MODIS Mosaic of Antarctica (MOA) Image Map

Summary

Staff from the National Snow and Ice Data Center (NSIDC) and the University of New Hampshire have assembled two digital image maps of surface morphology and optical snow grain size that cover the Antarctic continent and its surrounding islands. The MODIS Mosaic of Antarctica (MOA) image maps are derived from composites of 260 MODIS (Moderate-resolution Imaging Spectroradiometer) orbit swaths acquired between 20 November 2003 and 29 February 2004. The MOA provides a cloud-free view of the ice sheet, ice shelves, and land surfaces, and a quantitative measure of optical snow grain size for snow- or ice-covered areas. All land areas larger than a few hundred meters that are south of 60° S are included in the mosaic, as well as persistent fast ice regions and some grounded icebergs present near the coast in the 2003-2004 austral summer. The MOA surface morphology image map is derived from digitally processed MODIS Band 1 data. The optical snow grain size image is compiled using a normalized ratio of atmospherically corrected, calibrated band radiance data from Bands 1 and 2.

The data are available via FTP at two spatial grid scales: 750 m (112 MB) and 125 m (4 GB), and via a Web-based map server that can create manually-selected Joint Photographic Experts Group (JPEG) images (as of this writing, the Web-based map server has only a qualitative grain size image and not the final grain size data product). Image data on the FTP site include binary 16-bit images that preserve the full radiometric content of the scenes. A complete discussion of data selection, processing, and validation along with examples of data applications are discussed in Scambos et al., 2007.

Citing These Data

Overview Table

Category	Description
Data format	Data provided through the map-server are in JPEG format. Data obtained via FTP are provided as flat binary 16-bit integer files (.img files) in little-endian byte order, and as GeoTIFFs.
Spatial coverage	Southernmost Latitude: 90° S Northernmost Latitude: 60° S Westernmost Longitude: 180° W Easternmost Longitude: 180° E
Temporal coverage and resolution	The 260 image swaths were acquired between 20 November 2003 and 29 February 2004.
Tools for accessing data	As of this writing, the digitally smoothed red-light images are available via FTP at two spatial grid scales: 750 m (112 MB) and 125 m (4 GB), and via a Web-based map server capable of creating manually-selected JPEG images. The snow grain size images are available only at 750 m resolution via FTP.
File size	Uncompressed data files can range in size from approximately 56 MB to 4 GB. Compressed data files on the FTP site range in size from 2 MB to 1.2 GB.
Parameters	The MOA images are useful for studying ice sheet surface morphology and snow grain size variations on the Antarctic continent.
Procedures for obtaining data	Image data (16-bit) are available via FTP. Manually-selected JPEG images can be obtained via Web-based map server.

Table of Contents

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

1. Contacts and Acknowledgments

Investigators

Ted Scambos
National Snow and Ice Data Center
449 UCB, University of Colorado
Boulder, CO 80309-0449 USA

Terry Haran
National Snow and Ice Data Center
449 UCB, University of Colorado
Boulder, CO 80309-0449 USA

Jennifer Bohlander
National Snow and Ice Data Center
449 UCB, University of Colorado
Boulder, CO 80309-0449 USA

Tom Painter
National Snow and Ice Data Center
449 UCB, University of Colorado
Boulder, CO 80309-0449 USA

Mark Fahnestock
Institute for the Study of Earth, Oceans, and Space
University of New Hampshire
Morse Hall
39 College Road
Durham, NH 03824-3525 USA

Technical Contact

Acknowledgements

We thank R. Bindschadler, W. Abdalati, and J. Ferrigno for their interest and support in initiating this project. This work was supported by NASA grant NNG04GM10G and USGS funds for support (instigated by J. Ferrigno) of J. Bohlander during image selection and cloud clearing.

2. Detailed Data Description

Data Files and Format

The following files are available via FTP:

Table 1. MOA Data Files

File Name	Description
MOA_r1_readme.txt	ASCII text file containing an abbreviated explanation of the data available for download
Moa.mpp	ASCII text file containing the map projection parameters
Moa0125.gpd	ASCII text file containing the grid parameter definition for the 125 m images
Moa0750.gpd	ASCII text file containing the grid parameter definition for the 750 m images
moa125_r1_hp1.img.gz	125 m high-pass channel 1 image 16-bit unsigned integer little-endian flat binary
moa125_r1_hp1.img.hdr	ENVI header for moa125_r1_hp1.img
moa125_r1_hp1.tif.gz	125 m high-pass channel 1 image 8-bit gray scale GeoTIFF 16-bit to 8-bit stretch: 15096=>0, 17283=>255
moa750_r1_cnt.img.gz	750 m count of MODIS scenes contributing to grid cells 8-bit unsigned integer flat binary; min: 0, max: 38
moa750_r1_cnt.img.hdr	ENVI header for moa750_r1_cnt.img
moa750_r1_cnt.tif.gz	750 m count of MODIS scenes contributing to grid cells 8-bit gray scale GeoTIFF; min: 0, max: 38
moa750_r1_hls.tif.gz	750 m hue-lightness-saturation image converted to red-green-blue 24-bit color GeoTIFF;      hue: blue (8-bit value = 240),      lightness: moa750_r1_hp1.img, 12465=>0, 16200=>1,      saturation: moa750_r1_nds.img, -30=>0, 162=>1
moa750_r1_hp1.img.gz	750 m high-pass channel 1 image 16-bit unsigned integer little-endian flat binary
moa750_r1_hp1.img.hdr	ENVI header for moa750_r1_hp1.img
moa750_r1_hp1.tif.gz	750 m high-pass channel 1 image 8-bit gray scale GeoTIFF 16-bit to 8-bit stretch: 15096=>0, 17283=>255
moa750_r1_nds.img.gz	750 m normalized difference snow index grain size image where bright=coarser grains; 16-bit signed integer little-endian flat binary
moa750_r1_nds.img.hdr	ENVI header for moa750_r1_nds.img
moa750_r1_nds.tif.gz	750 m normalized difference snow index grain size image where bright=coarser grains; 8-BIT gray scale GeoTIFF 16-bit to 8-bit stretch: -586=>0, 239=>255
moa750_r1_wgt.img.gz	750 m weight image mean "quality" value of contributed scene pixels 16-bit unsigned integer little-endian flat binary
moa750_r1_wgt.img.hdr	ENVI header for moa750_r1_hp1.img
moa750_r1_wgt.tif.gz	750 m weight image mean "quality" value of contributed scene pixels 8-bit gray scale GeoTIFF 16-bit to 8-bit stretch: 0=>0, 49965=>255

