NOAA Polar Orbiter Data User's Guide

Section 5.1

Blue line drawn across page to separate text from document title.
Image denotes media you're currently using__Web or CD.Introduction Page, NOAA POD Guide TOC, Acronyms
Previous Section, Next Section

5.1 TOVS Sounding Product

NESDIS currently has the capability of producing a maximum of 100,000 soundings every 24 hours from two operational spacecraft. Up until March 9, 1992, the soundings were 250 km resolution and only those classified as "good" were archived. Beginning March 9, 1992, all the soundings were included ("good" and "redundant", not sampled) on the archive tape with their resolution increased to 80 km. These soundings are provided to SSB on a weekly basis.

Section 5.1.1 contains a description of the format used for the TOVS Sounding Product between January 1979 and March 8, 1992. As of March 9, 1992, a modified format for the TOVS Sounding Product was implemented because of a change in archive media - seven days of soundings are written to one IBM 3480 cartridge each week. A description of this format is contained in Section 5.1.2.

5.1.1 TOVS Sounding Product (January 1979 - March 8, 1992)

The data volume required for the TOVS Sounding Product for one week for one satellite is approximately one-third of a 9-track, 6250 BPI tape. Therefore, when there is only one operational satellite, three weeks worth of sounding data can be stacked on one CCT. However, when there are two operational satellites, the sounding data will be intermixed and in chronological order on one weekly CCT.

Individual soundings stored on tape are formatted according to the NMC/NESDIS Data Set Format which is described in detail later in this section. Each tape contains a data directory found at the beginning of the tape which will describe the contents of the tape. The data directory acts as a housekeeping file and consists of two parts: 1) an element containing directory information, and 2) the elements related to the data on the tape. The purpose of the NMC/NESDIS archive housekeeping file is to aid users in determining the location of time categorized, satellite-derived soundings on the tape. The end of each group of time-categorized soundings, as described by the individual data directory elements, is indicated by an end of file (EOF). A double EOF marks the end of data on tape as described by the last directory element in the housekeeping file. Together, the EOF markers and the housekeeping file are intended to provide the user with a means of selectively choosing soundings from the tape.

The format of the NMC/NESDIS archive housekeeping file is in two parts. The first part of the file is the Directory Information Element which contains 20 bytes or ten 2-byte words in the format shown in Table 5.1.1-1.

Table 5.1.1-1. Directory Information Element.
Byte # Contents
1-2 Number of data directory elements in housekeeping file
3-6 Total number of soundings (reports) on the tape
7-8 Processing year (2 digits)
9-10 Processing month (2 digits)
11-12 Processing day (2 digits)
13-20 Spares (denoted by "666610")

The second part of the housekeeping file is the Data Directory Element which contains 20 bytes per element in the format shown in Table 5.1.1-2.

Table 5.1.1-2. Data Directory Element.
Byte # Contents
1-2 Time category of reports stored
3-4 Number of reports in time category
5-6 Year of reports (Century x 256 + year)
7-8 Month and day of reports (Month x 256 + day)
9-10 Time of day (UTC) of earliest report (Hours x 256 + minutes)
11-12 Time of day (UTC) of latest report (Hours x 256 + minutes)
13-20 Spares (denoted by "666610")

Digital data may be selected from the TOVS Sounding Product tapes by time and/or area. Each record that contains any of the selection criteria will be included in its entirety. Note: if the TOVS Sounding Product selection software is run, the housekeeping file is not included with the selected data. Using the selection software, a 1600 BPI CCT holds approximately 131,000 soundings, while a 6250 BPI CCT holds approximately 455,000 soundings. Specific orbits cannot be selected with the existing software.

The time category (listed as the first two bytes of the Data Directory Element) is defined in Table 5.1.1-3. If the value of the time category number equals 10 + the category number, it will denote that the soundings for that data directory element are of bad quality.

Table 5.1.1-3. Time Category.
Category Number Time Range (UTC)
1 0000 - 0259
2 0300 - 0559
3 0600 - 0859
4 0900 - 1159
5 1200 - 1459
6 1500 - 1759
7 1800 -2059
8 2100 - 2359

The overall structure of each TOVS Sounding Product archive tape is shown in Figure 5.1.1-1. It shows the n+1 elements in the housekeeping file and the corresponding TOVS soundings files which are in the NMC/NESDIS Data Set Format.

