Education Center | Photo Gallery

Data Advanced Data Search More Data Discovery & Access Easy-to-use Data Products Data Collections Data Software & Programming Tools Data Submission Data News Centers & Programs Antarctic Glaciological Data Center Arctic System Science Data Coordination Center IARC Frozen Ground Data Center International Polar Year Data & Information Service (IPYDIS) NOAA at NSIDC NSIDC Distributed Active Archive Center (NSIDC DAAC) U.S. Antarctic Data Coordination Center World Data Center for Glaciology, Boulder Science Scientists Research Projects Publications NSIDC Notes Glaciological Data Series Special Reports Posters & Presentations Annual Reports Education Resources News & Events Press Room NSIDC In the News Data News Events About Expertise Data & Services NSIDC Sponsors Contacts Jobs Related Organizations

Greenland Ice Sheet Melt Characteristics Derived from Passive Microwave Data

Summary

The Greenland ice sheet melt extent data, acquired as part of NASA's Program for Arctic Regional Climate Assessment (PARCA), is a daily (or every other day, prior to August 1987) estimate of the spatial extent of wet snow on the Greenland ice sheet since 1979. It is derived from passive microwave satellite brightness temperature characteristics using the Cross-Polarized Gradient Ratio (XPGR) of Abdalati and Steffen (1997). It is physically based on the changes in microwave emission characteristics observable in data from the Scanning Multi-channel Microwave Radiometer (SMMR) and the Special Sensor Microwave/Imager (SSM/I) instruments when surface snow melts. It is not a direct measure of the snow wetness but rather is a binary indicator of the state of melt of each SMMR and SSM/I pixel on the ice sheet for each day of observation. It is, however, a useful proxy for the amount of melt that occurs on the Greenland ice sheet. The data are provided in a variety of formats including raw data in ASCII format, data in binary format gridded on a Greenland subset of the Northern Hemisphere polar stereographic projection, and annual and complete time series climatologies in binary and GeoTIFF format all at a resolution of 25 km. All data are available via FTP.

Citing These Data

To broaden awareness of our services, NSIDC requests that you acknowledge the use of data sets distributed by NSIDC. Please refer to the citation below for the suggested form, or contact NSIDC User Services for further information. We also request that you send us one reprint of any publication that cites the use of data received from our Center. This helps us to determine the level of use of the data we distribute. Thank you.

The following example shows how to cite the use of this data set in a publication: List the principal investigators, year of data set release, data set title and version number, dates of the data you used (for example, April 1992 - September 1992), publisher: NSIDC, and digital media.

Abdalati, Waleed. 2007. Greenland Ice Sheet Melt Characteristics Derived from Passive Microwave Data, [list the dates of the data used]. Boulder, Colorado USA: National Snow and Ice Data Center. Digital media.

Overview Table

Table of Contents

1. Contacts
2. Detailed Data Description
3. Data Access and Tools
4. Data Acquisition and Processing
5. References and Related Publications
6. Document Information

1. Contacts

Investigator

Waleed Abdalati
NASA Headquarters
Washington, DC 20546-0001 USA
waleed@icesat2.gsfc.nasa.gov

Technical Contact

2. Detailed Data Description

Format

Raw Files

The Greenland Ice Sheet melt characteristics raw data files are provided in ASCII text format and contain a list of pixel coordinates showing melt on the Greenland ice sheet for the date specified in the title of the file. The pixel coordinates are listed in each file in two columns of numbers; the first column is the X location, and the second column is the Y location. You need to be aware, however, that these X and Y values apply to a 60 x 109 sub-image of the standard 304 x 448 Polar Stereographic (PS) grid. To convert these numbers back to the standard PS grid, add 129 to all of the X values and 260 to all of the Y values. For example, using 0,0 as the upper left pixel, the raw file's 0,0 pixel is the 129th X value and the 260th Y value in the standard grid; and the raw file's 60,109 pixel corresponds to the 189th pixel in X and the 369th pixel in Y. Note: The index transformation depends on whether you prefer to call the upper left pixel on the PS grid 0,0 (as in C or IDL) or 1,1 (as in FORTRAN). (Abdalati and Steffen 1995).