Table 2. Antarctica's Coastline and Ice Sheet Grounding Line Files

File Name	Description
moa_coast_file.txt	ASCII text file of point locations for the Antarctic coastline
moa_coastline.dbf moa_coastline.shx moa_coastline.prj moa_coastline.shp	Shapefile for the Antarctic coastline
moa_coastline.evf	ENVI vector file for the Antarctic coastline
moa_coastline.gmt	GMT file for the Antarctic coastline
moa_groundingline.txt	ASCII text file of point locations for the Antarctic grounding line
moa_groundingline.db moa_groundingline.shx moa_groundingline.prj moa_groundingline.shp	Shapefile for the Antarctic grounding line
moa_groundingline.evf	ENVI vector file for the Antarctic grounding line
moa_groundingline.gmt	GMT file for the Antarctic grounding line
moa_islands.txt	ASCII text file of point locations for the Antarctic islands
moa_islands.dbf moa_islands.shx moa_islands.prj moa_islands.shp	Shapefile for the Antarctic islands
moa_islands.evf	ENVI vector file for the Antarctic islands
moa_islands.gmt	GMT file for the Antarctic islands

The following data products are currently in development:

• 125 m and 750 m optical mean summer snow grain size
• 750 m standard deviation of optical snow grain size
• 750 m mean image "weight" for the surface morphology composite
• 750 m contributing image count for the surface morphology composite.

Registered users will receive e-mail notification about any product changes and new data availability. Please complete the User Registration Form.

File Size

Uncompressed data files can range in size from approximately 56 MB to 4 GB. Compressed data files on the FTP site range in size from 2 MB to 1.2 GB.

Spatial Coverage

Southernmost Latitude: 90° S
Northernmost Latitude: 60° S
Westernmost Longitude: 180° W
Easternmost Longitude: 180° E

All land areas south of 60° S that are larger than a few hundred meters are included in the mosaic. (Image grids include areas north of 60° S but these regions are zero-filled. Ocean areas more than a few tens of kilometers from the coast are also masked with zero-fill.)

Spatial Resolution

The MOA mosaic uses an image stacking "super resolution" or "data cumulation" scheme to increase resolution in the final product beyond that of the individual scenes. The estimated spatial resolution of the final surface morphology composite image ranges between 150 m and 250 m, depending on the number of images used in stacking and the mean "weight" of the images used (see Compositing the image swaths via Data Cumulation). The input MODIS data from Bands 1 and 2 have a nominal resolution of approximately 250 m.

Projection

The MOA 125 m grid images are provided in the Antarctic mapping projection recommended by the Scientific Committee on Antarctic Research (SCAR). The projection is identical to the Radarsat Antarctic Mapping Project Antarctic Mapping Mission 1 Synthetic Aperture Radar (RAMP AMM-1 SAR) 125 m mosaic.

	125 m Grid	750 m Grid	Upper left corner of upper left cell (m)
x dimension	48333	8056	-3174450.0
y dimension	41779	6964	2406325.0

• Polar stereographic projection spheroid: WGS-84
• Longitudinal range of 180° W to 180° E with 0 longitude being at the top of the image
• Latitude of true scale: 71° S

Note: South Pole is not at the center of this projection grid.

Grid Description

The digitally smoothed red-light images are available via FTP at two spatial grid scales: 750 m (112 MB) and 125 m (4 GB). Presently, the snow grain size images are available only at 750 m resolution via FTP. Both 125 m and 750 m grid scales are available for all MOA data products on the Web-based map server.

Temporal Coverage

The image swaths were acquired between 20 November 2003 and 29 February 2004. Images were selected so that the sun is towards the upper right of the projection grid in all scenes across the entire continent, accomplished by limiting the acquisition time window to between 0500 GMT and 1330 GMT. To maintain a roughly uniform solar elevation across the composite, images selected for coverage of the region near the 135° W longitude coastline were acquired near austral summer solstice; for images near the 45° E longitude coastline, the majority of the scenes were acquired late January through February.

Temporal Resolution

See Sensor Description.

Parameter or Variable

Parameter Description

The MOA data set images report two parameters: the morphology of the surface from brightness variations in MODIS Band 1 red-light satellite images and the normalized difference radiance ratio of red and near-infrared light, which measures grain size variations.

