Harnessing the InSAR Data Revolution: GMTSAR

Overview

Satellite radar interferometry (InSAR) is a powerful technique for measuring small displacements (1- 10 cm) of the surface of the earth due to earthquakes, volcanoes, landslides, ice streams, grounding lines, uplift/subsidence from groundwater injection/withdrawal and underground nuclear tests.  The main advantage of this method is that one can monitor surface deformation anywhere on the Earth by using synthetic aperture radar (SAR) data provided by space agencies. Over the past decade we have developed freely available, open-source software for processing these data.  The software relies on the Generic Mapping Tools (GMT) for the back-end data analysis and display and is thus called GMTSAR. During our prior investigation funded by the NSF Geoinfomatics program, we accelerated the development of GMTSAR to include more satellite data sources and provide better integration and distribution with GMT. The most challenging aspect of this investigation was the development of a new type of image alignment algorithm that uses only the precise orbital information to geometrically align images to an accuracy of better than 10 cm.  This development was needed to process a new data type that is being acquired by the 20-year mission of the Sentinel-1 satellites (a-d). This combination of geometrically accurate software and open data is transforming radar interferometry from a tool for measuring displacement from events into a fully operational time series analysis tool.  

Intellectual Merit 

This proposed investigation will use standard software practices to harden the code, improve the geodesy, and make it more accessible to novice and advanced users on a wide array of UNIX platforms.  This will be achieved through improved and automated testing, partial redesign and simplification of the work flows, UNAVCO short courses, user feedback, and eventual migration of the code distribution and maintenance to a national facility (e.g., UNAVCO or Computing Infrastructure for Geosciences (CIG)).  Work will be performed by a postdoctoral researcher in collaboration with the GMTSAR and GMT development teams and with assistance from the XSEDE program. The expected outcome is to provide sustainable, open, geodetically-accurate software to move InSAR time series analysis from the intermediate-scale methods published today to large spatial and time scale analyses that are becoming possible using the new data streams.

Broader Impacts

The broader impacts of this work are to enable solid earth and cryospheric scientists to utilize the massive InSAR data sets that are becoming freely available to advance their interdisciplinary investigations.  Examples include monitoring volcanoes globally; estimating seismic hazard globally through strain-rate mapping; measuring co- and post-seismic deformation over spatial scales not resolved by GPS.  Geodetically accurate software with redundant observations, also provides new high spatial resolution maps of atmospheric and ionospheric processes.  Funding will be used to train a postdoc in best practices in software development.  The UNAVCO short courses, combined with thorough documentation, will educate students in the methods of large-scale data analysis and integration.