Gridded Files

The Greenland Ice Sheet melt characteristics gridded data files are provided in flat binary format (little-endian). The data were gridded to a 60 x 109 pixel subset of the Northern Hemisphere polar stereographic projection to make a product that could be easily displayed as an image.

Climatology Files

Both the complete time series and annual time series climatology files are provided in flat binary (little-endian) and GeoTIFF format. The binary files consist of a 60 x 109 pixel subset of the Northern Hemisphere polar stereographic projection. See the Processing Steps section of this document for more information about how the GeoTIFF files were created.

File and Directory Structure

The data are available on the FTP site in the nsidc0218_melt directory. Within this directory, there are four subdirectories as described in Table 1.

Figure 1 displays the FTP directory structure.

Figure 1. FTP Directory Structure

File Naming Convention

Raw Files

Raw files are named according to the following convention and as described in Table 2:

yyyydddiii.meltpts

where:

Gridded Files

The gridded files are named according to the following convention and as described in Table 3:

yyyydddiii.dat

where:

Climatology Files

Complete Time Series

The complete time series climatology binary and GeoTIFF files are named 19792007climatology_melt.dat and 19792007climatology_melt.tif, respectively.

Annual Time Series

The annual time series climatology files are named according to the following convention:

yyyyannual_melt.dat
yyyyannual_melt.tif

where:

File Size

Raw Files

The raw files range in size from zero to 16 KB per file. Note: zero KB files with no data in them indicate no melt for that day.

Gridded Files

The Gridded files are 13 KB each.

Climatology Files

Complete Time Series

The complete time series climatology files, 19792007climatology_melt.dat and 19792007climatology_melt.tif, are 13 KB and 4 KB, respectively.

Annual Time Series

The binary annual time series climatology files are 13 KB each, and the GeoTIFF annual time series climatology files are 2 to 4 KB each.

Spatial Coverage

The corner points of the Greenland subset of the EASE-Grid are listed in Table 5.

Note: Melt characteristics are only indicated for pixels completely within the boundaries of the Greenland ice sheet as digitized from the Quaternary Map of Greenland produced by the Geological Survey of Greenland (1:2,500,00 scale) (Abdalati and Steffen 1995).

Spatial Resolution

The resolution of all data files is 25 km.

Projection

The gridded data and the climatology files are gridded on a Greenland subset of the Northern Hemisphere polar stereographic projection. NSIDC chose the Hughes ellipsoid in constructing the polar stereographic grid. The Hughes ellipsoid assumes a radius of 6378.273 km and an eccentricity (e) of 0.081816153 (or e**2 = 0.006693883). For more information on polar stereographic projection, see the Polar Stereographic Projections and Grids Web page.

Grid Description

The Greenland subset of the 304 x 448 25 km Northern Hemisphere polar stereographic projection is a region starting at column 128, row 259 (0-based) that contains 60 columns and 109 rows.

Temporal Coverage

This data set currently spans 02 April 1979 - 31 October 2007. Processing is ongoing, and new data will be added as it becomes available.

Temporal Resolution

From 1979 to August 1987, data were recorded every other day; and from September 1987 to the most current processing, data were recorded every day.

Parameter

Snow melt is characterized by a threshold value derived from the XPGR equation. Refer to Equation 2. All values above this threshold were considered to be wet snow and therefore experiencing melt and those below the threshold were considered dry (that is, frozen).

Parameter Range

Raw Files

The range of the data for the raw files is zero to 59 for X values (left column) and zero to 108 for Y values (right column) indicating the X,Y location of the pixel showing melt.