For the morphology image, many processing steps were required to create a seamless and uniform image, and many images were combined to generate a single grid cell value. Thus, the data values no longer have a clear (quantifiable) relationship to the top-of-atmosphere red light reflectance data from which they are derived. Instead, the image provides a semi-quantitative, but highly consistent, representation of the surface shape and approximate reflectivity as illuminated by the sun across all surface types for the entire continent.

For the grain size image, pre-processing was reduced, sacrificing the seamlessness of the red-light image in favor of a truer quantitative image of the ratio of radiances in the two bands. This ratio may be approximately converted to mean snow grain size in areas of dust-free, non-shadowed snow, firn, and ice. Investigators derived lookup table values to correct for atmospheric effects and partial corrections for bi-directional reflectance distribution function (BRDF) effects. A series of model runs of the Santa Barbara Discrete Ordinate Radiative Transfer (DISORT) Atmospheric Radiative Transfer (SBDART) software provided corrections. The lookup tables were applied to image grids of the radiance ratio and solar elevation to generate images of snow optical grain size. Investigators combined these images using a weighting scheme similar to that of the surface morphology mosaic, favoring nadir-viewed scenes to generate the final mean summer optical snow grain size images. (Note: As of this writing, these final "quantitative" snow grain size images are not available the FTP or Web sites, but will be in early 2007. At present, a "proxy" relative snow grain size image based on a normalized difference ratio of the calibrated Band 1 and Band 2 data is provided instead.) The final grain size image data is partially validated by comparison with in situ snow spectra of snow over Antarctic sea ice in October 2003.

Registered users will receive e-mail notification about any product changes and new data availability. Please complete the User Registration Form.

Error Sources

Three types of error are considered here: geolocation error, surface obscurations, and snow grain size error.

Geolocation of the component MODIS scenes and the final MOA mosaic, is estimated to be within 50 m, which is considerably less than the satellite image pixel size or the final grid spacing. The estimate was checked by comparisons to coastlines, field camps, mountain peaks, and other mosaics and maps. Coastline and grounding line files are estimated to be accurate (tracking the best estimate of these features) to within 250 m, that is, two grid cells. (The coastline grounding line interpretation may not be accurate in all areas).

The mosaic composite is almost perfectly cloud-cleared. Some areas of thin clouds, cirrus cloud shadows, and fog or low-lying small clouds are present in the northeastern Ronne Ice Shelf region, which was persistently cloudy throughout the 2003-2004 austral summer. Other known sites of remaining clouds are at the grounding line near Bailey Ice Stream (79.67° S, 33.1° W), the ice tongue of Jutulstraumen Glacier (70° S, 0° E), and the Mobiloil Inlet and the adjacent Solberg Inlet (68.4° S, 66.5° W). In numerous areas, there are small patchy clouds and shadows less than approximately 1 km in size as a result of using the 750 m resolution images for cloud evaluation and masking. Investigators identified what they assume are blowing-snow features in several areas on a larger scale (up to hundreds of square kilometers), particularly in East Antarctica over the upper slopes of the ice sheet. Blowing snow appears as low-contrast mottlings of the surface, often arranged in quasi-linear bands oriented near the mean katabatic wind direction (for example, as mapped by Parrish and Bromwich, 1993). Also, widespread are low-contrast artifacts from hoar frost patches. These are regions where fog or emerging vapor from the snowpack have formed frost crystals on the surface. They generally appear as sharp-edged patches, often with a sawtooth or flame-like outline, that cross-cut the undulations or other topography on the ice surface. Selecting late spring or early summer images where possible reduces the number of hoar patches and averaging images acquired over a wide period of time (frost patches change on a scale of days) reduces their intensity in the final image.

Optical snow grain size error is estimated to be approximately +/- 50 micrometers based on comparison with in situ spectra of varying snow grain sizes with near-simultaneous MODIS images, processed in the same manner as the MOA grain size composite scenes. However, snow grain size varies greatly over the period of image acquisition for the MOA. So, in some areas, there are large ranges of snow grain size that were averaged together (for example; melting areas, or warm but sub-freezing areas that experienced numerous snowfalls followed by snow diagenesis).

Quality Assessment

Accuracy of the Level 1A geolocation data is estimated to be 50 m (Wolfe, et al., 2002), which is considerably better than the Level 1B ground-equivalent nadir pixel size, 250 m. Investigators conducted several tests of this geolocation accuracy and precision using known surface sites (for example, South Pole Station, Vostok Station, Siple Dome camp and traverse trail, Megadunes Camp runway, Dome Concordia camp) and areas of well-mapped coastline (for example, Ross Island and northern Peninsula).

Investigators did not find discrepancies of greater than 125 m in the projected location of a fixed object among the 260 scenes or relative to well-mapped coastline positions. Further, overlapped areas of separate images showed identical feature locations on the grid to within one grid cell. Thus, in areas where the mapped coast position differs from the MOA image, investigators conclude that the MOA coast is more accurate (relative to, for example, the CIA coastline database or the Antarctic Digital Database).

The "mean image weight" and "image count" data products are provided as a means for users to assess image quality in various parts of the MOA. In general, the MOA quality is higher in areas of both high count and high weight.

3. Data Access and Tools

Data Access

The mosaics are being distributed via a Web-based map server that permits instantaneous zoom, a series of contrast stretches, and geolocated image delivery. The data are available from both the University of New Hampshire and NSIDC Web sites. Image data (16-bit) are available via FTP. Manually-selected JPEG images can be obtained via Web-based map server.

