Go_LIVE! - Global Near-real-time Land Ice Velocity data from Landsat 8 at NSIDC
Abstract
The National Snow and Ice Data Center (NSIDC) is developing a processing and staging system under NASA funding for near-real-time global ice velocity data derived from Landsat 8 panchromatic imagery: Global Land Ice Velocity Extraction from Landsat (Go_LIVE). The system performs repeat image feature tracking using newly developed Python Correlation (PyCorr) software applied to image pairs covering all glaciers > 5km2 as well as both ice sheets. We correlate each Landsat 8 path-row image with matching path-row images acquired within the previous 400 days. Real-Time (RT) panchromatic Landsat 8 L1T images have geolocation accuracy of 5 meters and high radiometric sensitivity (12-bit), allowing for feature matching over low-contrast snow and ice surfaces. High-pass filters are applied to the imagery to enhance local surface texture and improve correlation returns. Despite the excellent geolocation accuracy of Landsat 8, the remaining error introduces an artificial offset in the velocity returns. To correct this error, we apply a shift to the x and y grids to bring the displacement field to zero over known stationary features such as bedrock. For ice sheet interiors where stationary features do not exist, we use near-zero (<10 ma-1) or slow-moving ice areas (10-25 ma-1) to refine velocities. Go_LIVE will eventually include Landsat 7, 5 and 4 imagery as well. Go_LIVE runs on the University of Colorado's supercomputer and Peta Library storage system to process 10,000 image pairs per hour. We are currently developing a web-based data access site at NSIDC. The data are provided in NetCDF (Network Common Data Format) as geolocated grids of x and y velocity components at 300 m spacing with accompanying error and quality parameters. Extensive data sets currently exist for Alaskan, Antarctic, and Greenlandic ice areas, and are available upon request to NSIDC. Go_LIVE's goal for 2017 is a system that updates global ice velocity at few-day or shorter latency.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFM.C13D0858K
- Keywords:
-
- 0720 Glaciers;
- CRYOSPHEREDE: 0758 Remote sensing;
- CRYOSPHERE