Gridded Files

Pixels with a value of 1 indicate snow melt, and pixels with a value of zero in the interior of the ice sheet indicate no snow melt. Other pixels, those outside the boundaries of the Greenland ice sheet, including areas along the coastline or mixed pixels only partly covering the ice sheet are not included in the assessment for melt conditions and are set to -999 in the files.

Climatology Files

Binary

The values of the data in the binary climatology data files represent the total number of days that melt occurred at that pixel during a single melt season. For example, if a pixel has a value of 85, that means that there was melt recorded at that pixel for 85 days that year.

GeoTIFF

The values of the data in the GeoTIFF climatology data files represent the total number of days that melt occurred at that pixel during a single melt season. For example, if a pixel has a value of 85, that means that there was melt recorded at that pixel for 85 days that year. In the image, the brighter the pixel the more days of melt there were for that pixel for the year stated in the file name.

Sample Data Image

Gridded Files

Figure 2 displays a sample image for one day of the gridded binary data.

		- Melt
		- No Melt
		- Outside Boundary
Figure 2. 28 July 1981 Sample Data Image(1981209smr.dat)

Annual Climatology Files

Figure 3 displays a sample image of an annual climatology melt data GeoTIFF file.

Figure 3. 2007 Annual Melt GeoTIFF Sample Data Image (2007annual_melt.tif)

3. Data Access and Tools

Data Access

Data are available via FTP. A Google Earth™ animation, "Greenland Annual Surface Melt, 1979 - 2007," was also created with the annual melt GeoTIFF files and is available from NSIDC's View NSIDC Data on Virtual Globes: Google Earth Web page.

Volume

Raw Files

The volume of the raw ASCII data files averages approximately 572 KB per year.

Gridded Files

The volume of the gridded binary data files averages approximately 1.9 MB per year.

Climatology Files

Complete Time Series

The complete time series climatology files, 19792007climatology_melt.dat and 19792007climatology_melt.tif, are 13 KB and 4 KB, respectively.

Annual Time Series

The binary annual time series climatology files are 13 KB each, and the GeoTIFF annual time series climatology files are 2 to 4 KB each.

Software and Tools

The tools needed to read the data differ as a function of format: ASCII, binary, or GeoTIFF.

ASCII Files

These are readable in any text editing program such as Notepad, WordPad, or Microsoft Word.

Binary Files

The FTP site has a tools directory that contains IDL procedures and geolocation tools for reading and plotting the binary data.

IDL Procedures

The IDL procedures in the tools directory are listed in Table 6.

Geolocation Tools

To geolocate this data, a map projection parameter (.mpp) file, a grid parameter definition (.gpd) file, and an ice mask file (.byte) are provided in the tools directory on the FTP site. Table 7 lists the geolocation tools available.

For information on using the .mpp and .gpd files see the Points, Pixels, Grids, and Cells: A Mapping and Gridding Primer Web page.

GeoTIFF Files

The GeoTIFF files can be opened and viewed with applications for displaying and analyzing geospatial data such as ArcMap, ENVI, or OpenEV. For more information on these applications, visit the ESRI ArcGIS Web site, the ITT ENVI Web site, and the OpenEV Web site.

Note: Not all image reading software, such as Microsoft Photo Editor or Adobe Photoshop, are able to correctly interpret or open the GeoTIFF files. NSIDC recommends only using an application for displaying and analyzing geospatial data.

Related Data Collections

• Snow Melt Onset Over Arctic Sea Ice from SMMR and SSM/I Brightness Temperatures
• Program for Arctic Regional Climate Assessment (PARCA): Web page listing other PARCA data sets.

4. Data Acquisition and Processing

Theory of Measurements

Due to the harsh environment and inaccessibility of the Greenland Ice Sheet, remotely sensed data is ideal as it allows the scientist access to an environment that would otherwise be unattainable. Passive microwave sensors are uniquely adequate for the study of melting snow for several reasons: the amount of change in microwave emission when liquid water forms in the snow is dramatic and therefore, easily detected; microwave radiation can pass through cloud cover; measurements can be taken at night; and coverage of the Greenland Ice Sheet is frequent with these wide-swath instruments (Abdalati and Steffen 1997).