Volume

The total compressed size of all files is about 1.8 GB, and the total uncompressed size is about 6.7 GB.

Software and Tools

See the Map Server User's Guide.

Related Data Collections

• Images of Antarctic Ice Shelves
• MODIS Data at NSIDC
• RAMP Data at NSIDC
• Radarsat Antarctic Mapping Project Digital Elevation Model Version 2

See Also

• British Antarctic Survey: Antarctic Digital Database
• MODIS Characterization Support Team
• MS2GT Software at NSIDC
• NOAA/ESRL Global Monitoring Division
• Santa Barbara DISORT Atmospheric Radiative Transfer
• Scripps Institution of Oceanography: Light Scattering Code Library

4. Data Acquisition and Processing

Theory of Measurements

The ability of visible and near-infrared (VIS-NIR) satellite sensors, notably the Landsat series and the Advanced Very High Resolution Radiometer (AVHRR), to reveal previously unknown features of the Antarctic continent and its coastline was broadly recognized in the 1980s by glaciologists and cartographers. Later, a host of studies showed how careful processing of image radiometry could provide unprecedented information about the ice sheet surface revealing details of ice flow, sub-ice bedrock structure, and wind-related features in the interior of the ice sheet by detailed portrayal of the subtle surface morphology at the approximately 1 km spatial resolution scale of AVHRR (Orheim and Lucchitta, 1988; Bindschadler and Vornberger, 1990; Scambos and Bindschadler, 1991; Seko et al., 1993). The U. S. Geological Survey created continent-wide mosaics using AVHRR (USGS, 1991; USGS, 1996); USGS and other groups then created regional ice feature image maps using Landsat (for example, Ferrigno et al., 1994; Swithinbank et al., 1988). Recognizing that a red-infrared band combination can provide grain size information, Winther et al., (2001) used the VIS-NIR data from the USGS/Ferrigno AVHRR mosaics to generate an approximate estimation of total blue ice area for the Antarctic continent. This followed earlier experiments at mapping snow grain size and blue ice extent from space (for example, Bourdelles and Fily, 1993; Orheim and Lucchitta, 1990). More recently, mosaics using synthetic aperture radar (SAR) have been compiled (for example, Fahnestock et al., 1993; Jezek, 1999) providing a unique view of the backscatter of the surface layer (to approximately 30 meters) over the continent. This active sensor type, operating at wavelengths nearly unaffected by clouds, permits a great deal of control over the assembly of the continent-wide image. However, surface morphology is muted and overprinted by a strong volume scattering component coming from deeper structures such as melt layers, coarse hoar crystals, or annual snow layering.

Compiling a uniform continent-wide satellite image mosaic using VIS-NIR data presents a number of challenges. The Antarctic continent combines low-contrast ice-dynamics-related features in the interior with high-contrast mountain and dry valley areas near the coast. Being a polar target, there are several considerations regarding image acquisition, solar elevation and azimuth, visible-light scattering off the snow surfaces, and the latitudinal limits and off-nadir degradations of the sensor. Clouds, cloud shadows, and blowing snow must be identified and removed, and the images combined seamlessly with uniform histograms.

The normalized difference band radiance ratio correlates with surface grain size, because snow reflectivity decreases in the infrared as grain size increases (Warren, 1982; Fily et al., 1997; Painter and Dozier, 2004). In detail, this reflectivity change is due to absorptive interactions between infrared light and ice crystal electronic structure. Decreasing reflectivity in the infrared region of the electromagnetic spectrum is the principal observable contributing to the capability to remotely determine snow grain size, and it is a function of the greater mean absorbing path length within a larger grain, that is, increased absorption with longer path length through the crystal structure. However, optical path within ice is also a function of grain shape; complex, feathery grains will have a small optical grain size even though individual grains may have a much larger maximum dimension. Moreover, smaller grain sizes have a larger single-scattering component off the crystal surface. For most albedo or energy
balance investigations, optical grain size is the desired parameter, because the fundamental physics of the study concerns ice interacting with light.

Sensor Description

The Earth Observing System (EOS) is a series of polar-orbiting, low-inclination satellites established by the National Aeronautics and Space Administration (NASA) for global observations of the land surface, biosphere, solid Earth, atmosphere, and oceans. The MODIS sensor is present on both the Terra (EOS AM) and the Aqua (EOS PM) satellites. The Terra satellite, which was launched on 18 December 1999, crosses Earth's equator at 10:30 a.m. traveling from north to south; the Aqua satellite, which was launched 4 May 2002, crosses Earth's equator at 1:30 p.m. traveling from south to north. Terra MODIS and Aqua MODIS view the entire Earth's surface every 1 to 2 days. These satellites have a 705 km sun-synchronous, near-polar, circular orbit.

MODIS contains 36 spectral bands. The spatial resolution of Bands 1 and 2 is 250 m, 3 through 7 is 500 m, and 8 through 36 is 1000 m. The swath dimension is 2330 km (cross track) by 10 km (along track at nadir). The sensor collects data at an orbital average rate of 6.1 Mbps. There are 12 bits per image cell of data.

Images in this data set were derived from MODIS Bands 1 and 2. The table below shows some specifications on these two high resolution bands.

