The MSFC SSM/I Brightness Temperature Data Sets

Table of Contents

• 1. Overview
• 2. Processing Steps
  • 2.1. Swath Data
    • 2.1.1 Data Acquisition
    • 2.1.2 Data Quality Control
    • 2.1.3 Data Processing
  • 2.2 Gridded Data
  • 2.3 Browse Images (HDF,GIF
• 3. Brightness Temperature (Tb) files
  • 3.1 File Names
  • 3.2 File Content
  • 3.3 Data Values
• 4. High Resolution Geolocations
  • 4.1 File Names
  • 4.2 File Contents
  • 4.3 Data Values
• 5. Low Resolution Geolocation (ln) files
  • 5.1 File Names
  • 5.2 File Names
  • 5.3 Data Values
• 6. Gridded Files
  • 6.1 File Names
  • 6.2 File Names
  • 6.3 Data Values
• 7. Browse Files (HDF,GIF)
• 8. File Access
  • 8.1 Sample Programs
• 9. HDF Library and Tools
• 10. References
• 11. Contact Information

1. Overview

IMPORTANT NOTICE: Due to the Navy's activation of a radiation/calibration beacon on the DMSP-F15 platform, the SSM/I 22V channel has become corrupted. This beacon was started on August 14, 2006. The corrupted 22V will affect all geophysical products which use it. This means the GHRC's IWV, CLW, and OWS (for F-15 only) are likely unusable, and will be discontinued from the time this beacon was activated. We're sorry for this inconvenience.

The Global Hydrology Resource Center (GHRC) has been processing and archiving Special Sensor Microwave / Imager (SSM/I) data since 1990. Prior to May 1995, the SSM/I data source for the GHRC was the National Environmental Satellite Data and Information Service (NESDIS). Since May, 1995, the source has been the Fleet Numerical Meteorology and Oceanography Center (FNMOC). Data is obtained from FNMOC and processed at the GHRC within hours of its reception. Each day full resolution or "swath" brightness temperatures (Tb's) and reduced resolution "gridded" data sets are generated. Browse images of the gridded files are created in both HDF raster and GIF formats. HDF represents the Hierarchical Data Format, the data format standard for NASA's Earth Observing System Data and Information System.

The GHRC is producing several geophysical products using oceanic SSM/I Tb data including: integrated water vapor, cloud liquid water, and oceanic wind speed. These products are also in swath and gridded formats.

The Department of Defense maintains a series of polar orbiting sun-synchronous meteorological satellites as a part of the Defense Meteorological Satellite Program (DMSP). To date, six of the DMSP satellites have flown an SSM/I (see Table 1).

Table 1

The SSM/I is a passive microwave sensor that detects microwave radiation at four frequencies: 19.35 GHz, 22.235 GHz, 37.0 GHz, and 85.5 GHz. The 19, 37, and 85 GHz channels have dual-polarization (vertical and horizontal), while the 22 GHz channel has only vertical polarization. This yields a total of seven channels. The instrument is a conical scanning device sweeping out an approximately 1400 km wide swath as it looks forward (F8 looks backward) at an angle of about 45 degrees from vertical. The instrument scans through 102 degrees for each scanning revolution. The scan period is 1.899 seconds.

The 85 GHz channels scan continuously, while the 19, 22, and 37 GHz channels sample every other element of every other scan. The scan on which all channels are sampled is termed the A-scan and the 85 GHz-only scan is called the B-scan. The A- and B-scans are referred to as a scan-pair. Because of these sampling characteristics, there are 64 low frequency samples and 128 high frequency samples in each A-scan. There are one-fourth as many elements in a low frequency (19, 22, or 37 GHz) swath than in a high frequency (85 GHz) swath.

Table 2 (from the SSM/I Calibration and Validation Final Report) shows footprint sizes and sampling resolutions for all four frequencies.

Table 2

The GHRC is currently processing and archiving SSM/I data from F13, F14, and F15. Table 3 shows typical values for some orbital characteristics of these DMSP satellites.

Table 3

UTC represents Universal Coordinated Time.

Since the orbital period for all DMSP satellites shown in Table 3 is about 102 minutes, each DMSP orbits the earth about 14.1 times per day. The GHRC processing breaks the SSM/I data stream into passes. A pass is defined as a pole-to-pole swath that is either ascending (south to north) or descending (north to south). A complete pass contains about 51 minutes of data or about 1610 A and B-scans. The first pass of a UTC day is defined as the first complete pass of the day. The last pass processed is the last complete pass beginning in the day being processed. Be aware that the last pass will almost always contain some data from the next UTC day. If there are any data for a pass, the output files are created and all missing scans are flagged. A typical day contains 28 or 29 passes.

