Close

Service Interruption

DMSP SSM/I Daily and Monthly Polar Gridded Sea Ice Concentrations

WARNING: NSIDC no longer supports or distributes this data set. This documentation is being archived and provided to those users who still have the NASA Team algorithm version of the DMSP SSM/I Daily and Monthly Polar Gridded Sea Ice Concentrations data. For the most current data and documentation, please see the Bootstrap algorithm version of the DMSP SSM/I Daily and Monthly Polar Gridded Sea Ice Concentrations guide document.

Summary

NSIDC SSM/I sea ice products in polar stereographic projection currently include DMSP-F8, F11 and F13 daily and monthly sea ice concentrations. Data, gridded at a resolution of 25 x 25 km, begin 25 June 1987.  Processing is ongoing. Two sets of SSM/I sea ice concentration grids have been formulated. The first data set was generated using the NASA Team algorithm and the second using the Bootstrap algorithm. The SSM/I-derived ice concentrations are daily total and monthly averaged ice fractions for both hemispheres. Multiyear ice fractions are provided by the NASA Team algorithm for the Northern Hemisphere only. Both the NASA Team and Bootstrap data are provided in Hierarchical Data Format (HDF), with accompanying Graphical Interchange Format (GIF) browse files.

Citation

To broaden awareness of our services, NSIDC requests that you acknowledge the use of data sets distributed by NSIDC. Please refer to the citation in this documentation 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 examples show how to cite the use of this data set in a publication. List the principal investigators, year of data set release, data set title, dates of the data you used, publishers (NSIDC), and digital media.

For the Bootstrap algorithm:
Comiso, J. 1990, updated current year. DMSP SSM/I daily and monthly polar gridded sea ice concentrations, [list dates used]. Edited by J. Maslanik and J. Stroeve. Boulder, Colorado USA: National Snow and Ice Data Center. Digital media.

For the NASA Team algorithm:
Cavalieri, D., P. Gloerson, and J. Zwally. 1990, updated current year. DMSP SSM/I daily and monthly polar gridded sea ice concentrations, [list dates used]. Edited by J. Maslanik and J. Stroeve. Boulder, Colorado USA: National Snow and Ice Data Center. Digital media.

Table of Contents

  1. Data Set Overview
  2. Applications
  3. Theory of Measurements
  4. Acquisition Materials and Methods
  5. Preparation and Description
  6. Notes and Plans
  7. Products and Access (Software)
  8. References
  9. Document Information

1.Data Set Overview

Data Set Identification

DMSP SSM/I Daily and Monthly Polar Gridded Sea Ice Concentrations

Data Set Introduction

Twenty years ago, changes in sea ice from season to season and year to year were not well known. As Robert Massom writes in Satellite Remote Sensing of Polar Regions (1991), "Before the advent of remote sensing from space, estimates of sea ice extent and associated meteorological and oceanic variables came from a combination of ship, island station and aircraft observations." Studies were "conducted on an opportunity basis and often concentrated in areas of logistic convenience during the summer navigation season...(yielding) data sets inherently biased and therefore of limited scientific value."

Fortunately, the situation has improved. Since 1972, three generations of space borne passive microwave imagers have been launched by the United States, including the Electrically Scanning Microwave Radiometer (ESMR), the Scanning Multichannel Microwave Radiometer (SMMR), and a series of the Defense Meteorological Satellite Program's (DMSP) Special Sensor Microwave/Imagers (SSM/I).

Today, one of the most important parameters provided by passive microwave data in the polar regions is sea ice concentration. Data acquisition is possible because passive microwave wavelengths are relatively unaffected by the frequent and extensive cloud cover that is prevalent in these regions. Sea ice concentration maps are used to track ice edges, estimate ice extent, ice type, actual ice area and the amount of open water within the ice pack. The latter is in turn needed to monitor occurrence, impact, and persistence of polynyas, to calculate heat and salinity fluxes between the ocean and the atmosphere in the polar regions, in addition to many other applications. Global data, immediately practical for use in shipping and petroleum development activities, have broader implications from the standpoint of adding to the meteorological foundations used in understanding and modeling climate change.

Several techniques have been developed to obtain ice concentration from passive microwave data (Svendson et al. 1983; Cavalieri et al. 1984; Swift et al. 1985; Comiso 1986; Comiso and Sullivan 1986; Smith 1996). A review of these techniques and a comparison of some ice concentration results are presented by Steffen et al. (1992) and Smith (1996). Although the various techniques are consistent in finding the location of the ice edge, there are differences in the derived fraction of open water within the ice pack, partly due to the use of different sets of SSM/I channels. Comparative studies with limited Landsat images did not yield conclusive evaluation of the merit of each technique. However, current work with synthetic aperture radar (SAR) data may provide new insights about the discrepancies. These discrepancies can be more than 20 percent as in the Bellingshausen/Amundsen Seas, where effects due to flooding, roughness and snow may be considerable (Comiso 1991). It is also important to mention that since these comparison studies, there has been an adjustment in the Bootstrap algorithm tie points to match ice concentrations derived from SAR during the melt period. This adjustment, as well as a comparison with ice concentrations derived with the NASA Team algorithm, can be found in Comiso et al. (1997).

This guide accompanies daily north and south polar sea ice concentration grids, and monthly averaged sea ice concentrations.

Related Data Sets

Investigators

Investigators' Names

Roger Barry, Jim Maslanik, and Julienne Stroeve
National Snow and Ice Data Center (NSIDC)
Cooperative Institute for Research in Environmental Sciences (CIRES)
University of Colorado, Boulder, CO, U.S.

Title of Investigation

Cryospheric Data Management System

Contact Information

NSIDC User Services
National Snow and Ice Data Center
CIRES, 449 UCB
University of Colorado
Boulder, CO 80309-0449  USA
phone: +1 303.492.6199
fax: +1 303.492.2468
form: Contact NSIDC User Services
e-mail: nsidc@nsidc.org

2. Applications

Sea ice concentration data are used primarily in the following applications:

NSIDC provides a suite of value-added products to aid in investigations of the variability and trends of sea ice cover. These products provide users with information about sea ice extent, total ice covered area, ice persistence, monthly climatologies of sea ice concentrations, and ocean masks.

3.Theory of Measurements

Sea ice concentrations are calculated using brightness temperatures mapped onto a polar stereographic grid. Two sets of sea ice concentration grids are derived from brightness temperatures for the F8, F11 and F13 instruments. The first set was derived using the NASA Team algorithm. The second set uses the Bootstrap algorithm.

The Bootstrap technique, as described by Comiso (1986) and Comiso and Sullivan (1986), uses basic radiative transfer equations and takes advantage of unique multichannel distributions of sea ice emissivity. To derive sea ice concentration using the Bootstrap algorithm, only two microwave channels are needed but additional channels may be required to mask the open ocean.

The NASA Team Algorithm uses three SSM/I channels to calculate sea ice concentration. These are the 19.4-GHz horizontally (H) and vertically (V) polarized channels and the vertically polarized 37.0-GHz channel. This algorithm is functionally the same as the Nimbus-7 SMMR algorithm described by Cavalieri et al. (1984), and Gloersen and Cavalieri (1986).

4.Acquisition Materials and Methods

Acquisition Equipment

Sensor/Instrument Description

The SSM/I is a seven channel, four frequency, orthogonally polarized, passive microwave radiometric system. The instrument measures combined atmosphere and surface radiances at 19.3, 22.2, 37.0 and 85.5 GHz. For more information please see the DMSP SSM/I Daily Polar Gridded Brightness Temperatures data set document. Also see the SSM/I instrument description for more details.

Collection Environment

Satellite

Source/Platform

DMSP-F8, DMSP-F11, and DMSP-F13

Source/Platform Mission Objectives

The first SSM/I was launched June 19, 1987 as a joint Navy/Air Force operational program to obtain synoptic maps of critical atmospheric, oceanographic, and selected land parameters on a global scale. Key geophysical variables observable by the SSM/I were rain storms over land, ocean surface wind speed, sea ice concentration, and ice/water boundaries.

Key Variables

The SSM/I instrument measures dual polarized radiances at 19.3, 37.0 and 85.5 GHz, and vertically polarized radiances at 22.2 GHz.

Principles of Operation

Comparison of Orbital Parameters, DMSP-F8, DMSP-F11 and DMSP-F13

The DMSP Block 5D-2 F8, F11 and F13 spacecraft flew in a near polar sun-synchronous orbit. Launched on 18 June 1987, the F8 satellite accomplished 14.1 revolutions per day, with the subsatellite ground track repeating approximately every 16 days. F8 coverage ended 31 December 1991. The F11 was launched on 28 November 1991 with coverage ending 30 September 1995. Processing continues with the launch of the F13, 5 May 1995.