Figure 5.1.1-1. Structure of TOVS Sounding Product Tape.
Directory Information Element
Data Directory Element #1
Data Directory Element #2
Data Directory Element #3
...
Data Directory Element #n EOF
Housekeeping File
TOVS Soundings described by Data Directory Element #1 EOF Data Files
TOVS Soundings described by Data Directory Element #2 EOF
...
TOVS Soundings described by Data Directory Element #n EOF
TOVS Quality information (since 1989) EOF
EOF

The TOVS Sounding Product tape has a logical record length of 280 bytes and a block size (physical record length) of 6,440 bytes. However, the housekeeping file has a variable length physical record. It ranges from a minimum of 280 bytes to a maximum of 3080 bytes because of the variable nature of the directory. The 6,440-byte block size holds 23 reports of 280 bytes each. Thus, each report occupies one logical record. A missing or undefined data value is denoted by "777710", a spare is "666610", and an end-of-report is indicated by "888810". The TOVS Sounding data files are written in the NMC/NESDIS Data Set Format. This format is in binary and contains 44 bytes of documentation and 236 bytes of Sounding Products/information for a maximum of seven data type categories. The format for the documentation portion is contained in Table 5.1.1-4.

Table 5.1.1-4. NMC/NESDIS Documentation Format.
Byte # Content
1-2 Satellite Identification (see Section 2.0.2)
3-4 Year (last two digits) x 256 + Month
5-6 Day x 256 + Hour (UTC)
7-8 Minutes x 256 + Seconds (UTC)
9-10 Latitude (degrees x 100, +N, -S, Range: +90 to -90)
11-12 Longitude (degrees x 100, +E, -W, Range: +180 to -180)
13-14 Solar Zenith Angle (degrees x 100, +Day, -Night, Range: +90 to -90)
15-16 Land/Sea Indicator (if land, surface elevation in meters; if sea, zero)
17-18 Surface (skin) temperature (K x 10)
19-20 Estimated Surface Model Pressure (mb x 10)
21-22 Instrument and/or Channel combination used to obtain various products
23-24 Retrieval Method
25-26 Standard Deviation of Low-level HIRS/2 Channel (7) (K x 100, range: 0 to 10)
27-28 Standard Deviation of Mid-level HIRS/2 Channel (5) (K x 100, range: 0 to 10)
29-30 Average value of N*, range: 0 to 1 (dimensionless x radiance values when there are some clouds obstructing the surface.)
31-32 Superswath Box counter and minibox (Superswath x 1000 + Box counter x 10 + minibox)
33-34 Sea Surface Temperature (K x 10). SST over ocean, skin temperature over land.
35-36 Day of Month and Hour (UTC) when Edit Flag is written to file. Scaling: (Day x 256) + Hour
37-38 Minute and Second (UTC) when Edit Flag is written to file. Scaling: (Minute x 256) + Seconds
39-40 TOVS Filter Flag: 0 = sounding is good; 1 = sounding is redundant
41-44 Special Counter for 7-Day Archive Tape (address of start on disk)
Note: A value of N* between 0 and 1000 indicates that the N* method was used. A value of "7777" indicates missing data and implies that the HIRS radiances are completely clear and the N* method was not used. A value of "9211" means that the HIRS radiances are completely cloudy and the user should check the channel combination flags to determine which channels were used.

A superswath is defined as the 40 scan lines of HIRS/2 data between calibration modes, while a box is 7 lines x 9 spots, and a minibox is 3 lines x 3 spots. There are a maximum of 26 superswaths per orbit.

Table 5.1.1-5 contains the format of the sounding products/information (arranged in categories). Category 6 contains radiometric information supplied in the form of equivalent blackbody or brightness temperatures. These can be transformed to radiances (mW/(m2-sr-cm-1)) by application of the Planck function. The Planck function requires channel central wave numbers and temperature-adjustment coefficients which are spacecraft/instrument dependent and available in Section 1.4.

