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

20010430: higher resolution global composite IR images (cont.)

>From: Edward Hindman <address@hidden>
>Organization: CCNY
>Keywords: 200104271758.f3RHwBL11872 McIDAS 

Ward,

re: non-Mollweide global composite
>The AREA8760 image looks like a remapped Mollweide with no higher 
>spatial resolution.

Actually, the Mollweide IR image in the Unidata-Wisconsin datastream
is created from this image, and the image is twice the resolution of
the UW stream Mollweide IR.  The following lists should illustrate
the point.

MYDATA/IMAGES.8760 is the image from the SSEC FTP site.  Its computed
resolution at the image's center is 33 km:

IMGLIST MYDATA/IMAGES.8760 FORM=EXP
Image file directory listing for:MYDATA/IMAGES
 Pos Satellite/         Date       Time      Center      Res (km)   Image_Size
     sensor                                 Lat  Lon    Lat   Lon
 --- -------------  ------------  --------  ---- ----  ----- ----- ------------
8760  G-8 IMG       30 APR 01120  21:00:00     0    0
    Band: 4  10.7 um Surface temp; longwave window       1.0   1.0   600 x 1200
     proj:    0 created: 2001120 212743  memo:
     type:PRD      cal type:BRIT

     offsets:  data=    1280 navigation=  256 calibration=  768 auxillary=    0
     doc length:   0   cal length:   0   lev length:   0 PREFIX=   0
     valcod:          0 zcor:  1 avg-smp: N
     lcor:    1  ecor:     1  bytes per pixel: 1  ss: 70
     Image Center Point Res (derived)  Lat:  33.36 (km)  Lon:  33.36 (km)

IMGLIST: done
 
This is a listing of the latest Mollweide-IR image from the Unidata-Wisconsin
datastream.  Its computed resolution at the image's center is 52 km in the
horizontal and 64 km in the vertical:

IMGLIST RTIMAGES/MOLL-IR FORM=EXP
Image file directory listing for:RTIMAGES/MOLL-IR
 Pos Satellite/         Date       Time      Center      Res (km)   Image_Size
     sensor                                 Lat  Lon    Lat   Lon
 --- -------------  ------------  --------  ---- ----  ----- ----- ------------
   2  G-8 IMG       30 APR 01120  18:00:00     0   -1
    Band: 4  10.7 um Surface temp; longwave window       1.0   1.0   350 x  640
     proj:    0 created: 2001120 193023  memo:
     type:VISR     cal type:BRIT
     offsets:  data=     768 navigation=  256 calibration=    0 auxillary=    0
     doc length:   0   cal length:   0   lev length:   0 PREFIX=   0
     valcod:          0 zcor:  1 avg-smp: N
     lcor: 9826  ecor:  9681  bytes per pixel: 1  ss: 70
     Image Center Point Res (derived)  Lat:  52.16 (km)  Lon:  64.34 (km)

IMGLIST: done

>It's terribly pixilated over Nepal.

Admitedly, neither image has the kind of resolution that either of us would
want, but the rectilinear one is better than the Mollweide remapped one.
I just displayed both of these at equal display resolutions and compared
them over Nepal.  It looks to me that the rectilinear image is much less
pixelated than the Mollweide one.  Try this yourself:

IMGDISP MYDATA/IMAGES.8760 1 LATLON=30 -92 EU=IMAGE SF=YES REFRESH='EG;MAP 
SAT;MAP X 5 FILE=OUTLHPOL'
IMGDISP RTIMAGES/MOLL-IR 2 LATLON=30 -92 EU=IMAGE SF=YES REFRESH='EG;MAP 
SAT;MAP X 5 FILE=OUTLHPOL' MAG=2

It may be possible that you are comparing a Mollweide remap of the
rectilinear image with a Mollweide remap of the Mollweide image.  After
doing that, I have to admit that the remapped Mollweide image looks
better.  I would not, however, compare the products of the two remaps,
but rather the remap of the Mollweide with the blow-up load of the
rectilinear one.

Here is what I did, so you can see what I am trying to point out:

1) rectilinear image remapped to Mollweide specifying output resolution of 8

IMGREMAP MYDATA/IMAGES.8760 MYDATA/IMAGES.1235 LATLON=30 -92 PRO=MOLL RES=8
IMGDISP MYDATA/IMAGES.1235 1 LATLON=30 -92 EU=IMAGE SF=YES REFRESH='EG;MAP SAT 
1;MAP X 5 FILE=OUTLHPOL'

2) Unidata-Wisconsin Mollweide IR remapped to Mollweide specifying output
   resolution of 8

IMGREMAP RTIMAGES/MOLL-IR MYDATA/IMAGES.1236 LATLON=30 -92 PRO=MOLL RES=8
IMGDISP MYDATA/IMAGES.1236 2 LATLON=30 -92 EU=IMAGE SF=YES REFRESH='EG;MAP SAT 
1;MAP X 5 FILE=OUTLHPOL'

3) display the rectilinear image at approx. the same resolution on frame 3

IMGDISP MYDATA/IMAGES.8760 3 LATLON=30 -92 MAG=2 EU=IMAGE SF=YES 
REFRESH='EG;MAP H;MAP X 5 FILE=OUTLHPOL'
 
4) make a loop of all three displays so they can be compared visually

LB 1 3
DR 3*20
TERM L ON

>The IMGREMAP 
>instructions you gave us on 27 April produced as good an image from the 
>standard Mollweide.

Please remember that when you remap images, you change data values.  The
remap might end up looking less pixelated, but this is a smoothing effect
that occurs in the remap.

>Where is the higher-resolution Mollweide imagery?

AREA8760 is the higher resolution image that I was referring to.  It has
twice the resolution as the one in the Unidata-Wisconsin datastream.

>It's criminal that India does not make Insat data available real-time or 
>otherwise.

I believe that this has changed.  The Indians at least talked about
making Insat imagery freely available in a non-encrypted form about 9
months to a year ago.  I guess that they figured that since China's
FY-2 would be imaging more-or-less the same area and making the data
available, it was a losing battle to keep the images encrypted.

What I have been trying to find for you is a site that has GMS data that
is willing to give you ADDE access.  I'll keep looking...

Tom