A comparison of instruments, and of the differences in orbital parameters (Abdalati et al. 1995) between the F8 and F11 using overlapping data indicated a high degree of correlation (greater than 0.98) between the F8 and F11 data sets. Small variations were attributed to the different orbital characteristics of the two satellites, especially to the differences in data collection times. When comparing F11 to F13, in terms of hemispheric averages of mean sea ice concentration, the biases introduced by the switch from F11 to F13 are slight and are not statistically significant; however, in some regions relatively large and significant differences are seen. In addition, differences in sea ice extent and total ice-covered area between the two platforms were found to be statistically significant (see NSIDC Special Publication 5 for more information).

Parameter DMSP F8 DMSP F11 DMSP F13
Nominal Altitude 860 km 830 km 850 km
Inclination Angle 98.8 degrees 98.8 degrees 98.8 degrees
Orbital Period 102 minutes 101 minutes 102 minutes
Ascending Node Equatorial Crossing (local time) approximately 6:00 a.m. approximately 5:00 p.m. approximately 5:43 p.m.

Based on the analysis, a set of corrections have been applied to F11 data to maximize consistency between the two data sets.

Sensor/Instrument Measurement Geometry

(adapted from Hollinger and Lo 1983, p. 1-3)

The SSM/I is a seven channel, four frequency, linearly polarized, passive microwave radiometric system. The instrument measures atmospheric/ocean surface antenna temperatures at 19.3, 22.2, 37.0 and 85.5 GigaHertz (GHz).

The instrument consists of a 24 x 26 inch offset parabolic reflector fed by a corrugated, broad-band, seven-port horn antenna. The reflector and feed are mounted on a drum which contains the radiometers, digital data subsystem, mechanical scanning subsystem, and power subsystem.

The reflector-feed drum assembly is rotated about the axis of the drum by a coaxially mounted bearing and power transfer assembly (BAPTA). All data, commands, timing and telemetry signals, and power pass through the BAPTA on slip ring connectors to the rotating assembly.

The SSM/I rotates at a uniform rate making one revolution in 1.9 seconds, during which time the satellite advances 12.5 km. The antenna beams are at an angle of 45 degrees to the BAPTA rotational axis, which is normal to the earth's surface. Thus, as the antenna rotates, the beams define the surface of a cone, and, from the orbital altitude of 833 km, make an angle of incidence of 53.1 degrees at the earth's surface.

The scene is viewed over a scan angle of 102.4 degrees centered on the ground track aft of the satellite, resulting in a scene swath width of 1394 km. The radiometer outputs are sampled differently on alternate scans. During the scene portion of the scans (Type A) the five channels are each sampled over 64 equal 1.6 degree intervals.

Sampling, to 12-bit precision, is accomplished by the "integrate, hold, and dump" method, with an integration period of 7.95 milliseconds for the five channels. Alternate 0.8 degree intervals are centered on the mid-points of the 1.6 degree intervals.

The five channels are sampled on an approximately 25 km grid along the scan and along the track.

Manufacturer of Instrument

Hughes Aircraft Company

Calibration

A small mirror and a hot reference absorber are mounted on the BAPTA and do not rotate with the drum assembly. They are positioned off-axis such that they pass between the feed horn and the parabolic reflector, occulting the feed once each scan. The mirror reflects cold sky radiation into the feed, thus serving, along with the hot reference absorber, as calibration references for the SSM/I.

This scheme provides an overall absolute calibration which includes the feed horn. Corrections for spillover and antenna pattern effects from the parabolic reflector are incorporated in the data processing algorithms.

Data Acquisition Methods

NSIDC produced brightness temperatures derived from the F8, F11 and F13 instruments to create the sea ice concentration data sets.

Observations

Data Notes

See Section 5, Errors, below for information about bad data values and file naming conventions.

Beginning with January 2000, processing of brightness temperature data was modified to include two additional quality control steps. The first performs a statistical analysis on the brightness temperature data to look for possible calibration errors. The second was an along-scan adjustment which corrects for interference by the cold-space reflector at scans of 100 or greater, and the difference between antenna temperature observations and the Wentz radiative transfer model. Corrections can be as large as 1 Kelvin. See Stroeve (1998) for more details. Brightness temperature data that included these additional quality control steps are only available for dates after January 2000.

5.Preparation and Description

Data Description

Spatial Characteristics

Spatial Coverage

Instrument coverage is global except for circular sectors centered over the pole, approximately 310 km in radius, located poleward of 87 degrees North and 87 degrees South, which are never measured due to orbit inclination. Data set coverage is of the polar regions and is defined by the spatial coverage map described below. The measurement footprint size (effective field of view) is as follows:

                19.3 GHz  70x45 km
                22.2 GHz  60x40 km
                37.0 GHz  38x30 km

Scan Geometry
SSM/I A/B Scan Geometry: SSM/I A/B Scan Geometry: Swath data consist of A/B scan
pairs. Each pair includes 256 scene stations (numbered). Scene station numbers (parameter
position numbers) are indicated. Large circles signify all channels, small circles stand for 85
GHz channels. Brackets indicate scene stations lost due to antenna pattern correction.

Spatial Coverage Map

SSM/I Polar Spatial Coverage Maps, Northern Hemisphere

Northern Hemisphere

 SSM/I Polar Spatial Coverage Maps, Southern Hemisphere

Southern Hemisphere

Spatial Resolution

The sea ice concentrations are provided at a resolution of 25 km.

Projection

The grids are in a polar stereographic projection. The polar stereographic projection used specifies a projection plane (i.e., the grid) tangent to the earth at 70 degrees. The planar grid is designed so that the grid cells at 70 degrees latitude are 25 km by 25 km. See the data files pixlarea.n and pixlarea.s provided with the tools (Section 7) for information on the true earth-area coverage of all grid cells. For more information on this topic please refer to Pearson (1990) and Snyder (1987).

The polar stereographic projection often assumes that the plane (i.e., the grid) is tangent to the Earth at the pole. Thus, there is a one-to-one mapping between the Earth's surface and grid (i.e., no distortion) at the pole. Distortion in the grid increases as the latitude decreases because more of the Earth's surface falls into any given grid cell, which can be quite significant at the edge of the northern SSM/I grid where distortion reaches 31 percent. For the South Pole, the SSM/I grid has a maximum distortion of 22 percent. To minimize the distortion, the projection is true at 70 degrees rather than the poles. This increases the distortion at the poles by three percent and decreases the distortion at the grid boundaries by the same amount. The latitude of 70 degrees was selected so that little or no distortion would occur in the marginal ice zone. Another result of this assumption is that fewer grid cells will be required as the Earth's surface is more accurately represented. This saves about 100 megabytes per year in data storage.

The polar stereographic formulae for converting between latitude/longitude and X-Y grid coordinates have been taken from Map Projections Used by the U.S. Geological Survey (Snyder 1982). Several different ellipsoids were compared to the Hughes ellipsoid and in each case, differences were less than 1 km over the SSM/I grids. However, differences of up to 9 km were found if a sphere rather than an ellipsoid was used. Thus, it is an explicit requirement that an ellipsoid be used in processing the data.

There are a variety of ellipsoids, for instance, SEASAT has its own ellipsoid. An ellipsoid is defined by equatorial radius and eccentricity. The ellipsoid used in the Hughes (1980, p.3-266) software assumes a radius of 3443.992 nautical miles or 6378.273 kilometers (km) and an eccentricity (e) of 0.081816153 (or e**2 = 0.006693883). The origin of this ellipsoid is not specified in the available Hughes documentation. However, the Hughes ellipsoid is similar to other ellipsoids quoted in the literature (Snyder 1982). To properly convert these coordinates to a polar stereographic grid (the projection of choice), the conversion should assume the Hughes ellipsoid.

Grid Description

Grid dimensions (column,row)

85.5 GHz All other
channels
North (608,896) (304,448)
South (632,664) (316,332)

Grid Coverage:

The origin of each x, y grid is the pole. The grids' approximate outer boundaries are defined in the following table. Values refer to the outside corner of corner pixels, and the outside edge of midpoint pixels. Apply values to the polar grids reading clockwise from upper left. Interim rows define boundary midpoints.

X(km) Y(km) Latitude (deg) Longitude (deg)
 North Polar: -3850 5850 30.98 168.35 corner
  0 5850 39.43 135.00 midpoint
  3750 5850 31.37 102.34 corner
  3750 0 56.35 45.00 midpoint
  3750 -5350 34.35 350.03 corner
  0 -5350 43.28 315.00 midpoint
  -3850 -5350 33.92 279.26 corner
  -3850 0 55.50  225.00 midpoint

 
X(km) Y(km) Latitude (deg) Longitude (deg)
South Polar -3950 4350 -39.23 317.76 corner
  0 4350 -51.32  0.00  midpoint
  3950 4350 -39.23 42.24 corner
  3950 0 -54.66 90.00 midpoint
  3950 -3950 -41.45 135.00 corner
  0 -3950 -54.66 180.00 midpoint
  -3950 -3950 -41.45 225.00 corner
  -3950 0 -54.66 270.00 midpoint

Temporal Characteristics

Temporal Coverage

The F8 data stream began 9 July 1987 through 18 December 1991; F11 began 3 December 1991 through 30 September 1995; and the F13 data stream began 5 May 1995, processing continues.

