SST

Sea surface temperature from AVHRR data


EnvironmentsPYTHON :: EASI :: MODELER
Quick linksDescription :: Parameters :: Parameter descriptions :: Details :: Examples :: Algorithm :: References :: Related

Back to top

Description


Computes the sea surface temperature using the three AVHRR thermal bands (channels 3, 4, and 5) and AVHRR sensor calibration information. This program uses the McClain method for atmospheric correction. For AVHRR disk formats, use FIMPORT to create a PCIDSK file containing the AVHRR bands and a text segment with the required sensor calibration information. The input image is expected to be either 2048 or 409 pixels wide (HRPT/LAC data or GAC data, respectively).
Note: SST currently supports only NOAA-9, NOAA-12, and NOAA-14 satellite data.
Back to top

Parameters


Name Type Caption Length Value range
FILI * String Input file name 1 - 192  
FILO * String Output file name 1 - 192  
AVHRRCHA * Integer Input thermal channels 3 - 3  
DBOC * Integer Output sea temperature channel 1 - 1  
AVHRRSEG Integer AVHRR calibration/orbit segment 0 - 1  
SENSFILE * String AVHRR sensor information file 1 - 192  
MCCLAIN Float McClain coefficients 0 - 4  
REPORT String Report mode 0 - 192 Quick links
MONITOR String Monitor mode 0 - 3 ON, OFF
Default: ON

* Required parameter
Back to top

Parameter descriptions

FILI

Specifies the name of the input file that contains the raw AVHRR image data (AVHRR thermal channels 3, 4, 5) and the AVHRRSEG text segment. The input file must be a PCIDSK file (created using FIMPORT) or an actual AVHRR disk file.

If the input and output file are the same, the input file must be a PCIDSK file (it cannot be an AVHRR disk file).

FILO

Specifies the name of the PCIDSK file which to receive the sea surface temperature results.

The output file must be specified and it must already exist; SST does not create a new output file. The output file can be the same as the input file, as long as the file type is PCIDSK.

AVHRRCHA

Specifies the three channels in the input file containing the AVHRR thermal channels (3,4,5).

For example, if AVHRR channels 3, 4, and 5 were read into channels 1, 2, and 3 of the input file by FIMPORT, specify this parameter as follows:
AVHRRCHA=avhrr3,avhrr4,avhrr5

DBOC

Specifies the output channel to receive the sea surface temperature information.

It is recommended that this channel be a 32-bit real channel to preserve maximum accuracy. It is possible, however, to use 8-bit and 16-bit channels, keeping the following in mind:

Temperatures less than -4.1 are scaled to 0. Temperatures greater than 21.4 are scaled to 255. A pseudocolor segment is also automatically generated so that the 8-bit sea surface temperature results can be viewed in a visually pleasing manner. See Pseudocolors in the Details section for more information.

AVHRRSEG

Specifies the text segment that contains the AVHRR calibration/orbital data. This text segment is automatically created when FIMPORT is used to read in an AVHRR image. If this parameter is not specified, SST will try to find a valid AVHRR text segment in the input file.

This text segment (type 140) contains image-specific information:

You can change the values in this segment using the TEXWRIT and TEXREAD functions. Be careful to maintain exactly the same format as the original segment.

For a detailed description of the AVHRRSEG text segment format, refer to AVHRRAD documentation.

SENSFILE

Specifies the satellite-specific sensor file. This is a text file containing the following information:

For NOAA-9/12/14, this sensor file is supplied by PCI Geomatics and resides in the "/etc" directory of your CATALYST Professional installation. For NOAA-14, the PCI-supplied sensor file is named "sstnoa14.dat".

By default, SST will look for the specified sensor file in the current working directory. If the file is not present, SST will look for the specified sensor file in the PCI etc directory.

MCCLAIN

Specifies the Multi-Channel Split coefficients. If this parameter is not specified, SST uses the values extracted from the sensor information file (SENSFILE).

See McClain Coefficients in the Details section for more information.

REPORT

Specifies where to direct the generated report.

Available options are:

MONITOR

The program progress can be monitored by printing the percentage of processing completed. A system parameter, MONITOR, controls this activity.

