GLIMS ASTER data list now available

Wed Jan 31 12:55:37 MST 2001

Dear GLIMS colleagues,

Now that GLIMS ASTER data are finally beginning to get acquired and
make their way to the USA, we want to offer a status report and some
guidance on finding, ordering, and using ASTER data.

STATUS
As of January 21,  over 50,090 data granules were available in the
ASTER database at EDG (EOS Data Gateway).  Several granules may
represent reprocessed versions of the same acquisition date and times,
so there is some repetition in the EDG database.  Approximately 2500
daytime ASTER granules over glaciers are listed.  Of the glacier
granules, only 76 are the cleaned up and geometrically corrected L1B
data.  Of those, 43 were acquired with the proper gain settings for snow

and ice.  Most of the glacier data were acquired in stereo mode (both
band 3N and 3B).

FINDING DATA
Thanks to extra efforts by EDC (EROS Data Center), the U.S.
Geological Survey in Flagstaff has begun to download weekly updates of
listing of all ASTER data received by the EDG.   We then parse and
spatially search the database within ArcView and IDL to produce a list
and maps of ASTER data over glaciers.  We have place the search results
on our GLIMS distribution protected website

(http://wwwflag.wr.usgs.gov/GLIMS/usgsdist/main.html)

The userid and passwd are:  glimsguest    ber1ng789

The queries are provided to our GLIMS colleagues as a
comma-delimited list of data granules sorted by regional center number
(RC#).  Information about the regional centers is listed in the
procrc.tab
on the above website or on the public GLIMS  website:

http://wwwflag.wr.usgs.gov/GLIMS/collaborators.html

At this stage, the cloud assessment provided by the ASTER processing
is still very unreliable over snow and ice.  We are working with EDC to
improve the cloud and quality assessment using MODIS data.  For
now,the lists we provide will be data dumps sorted by area that are
filtered by appropriate gains and sun angles.   You will see many
entries
with either 0% clouds or 100% clouds while the browse images will
show just the opposite situation.  It will be up to each Regional Center

and/or steward to search the EDG and assess the browse images before
ordering.

We've posted a table on the website (procrc.tab) that contains scene
statistics for all glacier areas and each Regional Center. The first is
a
block for all ASTER scenes for which we have been able to get
information from EDC and that fall in glacier areas(see below). This is
followed by a block for each RC.

 File=../GLIMS/searches/glacierbyregion_oct1-jan21_2001.csv
# scenes: total, 1A, 1B, other=  2499  2424   75    0   GLIMS=1104
        OFF  HIGH  NORM   LOW  Low1  Low2    ON
  Sub     0     1     2     3     4     5     6
 VNIR     9  1041  1409    40     0     0     0
 3aft     9    20  2430    40     0     0     0
 SWIR     9  1143  1335     0     0    12     0
  TIR     0     0     0     0     0     0  2499
----------------------------------------------------------

For each Region (separated by dashed line), the first four lines list
the region number, the geographic region and RC information. The 5'th
line contains the name of the inventory file used. The 6'th line lists
the total number of ASTER scenes in the region, the number that are
Level-1A, the number that are Level-1B, the number of other processing
levels (typically zero) and the number with the ASTER gains set for
GLIMS (VNIR='NORM', SWIR ='HIGH'. This is followed by a little
table of the statistics of Gain settings.

We have made individual files for each RC that list the ASTER images
in their areas.  Files contain one line for each image, with the data
from
EDC in comma delimited form.  The files have with names RCnn.tab
where nn is the RC number.  There are some images listed for every
Region.

ORDERING ASTER DATA
ASTER data is now publically available online, for free, at the EOS Data

Gateway:

http://edcimswww.cr.usgs.gov/pub/imswelcome/

You may log in as guest (though its better to register since it then
remembers your settings) and choose ASTER L1A and/or L1B as your
datasets. Then search on the start times provided in our GLIMS search
results or on the lat/long coordinates.  When you find  images that
match
your criteria and have acceptable browse images, add them to your
shopping cart and complete the order.  You may receive the data either
via FTP or on 8mm tapes from EDC.

USING ASTER DATA
The data you receive from EDC will be in HDF-EOS format, a fairly
complex NASA format that's described in several places on the web.
There are a variety of free and commercial applications to view the
HDF-EOS data.   You might get started with a free viewer called
"WebWinds" that may be downloaded from JPL (refer to
http://webwinds.jpl.nasa.gov/).  WebWinds will also let you save the
data in a simpler format.  The newest versions of ENVI, which is a
commercial package that uses IDL (see http://www.rsinc.com/), can also
read HDF-EOS (this is not an endorsement by the USGS).   Our group
(mostly Jim Torson)  is developing a useful image viewer with some basic

analysis tools that uses IDL on UNIX systems.  The viewer, CV, is
currently part of a larger set of image processing software called ISIS.

A new HDF-EOS compatible version of CV that does not require ISIS
is not yet available in the public release of ISIS.  We will notify you
when it's available.  More information is available at
http://wwwflag.wr.usgs.gov/ISIS

ASTER data is available in two basic forms: L1B and L1A.  For most
applications, L1B is ASTER data in its most useful form.   L1B data
have been cleaned up and geometrically corrected in Japan.  The most
recent examples that we have examined so far appear to have only minor
defects.  Overall, L1B is much easier to use than L1A since no
pre-processing is required.

L1A data represents a nearly raw version of the ASTER instrument data.
As a result, L1A data usually have image striping and band to band
misregistration.  VNIR bands 1 and 2 are usually reasonably good while
bands 3N and 3B have even/odd stripes.  SWIR bands 4-9 are not
registered in L1A data. TIR bands 10-14 are generally useable, though
somewhat striped in L1A.

If you find some useful images that are only available in L1A data,we
may be able to clean up and register a few scenes for you in Flagstaff
by
special request.  We have developed IDL applications that will destripe
and register the data.  This quick registration is a shift of an entire
band
image by an integral number of lines and samples; it does not involve
resampling, so the original resolution is not changed. This registration
will
leave parts of the images, probably the corners, out of registation by
up
to a few pixels.We do not have an efficient way to perform a geometric
correction yet.

If you have any questions, we would be glad to assist you through your
appropriate Regional Centers.

Thank you and happy hunting.

Sincerely,
 The US Geological Survey - Flagstaff GLIMS TEAM

+++++++++++++++++++++++++++++++++++++++++++++++++++++
Rick Wessels                rwessels at flagmail.wr.usgs.gov
U.S. Geological Survey      Phone  (520)-556-7022
2255 N. Gemini Drive        FAX             -7014
Flagstaff, AZ 86001         Main Office     -7000
+++++++++++++++++++++++++++++++++++++++++++++++++++++

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://nsidc.org/pipermail/glims/attachments/20010131/fb84b4d4/attachment.html>