For more information about the DMSP-F8, F10, F11, F12, F13, F14, and F15 satellites, refer to the GHRC Home Page's documentation at Dataset descriptions/Guide Document Pages.

For more information about the SSM/I instrument, refer to: SSM/I_Sensor.

2. Processing Steps

2.1 Swath Data

2.1.1 Data Acquisition

The source of the GHRC's SSM/I data is FNMOC. FNMOC is the Department of Defense's (DoD) center of expertise for DMSP passive microwave data processing. The DoD has an agreement with the National Oceanographic and Atmospheric Administration (NOAA) under which they share data via the Shared Processing Network (SPN). FNMOC generates SSM/I antenna temperature files known as Temperature Data Records (TDR's) and sends them to NOAA/NESDIS. NESDIS performs minimal file unpacking and makes the data available to the GHRC. Files are usually available within 1 hour (or less) of data acquisition by FNMOC.

2.1.2 Data Quality Control

Each day the GHRC performs a four step quality control process on the previous day's data. These steps are:

1. Date stamping
2. File merging
3. Navigation checking
4. Calibration checking

Step 1, date stamping, is necessary because each scan-pair of data has a "time" stamp, but not a "date" stamp. This step removes all data not from the day being processed.

Step 2, file merging, combines all of the data files from a specified day. During this process, data gaps are filled with flagged values.

Step 3, navigation checking, uses an orbit prediction model to determine the proper latitude and longitude of the 85 GHz mid-scan pixel at a given time. That element's latitude and longitude values are compared to the model results, and if they are not within a specified distance, the entire scan-pair is flagged as mislocated.

Internal checks are also performed in this step on the latitude and longitude values to look for "glitches" or dramatic changes. For instance, if two consecutive scan-pair longitude values show eastward motion, the second scan-pair's latitude, longitude, and surface type values are flagged as questionable. The antenna temperature values are left unchanged.

Step 4, calibration checking, checks the hot and cold load counts for each scan and each channel. If a bad hot or cold load count is found, and since 10-scan-pair averaging has been applied by FNMOC, ten scan-pairs of antenna temperatures are flagged for that specific channel, (the scan-pair at which the bad calibration occurred and the 9 previous scan-pairs). The three hot load thermistor temperatures are also checked, but if only these are bad, only the current scan-pair is flagged.

2.1.3 Data Processing

After these QC steps, data are read for each pass (ascending or descending), sequentially for an entire day. Antenna temperatures are then converted to brightness temperatures (Tb's) using Remote Sensing System's "DECODE-4A" intercalibration software. The output for each execution of the software is:

1. one pass file with brightness temperatures (Tb)
2. one pass file with high resolution geolocation information (hn)
3. one pass file with low resolution geolocation information (ln), associated with the low frequency channels.

In addition, on each execution an ASCII text file is written containing summary information of each pass. This provides a quick look of the day's processing.

2.2 Gridded Data

All of the swath files containing ascending passes are averaged into a 0.5 x 0.5 degree global grid (720 x 360). The same is done for descending passes. The global grid is centered on the Greenwich meridian. Each grid box value is the mean of the SSM/I brightness temperatures located within the half degree box centered at every xx.25 and xx.75 degrees. Only valid brightness temperatures are used. Table 4 lists some representative grid points and their geographic extents.

Table 4

2.3 Browse Images (HDF, GIF)

The HDF gridded data files are used to create HDF raster(8) images. These images do not contain scientific data, but instead contain only color values. GIF images are then created from the HDF raster images. There are 14 HDF raster files and 14 GIF files (ascending and descending files for each channel) generated per day.

3. Brightness Temperature Files

3.1 File Names

The brightness temperature pass file naming convention is:

fxx_Tb_yyddd_ppZ.hdf.gz
fxx_ln_yyddd_ppZ.hdf.gz
meta_fxx_yyddd.text

where

xx = the satellite ID number (13, 14, or 15)
yyddd = the date; year (yy) and day (ddd)
pp = the pass number (01-29)
Z = the pass direction (A-ascending or D-descending)
hdf = an HDF file
gz = it has been compressed using the "gzip" utility

For example, the file "f13_Tb_05008_06D.hdf.gz" contains F13 brightness temperature data from the 6th pass (descending) of day 05008. The corresponding high-resolution geolocation data for this pass is in "f13_hn_05008_06D.hdf.gz"

3.2 File Content

HDF files contain data grouped within the file called "objects". Table 5 lists the objects in the brightness temperature HDF file. The dimension 'N' represents the number of A- and B-scans in the pass, nominally 1570 - 1616, and can be retrieved with a simple HDF library call. (See section 7) All objects are filled with missing scans where needed. HDF labels are provided with each object.

