DOWNLOADING DATA VIA HTTPS
To learn more about Earthdata Login and register for an account, please see How to Register with Earthdata Login. Once you have logged in, data can be downloaded via a Web browser, command line, or client. For help with downloading data, please see Options Available for Bulk Downloading Data from HTTPS with Earthdata Login.
MEaSUREs Greenland Ice Sheet Velocity Map from InSAR Data, Version 2
This data set, part of the NASA Making Earth System Data Records for Use in Research Environments (MEaSUREs) program, contains seasonal (winter) ice-sheet-wide velocity maps for Greenland derived from Interferometric Synthetic Aperture Radar (InSAR) data obtained by RADARSAT-1, the Advanced Land Observation Satellite (ALOS), and the TerraSAR-X satellite.
See Greenland Ice Mapping Project (GIMP) for related data.
Note: These data are considered provisional pending a review by the MEaSUREs program. Once the data have been reviewed, this statement will be removed.
For Version 2, all velocity maps underwent additional screening for quality control and bad data points were removed. In addition:
- Improved baseline fits enhance consistency in the interior;
- Updated error estimates better represent the average behavior of the data;
- ALOS fine-beam data have been added to improve coverage;
- A new velocity map is included for 2009/2010
|Temporal Resolution:||12 month|
|Platform(s)||ALOS, RADARSAT-1, TSX|
|Data Contributor(s):||Ian Joughin, Ben Smith, Ian Howat, Ted Scambos|
|Metadata XML:||View Metadata Record|
As a condition of using these data, you must cite the use of this data set using the following citation. For more information, see our Use and Copyright Web page.Joughin, I., B. Smith, I. Howat, and T. Scambos. 2015. MEaSUREs Greenland Ice Sheet Velocity Map from InSAR Data, Version 2. [Indicate subset used]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. doi: http://dx.doi.org/10.5067/OC7B04ZM9G6Q. [Date Accessed].
As a condition of using these data, we request that you acknowledge the author(s) of this data set by referencing the following peer-reviewed publication.
Joughin, I., B. Smith, I. Howat, T. Scambos, and T. Moon. 2010. Greenland Flow Variability from Ice-Sheet-Wide Velocity Mapping, Journal of Glaciology. 56. 415-430.
Detailed Data Description
This data set contains Greenland ice sheet-wide velocity maps for seven winters derived from SAR data. The platform/sensor used as source data depends on the year. Table 1 lists the primary data source(s) for each winter.
|SAR Data Source||Temporal Coverage|
January 2013-March 2013
November 2012-March 2013
For maps of glacier outlet areas, some of which demonstrated profound velocity changes during the observation period, see the related data set MEaSUREs Greenland Ice Velocity: Selected Glacier Site Velocity Maps from InSAR.
Data are available in GeoTIFF (
.tif) format. Five GeoTIFF files are available for each data year: a velocity magnitude map; separate x- and y-component velocity maps (
vy); and separate x- and y-component error estimate maps (
The file naming convention used for this data set is:
greenland_vel_mosaic500_[yyyy_yyyy]_[vx OR vy]_v2.tif
greenland_vel_mosaic500_[yyyy_yyyy]_[ex OR ey]_v2.tif
The following table describes the variables in this data set's file naming convention.
||Greenland 500 m velocity magnitude mosaic|
||velocity x-direction, velocity y-direction|
||error x-direction, error y-direction|
.tif data file is approximately 64 MB The entire data volume (35 files) is approximately 2.2 GB.
This data set covers Greenland.
Southernmost Latitude: 60° N
Northernmost Latitude: 83° N
Westernmost Longitude: 75° W
Easternmost Longitude: 14° W
Data are provided in polar stereographic coordinates with a standard latitude of 70° N and rotation angle of -45° (sometimes specified as a longitude of 45° W). With this convention, the x-axis extends south from the North Pole along the 45° W meridian.
The data set currently provides velocity data for the following winters:
November 2012-March 2013
Velocities are reported in meters per year. The vx and vy files contain component velocities in the x and y directions defined by the polar stereographic grid. These velocities are true values and not subject to the distance distortions present in a polar stereographic grid. Small holes have been filled via interpolation in some areas. Interpolated values are identifiable as locations that have velocity data but no error estimates. Radar-derived velocities are determined using a combination of conventional InSAR and speckle tracking techniques (Joughin, et. al., 2002).
