autoradcorr

Description

Calculates a radiance correction for all 50x50 pixel boxes in the scene that meet the following criteria:

1. greater than 3.0 degree standard deviation in temperature
2. temperature of pixels must be greater than 225K
3. box must contain 50 percent usable pixels
4. box cannot overlap areas containing ghost signal
5. radiation correction for a box cannot have negative correction values

Boxes that meet all above criteria are given a Quality of 2 in the quality mask. Quality 1 boxes ignore criterium 5 - this allows for water cloud corrections.

Arguments and Return Values

Returns a structure containing:

• .avgdark 1x1x10 average radiance correction for entire image
• .blackmask XxYx1 mask of used pixels in calculations
• .dark (X/50)x(Y/50)x10 radiance corrections for every 50x50 box
• .darkcol 1xYx10 correction for entire image

Usage

Syntax: rc = autoradcorr(data [, bandlist = INT] [, b1 = INT] [, b2 = INT] [, sd = FLOAT] [, tempmin = FLOAT] [, pix_fraction = FLOAT] [, first_image = INT] [, ignore = FLOAT])

 'data'         is the full projected hdf.  You MUST supply this!
 'bandlist'     is the ordered list of THEMIS bands in the cube. Default is 1-10.
 'b1 and b2'    are the first and last z values to search for maxbtemp. Default is 3 and 9.
 'sd'           is the standard deviation of the temperature.  Default is 3K.
 'tempmin'      is the minimum brightness temperature.  Default is 225K.
 'pix_fraction' is the fraction of usable pixels in a box.  Default is 50 percent.
 'first_image'  = 1 is first 3600-line chunk of single data cube, = 0 is not. Default is 1.
 'ignore' is null value of the black space.  Default is -32768.

Examples

dv> a
345x1000x10 array of float, bsq format [13,800,000 bytes]

dv> emissivity_before = thm.themis_emissivity(a)
.struct, 2 elements
    emiss: 345x1000x10 array of float, bsq format [13,800,000 bytes]
    maxbtemp: 345x1000x1 array of float, bsq format [1,380,000 bytes]

dv> rc = autoradcorr(a)
...............................
struct, 4 elements
    avgdark: 1x1x10 array of float, bsq format [40 bytes]
    blackmask: 345x1000x1 array of byte, bsq format [345,000 bytes]
    dark: 7x20x10 array of float, bsq format [5,600 bytes]
    darkcol: 1x1000x10 array of float, bsq format [40,000 bytes]

dv> a[where a != -32768] = a - rc.darkcol
345x1000x10 array of float, bsq format [13,800,000 bytes]

dv> emissivity_after = thm.themis_emissivity(a)
.struct, 2 elements
    emiss: 345x1000x10 array of float, bsq format [13,800,000 bytes]
    maxbtemp: 345x1000x1 array of float, bsq format [1,380,000 bytes]

Images

The following images show dcs images of the radiance, emissivity before and emissvity after of an image that has been autoradcorr-ed.