Sensor or Instrument Description

The instruments used to acquire this data set were the SMMR instrument on the Nimbus 7 satellite and the SSM/I instruments on the Defense Meteorological Satellite Program (DMSP) F-8, F-11, and F-13 satellites.

See the SMMR Instrument Description for more information about that instrument. Also, see the SSM/I Instrument Description and the DMSP Satellite F8, DMSP Satellite F11, and DMSP Satellite F13 platform descriptions for further information on those platforms.

Derivation Techniques and Algorithms

Data are derived from passive microwave satellite brightness temperature characteristics using the XPGR equation of Abdalati and Steffen (1997). Refer to Equation 2. A mask was applied to the Greenland Ice Sheet data to extract only those pixels that lie fully on the ice sheet so that no mixed pixels of land and snow or ocean cause false melt results. Consequently, approximately 6 percent of the ice sheet along its perimeter is omitted (Abdalati and Steffen 1997).

Microwave brightness temperatures can be characterized by the Rayleigh-Jeans approximation as shown in Equation 1 and as described in Table 8.

where:

As dry snow melts and transitions to wet snow, there is a sharp increase in the brightness temperature. These increases are dependent on frequency and polarization such that as snow melts there is a greater increase in the horizontal brightness temperatures than the vertical ones at the same frequency. The XPGR technique is an enhanced gradient ratio technique. It is useful because it utilizes both frequency and polarization in its calculation. When snow melts, there is a depolarization effect; the differences between the vertical and horizontal signal lessen. When the vertical 37 GHz channel and the horizontal 19 GHz channel are subtracted, the difference is greater than when the same polarization is used, thus making the distinction between frozen and melting snow easier (Abdalati and Steffen 1995). Using this relationship, a threshold value can be determined; all values below this threshold represent dry snow (frozen) and those above it represent wet snow (melting). The XPGR values determined during this investigation are listed in Table 9 (Abdalati and Steffen 2001).

Equation 2 shows the cross-polarized gradient ratio (XPGR) of Abdalati and Steffen and is described in Table 10.

where:

Error Sources

As with any satellite data, atmospheric conditions and variability can introduce errors in the data. Water vapor in the atmosphere or severe weather can lessen the microwave signal, possibly causing melting snow to go undetected. However, the normalization in the XPGR technique helps to alleviate this (Abdalati and Steffen 1997). You are likely to find an occasional spurious melt pixel here and there due to bad weather, or a bad scan. These occur a couple of times a year on the ice sheet.

Another possible source of error occurs when the snow refreezes. Because two frequencies are used in this method, emission depths must be considered. Generally, when snow is melting, the emission depths of the two frequencies vary by only a few centimeters because the temperature of the emitting layers is close to 0 °C. In the case of refreeze, the surface temperature is lower than the temperature at depth, thus, possibly causing a melt signal even when surface is frozen. This source of error is not thought to impact the results for annual melt extent maps but may need further study when applied to surface melt shorter time scales (Abdalati and Steffen 1995).

For a discussion of error sources in the SMMR brightness temperatures, see the Nimbus-7 SMMR Pathfinder Daily EASE-Grid Brightness Temperatures guide document.

For a discussion of error sources in the SSM/I brightness temperatures, see the DMSP SSM/I Pathfinder Daily EASE-Grid Brightness Temperatures guide document.

Processing Steps

Raw Files

Since the data were acquired from multiple instruments, cross calibrations must be applied for consistency of the data. To account for the transition between instruments, the calibration coefficients of Jezek (1991) were employed for the SMMR data, those of Abdalati and Steffen (1995) were employed for SSM/I F-11 and F-13 data, and the SSM/I F-8 coefficients were used as the baseline standard.