Error estimates are provided for all non-interpolated, radar-derived velocity vectors in separate GeoTIFF files appended with
_ey.tif. These estimates include the statistical uncertainty associated with the phase and speckle tracking error. Formal errors agree reasonably well with errors determined by comparison with GPS data (Joughin, et al, 2002). The values, however, underestimate true uncertainty in several ways, and as such should be used more as an indication of relative quality rather than absolute error. Refer to the Error Sources section for more details.
Software and Tools
Data Acquisition and Processing
The velocity maps in this data set were created using SAR data. The methods include a combination of speckle tracking and conventional interferometry. Velocity maps were produced by mosaicking multiple strips of InSAR-derived data. Individual images were selected based on two criteria: images should come from the same time of year; and images for each individual year should be chosen from as short a span as possible. For more detail, refer to Joughin, et. al,. 2002.
Annual mosaics were created using data collected over an approximately 96-day period during the winter. Areas with no data correspond to either regions where no data were acquired or where the interferometric correlation was insufficient to produce an estimate, most often in areas with high snow accumulation. In Version 2, the missing data value for magnitude files is -0.1 and 2e^9 for vx, vy, ex, and ey files. Regions with data represent the average of between one and three estimates (larger numbers may occur in regions of swath overlap, especially at higher latitudes).
The data are posted on a 0.5-kilometer grid, but the true resolution varies between 0.5 and 1 km. Many small glaciers are resolved outside the main ice sheet, but it is important to remember that for narrow (<1km) glaciers, the velocity represents an average of both moving ice and stationary rock, so while the glacier may be visible in the map, its speed is likely underestimated. Also on some of the smaller glaciers, interpolation produces artifacts where the interpolated value is derived from nearby rock, causing apparent stationary regions in the middle of otherwise active flow. The data have been screened to remove most of these, but please proceed with caution.
The following sections briefly describe how each winter's velocity map was generated. For Version 2, all maps underwent additional screening for quality control and bad data points were removed.
In late 2000 and early 2001, during the RADARSAT-1 Modified Antarctic Mapping Mission, the Canadian Space Agency (CSA) acquired nearly complete coverage of Greenland with multiple passes suitable for InSAR (September 3 to January 24). All of the available data for Greenland were used to produce the 2000 to 2001 mosaic. In cases where the data quality was too poor, some products were discarded. All source data were obtained from the Alaska Satellite Facility (ASF). Version 2 reprocessing and screening removed a number of artifacts.
In 2005 and 2006, RADARSAT-1 imaged most of Greenland four consecutive times, producing three InSAR pairs. Once all of the data were processed, poor coherence passes were screened out and the remaining data were used to assemble the velocity maps. Version 2 reprocessing and screening removed a number of artifacts.
In Version 1, the 2006/2007 mosaic was produced with RADARSAT-1 fine beam data. Version 2 substantially improves coverage by including ascending ALOS quad-pol data, including new coverage in the southeast. The ionospheric errors are often large (>20 m/yr) in the ALOS data, so the PIs manually removed points where the errors were excessive. This approach was chosen to strike a balance between maximizing coverage and minimizing error. Nonetheless, care should be exercised when interpreting these data, particularly in the southeast.
The 2007/2008 mosaic was produced with RADARSAT-1 fine beam data, in the same manner as the 2006/2007 mosaic. Version 2 adds a substantial volume of ALOS fine-beam data, largely along the northwest coast.
Version 1 of the 2008/2009 map was the first to utilize TerraSAR-X data. Version 2 corrects a substantial error on Rink glacier where the time interval was off by a factor of 2. Version 2 also utilizes subtantially more ALOS data, making this the most complete single-year mosaic.
New for Version 2, the 2009/2010 map consists almost entirely of ALOS SAR data collected in Fine-Beam, Single-Polarization (FBS) mode. Because L-band is more subject to ionospheric distortion of speckle-tracked azimuth offsets, streak errors for some areas are large (>10 m/yr), often exceeding the magnitude of the accompanying error estimates. In other areas, these errors are barely perceptible. Some of the worst streaks were edited out. However, a number of lesser streaks were left in place to: a) preserve coverage; and b) illustrate the magnitude of these errors with obvious examples. Despite being more susceptible to the ionosphere, L-band data correlate well in areas with high accumulation. As a result, this map has better coverage in the southeast than many of the maps from other winters.
Finally, about 20 coastal sites in this map utilize 30 km x 50 km TerraSAR-X scenes. These X-band data greatly improve the results for many of the fast moving outlet glaciers.
