NOTICE OF PROCESSING ALGORITHM CHANGE:
The GHRC will begin using the latest software from Frank Wentz in our SSM/I automated processing. We will be using newer software in processing data from all current SSM/I platforms (F13, F14, & F15) beginning with January 1, 2002. This change is intended to provide the best possible SSM/I products for our customers.

We will reprocess as much data as possible from these 3 platforms. We plan to begin with F15 data and reprocess - working back in time - beginning with December 2001. Following this we'll reprocess F14, and then F13 - working back in time with each.

Our estimates of expected completion times are very rough - as time, money, and personnel issues have an affect. With this in mind, we hope to have F15 (2 years - back to January 2000) completed by the end of January 2002. The remaining F13 (4 2/3 years) and F14 (6 years) could be completed by June 2002.

The new code contains inter-calibration coefficients (calibrated to F8) for all other sensors (F10-F15). This DOES change the brightness temperature values of the low resolution channels.

We executed 2 days of F-13 and F-14 data and found that the new Tb's averaged roughly:

(F13)
V19 0.52K lower +/- 0.04K
H19 0.43K lower +/- 0.03K
V22 0.04K lower +/- 0.02K
V37 0.10K higher +/- 0.18K
H37 0.21K lower +/- 0.22K

(F14)
V19 0.83K lower +/- 0.04K
H19 0.20K lower +/- 0.02K
V22 0.19K lower +/- 0.02K
V37 0.42K lower +/- 0.40K
H37 0.04K lower +/- 0.23K

The high resolution brightness temperature channels (V85 & H85) DID NOT change.

This brief analysis is not expected to provide a detailed statistical picture of the changes in data values, but should give you an idea as to the changes to expect in product values (Oceanic Wind Speed, Cloud Liquid Water, & Integrated Water Vapor).

The new software generates data with the same filenames, so the only way you can tell which software was used is to check the HDF file attribute. Both the old and new software specifies the specific module used to create the file, but the new software gives a version number.

To be specific - the full resolution swath brightness temperature file attribute is:
tpass.f - version 14

the full resolution swath product file attribute is:
tprdx.f - version 24

the gridded brightness temperature file attribute:
tb_bro.f - version 19

the gridded products (iwv, clw, ows) file attribute:
px_bro.f - version 21

If you have questions about these changes, or want to order past data, please contact GHRC User Services at ghrc@eos.nasa.gov or 256-961-5932.

For the latest information about the reprocessing, visit Hydrology News.

Table of Contents

• 1. Introduction
• 2. Special Sensor Microwave Imager
  • 2.1 Instrument Description
  • 2.2 Scan Geometry
  • 2.3 Antenna Beam Characteristics
• 3. SSM/I TDR Data Set
  • 3.1 File Naming Conventions
  • 3.2 Data Set Format
• 4. References

1. Introduction

This README file contains information on the Special Sensor Microwave Imager (SSM/I) Temperature Data Record (TDR) data sets. Brief descriptions of the Defense Meteorological Satellite Program (DMSP) satellites, the SSM/I instrument, the computation of antenna temperatures and brightness temperatures and the format of the data sets are included. Included is the source code for three software programs. The first program, ssmitdrlatlon.c, extracts the portion of the daily file which contains a user specified latitude/longitude range. The second program, ssmitdrta.c, reads a NESDIS SSM/I TDR data file and outputs antenna temperatures. The third program, ssmitdrtb.f, converts the antenna temperatures produced by the second program into brightness temperatures. Pertinent scientific references are also included.

The Defense Meteorological Satellite Program (DMSP) is a Department of Defense program which is responsible for designing, building, launching and operating polar orbiting meteorological satellites. The SSM/I was first launched on June 19, 1987 aboard the DMSP Block 5D-2 Spacecraft F8. Since that time, additional launches have been made aboard DMSP Block 5D-2 Spacecraft F10 and DMSP Block 5D-2 Spacecraft F11. The F11 Spacecraft is still operational but the F8 Spacecraft completed it's mission on July 31, 1994 and F10 completed it's mission on November 14, 1997. DMSP Block 5D-2 Spacecraft F13 was launched in March 1995 and DMSP Block 5D-2 Spacecraft F14 was launched in April 1997.