Available options are:

Back to top

Details

SST computes the sea surface temperature, given 8-bit AVHRR data from the NOAA-9 satellite or 10-bit AVHRR data from the NOAA-12 or NOAA-14 satellite. The image data and calibration/orbital segments are created when FIMPORT is used to import a 10-bit AVHRR image. SST uses the McClain method for atmospheric correction.

SST computes the Celsius temperature for every pixel in an image, regardless of whether it represents land or water; however, only the temperature values for oceans will be precise. Temperature values for land, cloud, or inland water bodies can be used as rough indicators if necessary, but, in general, these should be ignored.

For mid-ocean sea surface temperatures, SST typically provides approximately 0.5 degree Celsius accuracy, given ideal data sets. Temperature accuracy is generally dependent on atmospheric conditions and land proximity. Water vapor content (the single most important factor), pollution, and strong winds are some of the conditions that can affect temperature accuracy. Temperature calculations near land are also of lesser accuracy due to the high land temperatures and extra atmospheric pollutants. Althought the McClain coefficients can be used to compensate for these conditions, it is often difficult or impossible to get precise McClain coefficients for a specific scene or set of circumstances.

The default McClain coefficients used by SST are for mid-ocean, sunny sky, low-wind conditions. This is a reasonable set of assumptions, because this is usually the only set of atmospheric conditions that allow you to see the ocean surface.

For 10-bit AVHRR data, SST expects the input image to be 2048 or 409 pixels wide (LAC/HRPT or GAC data, respectively). Any other image widths will result in an error.

McClain Coefficients

The four parameters A, B, C, and D specify the McClain coefficients (that is, the Multi Channel Split coefficients) used in the MCSST Split Equation:

       Temperature (Celsius) =
          A * (Temp. layer 4 - Temp. layer 5) +
          B * (Temp. layer 4) +
          C * ((Temp. layer 4 - Temp. layer 5) *
               (Secant of Satellite Zenith Angle - 1)) +
          D

For NOAA-14 data, the following McClain coefficients are used in the MCSST Split Equation, as specified in the NOAA Polar Orbiter Data User's Guide page E-33:

                        A       B        C        D
 Day coefficients:   2.139588 1.017342 0.779706 -278.43
 Night coefficients: 2.275385 1.029088 0.752567 -282.24

For NOAA-12 data, the following McClain coefficients are used in the MCSST Split Equation, as specified in the NOAA Polar Orbiter Data User's Guide page E-31:

                        A       B        C        D
 Day coefficients:   2.579211 0.963563 0.242598 -263.006
 Night coefficients: 2.384376 0.967077 0.480788 -263.94

For NOAA-9 data, the first three McClain coefficients are multipliers for the temperature values of AVHRR layers 3, 4, and 5. Note that the three input temperatures are expressed in degrees Kelvin, whereas the output temperature is in degrees Celsius (NOAA Polar Orbiter Data User's Guide, page E-11).

                       A     B       C      D
 Day coefficients:     0, 3.6569, -2.6705,-268.92
 Night coefficients:   0, 3.6836, -2.69,  -270.42

NOAA-9 Window Split equation:

         Temperature = A * Temperature (Kelvin) of Layer 3 +
         (Celsius)     B * Temperature (Kelvin) of Layer 4 +
                       C * Temperature (Kelvin) of Layer 5 +
                       D

For NOAA-9 data, it is possible to exclude a channel from the calculations (for example, if the channel is noisy) by setting the corresponding McClain coefficient to zero.

By default, McClain coefficients are extracted from the specified satellite sensor information file (SENSFILE) but you can specify different values.

The McClain coefficients vary slightly according to ocean location and for large inland water bodies; however, the default values are valid for almost all ocean bodies. Other McClain coefficient values can be obtained from various research papers.

Pseudocolors

If the specified output channel is an 8-bit channel, a pseudocolor segment is automatically created, allowing you to visualize each degree of temperature. After displaying the SST output channel in CATALYST Professional Focus, you can load the pseudocolor segment to display the temperatures.