Note to users of SSM/I polar stereographic data for 1994 through April 1995:

Substantial amounts of swath data over Alaska and the Canadian Prairies are missing beginning early in 1994 until May 1995. During this period the data tape recorder on the DMSP-F11 failed. As a result, it was necessary to download data to ground stations more frequently than usual. Data download and acquisition could not occur simultaneously, consequently data gaps exist in the SSM/I data for Alaska and the Canadian Prairies from early 1994 until data were available.

A complete summary of missing dates is available from the DMSP SSM/I Daily Polar Gridded Brightness Temperatures documentation.

Temporal Resolution

Daily and monthly averaged sea ice concentration grids.

Data Characteristics

Parameter/Variable

The data set consists of sea ice concentration derived from gridded brightness temperatures. Sea ice concentrations range from 0 to 100 percent.

Variable Description/Definition

Sea Ice: any ice found at sea which has originated from freezing of sea water. Sea Ice Concentration: fraction of a given area covered by sea ice irrespective of ice type; the ratio describing the areal density of ice in a given area.

Unit of Measurement

Sea ice concentrations are measured in percents or fractions of the pixel area covered by sea ice.

Data Source

DMSP-F8, F11 and F13 SSM/I

Data Range

Sea ice concentration data are stored in unsigned 1-byte arrays representing sea ice concentrations ranging from 0 to 100 percent. A stored value of 168 stands for land, and a stored value of 157 indicates missing data.

Monthly averaged sea ice concentration data files also contain sea ice concentration in percent ranging from 0 percent to 100 percent. A value of -88 (or 168 for unsigned byte data types) indicates a land pixel, and a value of -99 (or 157 for unsigned byte data types) indicates missing data.

Sample Data Record

SSM/I sea ice concentration grids produced by either algorithm are displayed as raster images. Each pixel contains eight bits. Each image for the north polar region contains 448 records (lines) with each line containing 304 pixels (304x by 448y). South polar images contain 332 records (lines) with each line containing 316 pixels (316x by 332y).

The same grid orientation is used for all SSM/I polar stereographic products. That is, the first data value in Northern Hemisphere files corresponds to 30.98 degrees latitude, 168.35 degrees longitude; and the first data value in Southern Hemisphere files corresponds to -39.23 degrees latitude, 317.76 degrees longitude. See page Grid Coverage above for grid coordinate information.

For each product (grid type) there are corresponding land and coastline masks in raster format. Each grid cell contains a flag for the cell type, either 1 or 0. In the land mask, 1 = non-water, 0 = water. In the coastline mask, 1 = coastline, 0 = all land or all water. The mask values are 8-bit bytes. See the Software section of this document for further details on the masks.

Browse images of daily and monthly sea ice concentration are also available in GIF format.

Data Organization

Data Granularity

A granule of data is the smallest aggregation of data that is independently managed (i.e., described, inventoried, retrievable). Granules may be managed as logical granules and/or physical granules.

 For the SSM/I data set, a data granule consists of the grids that are compiled for the daily and monthly averaged data.

Data Format

Daily and monthly SSM/I F8, F11 and F13 sea ice concentrations are provided in HDF format. Data and monthly browse images (in GIF format) are available via ftp.

Also see Importing SSM/I Daily and Monthly Sea Ice Concentration Data into ArcInfo.

File Naming Convention

Sea ice concentration data from the ftp site are compressed with the UNIX 'tar' utility and contain daily averaged data for the day, hemisphere and algorithm indicated. File naming conventions differ according to temporal resolution. The following table summarizes file naming conventions using the F13 platform as an example:

Compressed Uncompressed
Daily SSMI-F13-vvvyyyymmIHA.tar.Z SSMI-F13-vvvyyyymmdd.IHA
Monthly f13-yyyy-iha-pp.tar SSMI-F13-vvvyyyymm.iha.pp

where:

Data Manipulations

Derivation Techniques and Algorithms

NSIDC distributes two sets of daily and monthly sea ice concentration grids. The first set is generated using the NASA Team algorithm developed by Don Cavalieri, and the second is generated using the Bootstrap algorithm developed by Joey Comiso.

Data Processing Sequence

Daily sea ice concentrations are derived from brightness temperatures. For more details please see NSIDC's DMSP SSM/I Daily Polar Gridded Brightness Temperatures.

SSM/I daily averaged sea ice concentration grids for the Northern and Southern Hemispheres were generated using the NASA Team algorithm (Gloersen and Cavalieri 1986, Cavalieri 1996); and the Bootstrap algorithm (Comiso 1986, Comiso and Sullivan 1986).

It is important to note that in deriving the ice concentrations from the Bootstrap and NASA Team algorithms, different adjustments were made to the brightness temperatures before they were input into the ice algorithms. The brightness temperatures were adjusted to improve the consistency in ice concentrations between the different SSM/I platforms. However, it was found that different adjustments were needed for the different algorithms.

For the Bootstrap algorithm, we found that the brightness temperature adjustments as recommended by Wentz (1993) improved the consistency in total ice fraction, ice extent and ice-covered area between the successive SSM/I sensors. Note however, that even though the overall consistency in Bootstrap-derived sea ice products improved between the sensors, large and significant regional differences may remain as was found in the NASA Team-derived sea ice fractions (Stroeve et al. 1998).

Using the NASA Team algorithm, it was found that the Wentz adjustments did not improve the overall consistency in the sea ice concentrations. However, prior to this study, NSIDC decided to incorporate the adjustments from Abdalati et al. (1995) in the processing of the F11 sea ice concentrations. Thus, when making the comparisons in ice fractions between F11 and F13, the F11 sea ice data has been processed using the brightness temperature adjustments as recommended by Abdalati et al. (1995).

Beginning with January 2000, processing of brightness temperature data was modified to include two additional quality control steps. The first performs a statistical analysis on the brightness temperature data to look for possible calibration errors. The second was an along-scan adjustment which corrects for interference by the cold-space reflector at scans of 100 or greater, and the difference between antenna temperature observations and the Wentz radiative transfer model. Corrections can be as large as 1 Kelvin. See Stroeve (1998) for more details.

Comparisons between F8, F11 and F13 are summarized in the tables below. In making the comparisons, an enlarged land mask was used to eliminate the large differences found along the coastlines as a result of geolocation uncertainties and to false retrievals of ice cover in open-ocean coastal pixels due to mixed-pixel effects from adjoining land ("land contamination").

Summaries using the Bootstrap Algorithm

Comparisons for F8 and F11

Note, overlap period is from 3 December to 18 December 1991.

Comparison of the 16-day mean ice concentrations between F8, F11 and F11 adjusted via Wentz's brightness temperature adjustments. Given are mean F8 and F11 ice fractions, mean differences (F11 minus F8), standard deviation of the differences, correlation coefficient and the root mean square (rms) of the differences.

                Mean     Mean     Standard     Correlation    rms
                Ice   Difference Deviation    Coefficient    error
                (%)       (%)       (%)            (r)        (%)   
                
Northern Hemisphere
F8              24.36     
F11             24.61      0.25     1.29          0.999       1.29 
F11(wentz)      24.48      0.11     1.21          0.999       1.21

Southern Hemisphere
F8              16.15   
F11             16.18      0.03     1.25          0.999       1.25
F11(wentz)      15.92     -0.23     1.30          0.999       1.28

Comparison of the 16-day mean ice extent and ice-covered area between F8, F11 and F11 adjusted via Wentz's brightness temperature adjustments. Percent differences are given in parentheses. Given are the mean F8 and F11 ice extents and ice-covered areas, mean differences (F11 minus F8) and standard deviation of the differences.

Ice Extent:

                Mean            Mean Difference         Std. Dev
               x10**6 km**2          (F11-F8)           x10**6 km**2

Northern Hemisphere
F8              9.31
F11             9.48             0.17(1.91%)           0.16
F11(wentz)      9.35             0.04(0.42%)           0.10

Southern Hemisphere
F8              9.90
F11            10.03             0.13(1.25%)           0.05
F11(wentz)      9.91             0.01(0.09%)           0.04

Ice-Covered Area:

               Mean            Mean Difference)         Std. Dev
             x10**6 km**2      (F11-F8)                 x10**6 km**2

Northern Hemisphere
F8              8.45
F11             8.51             0.06(0.68%)            0.04
F11(wentz)      8.48             0.03(0.34%)            0.03

Southern Hemisphere
F8              7.06
F11             7.18             0.12(1.62%)            0.04
F11(wentz)      7.07             0.01(0.05%)            0.03

F11/F13 intercomparison

Note, overlap period is from 3 May through 30 September
Comparison of the 139-day mean ice concentrations between F11, F13 and F13 adjusted via Wentz's brightness temperature adjustments. Given are the mean F11 and F13 ice fractions, mean differences (F13 minus F11), standard deviation of the differences, correlation coefficient and the rms of the differences.

