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

20010201: remapping images in McIDAS-X

>From: address@hidden (Chris Hennon)
>Organization: Ohio State
>Keywords: 200101301841.f0UIfKX26078 McIDAS ADDE IMGREMAP IMGLIST

Chris,

Sorry I wasn't able to get to this last week when it came in.

>Thank you very much for your helpful reply to my question about how mcidas
>handles data.  I have been admiring my images for a few hours now.

Excellent.  So, since you are new to McIDAS, can you give me a brief
impression of how hard/easy it is to setup an ADDE dataset and begin
to use data?  What kind of information in our web documentation would
make this process easier for you?  Is a "Setting up ADDE Datasets" section
needed somewhere in the Installation and Configuration pages, or should
setting up the Unidata ADDE servers section be renamed and rewritten?
Any insight you can provide will help me to improve the documentation
for new users.  Thanks in advance...

>Now I have been trying to remap my image.  My raw file is a GOES-8, NH IR
>image (4km). I tried to remap it to a RECT projection using the following
>command:
>
>IMGREMAP ME.28 MYDATA/TEST-IMAGES.1 PRO=RECT RES=4 SIZE=ALL
>
>'ME' is an alias I created for my group.

OK. This command will remap the entire original image into a RECTilenear
projection.

>Anyways, you can see for yourself what it looks like:
>
>twister.sbs.ohio-state.edu/CHRIS1.GIF (original, with MAG=-2)
>twister.sbs.ohio-state.edu/CHRIS2.GIF (remapped, screwed up?)

One question is how you loaded the remapped image.  It appears that you
used a default load whould would put the upper left hand corner of the
image in the upper hand corner of the frame.  The left part of the
remapped image looks bad since it represents remapping of the image
information at the earth's limb.

What does the following look like:

IMGDISP MYDATA/TEST-IMAGES.1 LATLON=35 80 MAG=-2 EU=IMAGE REFRESH='EG;MAP H'

look like?

>I noticed that the remap resolution says 1.0 km, even though I specified
>RES=4.

What you are seeing is the vaguery of the resolution displayed in an
image load im McIDAS.  The number on the display is the multiplicative
factor for the base resolution.  The actual resolution of the image can
be seen from an IMGLIST invocation:

IMGLIST MYDATA/TEST-IMAGES.1 FORM=EXP

at the end of this listing, you will see the resolution of the pixel at
the center of the image.  This value is calculated from the navigation,
so it is reliable.

>The lat lon values seem out of whack too.  I'm not sure what this
>means.  

Your GIF (tm) images don't have enough information for me to gleen
anything about LAT LON values.  Did you setup your remote ADDE server?
If you did, I could take a look at your image(s) here at Unidata and be
able to tell more about what is going on.

>I also tried this with a mercator projection with similar results.
>Can you shed some light on a new mcidas user who is in step 2 of the
>program what is going on here?  Thanks.

Did the above help?  If not, we can work through an example that is
akin to what you are trying to do.  For instance, do the following
(this demonstrates the beauty of ADDE :-):

DATALOC ADD RTIMAGES ADDE.UCAR.EDU
IMGLIST RTIMAGES/GE-IR FORM=EXP
IMGREMAP RTIMAGES/GE-IR MYDATA/TEST-IMAGES.2 PRO=RECT RES=4 SIZE=ALL
IMGLIST MYDATA/TEST-IMAGES.2 FORM=EXP
IMGDISP MYDATA/TEST-IMAGES.2 1 EU=IMAGE SF=YES REFRESH='EG;MAP H'
IMGDISP MYDATA/TEST-IMAGES.2 2 EU=IMAGE STA=KSFO SF=YES REFRESH='EG;MAP H'

The DATALOC points you to adde.ucar.edu for images in the RTIMAGES
dataset.

The first IMGLIST will list out the information for the most recent
GOES-East IR (10.7 um) image from that dataset.

The IMGREMAP is the same as the one you are using.

The first IMGDISP display will have a section that looks incorrectly
remapped, but this is an artifact of remapping values at the limb of
the image.

The second IMGDISP is centered over San Francisco.  Note how it looks
like you would expect.

Tom

>From address@hidden Thu Feb  8 10:48:34 2001
>Subject: Re: 20010201: remapping images in McIDAS-X

Tom -

See comments below.

re: Any insight you can provide will help me to improve the documentation
for new users.  Thanks in advance...

Once you clarified to me what exactly what I had (mcidas area files not
named in mcidas format), things started to fall into place.  The key was
figuring out how to use the DSSERVE command to make mcidas look in that
particular directory.  I think the learning guide is very good at
explicitly laying out the group/descriptor relationship as well as data
types.  I just didn't realize that I had one type (IMAGE) and that I had
to use DSSERVE in that manner to get things to work right.  I think
sections of your last email to me would make a good reference page for
some new users.   

I haven't really had the chance to run through the rest of the learning
guide yet, but if I come across something that seems unclear to me I'll
pass it along. 

 
re: It appears that you used a default load whould would put the upper
left hand corner of the image in the upper hand corner of the frame.

I think you are right.

re: What does the following look like:
> IMGDISP MYDATA/TEST-IMAGES.1 LATLON=35 80 MAG=-2 EU=IMAGE REFRESH='EG;MAP H'

>It looks great!

re: "resolution" on loaded image is actually a multiplicative factor for
the actual resolution

Ah, I see.  The fog is slowly lifting.  Now that I kind of know what's
going on I'm going to go back and play a little bit more.  Thanks for your
help again.

Chris