SARSIGM

Description

Generates a calibrated radar backscatter image from a RADARSAT or ENVISAT ASAR image.

Parameters

Name	Type	Length	Value range
Input: Input raster channel *	Raster port	1 - 2
InputSAROffset: Input SAR offset segment *	Array port	1 - 1
InputSARGain: Input SAR gain segment *	Array port	1 - 1
InputSARIncAngle: Input SAR incident angle segment *	Array port	1 - 1
InputOrbit: Input satellite orbital ephemeris segment *	ORB port	1 - 1
InputGCP: Input GCP segment	GCP port	0 - 1
Output: Output backscatter image channel	Raster port	0 - 1
OutputOrbit: Output satellite orbital ephemeris segment	ORB port	0 - 1
OutputGCP: Output GCP segment	GCP port	0 - 1
Amplitude, Decibels, or Power	String	0 - 1	AMP | DB | POW Default: DB

* Required parameter

Parameter descriptions

Input: Input raster channel

For ASAR Alternating Polarization images, only one polarization can be processed at a time.

For Single Look Complex (SLC) images, two channels must be specified.

If two channels are specified, the first channel is taken as the "I" (in-phase or real) channel and the second as the "Q" (quadrature or imaginary) channel of an SLC image. For ASAR Alternating Polarization SLC images, the two specified channels must have the same polarization.

InputSAROffset: Input SAR offset segment

Specifies the segment that contains the input SAR Offset array, including the offset value to use to convert image pixel values.

This segment is created when the satellite image is read from CD using CDSAR or CDASAR. It is typically labeled "SAR Offset" or "A0Scale".

InputSARGain: Input SAR gain segment

Specifies the segment that contains the input SAR Gain array, consisting of a gain scaling table use to convert image pixel values.

This segment is created when the satellite image is read from CD using CDSAR or CDASAR. It is typically labeled "SAR Gain" or "AScaling".

For ASAR Alternating Polarization images, the input image will contain two SAR Gain channels, with the second layer identified as "SAR Gain, second polarization". Select the array channel that matches the polarization of the specified image layer(s) for processing.

InputSARIncAngle: Input SAR incident angle segment

Specifies the segment that contains the input table of incident angles.

For Radarsat images, this segment is created by running SARINCD prior to running SARSIGM. For ASAR images, the segment is created by CDASAR when the image is imported.

InputOrbit: Input satellite orbital ephemeris segment

Specifies the input segment that contains the satellite orbital ephemeris data.

InputGCP: Input GCP segment

Specifies the input segment that contains the ground control points.

Output: Output backscatter image channel

Specifies the output image channel to receive the generated backscatter image.

If the output file is to be created, the image is output to channel one and this parameter is ignored.

If the output file already exists and this parameter is specified, the output channel must be 32-bit real.

If the output channel is the same as the specified input channel (DBIC) of the same image file, the output image is processed correctly and overwrites the input image. This is not advised unless the input image channel is 32-bit real.

OutputOrbit: Output satellite orbital ephemeris segment

Specifies the output segment to receive the satellite orbital ephemeris data.

OutputGCP: Output GCP segment

Specifies the output segment to receive the ground control points.

Amplitude, Decibels, or Power

Specifies whether to output amplitude, power or decibel (dB) values.

Data may be converted from decibels to amplitude or power values (or from amplitude or power values to decibels) using the modeling functions or the MODEL function. For more information, see the Algorithm section.

Details

SARSIGM generates a calibrated radar backscatter channel from the input scaled radar channel (RADARSAT or ENVISAT ASAR) and an array of incident angles. For RADARSAT images, the array must be produced by the SARINCD function; for ASAR images, it is automatically created during import by CDASAR.

The input contains the scaled radar image, the offset and gain information stored in array segments (SAR Offset, SAR Gain), and the array of incident angles to use as input.

For ASAR Alternating Polarization images only one polarization can be processed at a time.

Algorithm

Decibels can be converted to amplitude values using the following model:

 model
     %1 = (exp10( %2 / 10.0 ))^0.5
 endmodel

Amplitude values can be converted to decibels using the following model:

 model
     %2 = 10.0 * log10( (%1)^2 )
 endmodel

Amplitude values can be converted to power values using the following model:

 model
     %2 = %1^2
 endmodel

Power values can be converted to amplitude values using the following model:

 model
     %2 = %1^0.5
 endmodel

Overview

When RADARSAT and ASAR data are written to media in 8 or 16-bit form for distribution, output scaling gain and offset are applied to the data to ensure optimum utilization of the available dynamic range. The scaling used can vary for each scene, making it difficult to directly relate information between scenes (that is, radiometric and temporal signatures).