                 Mean     Mean    Standard     Correlation    rms
                 Ice   Difference Deviation    Coefficient    error
                 (%)       (%)      (%)            (r)        (%) 

Northern Hemisphere
F11             16.67
F13             16.85      0.17     0.84          0.999       0.79
F13(Wentz)      16.90      0.23     0.81          0.999       0.80

Southern Hemisphere
F11             25.34
F13             25.77      0.43     0.84          0.999       0.72
F13(wentz)      25.75      0.41     0.84          0.999       0.70

Ice Extent:

                Mean            Mean Difference       Std. Dev
               x10**6 km**2     (F11-F8)              x10**6 km**2

Northern Hemisphere
F11             6.95
F13             7.04             0.09(1.36%)           0.07
F13(wentz)      7.04             0.09(1.37%)           0.05

Southern Hemisphere
F11            13.57
F13            13.71             0.14(1.02%)           0.06
F13(wentz)     13.66             0.10(0.70%)           0.06

Ice-Covered Area:

                Mean            Mean Difference       Std. Dev
               x10**6 km**2     (F11-F8)              x10**6 km**2
			   
Northern Hemisphere
F11             5.71
F13             5.79            0.08(1.47%)           0.07
F13(wentz)      5.77            0.06(1.11%)           0.05

Southern Hemisphere
F11            11.46
F13            11.65            0.19(1.63%)           0.06
F13(wentz)     11.64            0.17(1.53%)           0.06

Summaries using NASA Team algorithm:

Comparisons for F8 and F11

Comparison of the 16-day mean ice concentrations between F8, F11 and F11 adjusted via Wentz's and Abdalati's brightness temperature adjustments. Given are the mean F8 and F11 ice fractions, mean differences (F11 minus F8), standard deviation of the differences, correlation coefficient and the root mean square of the differences.

                 Mean     Mean    Standard     Correlation    rms
                 Ice   Difference Deviation    Coefficient    error
                 (%)       (%)      (%)            (r)        (%) 

Northern Hemisphere
F8              24.10
F11             24.29      0.19      1.87          0.999       1.86
F11(wentz)      24.38      0.28      1.86          0.999       1.84 
F11(abdalati)   24.33      0.23      1.85          0.998       1.85

Southern Hemisphere
F8              15.57
F11             15.44     -0.13      1.25          0.999       1.25
F11(wentz)      15.54     -0.02      1.29          0.999       1.27
F11(abdalati)   15.53     -0.04      1.29          0.999       1.27

Comparison of the 16-day mean ice extent and ice-covered area between F8, F11 and F11 adjusted via Wentz's and Abdalati's brightness temperature adjustments. Percent differences are given in parenthesis. Given are the mean F8 and F11 ice extents and ice-covered areas, the mean differences (F11 minus F8) and the standard deviation of the differences.

Ice Extent:

                Mean            Mean Difference         Std. Dev
             x10**6 km**2       (F11-F8)                x10**6 km**2
                                 
Northern Hemisphere
F8              9.619
F11             9.745            0.13(1.31%)           0.10
F11(wentz)      9.657            0.04(0.38%)           0.07
F11(abdalati)   9.672            0.05(0.55%)           0.08

Southern Hemisphere
F8             10.13
F11            10.17             0.04( 0.38%)           0.05
F11(wentz)     10.13            -0.00(-0.01%)           0.05
F11(abdalati)  10.14             0.01( 0.08%)           0.05

Ice-Covered Area

                Mean            Mean Difference         Std. Dev
             	x10**6 km**2    (F11-F8)                x10**6 km**2

Northern Hemisphere

F8              8.30
F11             8.32             0.02(0.25%)            0.04
F11(wentz)      8.38             0.08(0.97%)            0.03
F11(abdalati)   8.35             0.05(0.61%)            0.03

Southern Hemisphere
F8              6.80
F11             6.83             0.03(0.51%)           0.05
F11(wentz)      6.88             0.08(1.24%)           0.05
F11(abdalati)   6.87             0.07(1.11%)           0.07

F11/F13 intercomparison

Note, in these comparisons, the F11 sea ice concentrations are the ones that NSIDC currently archives, namely the see ice concentrations derived via the Abdalati (1995) brightness temperature adjustments.

Comparison of the 139-day mean ice concentrations between F11, F13 and F13 adjusted via Wentz's brightness temperature adjustments. Given are the mean F11 and F13 ice fractions, mean differences (F13 minus F11), standard deviation of the differences, correlation coefficient and the root mean square of the differences.

                 Mean     Mean    Standard     Correlation    rms
                 Ice   Difference Deviation    Coefficient    error
                 (%)       (%)      (%)            (r)        (%) 

Northern Hemisphere

F11             16.30
F13             16.38      0.08     0.75         0.999         0.74
F13(wentz)      16.35      0.24     0.80         0.999         0.78


Southern Hemisphere
F11             23.97
F13             23.98      0.01     0.67         0.999         0.66
F13(wentz)      24.23      0.26     0.78         0.999         0.76

Comparison of the 16-day mean ice extent and ice-covered area between F8, F11 and F11 adjusted via Wentz's and Abdalati's brightness temperature adjustments. Percent differences are given in parenthesis. Given are the mean F8 and F11 ice extents and ice-covered areas, the mean differences (F11 minus F8) and the standard deviation of the differences.

Ice Extent:

                Mean            Mean Difference         Std. Dev
             	x10**6 km**2    (F11-F8)                x10**6 km**2

Northern Hemisphere
F11             6.89
F13             6.95            0.06(0.97%)           0.03
F13(wentz)      6.96            0.07(1.12%)           0.03

Southern Hemisphere
F11            13.50
F13            13.57             0.07( 0.48%)           0.08
F13(wentz)     13.60             0.10( 0.74%)           0.09

Ice-Covered Area:

                Mean            Mean Difference         Std. Dev
             	x10**6 km**2    (F11-F8)                x10**6 km**2

Northern Hemisphere

F11             5.46
F13             5.46             0.00(0.08%)            0.03
F13(wentz)      5.49             0.03(0.53%)            0.03

Southern Hemisphere
F11            10.82
F13            10.83             0.01(0.04%)           0.05
F13(wentz)     10.87             0.05(0.46%)           0.06

Errors

In summer 1999, NSIDC was alerted to errors in latitude, longitude and pixel area files supplied with SSM/I polar stereographic gridded data. Please see the error notice explaining the steps taken to correct the problem.

Missing Pixels:
The links below provide textual and graphical summaries of sea ice data missing from the grids of this data set. This information allows users computing various statistics from the data such as ice extent, to quickly determine the viability of including a particular day's data in their calculations. For each daily image, NSIDC sums the total number of missing pixels for the entire image to determine the percentage of missing pixels.

F8 Northern
Hemisphere		 F8 Southern
Hemisphere
F8 N. Hemisphere Missing Pixels F8 S. Hemisphere Missing Pixels


F11 Northern
Hemisphere		 F11 Southern
Hemisphere
F11 N. Hemisphere Missing Pixels F11 S. Hemisphere Missing Pixels


F13 Northern
Hemisphere		 F13 Southern
Hemisphere
F13 N. Hemisphere Missing Pixels F13 S. Hemisphere Missing Pixels

Sources of Error

See also:

Geolocation Errors
Geolocation is a continuing problem for users of the SSM/I passive microwave data. Shortly after the SSM/I TDRs were released by FNMOC, independent investigations at the University of Massachusetts and at NSIDC determined that geolocation for the sensor was inaccurate. In addition, while processing the 1988 SSM/I data, RSS found a number of problems with the spacecraft and Earth locations computed at FNMOC, causing errors in excess of 20 to 30 km being routinely observed in the SSM/I data. Latitudes and longitudes on the DEF tapes produced by FNMOC were found to be in error due to the following problems (see Wentz's Users Manual for more detail).

There are some algorithm errors in the FNMOC data processing software (FNMOC and its contractors are continuing to improve their operational software).

The satellite ephemeris is sometimes incorrect due to spacecraft tracking errors and orbit prediction errors. This problem is particularly severe during periods of increased solar activity.

The boresite nadir angle and the alignment of the SSM/I instrument relative to the spacecraft are slightly misspecified.

In response to the problems identified, RSS developed their own routine for computing the latitude and longitudes rather than using the DEF values. The input into their geolocation routine is the satellite ephemeris for a 7-day period centered on the orbit being processed. The ephemeris is first subjected to quality control and then smoothed to remove any noise. The smoothed ephemeris is then used to compute the SSM/I cell latitudes and longitudes. This algorithm is believed to improve the geolocation accuracy to approximately 5 km (Wentz, personal communication).

For data collected prior to 1989, a correction algorithm developed by researchers at the University of Massachusetts' Department of Electrical and Computer Engineering was used that results in location accuracies of approximately eight kilometers both along and across the scan (Goodberlet 1990). This algorithm assumes all geolocation errors are a result of the pitch, yaw, and roll of the satellite. The three altitude angles were found to have latitude and time dependencies.