The pseudocolor segment contains the following information:

          Temperature    Gray      Color            R   G   B
           (Celsius)     Level
        
        < -0.6             < 35    white           255 255 255
          -0.6 -  0.3    35- 44    dark blue         0   0 130
           0.4 -  1.3    45- 54    blue              0   0 255
           1.4 -  2.3    55- 64    light blue        0 100 255   
           2.4 -  3.3    65- 74    aqua              0 255 255   
           3.4 -  4.3    75- 84    light green       0 255 160   
           4.4 -  5.3    85- 94    green             0 255   0    
           5.4 -  6.3    95-104    dark green        0 127   0   
           6.4 -  7.3   105-114    very dark green   0  95   0   
           7.4 -  8.3   115-124    brown           127 127   0   
           8.4 -  9.3   125-134    dark brown      160 150   0   
           9.4 - 10.3   135-144    yellow          255 255   0   
          10.4 - 11.3   145-154    orange          255 127   0   
          11.4 - 12.3   155-164    red             255   0   0   
          12.4 - 13.3   165-174    dark red        127   0   0
          13.4 - 14.3   175-184    very dark red   105   0   0
        > 14.4             >184    dark gray        60  60  60
Back to top

Examples

Create a sea surface temperature map from a NOAA-14 data set. Use the provided sensor file. All three channels of thermal data must be used to calculate the temperature. Print the report to the terminal. The output results will be written to the same file.

EASI>fili	=	'noaa14.pix'	! input file
EASI>filo	=	'noaa14.pix'	! use input file
EASI>avhrrcha	=	3,4,5	! input channels with AVHRR bands
EASI>dboc	=	1	! output channel
EASI>avhrrseg	=		! SST will look for AVHRR text segment
EASI>sensfile	=	'sstnoa14.dat'	! AVHRR sensor information file
EASI>mcclain	=		! default, use values in sensfile

EASI>run SST

Create a sea surface temperature map using only channel 5 from a NOAA-9 data set (by setting appropriate McClain coefficients to zero). Write the results to a second file.

EASI>fili	=	'noaa9.pix'	! input file
EASI>filo	=	'sst.pix'	! output file
EASI>avhrrcha	=	3,4,5	! input channels with AVHRR bands
EASI>dboc	=	1	! output channel
EASI>avhrrseg	=		! SST will look for AVHRR text segment
EASI>sensfile	=	'sstnoa9.dat'	! AVHRR sensor information file
EASI>mcclain	=	0, 0, 1.2, -268.92	! McClain coefficients

EASI>run SST
Back to top

Algorithm

SST extracts information from the AVHRRSEG text segment to calibrate the thermal channels. It uses platinum resistance temperature counts, internal target counts, and space counts to determine the average temperature of the internal blackbody target.

SST also reads other satellite and sensor-specific data from the file sstnoaxx.dat, where xx is the satellite number. The first few lines contain the coefficients used to compute the internal target temperature, followed by the McClain coefficients (to correct for atmospheric effects). These coefficients are followed by the information for each AVHRR thermal channel.

SST computes the GAIN and the INTERCEPT for each thermal channel to obtain a correspondence between gray-level values and radiance. After the table of gray level-to-radiance mappings is created, the radiances are transformed to temperatures using the inverse Planck function, then corrected for the non-linearity of the sensor. Finally, if the input data is 8-bit, the 10-bit information is compressed to 8-bit with the LUTs obtained from the input file's AVHRRSEG text segment.

Back to top

References

Condal, A.R., Le, H.V., Automated computer monitoring of sea-ice temperature by use of NOAA satellite date, 8th Canadian Symposium in Remote Sensing, Montreal, May 1983, 145-150, 1983.

McClain, E.P., Multiple atmospheric-window techniques for satellite-derived sea surface temperatures, Oceanography from space, J.F.R. Gower (ed.), Plenum Press, New York, 1981.

Mouchot, M.C. et Lambert, E., Exemples de restitution de la température de surface de l'océan à partir des données AVHRR de NOAA-9. Centre canadien de télédétection, Energie, Mines et Ressources Canada, 1986.

© PCI Geomatics Enterprises, Inc.®, 2026. All rights reserved.