Primary Use	Band	Bandwidth (nm)	Spectral Radiance (W/m2/µm/sr)	Required Signal-to-noise Ratio
Land/Cloud/Aerosols Boundaries	1 (Red)	620 - 670	21.8	128
	2 (NIR)	841 - 876	24.7	201

Data Acquisition Methods

Image Selection

MOA consists of 260 MODIS satellite image swaths. A specific range of Universal Time (05:30 to 13:30 UT) was used in selecting images. Thus, solar illumination direction was limited to a selected range of azimuths. The eight-hour UT range places the mean illumination direction towards the upper right in the image projection. Some exceptions to this UT range were made to gather Antarctic island data swaths for the composite. The restricted illumination range results in a more seamless representation of mountains and topography across the continent. Unfortunately, because of the timing of the Terra and Aqua satellite passes, the zone for which local noon is centered within the UT time range had a limited number of scenes available.

The selected range of UT times for image acquisition implies a broad range of local times across the continent. Over the northern West Antarctic coastline, the UT range occurs during local midnight. This means that VIS-NIR images must be acquired around the time of the austral summer solstice to provide an acceptable solar elevation ranging from 3° to 20°. Yet, at the opposite side of Antarctica in Enderby Land, a near-solstice image selection would result in a solar elevation of 35° to 45°. High solar elevation reduces the amount of topographic detail in the MODIS scenes. To moderate this, the chosen images were those taken late in the summer (late-January through February) for the regions near 45° E longitude and the entire continental coast, reducing the solar elevation by about 10° in this area. Mean solar elevation across the entire mosaic is 23.6° ± 7.7° (1 σ).

Processing Steps

Geolocation and Processing

Satellite image data swaths of Band 1 and Band 2 from MODIS Level 1B MOD02QKM (Terra) and MYD02QKM (Aqua) files, together with illumination and viewing angle data from Level 1A MOD03 and MYD03 files, were geolocated and resampled onto the projection grid using NSIDC's MS2GT software. The software interpolates the 1 km resolution latitude and longitude data from the Level 1A files to 250 m resolution and then resamples the Level 1B data to the grid using a forward elliptical weighted average (EWA) algorithm (Greene, et al., 1986).

Destriping of MODIS image data

The MS2GT algorithm was modified to remove striping artifacts incurred by the 40-detector whiskbroom scanner and the two mirror sides of the MODIS Band 1 and Band 2 sensor (Haran et al, 2002). These artifacts are a known problem with all Terra and Aqua MODIS Level 1B data at 250 m (MOD02QKM and MOY02QKM data). Inter-detector variations are as large as 1%, or 50 digital numbers (DN) in the 12-bit MODIS data, contributing to distinct horizontal striping in contrast-enhanced images. This primary striping pattern appears to be due to poor calibration among the 40 detectors that constitute a single scan of the 250 m data. A secondary variation in brightness appears between successive 40-line scans that is due to mirror side effects in the double-sided MODIS scan mirror. A third artifact appears as an every-fourth-pixel brightness shift in detectors 28 and 29 (a "stitching" artifact in appearance). These three artifacts limit the usefulness of MODIS over ice sheet interior images, because they induce brightness variations as large or larger than the shading due to subtle topography.

To correct the problems caused by artifacts, investigators conducted a Lambertian solar zenith angle normalization on the swath data for both bands. Telemetry noise and line drops in the MODIS scenes, having the appearance of "chads" in the projected images, were reset to zero (that is, treated as masked cloud areas). Though the mosaic composite is nearly cloud-free, some areas of thin cloud and small fog or cloud patches exist. Investigators are compiling a list of known cloud and other blemishes in the MOA images. Please notify NSIDC's User Services of any errors encountered while using these data by sending an email to nsidc@nsidc.org.

Two composite images were created from the geolocated Band 1 and Band 2 images: a high-pass filtered Band 1 image composite, which emphasizes the surface morphology, and a normalized-difference band-ratio image, which provides semi-quantative information about mean surface grain size over snow and ice surfaces.

Cloud Masking

The geolocated swath images were manually masked to remove clouds, cloud shadows, fog, blowing snow, and heavy surface frost. To do this, the images were compared to the RAMP AMM-1 SAR mosaic image (See RAMP Data.) and to initial versions of the MOA composite to identify cloud, fog, and blowing snow areas. Investigators conducted cloud and surface artifact masking using the 750 m images. They then applied the same mask to the 125 m scenes. Some small cloud and fog features at less than 750 m scale were not masked, and in a few persistently cloudy areas, some thin cloud and cloud-shadow contaminated images had to be used to cover the enitre continent with multiple scenes (for example, northeastern Ronne).

The mosaic composite is nearly perfectly cloud-cleared. Some areas of thin clouds, cirrus cloud shadows, and fog or low-lying small clouds are present in the northeastern Ronne Ice Shelf, which was persistently cloudy throughout the 2003-2004 austral summer. In numerous areas, there are small patchy clouds at spatial scales of less than 1 km as a result of using a 750 m-resolution image for the cloud clearing. More regionally, some areas are partly impacted by blowing snow, particularly in the East Antarctic. Investigators are compiling a list of known cloud cover and other blemishes in the MOA images on the distribution Web site. Please notify NSIDC's User Services of any errors encountered while using these data by sending an email to nsidc@nsidc.org.

Compositing the image swaths via Data Cumulation

Investigators created two composite images from the geolocated Band 1 and Band 2 images: a high-pass filtered Band 1 image composite, emphasizing the surface morphology; and a normalized-difference band-ratio image that provides quantitative information about mean surface grain size over snow and ice surfaces. The processing and assembly steps of these two composites is discussed separately.