Table of coefficients used for data collected before May 15, 1988

  *angle = C(l) + C(2)*JDAY + C(3)*abs(LAT) + (C(4)*(JDAY**2))/10000
            where: 
         angle = the pitch, yaw or roll correction angle in degrees
         JDAY = number of days since the beginning of 1987
         LAT = scene latitude in degrees

Although the correction algorithm appears suitable for most of the pre-1989 data, the algorithm does not bring the mislocation to within the 8 kilometer tolerance in later data. The geolocation correction seems to have a periodicity of about one year.

Researchers at Remote Sensing Systems, Inc. attributed the larger errors in the 1989 data to the following circumstances:

  1. An increase in solar activity during early 1989 caused up to 50 km errors in the orbit predictions;

  2. FNMOC implemented irregularities in the geolocation algorithm that occasionally caused "drifts" upwards of hundreds of kilometers. (FNMOC and its contractors are continuing to improve their operational software).

NASA Team Algorithm Error Analysis
Errors in the derived sea ice concentrations arise from several sources. In order of importance, these are (1) the inability of the algorithm to discriminate among more than two radiometrically different sea ice types, (2) seasonal variations in sea ice emissivity, (3) nonseasonal variations in sea ice emissivity, (4) weather effects at concentrations greater than about 15 percent, and (5) random and systematic instrument error.

The largest source of error is the inability of the algorithm to discriminate among more than two radiometrically different sea ice types (including different surface conditions). The broad categories of radiometrically different sea ice types are new and young ice, first year ice, and multi-year ice types. Since the algorithm allows for both first year and multi-year ice types, the largest source of error in total ice concentration is caused by the presence of newly forming sea ice. New and young ice, most commonly found in leads and coastal polynyas during winter, are characterized by polarization differences intermediate between open water and thick first year ice (Cavalieri et al. 1986). Normalized polarization ratio (PR) for thin ice will vary in proportion to ice thickness (Grenfell and Comiso 1986) and will increase in proportion to the fraction of new ice filling the SSM/I field of view. Larger areas of new ice within the sensor FOV will result in proportionally larger underestimates by the algorithm. Recently, a new thin ice algorithm has been developed (Cavalieri et al. 1994) which mitigates this problem in seasonal sea ice zones and also permits the mapping of new and young ice types. The thin ice algorithm of Cavalieri et al. (1994) has not however, been incorporated into the sea ice products described in this documentation.

Seasonal variations in sea ice emissivities can be extremely large. Multi-year ice, for example, loses its characteristic microwave spectral signature (negative gradient ratio (GR)) during spring and summer and becomes indistinguishable from first year ice. Another condition resulting in large errors in total ice concentration is the formation of melt ponds on the ice surface, making the ponded region indistinguishable from open water. While the areal extent of ponding is not well known, unpublished data reported by Carsey (1982) show that for the summer of 1975, 20 percent or less of the Arctic ice pack was covered by ponds and that ponding reached maximum areal extent in early July. For an area of the Beaufort Sea (AIDJEX triangle) during August 1975, Campbell et at. (1984) report that the average ponding was 30 percent. The percent coverage of melt ponds varies spatially and temporally across the Arctic and the extent to which they influence summer ice concentrations remains uncertain.

Nonseasonal variations in sea ice emissivity include local variations resulting from fluctuations in the physical and chemical properties of sea ice and overlying snow cover, and regional variations resulting from environmental differences. Regional and hemispheric variability may be considerable, as indicated by previous studies (Comiso 1983, Ackley 1979). Differences between Arctic and Antarctic sea ice microwave signatures noted above result in different sets of algorithm tie-points for each hemisphere. Algorithm errors can be reduced by using locally and seasonally chosen algorithm tie points.

While weather effects, resulting from atmospheric water vapor, cloud liquid water, rain, and sea surface roughening by near-surface winds on the calculated sea ice concentrations, are greatly reduced over open ocean at polar latitudes by the algorithm weather filter, they may nevertheless contribute to the sea ice concentration error at concentrations greater than about 15 percent. Presuming that the atmospheric contribution is nearly zero over consolidated first year ice and that the contribution at the open water end results totally from atmospheric effects estimated to be up to 15 percent, then the error resulting from atmospheric effects for any intermediate concentration may be estimated by a linear interpolation. While the effects of weather on high total ice concentrations are small, there is the potential for significant reductions in multiyear ice concentrations (Maslanik 1992).

Finally, errors in ice concentration also result from random and systematic instrument errors. Except for the 85-GHz channels, over the years of SSM/I operation no noticeable instrument drifts are apparent. Based on prelaunch measurements and on observed radiances over relatively stable targets where temporal and spatial geophysical variability is small, the error for each of the three SSM/I channels used in the algorithm is less than 1 K, and the absolute accuracy is estimated at 3 K (Hollinger 1989). Assuming a 1-K level of random instrument noise in each channel, an upper limit to the rss uncertainty in the calculated concentrations, which depends on surface type and concentration, ranges from about one percent to 1.8 percent for total ice concentration and from 4.5 percent to six percent for multi-year ice concentration.

Bootstrap Algorithm Error Analysis
Under ideal winter situations when only thick ice and open water are present, ice concentration can be derived with the Bootstrap technique at an accuracy of about five to 10 percent, based on standard deviations of emissivities as used in the formulation. Errors are higher in the seasonal sea ice region than in the central Arctic region because of higher standard deviations of consolidated sea ice in the 19 vs 37 GHz plots. This is partly because of spatial changes in surface temperature that are not as effectively accounted for by this set of data.

Constantly changing emissivities of some surfaces present unresolved problems. For example, when leads open up during winter, the open water is exposed to the cold atmosphere and grease ice quickly forms at the surface. The surface then metamorphoses from grease ice, to nilas, to young ice and then to first-year ice with snow cover. During these transitions, the emissivity of the surface can change considerably from one stage to another (Grenfell and Comiso 1986). Since such changes in emissivity are not taken into account in ice concentration algorithms, the derived fractions of open water are therefore not strictly those of open water and may include some mixtures of grease ice and new ice. In spring and summer, the emissivity of thick ice also changes with time, especially over the perennial ice region in the Arctic. The slopes and offsets of the consolidated ice line AD in the scatter plots (see Comiso 1986) are adjusted to automatically take this into account during onset of spring in June; original values should be restored during winter freeze-up. Despite this adjustment, the error is still substantial and can be larger than 20 percent due to spatial variations in melt and affects of meltponding.

Several field and aircraft experiments have been performed in both polar regions for algorithm evaluation. In some of these experiments, basic assumptions about ice types and interpretation of the cluster plots have been confirmed. However, validation of satellite ice concentration data using data from these experiments has not been easy. Field data are difficult to use because of limited coverage compared to the large footprint of the satellite sensors (about 30x30 km). While generally easier to interpret because of fine resolution and availability of ancillary measurements, aircraft data are useful but need to be validated by ground measurements.

Another strategy has been to utilize high resolution satellite data for validation. While the use of high resolution data has its advantages, such strategies degenerate into comparative analysis because the other satellite data also need to be validated. For example, unambiguous discrimination of open water, grease ice, small pancakes, and gray nilas in both visible and microwave channels may be impossible even with high resolution sensors. Generally, however, the passive microwave data provide valuable information about large scale characteristics of the ice cover as well as locations of ice edges, polynyas, and extensive leads. It is, however, useful to note that some of the comparative studies yielded high correlation coefficients.

For a comparison of the two algorithms, please see section 6., Notes and Plans, below.

6.Notes and Plans

Usage Guidance

Processing of the data is ongoing.

Difference Between the Bootstrap and NASA Team Techniques

Both the Bootstrap and NASA Team algorithm use a mixing formalism that assumes that only sea ice and open water exist within the satellite footprint. However, one difference between the two algorithms is that the NASA Team algorithm makes a distinction between two different ice types and open water, whereas the Bootstrap algorithm only assumes ice and open water. The two algorithms also utilize different channels. Both algorithms use the 19V and 37V channels, but the Team algorithm complements those with the 19H channel, whereas the Bootstrap adds the 37H channel. The use of both the 37 GHz channels in the Bootstrap method in the perennial ice region optimizes the resolution of the data. Another fundamental difference between the NASA Team and Bootstrap methods are that the Team algorithm uses two ratios defined by the three channels, whereas the Bootstrap algorithm uses only two channels at a time. Other differences between the two techniques include the use of different weather filters, which may cause differences in the location of the marginal sea ice region and the magnitude of residuals of weather in the open ocean, and differences in references brightness temperatures. The presence of new ice and melt ponds remain a problem for both algorithms.

Advantages of the Bootstrap algorithm include:

Advantages of the NASA Team algorithm include:

For a more detailed comparison between the two techniques, see Comiso et al. (1997) and T. Fink's comparison of passive microwave sea ice algorithms.

A comparison of instruments, and of the differences in orbital parameters (Abdalati et al. 1995) between the F8 and F11 using overlapping data indicated a high degree of correlation (greater than 0.98) between the F8 and F11 data sets. Small variations were attributed to the different orbital characteristics of the two satellites, especially to the differences in data collection times. When comparing F11 to F13, in terms of hemispheric averages of mean sea ice concentration, the biases introduced by the switch from F11 to F13 are slight and are not statistically significant; however, in some regions relatively large and significant differences are seen. In addition, differences in sea ice extent and total ice-covered area between the two platforms were found to be statistically significant (see NSIDC Special Publication 5 for more information).