Table 5.1.1-5. NMC/NESDIS Sounding Products/Information Format.
Byte # Content
Category 1 - Layer Mean Temperature (Maximum 15 layers)
45-46 Pressure at lower boundary (mb x 10)
47-48 Pressure at upper boundary (mb x 10)
49-50 Layer-mean temperature (K x 10)
51-52 Quality (K x 10)
53-164 Repeat of Category 1 data for 14 more layers
Category 2 - Layer Precipitable Water (Maximum 3 layers)
165-166 Pressure at lower boundary (mb x 10)
167-168 Pressure at upper boundary (mb x 10)
169-170 Layer Precipitable Water (mm)
171-172 Quality (percent)
173-188 Repeat of Category 2 data for two more layers
Category 3 - Tropopause Parameters
189-190 Pressure (mb x 10)
191-192 Temperature (K x 10)
193-194 Quality (mb x 10)
195-196 Spare
Category 4 - Ozone
197-198 Total ozone amount (Dobson units)
199-200 Quality (percent)
Category 5 - Cloud Cover
201-202 Pressure (mb x 10)
203-204 Amount (Percent)
Category 6 - Sounding Radiances
205-244 Twenty HIRS/2 Channel Equivalent Blackbody Temperatures (K x 64 except for Channel 20 which has K x 16)
245-252 Four MSU Channel Equivalent Blackbody Temperatures (K x 64)
253-258 Three SSU Channel Equivalent Blackbody Temperatures (K x 64)
259-260 Spare
261-276 New Category(s) Spare
277-278 Spare
279-280 End of Report

The documentation portion of the NMC/NESDIS Data Set Format contains two bytes (bytes 21-22) which indicate the instrument and/or channel combination used to obtain various products. ICC represents the value in this byte, and is defined as shown in Equation 5.1.1-1,

ICC=4096 Z + 256 Y + 16 X + 4 W + V

Table 5.1.1-6 describes the variables that comprise ICC.

Table 5.1.1-6. Variables that comprise ICC.
V denotes channel combinations used to obtain layer precipitable water for the layers: surface to 700 mb, 700 to 500 mb, and 500 to 300 mb. V can have the following values:
0 No Retrieval
1 HIRS/2 + MSU
2 HIRS/2
W denotes channel combinations used to obtain tropopause temperature and pressure. W can have the following values:
0 No Retrieval
1 HIRS' + MSU
2 HIRS/2
X denotes channel combinations used to obtain total ozone. X can have the following values:
0 No Retrieval
1 HIRS/2 (1, 2, 3, 8, 9, 10) + MSU (4)
2 HIRS/2 (1, 2, 3, 8, 9, 10)
3 HIRS/2 (1, 2, 3, 9, 10) + MSU (4)
4 HIRS/2 (1, 2, 3, 9, 10)
Y denotes channel combinations used to obtain mean temperature for the layers: surface to 850 mb, 850 to 700 mb, 700 to 500 mb, 500 to 400 mb, 400 to 300 mb, 300 to 200 mb, 200 to 100 mb. Y can have the following values:
0 No Retrieval
1 HIRS/2 + MSU
2 HIRS' + MSU
3 HIRS/2
4 MSU
5 HIRS' + MSU + SKINTK (ocean only)
6 MSU + SKINTK (ocean only)
Z denotes channel combinations used to obtain mean temperature for the layers; 100 to 70 mb, 70 to 50 mb, 50 to 30 mb, 30 to 10 mb, 10 to 5 mb, 5 to 2 mb, 2 to 1 mb, 1 to 0.4 mb. Z can have the following values:
0 No Retrieval
1 HIRS' + SSU + MSU (3,4)
2 HIRS' +MSU (3,4)
3 SSU + MSU (3,4)
4 HIRS' + SSU
5 HIRS'
6 MSU (3,4) (Output will terminate at 10 mb since there is no meaningful data above that level in this case.)
Note that HIRS' is equivalent to HIRS/2 Channels 1 (668 cm-1), 2 (679 cm-1), 3 (691 cm-1), and 17 (2360 cm-1).

The retrieval method is contained in bytes 23 and 24 of the documentation portion of the NMC/NESDIS Data Set Format. METREC represents the value in the two bytes, and is defined by Equation 5.1.1-2,

METREC=256 A + 16 B + C

Table 5.1.1-7 describes the variables that comprise METREC.

Table 5.1.1-7. Variables that comprise METREC.
A = 0 No HIRS/2 data.
1 Clear radiances are derived from completely clear spots.
2 Clear radiances derived from the N* approach.
B = 0 No HIRS/2 data.
1 All HIRS/2 channels were used in the retrieval.
2 The tropopause HIRS/2 channels were unusable due to clouds and only stratospheric channels were used in the retrieval.
C = 0 A statistical retrieval method was used.
1 The minimum information retrieval was used.
2 The minimum information retrieval was attempted, but the statistical retrieval was used.