High-Pass-Filtered Surface Feature Composite

After cloud masking, the geolocated, destriped Band 1 images were high-pass filtered to reduce non-Lambertian illumination and to reset the mean grayscale range to a common value for compositing. Investigators set the filter size for the images to 511 x 511 pixels, or the ground equivalent of 64 km x 64 km. The mean brightness of these filtered images was set to the same value (the integer 16000) to match gray levels for compositing.

Investigators then applied a weighting scheme to the masked and filtered image swaths creating a "weight image" for each gridded image that contains a scalar value for each non-masked pixel in the swaths. Weights for the pixels range from 0 to 50000. The weight was computed as the product of a fractional "scan weight", a fractional "mask weight", and a scale factor (50000). The scan weight was determined by the proximity to the nadir track, favoring near-nadir areas, and the mask weight by the proximity to an image edge or mask edge to "feather" the edges of the component images.

The scan weight image (wscan) is computed as follows:

Given:
Circular earth having radius: R = 6371 km
Circular orbit at altitude: A = 725 km
Maximum sensor zenith angle: seze_max = 66°
Sensor zenith angle image: seze

Computed:
Scan angle image: scan
scan = asin(R / (R + A) * sin(seze))
Maximum scan angle: scan_max
scan_max = asin(R / (R + A) * sin(seze_max)) = 55.1°
cos_scan_max_sq = cos(scan_max) * cos(scan_max)
Scan weight image: wscan
wscan = (cos(scan) * cos(scan) - cos_scan_max_sq) / (1 - cos_scan_max_sq)

The mask weight image (wmask) is computed as follows:

Given:
Land mask image: landmask
Cloud-masked Band 1 image: band1
Boxcar average smoothing width: mask_smooth_width = 43 pixels

Computed:
Mask image: mask
mask = band1
Where band1 > 0, mask = 1
Where landmask = 0, mask = 0
mask = smooth(mask, mask_smooth_width)
Mask weight image: wmask
wmask = (sqrt(mask) - sqrt(0.5)) / (1 - sqrt(0.5))

The weight image (weight) is computed as follows:

Computed:
Weight image: weight
weight = wscan * wmask * 50000

The high-pass filtered Band 1 and weight images, were then combined using "stacking" techniques (also called "image super-resolution" or "data cumulation"; see Scambos et al., 1999) to improve spatial and radiometric detail beyond the 250-meter and 12-bit characteristics of single Band 1 and Band 2 MODIS images. This improvement in part is due to reduction of random image noise by cancellation and improvement of relative radiometric resolution by combining repeated measurements of the surface (in the form of multiple digital images). The additional spatial resolution in the stacked image composite is a result of knowing the pixel center locations to a high precision (50 m), a precision that is smaller than the size of the pixel sample area (250 m).

Images were added to the mosaic, with weighting applied, by the following scheme:

Given a set of n high-pass filtered Band 1 images (B_i) and a corresponding set of n weight images (W_i), compute the composited band image (B_c), composited weight image (W_c), and count image (N_c).

B_c, W_c, and N_c are initially all zero, and N_i is set to 1 for each pixel.

For each i from 1 to n, where B_i is not zero, and W_i is not zero:
N_{c_old} = N_c
N_c = N_{c_old} + N_i
W_{c_0} = N_{c_old} * W_c / N_c
W_{c_1} = N_i * W_i / N_c
W_c = W_{c_0} + W_{c_1}
B_c = (W_{c_0} * B_c + W_{c_1} * B_i) / W_c

A set of n intermediate composites can themselves be composited into a single composite by setting the values of B_i, W_i, and N_i to the corresponding values of B_c, W_c, and N_c for each composite (in this case N_i is not set to 1), and then applying the given algorithm.

The image stacking allows multiple images to contribute to the representation of a single grid cell in the MOA composite. Image count ranges from 38 to 1; 98.53% of the imaged area is made up of 6 or more contributing images. Mean image count is 14.7. Areas of low image count are the northeastern and far northwestern Ronne Ice Shelf (which had persistent cloud cover throughout the summer of 2003-2004), the ice sheet region between the Executive Committee Range and Crary Mountains in West Antarctica, and the area between the Dome Fuji and Plateau Station camps in the East Antarctic Plateau (This is in the region of local noon for the UT time range; therefore, it has fewer near-nadir passes). Using the equations and models developed in Scambos et al., 1999 investigators infer that regions with 5 or more scenes contributing have resolutions of approximately 200 m or better ranging to a best resolution value of approximately 150 m (for images composed of 10 or more scenes). Mean weight of the image pixels for the MOA grid cells is 30322, ranging from 4800 to 49980. A large region of lower weights (5000 - 9000) is centered on the South Pole, because all images are (must be) significantly off-nadir in this area.

In general, image quality is best in regions of high count and high composited weight.

Optical Mean Summer Snow Grain Size

Two simple grain-size composite images were generated by applying a normalized difference algorithm to Band 1 and Band 2 of the composite scenes.

[Band1 - Band2]/[Band1 + Band 2]

This ratio takes advantage of the decreasing reflectivity of snow in the infrared range, creating an image that is sensitive to grain-size variations (Warren, 1982; Fily et al., 1997). To maintain a quantitative ratio, these images were not processed beyond geolocation, calibration, and destriping. Calibration is based on the information provided in the MOD02QKM and MYD02QKM files, as applied by the MS2GT software. The image is provided as a 16-bit image of relative grain size.