Two on-line documents that may also be helpful to users are the SSM/I Brightness Temperature and Sea Ice Concentration Grids Frequently Asked Questions (FAQ) and a selected bibliography of published results using SSM/I data.

Future Modifications and Plans

Processing of data is ongoing. Input data are received by NSIDC approximately three to six months after they are collected. Processing and release of the ice concentrations typically occur within another three to six months.

7.Products and Access

Data Center Identification

National Snow and Ice Data Center

Procedures for Obtaining Data

Data are available via FTP.

Data Center Status/Plans

Data processing is ongoing.

Software Description

Data are in HDF Raster Image Format for sea ice, or in HDF Scientific Data Set (SDS) for brightness temperatures. HDF software and libraries are developed and maintained by The HDF Group and may be accessed from The HDF Group Web site. Browse images of daily and monthly sea ice concentration are also provided in GIF format. Finally, NSIDC has developed the tools described below to work with this data set.

Software for reading and displaying the sea ice concentration files is provided via ftp. Included are tools to extract the sea ice concentration files and geolocate the data, as well as masking tools that limit the influence of non-sea ice brightness temperatures.

Sample Interactive Display Language (IDL) commands to read and display sea ice grids and latitude/longitude grids are provided here.

Data Extraction and Display

Source code to extract images from DMSP SSM/I  ice concentration products are available via ftp:

hdftor8_alpha -  A compiled C executable (at The HDF Group) to extract 8-bit raster image data from HDF files.

hdftor8_dec - A compiled C executable (DECstation ULTRIX4.3) to extract 8-bit raster image data from HDF files.

hdftor8_hp - A compiled C executable (at The HDF Group) to extract 8-bit raster image data from HDF files.

hdftor8_pc - A compiled C executable (at The HDF Group) to extract 8-bit raster image data from HDF files.

hdftor8_sgi - A compiled C executable (SGI Challenge IRIX5.2) to extract 8-bit raster image data from HDF files.

hdftor8_sun -A compiled C executable (at The HDF Group) to extract 8-bit raster image data from HDF files.

The extraction executables listed above are from The HDF Group.

Interactive Data Language (IDL) is a commercial data visualization and analysis software package available from Research Systems, Inc. The package is widely used at NSIDC, though no endorsement of the vendor or product is implied. We have provided procedures developed at NSIDC for the convenience of IDL users.

The IDL tools provided with this data set are available via ftp and consist of procedures that allow the user to read, display and export F8, F11, and F13 SSM/I  sea ice grids. These were revised in April 2001 to provide display and export capability for the Near-Real-Time DMSP SSM/I Daily Polar Gridded Sea Ice Concentrations. Examples of running the IDL programs are also provided.

extract_ice.pro - An IDL program that extracts a time series of SSM/I polar sea ice data.  The extraction routine works on both the  daily and monthly sea ice files allowing users to select and display sea ice  concentrations from either the NASA Team or Bootstrap algorithms. For monthly  data, users may select which image threshold to view (e.g. either zero, five, 10 or 15 percent ice cut off). Typing:
            extract_ice,array
in IDL compiles the program, and stores the data extracted in the array.

disp_ssmi_ice_xa.pro - An IDL program that will display animations of the sea ice concentration grids (automatically). The routine works on both the daily and monthly sea ice  files and allows users to display sea ice concentrations derived from either the  NASA Team or Bootstrap algorithm. For monthly data, users may select which image  threshold to view (e.g. either zero, five, 10 or 15 percent ice cut off). Typing:
        disp_ssmi_ice_xa
in IDL compiles and runs the program, displaying the images for however many days the user requested when prompted by the program.

Users who wish to overlay coastline, latitude/longitude lines, or land masks on the data must download them via ftp or from the "tools" directory on the brightness temperature CD-ROMs.

Useful IDL commands:

    openw       - a procedure that opens a file for write.

                        IDL> openw,1,'n199809av.ic'

    writeu      - a procedure that writes data (ex: 'seaice') into an opened file.

                        IDL> writeu,1,seaice

    loadct      - a procedure that loads a color palette by providing a  list or loading the palette indicated by the
                     argument.

    xloadct     - a procedure that opens a new window giving the user a visual choice of colors to use.

For a complete description of the Interactive Display Language (IDL), see IDL Reference Guide or IDL User's Guide. Research Systems, Inc., 2995 Wilderness Pl., Suite 203, Boulder, CO 80301, (303) 786-9900.

Geo-Coordinate and Pixel Area Tools

The geolocation and pixel area tools consist of a FORTRAN routine called, "locate," a latitude/longitude grid, and a pixel area grid. "locate" allows the user to enter an i,j coordinate and get the corresponding latitude/longitude coordinate, and vice versa.

Sample Interactive Display Language (IDL) commands to read and display latitude/longitude grids are provided here.

Geo-Coordinate and Pixel Area Tools are available via ftp.

locate.exe_pc -  An compiled fortran executable (PC DOS) that allows the user to enter an i,j coordinate and get the corresponding latitude/longitude coordinate, and vice versa.

locate.for - A fortran executable that allows the user to enter an i,j coordinate and get the corresponding latitude/longitude coordinate, and vice versa.

mapll.for and mapxy.for - These subroutines are associated with the locate.for program. These programs need to be compiled, but are not run explicitly.  They are called by locate.for. Thus, the user should compile these programs with locate.for and then use locate to do the conversions.

psn12lats.dat (pss12lats.dat) - Grids that can be used to determine the latitude of a given pixel for the 12.5 km grids (85 Ghz data) for either hemisphere. These latitude grids are in binary format and stored as long word integers (4 byte) scaled by 100,000. Each array location (i,j) contains the latitude value at the center of the corresponding data grid cells.

psn12lats.dat : 608 columns x 896 rows, range = [31.0967, 89.8363]
pss12lats.dat : 632 columns x 664 rows, range = [-39.3649, -89.8368]

psn12lons.dat (ps12lons.dat) - grids can be used to determine the longitude of a given pixel for the 12.5 km grids (85 Ghz data) for either hemisphere. These longitude grids are in binary format and stored as long word integers (4 byte) scaled by 100,000. Each array location (i,j) contains the longitude value at the center of the corresponding data grid cells.

psn12lons.dat : 608 columns x 896 rows, range = [00.0000, 360.0000]
pss12lons.dat : 632 columns x 664 rows, range = [000.1651, 359.8350]

psn25lats.dat (pss25lats.dat) - Grids that can be used to determine the latitude of a given pixel for the 25 km grids for either hemisphere. These latitude grids are in binary format and stored as long word integers (4 byte) scaled by 100,000. Each array location (i,j) contains the latitude value at the center of the corresponding data grid cells.

psn25lats.dat : 304 columns x 448 rows, range = [31.0967, 89.8363]
pss25lats.dat : 316 columns x 332 rows, range = [-39.3649, -89.8368]

psn25lons.dat (ps25lons.dat) - grids can be used to determine the longitude of a given pixel for the 25 km grids for either hemisphere. These longitude grids are in binary format and stored as long word integers (4 byte) scaled by 100,000. Each array location (i,j) contains the longitude value at the center of the corresponding data grid cells.

psn25lons.dat : 304 columns x 448 rows, range = [00.0000, 360.0000]
pss25lons.dat : 316 columns x 332 rows, range = [000.1651, 359.8350]

Please note that the data ranges given here are latitude and longitude values for the center of each grid cell. The range covered by the full grid extends to the pole (90 degrees latitude) and all longitudes (0 to 360 degrees longitude).

To determine the lat/lon values of corresponding (i,j) data grid cells:
1) read in the array as long (4-byte) integers.
2) divide these values by 100,000. The resulting array gives the lat/lon values for the data grid cells in decimal degrees.

psn12area.dat ( pss12area.dat) - grids that can be used to determine of the area of a given pixel for the 12.5 km grids (85 Ghz data) for either hemisphere.  The arrays are in binary format and stored as long word integers (4 byte) scaled by 100. Each array location (i,j) contains the real value of the corresponding grid cell.

psn12area.dat and pss12area.dat: 608 columns x 896 rows

psn25area.dat ( pss25area.dat) - grids that can be used to determine of the area of a given pixel for the 25 km grids for either hemisphere.  The arrays are in binary format and stored as long word integers (4 byte) scaled by 100. Each array location (i,j) contains the real value of the corresponding grid cell.

psn25area.dat and pss25area.dat: 304 columns x 448 rows

Land Masks

Beginning with the SMMR era, the first masks to be developed were "landmask.ntb" and "landmask.stb". Other masks were later added, for use with SSM/I F8 data ("n3a" and "s3a"). The next set of masks added are those beginning with "gsfc," which were originally developed for use with SSM/I F11 and F13. The last set of masks added are those beginning with "amsr." All of the masks provided here will work across all platforms (SMMR through SSM/I F13).