A group of PCI functions and procedures are available to convert distributed RADARSAT and ASAR digital image data to radar brightness (Beta naught) and calibrated radar backscatter (Sigma naught). These conversions yield a 32-bit real image which is radiometrically calibrated and which more accurately represents the original signal amplitudes. These functions are outlined in Program Flow, below.

In addition to the conversion functions, a suite of utility functions are available to convert between ASCII text and array segments (ARRREAD, ARRWRIT), generate an array segment from row/column data (RCSTATS), and create a visible (detected) channel from single-look complex data (SLC2IMG).

Program flow

There are ten functions and procedures associated with the conversion of RADARSAT and ASAR digital images to radar brightness (Beta naught) and calibrated radar backscatter (Sigma naught). Five of these (SARBETA, SARINCD, SARSIGM, and SLC2IMG) are contained in the Radar Analysis component. The ARRREAD, ARRWRIT, and RCSTATS functions are contained in the Kernel component. The CDASAR function is contained in the Satellite Ortho component.

The following diagram illustrates the normal function flow for this group of functions:

                CDASAR (CD)   - read CDASAR data to a PCIDSK file
                  |
                  |
         --------------------
        |                    |
   (brightness)        (backscatter)
     SARBETA              SARINCD    - not required for ASAR
                             |
                          SARSIGM

The remaining functions are utilities which can be run any time, and do not fit into any natural flow.

Array segments

When images are read using CDSAR, two array segments may be appended to the output file that is created. For ASAR images, CDASAR is used, and three array segments are created. For ASAR Alternating Polarization, Moderate Resolution and Wide Swath Mode images, four array segments are created. (The segments are only created if all required information is found in the raw image file.)

For ASAR images, the equivalent information is extracted from image headers and auxiliary Annotation Data Sets, or derived in CDASAR from other available values.

For RADARSAT images, the fields for the second array segment "SAR Gain" or "AScaling" are the output scaling gain table read from Bytes 89 - 8265 of the "Radiometric Data" record. This table is expanded by CDSAR to the number of pixels in the scanline before it is output to the segment. The expanded table is used in SARBETA and SARSIGM. A value is appended to the end of the table, specifying the polarization of the corresponding image channel(s). The value 11 represents H/H polarization, 12 represents H/V, 21 represents V/H, and 22 represents V/V.

For ASAR images, the scaling values are constant for all pixels. To maintain consistency with other modules, however, the constant-valued scaling arrays are created for ASAR images as well by CDASAR. For ASAR Alternating Polarization images, two SAR Gain arrays are created, one for each polarization. The polarization corresponding to each array can be established by verifying the last value of the array, as described above.

For ASAR images, the IncidAng segment is also created in CDASAR, from incident angle values provided for tie point lines and pixels in the Geolocation Grid ADSRs. The incident angle values at tie points are interpolated to full lines, then averaged across all tie lines to arrive at a single average incident angle line.

ASAR images may also contain an additional array segment SRGR (Slant Range to Ground Range Coefficients), created by CDASAR. This segment is only created if the input image has more than one set of coefficients (more than one record in the SRGR ADS). This is typically the case for Moderate Resolution and Wide Swath Mode (ScanSAR) images. This array segment is not used in the SARSIGM and SARBETA functions.

Acknowledgements

The "SAR Sigma" functions were added to PCI software through RADARSAT User Development Program (RUDP) funding from the Canada Centre for Remote Sensing (CCRS), Natural Resources of Canada. Thanks are given to Mike Manore at CCRS for his involvement in the project. Special thanks are given to Tom Lukowski at CCRS who was invaluable in verifying the output from the "SAR Sigma" functions SARBETA, SARINCD, and SARSIGM.

References

N.W.Shepherd and Associates, (February 26, 1996). 4.2 Calibration Equation for Detected Products. In Technical Note No. 4.2 CDPF Output Data Calibration (p. 9)

N.W.Shepherd and Associates, (February 26, 1996). 4.3 Calibration Equation for SLC Products. In Technical Note No. 4.2 CDPF Output Data Calibration (p. 9)

The algorithm for computing radar backscatter values for ASAR images is provided on the ESA web site: http://envisat.esa.int/dataproducts/asar/CNTR2-11-5.htm.