The satellites can broadcast visual, infrared and microwave imagery directly to transportable tactical sites around the world. The data is also stored for transmission to the Navy's Fleet Numerical Meteorology and Oceanography Center (FNMOC) and to the Air Force Global Weather Central (AFGWC). The Naval Oceanography Command (NOC) and the Air Weather Service process the SSM/I data to obtain near real-time global maps of cloud water, rain rates, water vapor over oceans, marine wind speeds, sea ice location, age and concentration, snow water content and land surface type, moisture and temperature.

Each of the DMSP satellites fly in a sun-synchronous, near-polar orbit. For a satellite in sun synchronous orbit, the ascending equatorial crossing time remains relatively constant with respect to the local time throughout the lifetime of the satellite. The table below summarizes the orbits for each spacecraft (valid for the orbit shown).

2. Special Sensor Microwave Imager

The information in this section is quoted from the DMSP Special Sensor Microwave /Imager Calibration/ Validation Report, Volume I, Hollinger, et al. (1989) and the DMSP SSM/I Calibration/Validation Report, Volume II, Hollinger, et al. (1990).

The SSM/I is a seven-channel, four-frequency, linearly-polarized, passive microwave radiometric system. The SSM/I receives both horizontally and vertically linearly polarized radiation at 19.3,37.0, and 85.5 GHz and vertical only at 22.2 GHz. The following table summarizes satellite specific instrument details.

2.1 Instrument Description

The information in this section is quoted from the DMSP Special Sensor Microwave /Imager Calibration/ Validation Report, Volume I, Hollinger, et al. (1989) and the DMSP SSM/I Calibration/Validation Report, Volume II, Hollinger, et al. (1990).

The SSM/I instrument consists of an offset parabolic reflector with dimensions of 61 by 66 cm. This reflector is illuminated by a corrugated, broad-band, seven-port horn antenna. The feed-horn antenna and the reflector are mounted on a drum which contains the radiometers, digital data subsystem, mechanical scanning subsystem, and power subsystem. The entire feed-horn, reflector and drum assembly rotate about the axis of the drum by a coaxially mounted bearing and power transfer system (BAPTA). All data, commands, power, timing and telemetry signals pass through the BAPTA on slip ring connectors to the rotating assembly.

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 horn once each scan. The mirror reflects cosmic background radiation (3 K) into the feed horn thus serving, along with the hot reference absorber, as calibration references for the SSM/I. This scheme provides an overall end-to-end absolute calibration which includes the feed horn. The combination of the calibration scheme and the use of total power radiometers greatly improve the SSM/I performance in comparison with previous spaceborne radiometeric systems. Corrections for spillover and antenna pattern effects from the parabolic reflector are incorporated in the data processing algorithms.

The SSM/I employs a total-power radiometer configuration. A balanced mixer down-converts the output of the feed horn and amplifies it. Then a square-law detector converts that output to a video voltage. A bandpass filter defines the receiver passband and improves the out-of-band rejection. The component of receiver output due to receiver noise is removed through the amplification and offset of the detected video signal. The output of the video amplifier is integrated by an integrate and dump filter for 7.95 msec for all frequencies, except the 85.5 GHz frequency which is integrated for 3.89 msec. Following this integration, the output is delivered to the data processing system. The time between samples is the same for all frequencies (12.5 msec) except for the 85.5 GHz frequency (4.22 msec).

The data processor uses an analog multiplexer to multiplex the seven radiometer output signals. The data processor then samples and holds the signals prior to them being digitized into 12-bit words. Twelve channels, which contain three hot target temperature measurements, two temperature sensor measurements within the radiometer, reference voltage, and reference return data, are multiplexed with radiometer data.

A microprocessor supervises instrument timing, control, and data buffering with the DMSP Operational Line Scanner (OLS) instrument (colocated on the satellite) which records all SSM/I data. The average data rate of the SSM/I including zeros required to match the OLS interface is 3276 bps.

2.2 Scan Geometry

The information in this section is quoted from the DMSP Special Sensor Microwave /Imager Calibration/ Validation Report, Volume I, Hollinger, et al. (1989) and the DMSP SSM/I Calibration/Validation Report, Volume II, Hollinger, et al. (1990).

