Package noaa.coastwatch.tools

Class cwregister

java.lang.Object

noaa.coastwatch.tools.cwregister

public final class cwregister
extends Object

The registration tool resamples gridded earth data to a master projection.

Name

cwregister - resamples gridded earth data to a master projection.

Synopsis

cwregister [OPTIONS] master input output

Options:

-f, --srcfilter=TYPE
-h, --help
-m, --match=PATTERN
-M, --method=TYPE
-O, --overwrite=TYPE
-p, --polysize=KILOMETERS
-r, --rectsize=WIDTH/HEIGHT
-s, --srcexpr=EXPRESSION
-v, --verbose
--version

Description

THIS TOOL HAS BEEN DEPRECATED AND REPLACED WITH CWREGISTER2. IT ONLY EXISTS TO HELP WITH BACKWARDS COMPATIBILITY. USERS SHOULD TRANSITION TO CWREGISTER2 FOR BETTER TIME PERFORMANCE AND ACCURACY.

The register tool resamples gridded earth data to a master projection. A master projection specifies the translation between grid row and column coordinates and earth latitude and longitude coordinates. The master projection file is any valid earth data file from which a set of row and column dimensions and earth transform parameters may be extracted. This includes standard CoastWatch product files as well as master files created using the cwmaster tool.

Parameters

Main parameters:

master

The master projection file name. Note that the master file is not altered in any way. It is simply accessed in order to determine grid and earth transform parameters.

input

The input data file name.

output

The output data file name.

Options:

-h, --help

Prints a brief help message.

-f, --srcfilter=TYPE

Specifies a filter used to determine whether source pixels at a given location should be used in registration. The only filter type currently supported is 'viirs', which filters out pixels from the VIIRS sensor that have been omitted due to bow-tie overlap. By default all valid source pixels are used in registration. See also the --srcexpr option to filter using an expression.

-m, --match=PATTERN

The variable name matching pattern. If specified, the pattern is used as a regular expression to match variable names. Only variables in the input file matching the pattern will be registered. By default, no pattern matching is performed and all variables are registered.

-M, --method=TYPE

The registration resampling method. Valid methods are 'inverse', 'mixed', and 'direct':

• Inverse - This method divides the destination grid into a series of rectangles of physical size no larger than that specified by the --polysize option (100 km by default), and computes polynomial coefficients for approximating the coordinate transform within each rectangle. Each polynomial approximation computes a source coordinate in the source grid for each destination coordinate in the destination grid rectangle. This is the default method of registration and is recommended when the source earth location data is smooth and continuous such as with AVHRR LAC swath data.
• Mixed - This method divides the source grid into rectangles of a size specified by the --rectsize option (50x50 pixels by default), and computes polynomial coefficients on each rectangle that are used to approximate a source coordinate in the source grid rectangle for each destination coordinate in the corresponding destination grid rectangle. The method follows up with a single pixel interpolation to fill in pixels that fell between destination rectangles bounds. This method is recommended when the source earth location data is discontinuous at regular intervals, such as with MODIS swath data. In this case, the --rectsize option must be used to specify rectangles that are compatible with the discontinuity. For example if the source earth location data is discontinuous at 16 pixel intervals along the row direction, then the rectangle size should be set to 16x16.
• Direct - This method is the simplest and performs a direct lookup of the source coordinate in the source grid for each destination coordinate in the destination grid. This method is available mainly as a comparison for testing the accuracy and speed of the other methods, since it is expected to run more slowly but have the highest accuracy. Note that the direct method cannot currently be used if the source grid contains discontinuous earth location data (such as MODIS swath data).

-O, --overwrite=TYPE

ADVANCED USERS ONLY. Sets the overwrite policy for 'mixed' mode resampling: either 'always' (the default), 'never', or 'closer'. If during resampling, more than one source pixel maps to a single destination pixel, this option is used to determine if the new value should overwrite the old value. By default, the new value always overwrites the destination pixel (the 'always' mode). If 'never' is specified, the first written value is never overwritten. If 'closer' is specified, the destination pixel is only overwritten if the source pixel is closer in physical location to the destination than any previous source pixel.

-p, --polysize=KILOMETERS

The polynomial approximation rectangle size in kilometers. This option is only used by the inverse resampling method (see the --method option). The inverse resampling method employs a polynomial approximation to speed up the calculation of data locations. The polynomial rectangle size determines the maximum allowed size of the resampling rectangles in the destination. The default polynomial size is 100 km, which introduces an error of less than 0.15 km for AVHRR LAC data.

-r, --rectsize=WIDTH/HEIGHT

The polynomial approximation rectangle size in pixels. This option is only used by the mixed resampling method (see the --method option). The mixed resampling method employs a polynomial approximation to speed up the calculation of data locations. The polynomial rectangle size determines the exact dimensions of the resampling rectangles in the source. The default polynomial rectangle size is 50/50, which introduces only a small error for AVHRR LAC data.

-s, --srcexpr=EXPRESSION

Specifies that an expression should be used to determine if pixels from the source should be used in registration. If the result of the expression is true (in the case of a boolean result) or non-zero (in the case of a numerical result), the source pixel at the given location is used, otherwise it is ignored. The syntax for the expression is identical to the right-hand-side of a cwmath expression (see the cwmath tool manual page). By default all valid source pixels are used in registration.

-v, --verbose

Turns verbose mode on. The current status of data conversion is printed periodically. The default is to run quietly.

--version

Prints the software version.

Exit status

0 on success, > 0 on failure. Possible causes of errors:

• Invalid command line option
• Invalid master, input or output file names
• Unsupported master or input file format

Examples

The following shows the registration of NOAA-17 AVHRR channel 2 swath data to a southern California master:

   phollema$ cwregister -v --match avhrr_ch2 ws_master.hdf 2002_318_1826_n17_mo.hdf
     2002_318_1826_n17_ws.hdf

   [INFO] Reading master ws_master.hdf
   [INFO] Reading input 2002_318_1826_n17_mo.hdf
   [INFO] Creating output 2002_318_1826_n17_ws.hdf
   [INFO] Adding avhrr_ch2 to resampled grids
   [INFO] Found 1 grid(s) for resampling
   [INFO] Resampling 4788x2048 to 1024x1024
   [INFO] Creating location estimators
   [INFO] Computing row 0
   [INFO] Computing row 100
   [INFO] Computing row 200
   [INFO] Computing row 300
   [INFO] Computing row 400
   [INFO] Computing row 500
   [INFO] Computing row 600
   [INFO] Computing row 700
   [INFO] Computing row 800
   [INFO] Computing row 900
   [INFO] Computing row 1000
   [INFO] Closing files
 

Since:

3.1.2

Author:

Peter Hollemans

Method Summary

Modifier and Type	Method	Description
static void	main(String[] argv)	Performs the main function.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

main

public static void main(String[] argv)

Performs the main function.

Parameters:

argv - the list of command line parameters.