Table 5

In the Spacecraft Position Vector array, missing scans are flagged with -999.0. Since the values are floating point, no scaling is done. Times are in seconds (to the nearest half second), latitudes range from 0.00 to 180.00 degrees, longitudes are positive east from -180.00 to 180.00 degrees, and altitudes are in whole meters. The two-line element set (object #12) contains a two-line formatted group of ephemeris values closest in time to 1200 UTC of the day being processed. The GHRC retrieves two-line element sets daily. The McIDAS Navigation directory (object #13) is for use with McIDAS software. NOTE: See Pass Metadata object for information about the Pass Metadata object (object #14).

3.3 Data Values

The brightness temperature data in the HDF files are integer*2, ranging from -32768 to 32767. Table 6 shows possible values of brightness temperatures.
Note: Time of day values are negated for several conditions – see Table 8-A.

Table 6

4. High Resolution Geolocation (hn) Files

4.1 File Names

The high resolution geolocation pass file naming convention is:

fxx_hn_yyddd_ppZ.hdf.gz

where

xx = the satellite ID number (13, 14, or 15)
yyddd = the date; year (yy) and day (ddd)
pp = the pass number (01-29)
Z = the pass direction (A-ascending or D-descending)
hdf = an HDF file
gz = it has been compressed using the "gzip" utility

For example, the file "f13_hn_95165_27D.hdf.gz" contains F13 high resolution geolocation data from the 27th pass (descending) of day 95165. It corresponds to the brightness temperature data in "f13_Tb_95165_27D.hdf.gz"

4.2 File Content

Table 7 lists the objects in the high resolution geolocation HDF files. The dimension 'N' represents the number of A- and B-scans in this pass (nominally 1570 - 1616) and can be retrieved with a simple HDF library call. All objects are filled with missing scans where needed. Labels are provided with each object.
 

Table 7

4.3 Data Values

All latitude and longitude data are scaled by 100. Latitude is positive north (-9000 to 9000) and longitude is positive east (-18000 to 18000).
Tables 8-A and 8-B show Latitude, Longitude, & Surface type values in “hn” files. 

* Time of day values are negated for these situations.

Table 8-A

Table 8-B

5. Low Resolution (ln) Files

5.1 File Names

Low resolution geolocation files are created for users who either do not need high resolution information or are using geophysical products, most of which are generated at the resolution of the low frequency channels (19, 22, and 37 GHz). The low resolution geolocation pass file naming convention is:

fxx_ln_yyddd_ppZ.hdf.gz

where

xx = the satellite ID number (13, 14, or 15)
yyddd = the date; year (yy) and day (ddd)
pp = the pass number (01-29)
Z = the pass direction (A-ascending or D-descending)
hdf = it is an HDF file
gz = it has been compressed using the "gzip" utility

For example, the file "f13_ln_95299_27A.hdf.gz" contains F13 low resolution geolocation data from the 27th pass (ascending) of day 95299. It corresponds to the brightness temperature data (B-scan only) in "f13_Tb_95299_27A.hdf.gz"

5.2 File Content

Table 9 lists the objects in the low resolution geolocation HDF files. The dimension 'N' represents the number of A-scans in this pass (nominally 785 - 808) and can be retrieved with a simple HDF library call. All objects are filled with missing scans where needed. Labels are provided with each object.
 

Table 9

5.3 Data Values

All latitude and longitude data are scaled by 100. Latitude is positive north (-9000 to 9000) and longitude is positive east (-18000 to 18000).
Tables 8-A and 8-B show Latitude, Longitude, & Surface type values in “ln” files. 

6. Gridded Files

6.1 File Names

The files are in the Hierarchical Data Format (HDF). Each HDF file has also been compressed using the GNU gzip utility. The file naming convention is:

fxx_Tb_yyddd_dayAD.hdf.gz

where

yyddd = the date; year (yy) and day (ddd)
xx = the satellite ID number (13, 14, or 15)

For example, the file "f14_Tb_04219_dayAD.hdf.gz" contains gridded F14 brightness temperatures for day 04219. The 'AD' in the filename represents the fact that the ascending and descending passes are computed separately and placed in two images.

6.2 File Content

Table 10 lists the objects in the HDF files and their contents.

Table 10

Note: See Gridded Metadata object for information about the Gridded Metadata object (object #16).