Four sets of land masks are supplied via ftp. The first two are "built in" to SMMR and SSM/I sea ice concentrations, respectively; therefore, users should be aware of effects of land mask differences in time-series analyses from multiple data sets. The following table summarizes the history of land mask development at NSIDC; all are currently provided via ftp:

Distributed beginning with data set File name Characteristics
SSM/I F-8 series n3acoast.hdf
n3altln.hdf
n3amask.hdf
n3bcoast.hdf
n3bltln.hdf
n3bmask.hdf
s3acoast.hdf
s3altln.hdf
s3amask.hdf
s3bcoast.hdf
s3bltln.hdf
s3bmask.hdf		 North 12.5 km coastline mask
North 12.5 km lat/lon mask
North 12.5 km land mask
North 25 km coastline mask
North 25 km lat/lon mask
North 25 km land mask
South 12.5 km coastline mask
South 12.5 km lat/lon mask
South 12.5 km land mask
South 25 km coastline mask
South 25 km lat/lon mask
South 25 km land mask
SSM/I F-11 and F-13 series gsfc_25n.hdf
gsfc_25s.hdf		 North 25 km land mask
South 25 km land mask
Snow Melt Onset Over Arctic Sea Ice from SMMR and SSM/I Brightness Temperatures n3blcmsk.hdf North 25 km land mask. Ocean pixels adjacent to land or one pixel removed from land are designated as "land-contaminated."
AMSR-E/Aqua Daily L3 12.5-km and 25-km sea ice polar grids amsr_gsfc_12n.hdf
amsr_gsfc_12s.hdf
amsr_gsfc_25n.hdf
amsr_gsfc_25s.hdf		 The North 25 km AMSR-E land mask is identical to "gsfc_25n.hdf," while the South land masks are based on data from the 1997 RADARSAT Antarctic campaign. Although the AMSR-E land masks are more current than those NSIDC uses for SSM/I data, they have not been updated since 1997 to account for many iceberg calvings.

The following table summarizes which land masks are used for NSIDC's three primary sea ice products:

Data set Land mask used Differences
DMSP SSM/I Daily and Monthly Polar Gridded Sea Ice Concentrations NASA Team algorithm: n3bmask.hdf and s3bmask.hdf

Bootstrap algorithm: gsfc_25n.hdf and gsfc_25s.hdf		 North polar grid:
There are 591 additional pixels that "gsfc_25n.hdf" classifies as ocean and "n3bmask.hdf" classifies as land.

There are 1305 additional pixels that "n3bmask.hdf" classifies as ocean and "gsfc_25n.hdf" classifies as land.

South polar grid:
There are 223 additional pixels that "gsfc_25n.hdf" classifies as ocean and "n3bmask.hdf" classifies as land.

There are 232 additional pixels that "n3bmask.hdf" classifies as ocean and "gsfc_25n.hdf" classifies as land.
Bootstrap Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I gsfc_25n.hdf and gsfc_25s.hdf  
Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I Passive Microwave Data gsfc_25n.hdf and gsfc_25s.hdf South polar grid: "gsfc_25s.hdf" has seven additional pixels of land area, compared with the "gsfc_25s.hdf" land mask in the Bootstrap Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I product.


Data set Land mask Source
Nimbus-7 SMMR GSFC-1 CIA World Shoreline Data Bank II
DMSP SSM/I JPL-1
GSFC-2		 CIA World Shoreline Data Bank I
USGS Digital Chart of the World

Note on Land Masks

The relatively slight differences in numbers of SSM/I-grid pixels masked as land in the three grids noted above can introduce discrepancies in analyses of time series spanning the SMMR and SSM/I period. One method of addressing this issue is to generate a composite mask in which all pixels mapped as land in any of the masks are coded as land pixels in the composite mask. Use of such a composite mask improves the consistency of the SMMR and SSM/I record at the expense of masking additional ocean areas as land. A composite of the CIA World Data Bank I and II (GSFC-1 and JPL-1) has been produced by J. Maslanik at NSIDC.

An additional issue concerns effects on coastal ocean pixels of contamination by proximity to land. Such proximity can modify the brightness temperatures of coastal ocean pixels, producing false ice concentration values along some coasts. These "pixel mixing" errors are considered in the summer 1996 issue of NSIDC Notes (Issue no. 18). Maslanik et al. discuss the effects of such land contamination in introducing differences between SMMR and SSM/I time series, and describe the use of a modified land mask where land areas are extended to mask substantial contamination. The Snow Melt Onset Over Arctic Sea Ice from SMMR and SSM/I Brightness Temperatures land mask ("N3BCLMSK.HDF") alleviates this problem whereby ocean pixels adjacent to land or one pixel removed from land are designated as "land-contaminated."

Regional Masks

The regional masks "sectmask.n" and "sectmask.s," available via ftp, are described further in: "Arctic and Antarctic Sea Ice, 1978-1987 Satellite Passive Microwave Observations and Analysis," NASA SP-511.

The files contain standard 300-byte headers, followed by 2-dimensional byte arrays of 448 rows x 304 columns (332 x 316) stored by rows in column order. Regions are assigned different byte values as follows:

               Arctic
Region                                   Byte Value
     Lakes                                    0
     Non - region Oceans                      1
     Sea of Okhotsk and Japan                 2
     Bering Sea                               3
     Hudson Bay                               4
     Baffin Bay/Davis Strait/Labrador Sea     5
     Greenland Sea                            6
     Kara and Barents Seas                    7 
     Arctic Ocean                             8 
     Canadian Archipelago                     9 
     Gulf of St. Lawrence                   10 
     Land                                    11 
     Coastline                               12

                        Antarctic
Sector                                   Byte Value
     Weddell Sea                              2
     Indian Ocean                             3
     Pacific Ocean                            4
     Ross Sea                                 5
     Bellingshausen Amundsen Sea              6 
     Land                                    11 
     Coastline                               12

Ocean Masks and images of maximum ice extent

Please see Sea Ice Trends and Climatologies from SMMR and SSM/I for details.

Also see Importing SSM/I Daily and Monthly Sea Ice Concentration Data into ArcInfo.

8.References

Also see Selected Bibliography: SSM/I Brightness Temperature and Sea Ice Concentration Grids for the Polar Regions.

Abdalati, W., K. Steffen, C. Otto, and K. C. Jezek. 1995. Comparison of brightness temperatures from SSMI instruments on the DMSP F8 and F11 satellites for Antarctica and the Greenland Ice Sheet. International Journal of Remote Sensing 16(7):1223-1229.

Ackley, S. F. 1979. Mass balance aspects of Weddell Sea pack ice. Journal of Glaciology 24(90):391-406.

Ackley, S. F., A. J. Gow, K. R. Buck, and K. M. Golden. 1980. Sea ice studies in the Weddell Sea aboard USCGC Polar Sea. Antarctic Journal of U. S. 15(5):84-86.

Bonbright, D. I. 1984. PODS SSM/I Functional Requirements (Version 1.0). Jet Propulsion Laboratory Document 715-63.

Bonbright, D. I., J. W. Brown, J. E. Hilland, I. T. Hsu, J. A. Johnson, T. L. Kotlarek, R. A. Lassanyi, C. L. Miller, C. S. Morris, and F. J. Salamone. 1987. NASA ocean data system version 3.0. User handbook. Jet Propulsion Laboratory. Document 715-66, 50pp. + appendices.

Campbell, W. J., P. Gloersen, and H. J. Zwally. 1984. Aspects of Arctic sea ice observable by sequential passive-microwave observations from the Nimbus 5 satellite. In Arctic technology and policy. I. Dyer and C. Chryssostomidis, eds. New York: Hemisphere Publishing. p.197-222.

Carsey, F. D. 1982. Arctic sea ice distribution at end of summer 1973-1976 from satellite microwave data. Journal of Geophysical Research 87:5809-5835.

Cavalieri, D. 1996. NASA Team Sea Ice Algorithm. http://nsidc.org/data/docs/daac/nasateam/index.html

Cavalieri, D. J. and P. Gloersen. 1984. Determination of sea ice parameters with the Nimbus 7 SMMR. Journal of Geophysical Research 89(D4):5355-5369.

Cavalieri, D. J., J. Crawford, M. Drinkwater, W. J. Emery, D. T. Eppler, L. D. Farmer, M. Goodberlet, R. Jentz, A. Milman, C. Morris, R. Onstott, A. Schweiger, R. Shuchman, K. Steffen, C. T. Swift, C. Wackerman, and R. L. Weaver. 1992. NASA sea ice validation program for the DMSP SSM/I: final report. NASA Technical Memorandum 104559. National Aeronautics and Space Administration, Washington, D.C. 126 pages.

Cavalieri, D. J., J. Crawford, M. R. Drinkwater, D. Eppler, L. D. Farmer, R. R. Jentz and C. C. Wackerman. 1991. Aircraft active and passive microwave validation of sea ice concentration from the DMSP SSM/I. Journal of Geophysical Research 96(C12):21,989-22,009.