5.1.1.1 TOVS Quality Information File (Sept. 1989 - March 8, 1992)

Since September 1989, TOVS quality information as required by NOAA's Climate and Global Change Program, has been generated and written to the last file (usually file 58) of the TOVS Sounding Product archive tape. This quality information is cumulative (e.g., the quality information written during the week of January 8, 1990, contains quality statistics from July 1989 through January 7, 1990). Statistics were generated beginning in July 1989 but were not written to the archive tape until September 1989.

Table 5.1.1.1-1 shows the general structure of the TOVS quality information file. The housekeeping record (record 1) has the format shown in Table 5.1.1.1-2, while the directory records (records 2 to 8) have the format shown in Table 5.1.1.1-3. Data records (records 9 to 848) have the format shown in Table 5.1.1.1-4. Each directory record contains 120 directory elements. Each element is a group of 12 words containing information on the corresponding data record (described in Table. 5.1.1.1-4).

Table 5.1.1.1-1. General structure of the TOVS quality information file.
Record # Contents
1 Housekeeping record (see Table 5.1.1.1-2)
2-8 Directory records (see Table 5.1.1.1-3)
9-848 Data records (see Table 5.1.1.1-4)

Table 5.1.1.1-2. Format of the housekeeping record for the TOVS quality information file.
Word # (I*4) Contents
1 Total number of records in the file (848)
2 Last record updated
3 Last record updated at the time the file was archived on tape
4 Beginning record number of directory (2)
5 Beginning record number of data (9)
6-720 Spares (fill = -110)

Table 5.1.1.1-3. Format of directory record for the TOVS quality information file.
Record # (I*2) Contents
2 Directory of data records 9 - 128 (fill = -110)
3 Directory of data records 129 - 248
4 Directory of data records 249 -368
5 Directory of data records 369 - 488
6 Directory of data records 489 - 608
7 Directory of data records 609 - 728
8 Directory of data records 729 - 848

Table 5.1.1.1-4. Format of the data record for TOVS quality information file.
Word # Contents
1 Start month
2 Start day
3 Start year
4 End month
5 End day
6 End year
7 Plot title code
8 Data type code
9 Satellite ID
10 Record number
11 Thickness/Virtual temperature indicator
12 Spare (fill = -110)

Table 5.1.1.1-5 contains the format of the TOVS quality information data records. The format included in Table 5.1.1.1-5 is for records 9 to 848, which are data records (all words are INTEGER*2 with fill = -110). These 1440 words are repeated for each week on the file.

Table 5.1.1.1-5. Format of TOVS quality information data records.
Word # Description Scaling
1-240 Zone 1 (90N - 60N) n/a
1-60 Mean (retrieval - radiosonde) n/a
1-30 Temperature (1000 - 10 mb) 100
31-44 Thickness/Virtual Temperature (1000 - 10 mb) 10
45 Tropopause temperature 100
46 Tropopause pressure 10
47-60 Mixing ratio (1000 - 350 mb) 100
61-120 Root Mean Square (Retrieval -radiosonde) n/a
61-90 Temperature (1000 - 10 mb) 100
91-104 Thickness/Virtual Temperature (1000 - 10 mb) 10
105 Tropopause temperature 100
106 Tropopause pressure 10
107-120 Mixing ratio (1000 - 350 mb) 100
121-180 Standard deviation (σ) n/a
121-150 Temperature (1000 - 10 mb) 100
151-164 Thickness/Virtual Temperature (1000 - 10 mb) 10
165 Tropopause temperature 100
166 Tropopause pressure 10
167-180 Mixing ratio (1000 - 350 mb) 100
181-240 Sample size (N) n/a
181-210 Temperature (1000 - 10 mb) 1
211-224 Thickness/Virtual Temperature (1000 - 10 mb) 1
225 Tropopause temperature 1
226 Tropopause pressure 1
227-240 Mixing ratio (1000 - 350 mb) 1
241-480 Zone 2 (60N to 30N). Format same as for Zone 1. n/a
481-720 Zone 3 (30N to 0). Format same as for Zone 1. n/a
721-960 Zone 4 (0 to 30S). Format same as for Zone 1. n/a
961-1200 Zone 5 (30S to 60S). Format same as for Zone 1. n/a
1201-1440 Zone 6 (60S to 90S). Format same as for Zone 1. n/a

5.1.2 TOVS Sounding Product (March 9, 1992 - May 31, 1998)

Beginning with the data for March 9, 1992, the format for the TOVS Sounding Product has changed. This section describes the new format and how it relates to the old format.