The raw files were processed using the following steps:

1. Data from the SMMR instrument were smoothed with a 5-day running mean.
    
2. The brightness temperature on a given day is an average of the scans from two days before and two days after that day.
    
3. The XPGR value is then computed and a pixel is determined to be showing melt or no melt based on that value.
    
4. If a pixel showed melt, it was recorded to the file.
   
   Note: During years when two instruments were acquiring measurements, for example, in 1987 both SMMR and SSM/I f08 were acquiring data, files for both instruments are provided.

Gridded Files

The gridded files were processed using the following steps:

1. The raw ASCII files were read with an IDL procedure and converted into PS gridded data.
    
2. A subset of the data is taken for the Greenland ice sheet using the mask file icemask_60x109.byte.
    
3. Pixels were given a value depending on if they are showing melt, no melt, or were outside the Greenland ice sheet boundary.
    
4. Then the program writes each gridded file to the .dat files.
   
   Note: During years when two instruments were acquiring measurements at the same time, for example, in 1987 both SMMR and SSM/I f08 were acquiring data, files for both instruments are provided.

Climatology Files

Binary

The binary climatology files were processed using the following steps:

1. The files were derived straight from the 60 x 109 gridded files.
   
   Note: Since the SMMR instrument only acquired measurements every other day, dummy files were created to make a complete daily time series. The dummy files are copies of the data file for the previous day.
    
2. Each gridded file for a particular year was queried to determine how many days of melt were recorded that year for a certain pixel.
    
3. These values were then written to the binary files.
   
   Note: During years when two instruments were acquiring measurements, for example, in 1987 both SMMR and SSM/I f08 were acquiring data, the data acquired from the newer instrument was used.

GeoTIFF

The GeoTIFF climatology files were processed using the following steps:

1. A data format converter was used to convert the binary climatology files to GeoTIFF format.

5. References and Related Publications

Abdalati, Waleed and Konrad Steffen. 2001. Greenland Ice Sheet Melt Extent: 1979-1999. Journal of Geophysical Research (Atmospheres) 106(D24): 33983-8.

Abdalati, Waleed and Konrad Steffen. 1997. Snowmelt on the Greenland Ice Sheet as Derived from Passive Microwave Satellite Data. Journal of Climate 10: 165-75.

Abdalati, Waleed and Konrad Steffen. 1995. Passive Microwave-derived Snow Melt Regions on the Greenland Ice Sheet. Geophysical Research Letters 22: 787-790.

Abdalati, Waleed and J. Stroeve. 1993. Climate Sensitivity Studies of the Greenland Ice Sheet Using Satellite AVHRR, SMMR, SSM/I, and In Situ Data. Meteorology and Atmospheric Physics 51: 239-258.

Jezek, K. C., C. Merry, D. Cavalieri, S. Grace, J. Bedner, D. Wilson, and D. Lampkin. 1991. Comparison Between SMMR and SSM/I Passive Microwave Data Collected Over the Antarctic Ice Sheet. Byrd Polar Research Center Technical Report 91-03, Ohio State University, Columbus OH, 62 pp.

Mätzler, C. H. and R. Hüppi. 1989. Review of Signature Studies for Microwave Remote Sensing of Snowpacks. Advances in Space Research 9: 253-265.

Ulaby, F. T., R. K. Moore, and A. K. Fung. 1982. Radar Remote Sensing and Surface Scattering and Emission Theory. Microwave Remote Sensing. Vol. 2, Artech House.

Ulaby, F. T., R. K. Moore, and A. K. Fung. 1986. From Theory to Applications. Microwave Remote Sensing. Vol. 3, Artech House.

Table 11 lists related documents that are available on NSIDC's Web site.

6. Document Information

Acronyms

Table 12 list acronyms used in this document.

Document Creation Date

August 2007

Document URL

http://nsidc.org/data/docs/daac/parca/nsidc0218_greenland_melt_pm/index.html