[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20010905: McIDAS - IMGCOPY questions..... (cont.)



>From: Giovanni Leoncini <address@hidden>
>Organization: SJSU
>Keywords: 200108291527.f7TFR3115078 McIDAS IMGCOPY

Giovanni,

>once again, you were right: the calibrations unit of my images are BRIT,

I figured that would be the case.

>but this is true for wv and ir, as well as vis. is this normal?

If the images are pre-calibrated and 1-byte, yes.

>is there a way to get albedo?

I went to SSEC and did a search on questions asked about albedo.  Here is
information from their support logs:

  ... the ALBEDO in the gvar images is more like a REFLECTANCE. The true
  meaning of albedo includes solar angles which are not accounted for
  in the gvar calibrations. 'Albedo' terminology is used in gvar for
  historical reasons.

  The brightness is a gray shade that represents a physical
  quanitity. It's values are 0-255, which are stored as a 1-byte
  integer (less precision needed).

  Albedo (vis)/ Temperature (ir) are physical quantities that are stored in
  2-byte integers to retain precision of the value derived from the
  RAW satellite counts.

  VIS: RAW -> ALB -> BRIT IR: RAW -> RAD -> TEMP -> BRIT

  To get brightness from albedo in GVAR and POES, this equation is
  used: BRIT=25.5*SQRT(ALB[%])

  ... asked about how the RAW counts are converted into
  albedo. I pointed him to the Programmer's manual Chapter 6, where
  the calibration is mentioned. The module he would look at is
  kbxgvar.dlm.

A second, much older entry in their tracking system had this comment on
calculating albedo:

  Currently we have no conversions for 1 byte vis counts to albedo.

Now, it seems to me that one could invert the BRIT from ALB equation listed
above to get albedo values, but I have never tried it.  These values,
however, would be quantized into integers.

What I have in mind is inverting:

BRIT=25.5*SQRT(ALB[%])

into:

ALB[%] = (BRIT/25.5)**2

which is equivalent to:

ALB[%] = 0.00153787 * BRIT**2

This equation makes some sense since a brightness of 0 will map into an
albedo (actually reflectance as per comments above) of 0; a brightness
of 255 will map into an albedo of 100 (0.00153787 * 255**2 = 100).

The IMGOPER command that will calculate the right hand side of this
equation using an input VIS image from the RTIMAGES dataset is:

IMGOPER RTIMAGES/GW-VIS MYDATA/IMAGES.1234 SIZE=ALL MCON=0.00153787 POW=2 
PROD=ALB

In order for this to work, you would have had to define the dataset
MYDATA/IMAGES.  If you havn't already done this, you can easily do so
from your McIDAS session by running:

BATCH MYDATA.BAT

Since MYDATA/IMAGES is setup to have a one-to-one mapping between dataset
positions and AREA file numbers:

MYDATA/IMAGES.1    <-> AREA0001
 ...
MYDATA/IMAGES.100  <-> AREA0100
 ...
MYDATA/IMAGES.9999 <-> AREA9999

The output image file created would be AREA1234.

NOTE: you have to be careful when selecting which image to write out
to.  If your site is ingesting the Undiata-Wisconsin datastream and
filing the images in the default manner, then you will already have
images in the AREA file number range of 60 to about 1200.  You should
not, therefore, use output AREA file numbers in this range.  In
addition, the McIDAS distribution comes with topography images in the
AREA file number range of 9000 - 9020, so you should not use this range
either.  To be safe, it is best to use AREA file numbers in the 2000,
3000, ..., 8999 range.

After doing the above calculation, you can use IMGPROBE and see that the
values stored in the image are all ALBedo.  

The next step, as I understand what you are attempting to do, is to
write these values into an output netCDF version of the image.  You
may first want to change the directory header word 52 to "ALB ":

LWU POKE AREA1234 'ALB ' 52

This will give you the entry that ncdfaput.c is searching on at least:

   if (strncmp((char *) &areaDir[CAL_TYPE], "ALB ", 4) == 0) {

>what kind of images should I get?

Typically, RAW images.

>thanks!

I hope that this helped.

Tom

>From address@hidden Thu Sep  6 15:34:12 2001

re: I hope that this helped.
are you kidding? it's the most detailed and complete I ever received!! ;)
you did 3/4 of the work for me!!

thanks a lot, I'll get back to you in a few days. 

*****************************************************************************
Giovanni Leoncini
Graduate Student
Meteorology Department                     
San Jose State University
One Washington Square,                         tel :  408-924-5199
San Jose, CA, 95112                           email:  address@hidden
*****************************************************************************


NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.