The TOVS sounding product currently consists of a full resolution file, which replaces the former archive of a sampled dataset. Some of the changes which have been implemented in this new format include the following:

  1. The number of words in each data record remains the same at 140 words. Each variable is INTEGER*2 and the logical record length is fixed at 280 bytes.
  2. The new format consists of one large file for all the TOVS data (instead of 57 small files) plus a separate file containing the TOVS Quality information file.
  3. There is no longer a housekeeping file at the beginning of the data.
  4. The TOVS quality information file is contained in a separate file from the normal TOVS Sounding data and must be specifically ordered from SSB if desired.
  5. The archive media changed from 9 track 6250 bpi magnetic tape to IBM 3480 cartridges.

The number of records on the cartridge varies from week to week. Before Sept. 21, 1992, the data were blocked into 6440 byte physical records. Each physical record contained 23 logical records of 280 bytes each. To minimize space on the cartridge, the block size was changed to 31,920 bytes beginning with the data on Sept. 21, 1992. Currently each physical record (block) contains 114 logical records of 280 bytes each. Each logical record comprises a report and is composed of 140 INTEGER*2 words (280 bytes). The data are separated into three-hour time periods, thus there are eight possible time periods in a day. The last two data records of each time period are fillers, i.e., every byte in the record is set to a value of "-33310". Although the format of the old and the new TOVS Sounding data records are very similar, for clarification purposes, the format for the new data record is repeated in Tables 5.1.2-1.

Table 5.1.2-1. Format of new TOVS Sounding Product data record.
Word # Contents
1 Satellite Identification (see Section 2.0.2)
2 Year x 256 + Month
3 Days x 256 + Hours (UTC)
4 Minutes x 256 + Seconds
5 Latitude (degrees x 100, +N, -S, range: +90 to -90)
6 Longitude (degrees x 100, +E, -W, range: +180 to -180)
7 Solar Zenith Angle (degrees x 100, range: 0 to 90, 90=night)
8 Land/Sea indicator (if land, surface elevation in meters; if sea, zero)
9 Surface temperature (K x 10)
10 Estimated pressure at base of sounding (mb x 10)
11 Instrument and/or channel combination
12 Retrieval method
13 Standard deviation for low-level channel (K x 100, range: 0 to 10)
14 Standard deviation for medium-level channel (K x 100, range: 0 to 10)
15 Mean value of N* (N* x 1000, range: 0 to 1)
16 Superswath, box and minibox counter (superswath x 1000 + box counter x 10 + minibox)
17 Sea Surface Temperature (K x 10)
18 Day of month and hour (UTC) when edit flag is written to file (Day x 256 + hour)
19 Minute and second when edit flag is written to file (Minute x 256 + seconds)
20 TOVS filter flag (range: 0 to 3)
21-22 Spare
Words 23-82 are Layer mean temperatures
23 Pressure at low boundary (mb x 10)
24 Pressure at upper boundary (mb x 10)
25 Layer-mean temperature (K x 10)
26 Quality (K x 10)
27-82 Same as words 23-26 for the other 14 layers
Words 83-94 are for layer precipitable water:
83 Pressure at lower boundary (mb x 10)
84 Pressure at upper boundary (mb x 10)
85 Layer precipitable water (mm)
86 Quality (percent)
87-94 Same as words 83-86 for the other 2 layers
95 Tropopause pressure (mb x 10)
96 Tropopause temperature (K x 10)
97 Tropopause quality (percent)
98 Spare
99 Total ozone amount (Dobson units)
100 Total ozone quality (percent)
101 Cloud cover pressure (mb x 10)
102 Cloud cover amount (percent)
Words 103-130 are for sounding radiances:
103-121 Equivalent Blackbody temperatures used to obtain HIRS/2 Ch. 1-19, (K x 64)
122 Equivalent Blackbody temperatures used to obtain HIRS/2 Ch. 20, (K x 16)
123-126 Equivalent Blackbody temperatures used to obtain MSU Ch. 1-4 (K x 64)
127-129 Equivalent Blackbody temperatures used to obtain SSU Ch. 1-3 (K x 64)
130 Spare
131 Stability departure
132 Stability departure time difference
133-139 Spares
140 End of report indicated by "8888"
Note: A value of N* between 0 and 1000 indicates that the N* method was used. A value of "7777" indicates missing data and implies that the HIRS radiances are completely clear and the N* method was not used. A value of "9211" means that the HIRS radiances are completely cloudy and the user should check the channel combination flags to determine which channels were used.