The 2012/2013 data were collected during the last few months in the life of RADARSAT 1 from January 2013 to March 2013. These data were combined with TerraSAR-X (TSX) winter data (November 2012 to March 2013). Processing is consistent with the Version 2 releases for the other mosaics.
Each image-pair that is used in the mosaic requires a 4-to-6 parameter fit for the baseline parameters (in other words, the separation between satellite tracks). For Version 1, the PIs fitted the baseline to a sparse, common set of ground control points as described by Joughin et al., 2010. This lead to errors exceeding 10 m/yr being misinterpreted as actual change (Phillips et al., 2013). In Version 2, for a year where the data were not well controlled, the PIs used control points from other years with adequate controls. This greatly improves consistency of the data from year to year. While this could mask some true change, the errors without this procedure were far larger than any change likely to occur.
As a result, these data should not be used to determine inter-annual change for interior regions of the ice sheet (roughly defined as areas above 2000 meters). In outlet glaciers close to the coast, where the baselines are well-constrained by bedrock, the velocity maps are well suited to this task. However, care should be exercised in interpreting any change observed in intermediate regions (roughly 1000 m to 2000 m). For example, be careful to avoid areas where the observed changes seem to follow swath boundaries. Refer to Figure 5 in Phillips et al., 2013 for more information. Note that baseline errors are not included in the formal error estimates and thus actual errors can always be substantially larger than stated. In particular, where they are derived largely from phase, the reported errors are extremely low (<0.5 m/yr). With baseline errors included, the actual error is more likely in the 1 to 3 m/yr range in most cases.
Version 1 of this data set reported error estimates based on local statistics. At some locations, however, the ionosphere introduced streak-like distortions in the azimuth offsets, resulting in much larger errors. In many cases, particularly in the interior, these errors are clearly visible as streaks running across the individual swaths that were used for the mosaic. Because these errors depend on the solar cycle, they are worse near solar max (in other words, the 2000/2001 velocity data). For each year, the PIs have estimated the average error across all data and added it to the Version 1 estimates. In addition, the actual errors are spatially variable and thus may be considerably larger than indicated in some areas, due to the assumption of spatially uniform ionospheric distortion.
In general, the error maps represent the average behavior of the data. This means that errors could be much lower than reported in some areas and much greater in others; care should be taken when assigning statistical significance based on the errors, especially given that the errors can be correlated over large areas. For example, even if the errors are correct in a global sense, one might compare two mosaics and find a large difference over 5% of the ice sheet. However, because errors can be spatially correlated over broad areas, one can not assume significance at the 95% confidence level; this might be precisely the 5% that statistically should exceed the errors because the errors are not uniformly distributed. By contrast, if the errors were completely uncorrelated, one could average over neighborhoods to reduce the error.
Compared with Version 1, more phase data (as opposed to speckle tracked) have been used for the x- and y-component of motion in the across-track direction, improving the accuracy in areas with slow-moving ice. The phase data also have been far more rigorously screened than in Version 1. In addition, some mosaics have more crossing orbit data (ascending and descending). This substantially improves the accuracy of both components in some areas by minimizing the use of noisy azimuth offsets.
Small holes in the final maps have been filled via interpolation. These points can be identified as those which have valid velocity data but no corresponding error estimate.
See Joughin, et al, 2002 for more detail on errors and how they were computed.
Version 2 was released in December, 2015. Refer to Table 3 for this data set's version history.
|V2||New map added for 2009/2010; baseline fits improved for consistency in the interior; error estimates revised (December, 2015).|
|V1.1||Binary data file format discontinued; data available in GeoTIFF only (August, 2015).|
|V1||Initial version (September, 2010)|
References and Related Publications
Contacts and Acknowledgments
Dr. Ian Joughin
University of Washington
Applied Physics Laboratory
1013 NE 40th Street
Seattle, Washington 98105
Dr. Ben Smith
University of Washington
Polar Science Center Applied Physics Laboratory
Department of Earth and Space Sciences
Seattle, Washington 98195
Dr. Ian Howat
Ohio State University
Byrd Polar Research Center
Scott Hall Room 108
1090 Carmack Road
Columbus, Ohio 43210
Dr. Ted Scambos
University of Colorado
Cooperative Institute for Research in Environmental Science
National Snow and Ice Data Center
CIRES, 449 UCB
Boulder, Colorado 80309
This project was supported by a grant (MEAS-12_0006) from the NASA Making Earth System Data Records for Use in Research Environments (MEaSUREs) Program.
DOCUMENT CREATION DATE