A composite image of optical surface grain-size was generated by applying a model-derived lookup table to images of normalized difference MODIS Band 1 (red light; 620 nm - 670 nm) and MODIS Band 2 (near-infrared; 841 nm - 876 nm) radiances and solar zenith angle:

G_i,j = (b1_i,j - b2_i,j) / (b1_i,j + b2_i,j)
S_i,j = f(G_i,j, Z_i,j)

G_i,j represents a normalized radiance ratio image, b1_i,j and b2_i,j are the calibrated band radiance values at grid location (i,j) for the respective MODIS bands, Z_i,j is the solar zenith angle image for each MODIS pixel in the scene, f(G_i,j, Z_i,j) is the lookup table, and S_i,j is the optical grain radius image. The final grain size image is composited from the 260 derived snow-grain images in a similar manner as the surface morphology composite. Calibration scale factors are provided in the MOD02QKM and MYDO2QKM data for the Terra and Aqua MODIS sensors, respectively.

Lookup table values for the grain size conversion were derived from runs of the Santa Barbara DISORT Atmospheric Radiative Transfer (SBDART) software (Ricchiazzi et al., 1998; http://www.crseo.ucsb.edu/esrg/sbdart/). These lookup tables were then applied to the 260 MOA scenes to produce a corresponding set of grain size images. The grain size images were then composited together using the same weighting and compositing scheme as that used for the surface morphology image, with the exception that (in addition to masked areas) values of 5 or 1105 µm were treated as missing data. (Values that led to unmodeled grain sizes, that is less than 10 µm or greater than 1100 µm, had the associated grain size value set to "marker" values of 5 µm and 1105 µm, respectively.) Count images for the grain size composite are slightly different from the surface morphology count images because of these additional missing values. In cases where all images gave out-of-range grain size results, the grid cell value was set to either 5 µm or 1105 µm. Additionally, the grain size images from the 260 MOA contributing scenes were used to create a grain size standard deviation image.

Runs of the SBDART software provided predictions of MODIS top of atmosphere (TOA) Band 1 and Band 2 radiance values for a series of snow grain sizes (10 µm to 1100 µm at 10 µm increments) at a series of solar illuminations in 0.1° increments of solar zenith angle, 0° to 89°. Investigators provided SBDART with snow optical properties, that is, BRDF spectral reflectance information from a recent snow reflectance model (Painter and Dozier, 2004), and created files (one for every 10 µm of grain size) that used the data to represent the reflecting substrate in the atmospheric radiative transfer model. Investigators held the sensor viewing angle at nadir for all SBDART evaluations. The weighting scheme favoring nadir-viewing conditions was appropriate to reduce off-nadir BRDF effects in the final optical snow grain size composite.

Validation data for the optical grain size measurement are shown in the following table. During three locally clear-sky days in October 2003, investigators collected 6 surface snow spectral measurements using a field spectrometer as part of the Antarctic Remote Ice Sensing Experiment (ARISE; Massom et al., 2006 in press). Investigators averaged 60 separate spectra of snow (spectral range 0.4 to 2.5 µm), normalized to 40 interspersed spectra of an Etalon calibration target. They inverted the spectral data to the optically equivalent grain size radius using conversions developed by Nolin and Dozier (2000) with 50 µm - 100 µm grain radius uncertainties. The same sites were imaged by 6 MODIS scenes, 5 from the Terra platform and 1 from Aqua. In situ grain size radius ranged from 66 µm, for snow immediately following a storm accumulation, to 170 µm, for wet snow in warmer conditions. Comparison of predicted grain sizes from the method used for these images and the in situ spectra generally show an underestimate of grain size from the satellite algorithm, ranging from 15 µm to 151 µm.

5. References and Related Publications

Bindschadler, R. A. and P. L. Vornberger. 1990. AVHRR Imagery reveals Antarctic ice dynamics. EOS 71(23), 741-742.

Bohlander, J., T. Scambos, T. Haran, M. Fahnestock. 2004. A New MODIS-based Mosaic of Antarctica: MOA. EOS, Transactions, American Geophysical Union 85(47). F452.

Bourdelles, B., and M. Fily. 1993. Snow grain-size determination from Landsat imagery over Terre Adélie, Antarctica. Annals of Glaciology 17, 86-92.

Cooper, A. P. R. 1994. A simple shape-from-shading algorithm applied to images of ice-covered terrain. IEEE Transactions on Geoscience and Remote Sensing 32(6), 1196–1198.

Dowdeswell, J. A., and McIntyre, N. F. 1987. The surface topography of large ice masses from Landsat imagery. Journal of Glaciology 33(133), 16–33.

Greene, N. and P. S. Heckbert. 1986. Creating Raster Omnimax Images from Multiple Perspective Views Using the Elliptical Weighted Average Filter. IEEE Computer Graphics and Applications 6(6). 21-27.

Fahnestock, M. A., R. Bindschadler, R. Kwok, and K. Jezek. 1993. Greenland Ice Sheet surface properties and ice dynamics from ERS-1 SAR imagery. Science 262, 1530–1534.

Ferrigno, J. G., J. L. Mullins, J. A. Stapleton, R. A. Bindschadler, T. A. Scambos, L. B. Bellisime, J. A. Bowell, and A. V. Acosta. 1994. Landsat TM image maps of the Shirase and Siple Coast ice streams, West Antarctica. Annals of Glaciology 20, 407-412.