Weekly (on a seven day cycle beginning with Monday), sounding data for two satellites are written onto an IBM 3480 cartridge (which is full). SSB has selection software for this version of the TOVS Sounding Product. Digital data may be selected from the TOVS Sounding Product tapes by time and/or area. Each record that contains any of the selection criteria will be included in its entirety.

The instrument and/or channel combination described in word 11 is represented by ICC, which is defined by Equation 5.1.2-1:

ICC=4096 Z + 256 Y + 16 X + 4W + V

Table 5.1.2-2 describes the variables that comprise ICC.

Table 5.1.2-2. Variables that Comprise ICC.
V denotes channel combinations used to obtain layer precipitable water for the layers: surface to 700 mb, 700 to 500 mb, and 500 to 300 mb. V can have the following values:
0 No Retrieval
1 HIRS + MSU
2 HIRS
W denotes channel combinations used to obtain tropopause temperature and pressure. W can have the following values:
0 No Retrieval
1 HIRS' + MSU
2 MSU
X denotes channel combinations used to obtain total ozone. X can have the following values:
0 No Retrieval
1 HIRS (1, 2, 3, 8, 9, 10) + MSU (4)
2 HIRS (1, 2, 3, 8, 9, 10)
3 HIRS (1, 2, 3, 9, 10) + MSU (4)
4 HIRS (1, 2, 3, 9, 10)
Y denotes channel combinations used to obtain mean temperature for the layers: surface to 850 mb, 850 to 700 mb, 700 to 500 mb, 500 to 400 mb, 400 to 300 mb, 300 to 200 mb, 200 to 100 mb. Y can have the following values:
0 No Retrieval
1 HIRS + MSU
2 HIRS' + MSU
3 HIRS
4 MSU
5 HIRS' + MSU + SKINTK (ocean only)
6 MSU + SKINTK (ocean only)
Note: SKINTK is no longer used, so Y values of 2 and 5 are identical, as well as Y values of 4 and 6.
Z denotes channel combinations used to obtain mean temperature for the layers; 100 to 70 mb, 70 to 50 mb, 50 to 30 mb, 30 to 10 mb, 10 to 5 mb, 5 to 2 mb, 2 to 1 mb,

1 to 0.4 mb. Z can have the following values:

0 No Retrieval
1 HIRS' + SSU + MSU (3,4)
2 HIRS' + MSU (3,4)
3 SSU + MSU (3,4)
4 HIRS' + SSU
5 HIRS'
6 MSU (3,4) (Note: If Z = 6, output will terminate at 10 mb, since there is no meaningful information above that level in this case.)
Note that HIRS' is equivalent to HIRS/2 Channels 1 (669 cm-1), 2 (679 cm-1), 3 (690 cm-1), and 17 (2358 cm-1).

The retrieval method in word 12 is indicated by MR, where MR is defined by Equation 5.1.2-2:

MR=256 X + 16 Y + Z

Table 5.1.2-3 describes the variables that comprise MR.

Table 5.1.2-3. Variables that comprise MR.
X indicates: 0 No HIRS/2 data.
1 Clear radiances are derived from completely clear spots.
2 Clear radiances are derived from the N* method.
Y indicates: 0 No HIRS/2 data.
1 All HIRS/2 channels were used.
2 Tropospheric HIRS/2 channels were unusable due to clouds only stratospheric channels were used.
Z indicates: 0 Statistical retrieval method was used.
1 Minimum information retrieval was used.
2 Minimum information retrieval attempted, but statistical retrieval used.
3 No HIRS.

Radiometric information are supplied in the form of equivalent blackbody (or "brightness") temperatures (K). These can be transformed into radiances (mW/(m2-sr-cm-1)) by application of the Planck function. Parameters required for such transformation such as channel central wave numbers and temperature adjustment coefficients are spacecraft instrument dependent and are supplied in respective subsections of Section 1.4 for each satellite.