Cavalieri, D.J., P. Gloersen, and W. J. Campbell. 1984. Determination of sea ice parameters with the NIMBUS-7 SMMR. Journal of Geophysical Research 89(D4):5355-5369.

Cavalieri, D. J., P. Gloersen, and T. T. Wilheit. 1986. Aircraft and satellite passive microwave observations of the Bering Sea cover during MIZEX West. IEEE Transactions on Geoscience and Remote Sensing GE-24:368-377.

Cavalieri, D. J., K. M. St. Germain, and C. T. Swift. 1995. Reduction of weather effects in the calculation of sea iIce concentration with the DMSP SSM/I. Journal of Glaciology 41(139): 455-464.

Comiso, J. C. 1991. Satellite remote sensing of the polar oceans. Journal of Marine Systems 2:295-434.

Comiso, J. C. 1990. Arctic multiyear ice classification and summer ice cover using passive microwave satellite data. Journal of Geophysical Research 95(C8):13411-13422.

Comiso, J. C. 1986. Characteristics of Arctic winter sea ice from satellite multispectral microwave observations. Journal of Geophysical Research 91(C1): 975-994.

Comiso, J.C. 1983. Sea ice effective microwave emissivities from satellite passive microwave and infrared observations. Journal of Geophysical Research 88(C12):7686-7704.

Comiso, J. C., and C. W. Sullivan. 1986. Satellite microwave and in-situ observations of the Weddell Sea ice cover and its marginal ice zone. Journal of Geophysical Research 91(C8):9663-9681.

Comiso, J. C., S. F. Ackley, and A.L. Gordon. 1984. Antarctic sea ice microwave signatures and their correlation with in-situ ice observations. Journal of Geophysical Research 89(C1):662-672.

Comiso J.C., D.J. Cavalieri, C.L. Parkinson, and P. Gloersen. 1997. Passive Microwave Algorithms for Sea Ice Concentration: A Comparison of Two Techniques, Remote Sensing of the Environment 60:357-384.

Comiso, J. C., T. C. Grenfell, D. Bell, M. Lange, and S. Ackley. 1989. Passive microwave in-situ observations of Weddell Sea ice. Journal of Geophysical Research 88(C12):7686-7704.

Comiso, J. C., T. C. Grenfell, M. Lange, A. Lohanick, R. Moore, and P. Wadhams. 1992. Microwave remote sensing of the Southern Ocean Ice Cover. Chapt. 12 In Microwave remote sensing of sea ice. Frank Carsey, editor. American Geophysical Union. Washington, D.C. 243-259.

Comiso, J. C., P. Wadhams, W. Krabill, R. Swift, J. Crawford, and W. Tucker. 1991. Top/bottom multisensor remote sensing of Arctic sea ice. Journal of Geophysical Research 96(C2):2693-2711.

Gloersen P., W. J. Campbell, D. J. Cavalieri, J. C. Comiso, C. L. Parkinson, H.J. Zwally. 1992. Arctic and Antarctic Sea ice, 1978-1987: Satellite Passive Microwave Observations and Analysis. NASA Special Publication 511.

Gloersen P. and D. J. Cavalieri. 1986. Reduction of weather effects in the calculation of sea ice concentration from microwave radiances. Journal of Geophysical Research 91(C3):3913-3919.

Goodberlet, M. A. 1990. Special Sensor Microwave/Imager Calibration/Validation. Ph.D. dissertation submitted to the University of Massachusetts.

Grenfell, T. C. and J. C. Comiso. 1986. Multifrequency passive microwave observations of first-year sea ice grown in a tank. IEEE Transactions on Geoscience and Remote Sensing GE-24:826-831.

Hollinger, J. P. 1989. DMSP Special Sensor Microwave/Imager Calibration/Validation. Naval Research Labs. Washington, D. C.

Hollinger, J. P. and R. C. Lo. 1983. SSM/I Project Summary Report. Naval Research Laboratory. NRL Memorandum Report 5055. 106 pp.

Hollinger, J. P., J. L. Pierce, G. A. Poe. 1990. SSM/I instrument evaluation. IEEE Transactions on Geoscience and Remote Sensing 28(5):781-790.

Hollinger, J. P., B. E. Troy, R. O. Ramseier, K. W. Asmus, M. F. Hartman, and C. A. Luther. 1984. Microwave emission from high Arctic sea ice during freeze-up. Journal of Geophysical Research 89(C5): 8104-8122.

Hughes Aircraft Company. 1986. Data requirements document for Fleet Numerical Oceanography Center, Rev. B.

Hughes Aircraft Company. 1980. Special Sensor Microwave Imager (SSM/I), computer program product specification (Specification for FNMOC). Vol. II, Sensor Data Processing, Computer Program Component (SMISDP).

Martino, M. G., D. J. Cavalieri and P. Gloersen. 1995. An improved land mask for the SSM/I Grid. NASA Technical Memorandum. TM104625.

Maslanik, J. A. 1992. Effects of weather on the retrieval of sea ice concentration and ice type from passive microwave data. International Journal of Remote Sensing 13(1):37-54.

Massom, R. A. 1991. Satellite remote sensing of polar regions. Boca Raton: Lewis Publishing.

National Center for Supercomputing Applications. 1993. Getting started with HDF. Draft Version 3.2, pp. 44.

Pearson, F. 1990. Map projections: Theory and applications. CRC Press. Boca Raton, Florida. 372 pages.

National Snow and Ice Data Center. 1995. DMSP SSM/I Brightness Temperature and Ice Concentrations Grids for the Polar Regions. User's Guide. Revised Edition. NSIDC Distributed Active Archive Center, University of Colorado, Boulder.

Poe, G. A. and R. W. Conway. 1990. A study of the geolocation errors for the Special Sensor Microwave/Imager (SSM/I). IEEE Transactions on Geoscience and Remote Sensing 28(5):791-799.

Smith, D.M. 1996. Extraction of winter total sea-ice concentration in the Greenland and Barents Seas from SSM/I data. International Journal of Remote Sensing 17(33): 2625-2646.

Snyder, J.P. 1987. Map projections - a working manual. U.S. Geological Survey Professional Paper 1395. U.S. Government Printing Office. Washington, D.C. 383 pages.

Snyder, J. P. 1982. Map Projections Used by the U.S. Geological Survey. U.S. Geological Survey Bulletin 1532.

Steffen, K. and A. Schwieger. 1991. NASA Team algorithm for sea ice concentration retrieval from Defense Meteorological Satellite Program Special Sensor Microwave/Imager: Comparison with Landsat satellite imagery. Journal of Geophysical Research 96(C12):21,971-21,988.

Steffen, K., D. J. Cavalieri, J. C. Comiso, K. St. Germain, P. Gloersen, J. Key, and I. Rubinstein. 1992. The estimation of geophysical parameters using passive microwave algorithms. Chapt 10 In Microwave remote sensing of sea ice. Frank Carsey, editor. American Geophysical Union. Washington, D.C. 243-259.

Stroeve, J. 1998. Impact of various processing options on SSM/I-derived brightness temperatures. NSIDC Special Report-7. http://nsidc.org/pubs/special/7/index.html

Svendsen, E., K. Kloster, B. Farrelly, O. M. Johannessen, J. A. Johannessen, W. J. Campbell, P. Gloersen, D. Cavalieri, and C. Matzler. 1983. Norwegian Remote Sensing Experiment: Evaluation of the Nimbus 7 scanning multichannel microwave radiometer for sea ice research. Journal of Geophysical Research 88(C5):2781-2791.

Swift, C. T., D. J. Cavalieri. 1985. Passive microwave remote sensing for sea ice research. EOS 66(49):1210-1212.

Swift, C. T., Fedor, L. S. and Ramseier, R. O. 1985. An algorithm to measure sea ice concentration with microwave radiometers. Journal of Geophysical Research 90(C1):1087-1099.

Wadhams, P., M. A. Lange, and S. F. Ackley. 1987. The ice thickness distribution across the Atlantic sector of the Antarctic ocean in midwinter. Journal of Geophysical Research 92(C13):14,535-14, 552.

Wentz, F. J. 1993. User's manual: SSM/I antenna temperature tapes. rev. 2. Remote Sensing Systems, Inc. Santa Rosa, CA. RSS Technical Report 120193.

Wentz, F. J. 1992. Final report, production of SSM/I data sets. Remote Sensing Systems, Inc., Santa Rosa, CA. RSS Technical Report 090192.

Wentz, F. J. 1991. User's manual: SSM/I antenna temperature tapes. Remote Sensing Systems, Inc., Santa Rosa, CA. RSS Technical Report 032588.

 

9.Document Information

Document Revision Date

March 2006

Document URL

http://nsidc.org/data/docs/daac/nsidc0002_ssmi_seaice.gd.html

Contact Us | NSIDC Web Policy | Use & Copyright | Our Sponsors

The National Snow and Ice Data Center

Supporting Cryospheric Research Since 1976
449 UCB  University of Colorado  Boulder, CO 80309-0449

Cooperative Institute for Research in Environmental Sciences (CIRES)    University of Colorado Boulder