20010103: IMGREMAP: MAG= vs RES= (cont.)

>From: Jason Allard <address@hidden>
>Organization: PSU
>Keywords: 200010240345.e9O3jC425238 McIDAS-X IMGREMAP MAG RES

Jason,

re: projection and probing
>Okay, this makes sense to me... I've gotten the picture now.  With a different
>projection, it could give slightly different values... but changing projections
>can also change what values are in the image, so while using the original GOES
>7 projection might make it harder to come up with the coordinates I want, it
>is the most 'pristine' data set from which to work, and it is close to other
>projections... I'll just have to use a little trial and error with changing
>line and element numbers, and using Alt E to get as close to the desired 
>coordinates as possible.  Thanks for the information.
>
>On a related note... is there a was to send the values obtained from the
>IMGPROBE to a text file instead of the text and command window.  I'm going to
>have to perform some analyses on the data, and it would be very time consuming
>to even attempt to transfer the values from the text window to a file by hand.
> ..
>if I can get the values saved into an ascii text file, I can open them in excel
>and work with them.

Yes.  Any textual output from McIDAS commands can be "redirected" to a
text file by using the global keyword DEV=.  Here is an example:

with no "redirection":

IMGDISP RTIMAGES/GE-IR
Image file directory listing for:RTIMAGES/GE-IR
 Pos Satellite/         Date       Time      Center   Band(s)
     sensor                                 Lat  Lon
 --- -------------  ------------  --------  ---- ---- ------------
 363  G-8 IMG        4 JAN 01004  00:15:00    23   71 4
IMGLIST: done

with "redirection":

IMGLIST RTIMAGES/GE-IR DEV=T imglist.out
IMGLIST: done
 
SEE imglist.out
Image file directory listing for:RTIMAGES/GE-IR
 Pos Satellite/         Date       Time      Center   Band(s)
     sensor                                 Lat  Lon
 --- -------------  ------------  --------  ---- ---- ------------
 363  G-8 IMG        4 JAN 01004  00:15:00    23   71 4
SEE: Done...EOF Encountered.

imglist.out is an ASCII text file.

>Also, where would I find what units the IR images are in... from probing the
>images, I get values for cloud tops of 190 to 210 for fairly high clouds...
>from the values I get, it doesn't seem that the units are kelvin... and I'd
>like to be able to convert the values I extract (with IMGPROBE as we discussed
>above) into kelvin.

IR temperatures in McIDAS are in Kelvin.

>Near as I can tell, the IMGPROBE is returning raw values,
>not in kelvin...

IMGPROBE will return all values related to the pixel being interrogated.
In the case of an IR image, you should get back RAW, BRIT, and TEMP.
TEMP will be in K.

Here is an example using realtime GOES-West IR (10.7 um channel) data:

IMGDISP RTIMAGES/GW-IR STA=KGJT REFRESH='EG;MAP H'
Beginning Image Data transfer, bytes= 616496
IMGDISP: loaded frame  13
EG;MAP H
IMGDISP: done
Erased graphic frame(s) 13-13
MAP: Completed frame 13
 
<move the cursor to a point on the screen>

IMGPROBE MODE=N

     Image Name           Day      Nominal Time   Scan Time    Band
  ----------------      -------    ------------   ---------    ----
  RTIMAGES/GW-IR.362   4 Jan 01004   00:00:00      MISSING       4

                        File     Nominal  Image     RAW     TEMP    BRIT
       Lat/Lon      Line/Element  Line/Element               K
 39:12:23 / 107:53:50  387/ 1376    4181/19005       127     266.5    127

IMGPROBE: Done

(Your listing will be slightly different.  You are using McIDAS-X 7.6x
while I am using 7.7x).

You can see from the listing that IMGPROBE should give you back all
possible interpretations of the pixel and tell you the units for TEMP.

>when I do an Alt D, it give me both the raw (as I got with
>the IMGPROBE LIST BOX command) and the desired kelvin units.

OK.

>I tried to
>specify a band number in the just mentioned command (band 8 for GOES 7), but
>it gave me the raw value.

Is the image you are probing multibanded?  Does it contain band 8?  I
am not really sure just what band 8 is supposed to be.  Hmm...  Is the
image you are probing the result of a remap into a GOES-7 image?  If
so, then you will need to change the band number in the remapped copy
to match that from the original.  Then you can successfully problem the
image.

You can change the band number in an image (and other things) using
the IMGCHA command:

IMGCHA group/descriptor.pos BAND=band_number

(you have to substitute for 'group/descriptor.pos' and 'band_number')

>Thanks so much for the help... I'm getting close to knowing everything I
>need to know to perform my analyses... I'm not sure I'd even have come
>close to figuring it out without you... thanks,

You are welcome, and I really do think that you are very close.

Tom