Spares are indicated by "666610", missing or undefined values are indicated by "777710" and end of report is denoted by "888810".

The TOVS filter flag in word 20 indicates that one of the following conditions is true:

0 Sounding is good
1 Sounding is redundant

Word 17 contains the sea surface temperature value when over ocean and the skin temperature when over land.

5.1.2.1 TOVS Sounding Quality Information File (March 9, 1992 - Present)

The format of the TOVS sounding quality information file is drastically different from the old format. This file is an on-going accumulation of weekly statistics for the TOVS Sounding Product. It contains a housekeeping record which has 720 INTEGER*4 words. The first five words of the housekeeping record are contained in Table 5.1.2.1-1. Words 1 and 3 are the same since the entire file is archived. Word 4 is set to a value of two since the housekeeping record is the first record on the file and the directories follow.

Table 5.1.2.1-1. Format of the housekeeping record.
Word # (I*4) Contents
1 Total number of records on file
2 Last record updated
3 Last record updated at time of archive
4 Beginning record number of directory
5 Beginning record number (i.e., start of data for first directory)
6-720 Zero filled

The directories are dimensioned as 12 x 120 INTEGER*2 arrays. The number of records containing directories can be computed from the housekeeping record: word 5 minus word 4. Table 5.1.2.1-2 contains the format of the directories.

Table 5.1.2.1-2. Format of the directory record.
Word # (I*2) Contents
1 Starting month of data period (period is a week)
2 Starting day
3 Starting year
4 Ending month of data period
5 Ending day
6 Ending year
7 Plot title code
8 Data type code
9 Satellite ID
10 Record number of data in file
11-12 Spares ("FFFFZ")
13-1440 Repeat of word numbers 1-12 for each directory record

Table 5.1.2.1-3 contains the format for the quality information record. Each week, one record of data is written for each operational spacecraft for the clear, N* and cloudy cases. Therefore, in one week for two operational satellites, six data records are created. The data records contain 1440 INTEGER*2 words. Each record contains data for six latitude zones: 90N-60N, 60N-30N, 30N-0, 0-30S, 30S-60S, and 60S-90S. Unless otherwise specified when ordering, the current quality information file will not be included with a customer's order of TOVS Sounding Product data.

Table 5.1.2.1-3. Format of new TOVS quality information record.
Word # (I*2) Description Scaling
1-240 Zone 1 (90N-60N)
1-120 Retrieval temperature (at 30 TOVS levels)
1-30 Mean (K) 1000
31-60 Root mean square 1000
61-90 Standard deviation 1000
91-120 Sample size 1000
121-176 Thickness (at 14 levels)
121-134 Mean (m) 100
135-148 Root mean square 100
149-162 Standard deviation 100
163-176 Sample size 100
177-180 Tropopause temperature
177 Mean (K) 1000
178 Root mean square 1000
179 Standard deviation 1000
180 Sample size 1000
181-184 Tropopause pressure
181 Mean (mb) 100
182 Root mean square 100
183 Standard deviation 100
184 Sample size 100
185-240 Mixing ratio at 14 levels
185-198 Mean (g/kg) 1000
199-212 Root mean square 1000
213-226 Standard deviation 1000
227-240 Sample size 1000
241-480 Zone 2 (60N-30N). Format same as Zone 1.
481-720 Zone 3 (30N-0). Format same as Zone 1.
721-960 Zone 4 (0-30S). Format same as Zone 1.
961-1200 Zone 5 (30S-60S). Format same as Zone 1.
1201-1440 Zone 6 (60S-90S). Format same as Zone 1.

5.1.3 RTOVS Sounding Product (October 22, 1997 - present)

RTOVS is the Revised TOVS system introduced on October 22, 1997 for NOAA-14 processing and on November 19, 1997 for NOAA-11 processing. RTOVS uses the same instruments as TOVS but newer software designed to easily transition to the new NOAA KLM series of satellites. The format of the archive has been kept the same as TOVS (fully described in Section 5.1.2.), to ease data processing by users, although the data volume is higher. There is a parallel archive of TOVS and RTOVS data through May 1998.

Amended March 7, 2001
Amended April 11, 2001



Previous Section Top of Page Next Section