Hazard monitoring from Space: Next generation InSAR time series analysis (part of the SENSE Centre for Doctoral Training)

Radar Interferometry (InSAR) is a technique that provides measurements of surface displacement from Space, potentially with millimetric accuracy. These measurements are used in the natural hazards community for earthquake analysis and monitoring of volcanoes and landslides, as well as for monitoring anthropogenic activities such as oil and gas extraction, and drawdown of underground water storage.

In this project the student will work with leading scientists at Leeds, NASA and SatSense Ltd to develop new algorithms for extracting the measurements more accurately from the data. The student will then apply the new approach to selected natural hazards, e.g deforming volcanoes and landslides, and also to a case study of anthropogenic deformation.

A single “interferogram” provides a map of the surface displacement between two image acquisition dates, but also includes noise terms due to variable propagation delays through the atmosphere, changes of scattering properties of the surface, and data processing issues. Time series analysis techniques reduce these error terms to some extent by processing multiple interferograms together (Hooper et al, 2012). However, these techniques were designed to deal with sequences of 10’s of acquisitions rather than the 100’s of acquisitions that are possible with modern sensors, due to their short revisit times. In addition, hazard monitoring in close-to-real time requires rapid ingestion of new images without complete reanalysis of the time series. Progress in this direction has been made (Spaans and Hooper, 2016) but challenges still remain.

Other issues include: 1) Decorrelation events; InSAR only works when the ground scattering properties do not change significantly, so new construction and farming practices can lead to a complete loss of measurement for some areas. 2) Changes in the scattering properties of the ground due to changes in moisture content and vegetation (De Zan et al, 2015); this effect was previously assumed to average out over time, but it has been shown recently that this noise source can accumulate systematically when long times series are built from interferograms of shorter length.

In this project the student will develop a novel time series analysis algorithm that:

- Ingests new acquisitions rapidly;
- Works at a variety of resolutions;
- Handles variation in soil moisture and vegetation;
- Handles decorrelation events.

Selected natural hazards, such as deforming volcanoes and landslides, will be used as case studies, together with an example of anthropogenic deformation supplied by SatSense Ltd.

This PhD is part of the NERC and UK Space Agency funded Centre for Doctoral Training "SENSE": the Centre for Satellite Data in Environmental Science. SENSE will train 50 PhD students to tackle cross-disciplinary environmental problems by applying the latest data science techniques to satellite data. All our students will receive extensive training on satellite data and AI/Machine Learning, as well as attending a field course on drones, and residential courses hosted by the Satellite Applications Catapult (Harwell), and ESA (Rome). All students will experience extensive training on professional skills, including spending 3 months on an industry placement. See http://www.eo-cdt.org