6.3 Data Values

The brightness temperature data in the gridded HDF files are integer*2. Table 11 shows possible values.

Table 11

7. Browse Files (HDF, GIF)

Browse files are generated from the daily gridded data in two formats, HDF raster-8 and GIF. There is only one image per file. The file naming conventions are:

fxx_ccc_yyddd_dayZ_ras8.hdf
fxx_ccc_yyddd_dayZ_ras8.gif

where

xx = the satellite ID number (13, 14, or 15)
ccc = the channel (V19, H19, V22, V37, H37, V85, H85)
yyddd = the date; year (yy) and day (ddd)
Z = the pass direction (A-ascending or D-descending)
hdf = an HDF raster image file
gif = a GIF image file

For example, the file "f15_H37_05003_dayD.ras8.hdf" contains an HDF raster8 image of the F15 H37 brightness temperatures for the descending passes of day 05003. HDF-Raster8 files are in the HDF 8-bit raster image format. They each contain one 8-bit raster image, a palette, a file description, and a legend.

GIF files are in the Compuserve GIF image format. They each contain one annotated image.

8. Data and Browse Access

SSM/I data and browse processed by the GHRC are available for online ordering via the Global Hydrology Resource Center. (click "Dataset List", then "DISCOVER") Additionally, the most recent two years of SSM/I data are available for immediate download via our Data Pool at http://datapool.nsstc.nasa.gov/.

Note: The SSM/I Tb data and Navigational data on our Data Pool are not compressed, nor are they stored together in the same directory as on our GHRC anonymous FTP server. Please refer to the Data Pool Navigational Documentation found here: http://moby.itsc.uah.edu/ftpdata/doc/ssmi_tb/readme_1st.html for directions on how to locate all the necessary data. The Data Pool has the additional capabilities of subsetting, packaging, and compressing the data when you use the preferred "shopping cart" method of data ordering.

GHRC SSM/I Tb Ordering Options

8.1 Sample Programs

Two sample FORTRAN program have been provided to read swath and gridded data. read_ssmi_tb.f is for reading Tb swath files while read_ssmi_tb_grid.f is for reading Tb grid files. They will read a file and fill arrays containing various data and metadata. The HDF software library is required (see below).

9. HDF Library and Tools

HDF is a library and platform independent data format for the storage and exchange of scientific data. It includes Fortran and C calling interfaces, and utilities for analyzing and converting HDF data files. HDF is developed and supported by the National Center for Supercomputing Applications (NCSA) and is available in the public domain (http://hdf.ncsa.uiuc.edu/). HDF stands for Hierarchical Data Format. It is a multi-object file format for the transfer of graphical and numerical data between machines. HDF is a portable file format. HDF files can be shared across platforms. An HDF file created on one computer, say a Silicon Graphics (SGI) workstation, can be read on another system, say IBM PC, without modification.

9.1 How to Obtain the HDF Library

The HDF library and tools can be down loaded from the World Wide Web at: http://hdf.ncsa.uiuc.edu/. Follow the instructions given there.

9.2 Visualization Software

The HDF files were created using HDF version 4.1, release 3. Numerous useful visualization tools can be found at http://hdf.ncsa.uiuc.edu/hdftools.html. The GIF images may be viewed by most popular GIF image viewers.

10. References

Hollinger, J., et al., DMSP Special Sensor Microwave / Imager Calibration / Validation Final Report Volume I, Naval Research Laboratory, Washington, D. C.,20 July 1989.

Wentz, Frank J., 1991: User's Manual SSM/I Antenna Temperature Tapes Revision 1, RSS Technical Report 120191, Dec. 1, 1991, Remote Sensing Systems, Santa Rosa, CA.

Wentz, Frank J., 1993: User's Manual SSM/I Antenna Temperature Tapes Revision 2, RSS Technical Report 120193, Dec. 1, 1993, Remote Sensing Systems, Santa Rosa, CA.

11. Contact Information

Data can be ordered and questions addressed at http://ghrc.nsstc.nasa.gov/.

To order this data or for further information, please contact: Global Hydrology Resource Center User Services 320 Sparkman Drive Huntsville, AL 35805 Phone: 256-961-7932 E-mail: ghrc@eos.nasa.gov

---

NASA Information Contact: Michael Goodman, Global Hydrology and Climate Center GHRC Web Curator: GHRC Web Team Last update: Tuesday, 17-Oct-2006 14:16:42 CDT If you have trouble viewing or navigating this page, please contact GHRC User Services. U.S. Government Compliance report.