WATEXT

Determine pixels in an image that correspond to water


EnvironmentsPYTHON :: EASI
Quick linksDescription :: Parameters :: Parameter descriptions :: Details :: Example :: Acknowledgements :: Related

Back to top

Description


WATEXT determines the pixels in the input raster that correspond to water based on backscatter thresholding of the HH channel intensity. This function supports only C-Band RADARSAT-2 data.
Back to top

Parameters


Name Type Caption Length Value range
FILI * String Input SAR image 1 - 192  
FILO * String Output classified image 1 - 192  
WNSPD Float Wind speed 0 - 1  
WNDIR Float Wind direction 0 - 1  
MONITOR String Monitor mode 0 - 3 ON, OFF
Default: ON

* Required parameter
Back to top

Parameter descriptions

FILI

Specifies the name of the RADARSAT-2 data set that contains the input matrix data.

FILO

Specifies the output image file to receive the classified pixels.

The file must not already exist. The output file has 1 channel.

WNSPD

Optionally specifies, in m/s, the average wind speed at the time of image acquisition. If this parameter is specified, WNDIR must also be specified.

If this parameter is specified, the HH backscatter intensity threshold value will be based on this value rather than on a maximum-likelihood statistics threshold.

WNDIR

Optionally specifies, in degrees from North ranging from -180 to +180, the average wind direction at the time of image acquisition. If this parameter is specified, WNSPD must also be specified.

If this parameter is specified, the threshold value will be based on this value rather than on a maximum-likelihood statistics threshold.

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

WATEXT automatically identifies bodies of water contained in the input SAR image. The function requires as input SAR C-band data calibrated to sigma nought with horizontal transmit and horizontal receive (HH) data. The specular return from open water, under calm conditions, produces areas of very low backscattering in SAR imagery. Wind can cause an increase in the amount of radar backscatter due to reflections from capillary waves. The amount of backscattering from water at various wind speeds and incident angles has been modeled for C-band data.

Based on the image histogram, an initial estimate of the land/water backscattering threshold and wind speed is determined. The nominal wind speed is used to adjust the threshold values for incident angle. The 'water body mask' is determined by identifying pixels with a backscattering value below the threshold for the calculated wind speed and incident angle.

WATEXT requires calibrated C-Band HH data from the RADARSAT-2 sensor as input. The input data may be in either complex or detected format. In general, the best results are achieved when the image is acquired during low wind speed conditions. A water body mask extracted from scenes with high wind conditions and with a steep incidence angle range or covering mountains (shadows) should be interpreted with care.

Note: WATEXT does not verify the polarization of the radar band and calibration type. Only C-band data calibrated to sigma nought should be specified as input.

The input file must contain an incidence angle array segment and must be calibrated to sigma0 on ingest (using SARINGEST). It is recommended that the input file be filtered using PSBOXCAR, PSPOLFIL, or SAR filters for detected products for better results.

Back to top

Example

Determine the water pixels of a RADARSAT-2 image using the maximum-likelihood calculation.

        EASI>FILI = "rsat2_s4c.pix"
        EASI>FILO = "rsat2_water.pix"
        EASI>WNSPD =
        EASI>WNDIR =

        EASI>run WATEXT
      
Back to top

Acknowledgements

PCI Geomatics gratefully acknowledges the financial support provided by the Canadian Space Agency through the Earth Observation Application Development Program (EOADP), contract number 9F028-034946.

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