The SSM/I continuously rotates about an axis parallel to the local spacecraft vertical at 31.6 rpm. The SSM/I measures, over an angular section of 102.4 degrees about the sub-satellite track, the upwelling scene brightness temperatures. When looking in the forward direction of the spacecraft, the scan is directed from left to right with active scene measurements lying 51.2 degree about the forward direction. The resulting swath width is 1400 km, which results in 24 hour global coverage. The spacecraft sub-satellite point travels 12.5 km during the 1.9 second period. For each scan, 128 uniformly spaced 85.5 GHz scene data are taken over a 102.4 degree scan region. The sampling interval is 4.22 msec and equals the time for the beam to travel 12.5 km in the cross track direction. Radiometer data at the remaining frequencies are sampled every other scan with 64 uniformly spaced samples being taken and have an 8.44 msec interval. Scan A denotes scans in which all channels are sampled and Scan B denotes scans in which only 85.5 GHz data are taken. The start and stop times of the integrate and dump filters at 19.35, 22.235, and 37.0 GHz are selected to maximize the radiometer integration time and achieve concentric beams for all sampled data. The effect of the radiometer integration times is to increase the effective along scan beam diameter and make the beams at 37 and 85 GHz nearly circular.

There are two attachments to this README. Attachment 1 contains the SSM/I scan geometry. Attachment 2 contains the coverage of SSM/I in 24 hours. Both attachments are in GIF format.

2.3 Antenna Beam Characteristics

The information in this section is quoted from the DMSP Special Sensor Microwave /Imager Calibration/ Validation Report, Volume I, Hollinger, et al. (1989) and the DMSP SSM/I Calibration/Validation Report, Volume II, Hollinger, et al. (1990).

The data in Table 2.1 apply to the S/N 002 Sensor on F-10 and F-11 and represents measurements of the instantaneous field of view (IFOV)3 dB beamwidths of the secondary radiation patterns as a function of the channel frequency and polarization for SSM/I. The data are based on measurements, taken prior to launch, of antenna temperatures which were averaged over the RF bands. An effective field of view (EFOV) can be defined for each sampled radiometer brightness temperature that takes into account the integration time. For the along-track direction, the IFOV and the EFOV are basically the same but in the cross track direction, or H-plane direction, the IFOV is significantly smaller than the EFOV. Table 2.1 provides information on both the EFOV 3-dB beamwidths and the IFOV beamwidths to allow for intercomparison. The table also provides the along-track and cross track dimensions of the EFOV beamwidths as projected onto the surface of the Earth.

The main beam efficiency is an important parameter in the antenna performance. The main beam efficiency is defined as the percentage of energy received within the main beam of the far-field radiation pattern in the desired polarization within the prescribed bandwidth to the total energy received.

Another important antenna performance parameter is the main beam efficiency which is defined as the percentage of energy received within the main beam of the far-field radiation pattern in the desired polarization within the prescribed bandwidth to the total energy received. The far-field antenna pattern is the combination of the radiation patterns of the feedhorn antenna and the parabolic reflector antenna. The antenna beam efficiencies as a function of channel frequency and polarization for the S/N 002 instrument are provided in Table 2.2. The data are based upon the S/N 002 instrument antenna range measurements of both the feedhorn patterns and the radiation patterns from the reflector. The antenna sidelobe column denotes the percentage of energy lying outside 2.5 times the 3-dB beamwidth of the far-field pattern when normalized to the sum of the co- and cross-polarization energies. The cross-polarization column is defined as the percentage of cross-polarized energy appearing at the output of the feedhorn and includes contributions from the reflector and feedhorn. The feedhorn spillover factor refers to the loss of the energy in the far-field pattern which is not intercepted by the reflector. Thus, in the computation of beam efficiency, the feedhorn spillover loss is a multiplicative factor.

The loss in beam efficiency due to small scale surface roughness is very small at all frequencies and was therefore not included in Table 3.2. The rms surface roughness, which is less than 0.025 mm, translates to a loss of less than 0.15% all frequencies except the 85.5 GHz frequency which has a loss of less than 0.8%.

