Table of contents Previous: Creating an ADDE Dataset Next: Image Remapping

15.1 IR Channel Differencing

The strength of McIDAS as a meteorological data analysis/display packages is it's ability to display and manipulate satellite imagery. In this lesson, you will use the IMGOPER command to create new image products.

IMGOPER generates a new image by applying mathematical functions to data from one or more source images. The following equation computes the data value for each line/element pair in the destination image. This operation is performed repeatedly using data from source image line/element pairs as input values (input) until the entire image is completed.
output data value=  
FUNC[ACON+(MCON*(FORM((COEF1*(OFF1+((SIGN1)*input1)**POW1))
 (COEF2*(OFF2+((SIGN2)*input2)**POW2)) . .
 (COEFn*(OFFn+((SIGNn)*inputn)**POWn)))))]

The part of the equation operating on a single input data value, (COEFn*(OFFn+((SIGNn)*inputn)**POWn)), is referred to as a term. n represents the number of sdataset images. It may not be larger than 100. input1 . . n represents the individual data values from each of the source images. The FORM determines how the terms are combined. For example, if you specify FORM=MULT, the terms are multiplied.

All source images must have the same projection, resolution and navigation. You can use the IMGCOPY or IMGREMAP before using IMGOPER.

  1. If a McIDAS session is not already running, start one with at least 6 frames.

    Type: mcidas

    at the Unix prompt.

Fog Product

The identification of fog and stratus at night is a new application of the GOES Imager data. A night-time image product is made by subtracting the 3.9µm scene temperatures from those at 10.7 µm, and scaling the results to show (-/+) differences. The formula is:

For more information on this product, see the GOES 3.9 µm Channel Tutorial.

Images Needed:

The Unidata/Wisconsin datastream contains half-hourly GOES IR Channel 4 and GOES IR channel 2 images.

  1. Copy today's 9:15 UTC GOES-East IR Channel 4 image to your local dataset.

    Type: DATALOC ADD RTIMAGES ADDE.UCAR.EDU
    Type: IMGCOPY RTIMAGES/GE-IR MYDATA/IMAGES.3000 TIME=9:15 SIZE=ALL

  2. Display the image on frame 1. Reduce the resolution by a factor of 2 so most of it will fit in the frame and draw a map on the image.

    Type: IMGDISP MYDATA/IMAGES.3000 1 MAG=-2 SF=YES REFRESH='EG;MAP'

  3. Copy today's 9:15 UTC GOES-East 3.9 um image to your local dataset.

    Type: IMGCOPY RTIMAGES/GE-39 MYDATA/IMAGES.3001 TIME=9:15 SIZE=ALL

  4. Display the image on frame 2. Reduce the resolution by a factor of 2 so it will fit in the frame and draw a map on the image.

    Type: IMGDISP MYDATA/IMAGES.3001 2 MAG=-2 SF=YES REFRESH='EG;MAP'

  5. Loop between the two images.

    Type: LB 1 2;DR 10
    Press: Alt L

    Note that the two images cover identical geographical regions and are the same size. If they were of different sizes, you would need to remap them to the same projection and size.

  6. Stop the loop.

    Press: Alt L

  7. List out the image directories.

    Type: IMGLIST MYDATA/IMAGES.3000 3001 FORM=BAND

  8. Create a FOG product from the two images. Using the FOG equation and the format above for IMGOPER, 150 would be the additive constant (ACON), 10 would be the multiplicative constant (MCON). Using FORM=ADD, we can use 1 for COEF1 and -1 for COEF2 to create a subtraction of the bands. Since the equation is in terms of temperature, use the UNIT= keyword.

    Type: IMGOPER MYDATA/IMAGES.3000 MYDATA/IMAGES.3001 MYDATA/IMAGES.3002 SIZE=ALL FORM=ADD ACON=150 MCON=10 COEF=1 -1 UNIT=TEMP

  9. Display the images in frames 1-3. Load them centered on Atlanta, GA.

    Type: IMGDISP MYDATA/IMAGES.3000 1 STA=KATL REPEAT=3 REFRESH='EG;MAP' SF=YES

  10. Step through the images.

    Type: LB 1 3
    Press: Alt A

    Note that the bright areas correspond to the warm, low clouds and fog. Cold, wispy cirrus shows up as black areas.


Table of contents Previous: Creating an ADDE Dataset Next: Image Remapping