Fily, M, B. Bourdelles, J. P. Dedieu, and C. Sergent. 1997. Comparison of in situ and Landsat Thematic Mapper derived snow grain characteristics in the Alps. Remote Sensing of Environment 59. 452-460.

Haran , T. M., M. A. Fahnestock, and T. A. Scambos. 2002. De-striping of MODIS optical bands for ice sheet mapping and topography. EOS, Transactions, American Geophysical Union 88(47). F317.

Jezek, K. 1999. Glaciological properties of the Antarctic ice sheet, from Radarsat-1 Synthetic Aperture Radar Imagery. Annals of Glaciology 29, 286–290.

Merson, R. H., 1989. An AVHRR mosaic image of Antarctica. International Journal of Remote Sensing 10, 669.

Nolin, A. W., and J. Dozier. 2000. A hyperspectral method fro remotely sensing the grain size of snow. Remote Sensing of Environment 74(2), 207-216.

Orheim, O. and B. Lucchitta. 1988. Numerical analysis of Landsat thematic mapper images of Antarctica: surface temperatures and physical properties. Annals of Glaciology 11, 109.

Orheim, O., and B. Lucchitta. 1990. Investigating climate change by digital analysis of blue ice extent on satellite images of Antarctica. Annals of Glaciology 14, 211-215.

Painter, T., and J. Dozier. 2004. Measurements of the hemispherical-directional reflectance of snow at fine spectral and angular resolution. Journal of Geophysical Research 109, D18115, doi:10.1029/2003JD004458.

Ricchiazzi, P., S. Yang, C. Gautier, and D. Sowle. 1998. SBDART: A research and teaching software tool for plane-parallel radiative transfer in the Earth’s atmosphere. Bulletin of American Meteorological Society 79(10), 2101-2114.

Scambos, T. A., and R. A. Bindschadler. 1991. Feature map of Ice Streams C, D, and E, West Antarctica. Antarctic Journal of the United States 26(5), 312-314.

Scambos, T., G. Kvaran, and M. Fahnestock. 1999. Improving AVHRR resolution through data cumulating for mapping polar ice sheets. Remote Sensing of Environment 69. 56-66.

Scambos, T. A., and M. A. Fahnestock. 1998. Improving digital elevation models over ice sheets using AVHRR-based photoclinometry. Journal of Glaciology 44, 97–103.

Scambos, T. A., Dutkiewitcz, M. J., Wilson, J. C., and R. A. Bindschadler. 1992. Application of image cross-correlation software to the measurement of glacier velocity using satellite image data. Remote Sensing of Environment 42, 177-186.

Scambos, T., T. Haran, M. Fahnestock, T. Painter, and J. Bohlander. 2007. MODIS-based Mosaic of Antarctica (MOA) data sets: continent-wide surface morphology and snow grain size. Remote Sensing of Environment 111(2): 242-257, doi:10.1016/j.rse.2006.12.020.

Seko, K., Furukawa, T., Nishio, F., and Watanabe, O. 1993. Undulating topography on the Antarctic ice sheet revealed by NOAA AVHRR images. Annals of Glaciology 17, 55–62.

Swithinbank, C., K. Brunk, and J. Sievers. 1988. A glaciological map of Filchner-Ronne Ice Shelf, Antarctica. Annals of Glaciology 11, 150–155.

USGS. 1991. Satellite Image Map of Antarctica, 1:5,000,000. Miscellaneous Map Investigation Series I-2284.

USGS. 1996. Satellite Image Map of Antarctica, 1:5,000,000. Miscellaneous Map Investigation Series I-2560.

Warren, S. 1982. Optical properties of snow. Reviews of Geophysics and Space Physics 20(1), 67-89.

Winther, Jan-G., M. N. Jespersen and G. E. Liston. 2001. Blue-ice areas in Antarctica derived from NOAA AVHRR satellite data. Journal of Glaciology 47(157), 325–334.

Wolfe, R. E., M. Nishihama, A. J. Fleig, J. A. Kuyper, D. P. Roy, J. C. Storey and F. S. Patt. 2002. Achieving sub-pixel geolocation accuracy in support of MODIS land science. Remote Sensing of the Environment 83 (1-2), 31-49.

6. Document Information

Acronyms and Abbreviations

The following acronyms and abbreviations are used in this document:

AMM	Antarctic Mapping Mission
ARISE	Antarctic Remote Ice Sensing Experiment
BRDF	Bi-directional reflectance distribution function
DEM	Digital Elevation Model
DISORT	Discrete Ordinate Radiative Transfer
DN	Digital number
EOS	Earth Observing System
EWA	Elliptical weighted average
FTP	File Transfer Protocol
JPEG	Joint Photographic Experts Group
MOA	Mosaic of Antarctica
MODIS	Moderate Resolution Imaging Spectroradiometer
Mbps	Megabytes per second
NASA	National Aeronautics and Space Administration
NIR	Near infrared
NSIDC	National Snow and Ice Data Center
RAMP	RADARSAT Antarctic Mapping Project
SAR	Synthetic Aperture Radar
SBDART	Santa Barbara DISORT Atmospheric Radiative Transfer
SCAR	Scientific Committee on Antarctic Research
TOA	Top of atmosphere
URL	Uniform Resource Locator
UT	Universal Time
VIS	Visible

Document Creation Date

February 2006

Document Revision Date

November 2008

Document URL

http://nsidc.org/data/docs/aged/nsidc0280_scambos/index.html