1.8cm

Deformation of 8.5cm/yr at Campi Flegrei detected after 60 days

Abstract

Automated systems for detecting deformation in satellite InSAR imagery could be used to develop a global monitoring system for volcanic and urban environments. Here we explore the limits of a CNN for detecting slow, sustained deformations in wrapped interferograms. Using synthetic data, we estimate a detection threshold of 3.9cm for deformation signals alone, and 6.3cm when atmospheric artefacts are considered. Over-wrapping reduces this to 1.8cm and 5.0cm respectively as more fringes are generated without altering SNR. We test the approach on timeseries of cumulative deformation from Campi Flegrei and Dallol, where over-wrapping improves classication performance by up to 15%. We propose a mean-filtering method for combining results of different wrap parameters to flag deformation. At Campi Flegrei, deformation of 8.5cm/yr was detected after 60days and at Dallol, deformation of 3.5cm/yr was detected after 310 days. This corresponds to cumulative displacements of 3 cm and 4 cm consistent with estimates based on synthetic data.