3. SSM/I TDR Data Set Summary

3.1 File Naming Conventions

The FNMOC TDRs provided by the GHRC are in a file containing a day's worth of orbit files. The daily files are named tallmiyy.jd_ds_daily.tar, where ta = abbreviation for antenna temperatures, ll=satellite number (10 , 11, 13, 0r 14), mi = abbreviation for sensor SSM/I, yy = the year of the century, jd = day of the year, ds = data source (i.e., fnmoc), and daily.tar = the file type where all information has been put together in one file and tarred. A sample file name for the data from the F-10 sensor for day 364 in 1994 would be: ta10mi94.364_fnmoc_daily.tar.

Each orbit file concatenated into the daily tar file follows the NESDIS file naming convention of NSS.TDRR.Sx.Dyyddd.Shhmm.Ehhmm.Annnnnnn, where S = "satellite", x = the satellite identification number minus 6, D = "day", yy = the year of the century, ddd = the day of the year, S = "orbit file start time", hh = start hour of orbit file, mm = start minute of orbit file, E = "orbit file end time", hh = end hour of orbit file, mm = end minute of orbit file, and A = the processing sequence designator. The satellite identification numbers for F-8, F-10, and F-13 are, respectively, 2, 4, and 7. The processing sequence designator in either an "A" or a "B".

Note that the orbit file names are generated by NESDIS using the information found in the file header. The file header is created by FNMOC and does not always accurately reflect the start and end times of the data in the file.

3.2 Data Set Format

Please note: NESDIS TDR format has changed. Each A-B scan pair now occupies one record of 3604 bytes. The data records are preceded by a 3604-byte header record. Here are the record layouts. Lengths are in bytes. At the GHRC, the new format begins with the May 1997 data (D97121).

The TDR files are stored in the Shared Processing Network Data Exchange Format (SPN DEF). This format consists of data description (DD) blocks and data blocks. All blocks begin with a block identification and a block length and end with a checksum. The data description blocks are used to store acronyms, start bytes, lengths, units, signs, and exponents for the corresponding data blocks. There are 7 non-repeating header blocks, three repeating scan blocks, and a non-repeating trailer block. Table 3.1 lists the blocks in the order that they appear in the file. The data blocks will be described in subsequent sections. For detailed information on all the blocks, please refer to the SSM/I Data Requirements Document (Hughes Aircraft, 1986) and the software in Appendices A, B, and C.

3.2.1 Revolution Header Data Block

This block contains the spacecraft identification, the revolution number, the data start time, the data end time, and the nearest ascending node time. The times are stored as day of year, hours, minutes, seconds.

3.2.2 Scan Header #1 Data Block

This block contains the parameters listed in Table 3.2 for the respective scan data block. The Scan Header #1 block is repeated at the beginning of every new scan A/B pair.

3.2.3 Scan Header #2

This block contains the parameters listed in Table 3.3 for the respective scan data block. The Scan Header #2 block is repeated at the beginning of every new scan A/B pair.

3.2.4 Scan Data Block

This block contains the latitude (LAT), longitude (LON), surface type (SFT), pixel number, and antenna temperature data listed in Table 3.4 for an A/B scan pair.

4. References

Special Sensor Microwave/Imager (SSM/I) Data Requirements Document, Department of the Air Force, Headquarters, February, 1986.

Hollinger, J., et al., Special Sensor Microwave / Imager User's Guide, Naval Research Laboratory, Washington, D. C., 14 September 1987.

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

Lee, S. O., SSM/I Level 1b Interface Control Document, NOAA/NESDIS, Suitland, Maryland, 20233, June 15, 1993.

Wentz, F. J., Measurement of Oceanic Wind Vector Using Satellite Microwave Radiometers, RSS Technical Report 051591, 33 pp, Remote Sensing Systems, Santa Rosa, CA, 15 May 1991.

Wentz, F. J., User's Manual SSM/I Antenna Temperature Tapes (Revision 1), RSS Technical Report 120191, 70 pp, Remote Sensing Systems, Santa Rosa, CA, 1 December 1991.

You can contact user and data services by phone or email using the number or address below:

Global Hydrology Recource Center
320 Sparkman Drive
Huntsville, AL 35805

Phone: 256-961-7932
FAX: 256-961-7723
E-mail: ghrc@eos.nasa.gov