Storing glacier-identifying imagery in the GLIMS DB

[Bruce Raup]

On 2001-02-15 14:55 -0700,  Luke Copland wrote:

> Hi Bruce,
> 	One thing that we've been chatting about here is the possibility of
> storing an actual image (probably from satellite, but could also be from
> an aerial photo and even from the ground) of each glacier in the GLIMS
> database. This would make it much easier for future researchers to
> locate the original imagery and make comparisons with it, and to make
> sure that the correct glacier is being identified. When an aerial photo
> or satellite image is referenced, it isn't a particularly simple process
> to get hold of that imagery and find exactly which glacier is being
> described in an area which is heavily ice covered (i.e., the user has to
> get hold of the image, georeference it, and then use the lat/long
> coordinates to identify the glacier - an expensive and long-winded
> process if you just want a quick look of one glacier).
> 	We realise that there are potential issues about copyright, although
> this may not be too much of a problem if the ASTER imagery is free and
> Landsat 7 allows users to pass on copies to others. There's also the
> question of storage space, although I'm only envisaging that the stored
> files will be a few hundred kb at most (and in a standard format such as
> jpg). Essentially they will just be screen captures created as a
> regional center processes the satellite images. Doing this would also
> make the whole inventory a lot more accessible to users on the web.
> 	What are your thoughts/comments?
>
> Cheers,
> Luke

I think this is an excellent idea, and we at NSIDC talked about this very
thing yesterday.

I don't know much about the copyright issue, although as you say, it
shouldn't be a problem for ASTER or Landsat 7, particularly if what is
stored is a sub-image of the full original scene, in JPEG format, which
uses lossy compression.  Perhaps short "photo credits" should be added to
the bottom of each image, such as "ASTER image courtesy of METI/NASA".

I've attached a jpg image, 7727 bytes, that probably represents the
minimum quality image we would want to store.  It is a subset of a TM
image (of the Knik Glacier in the Chugach, AK, I think), saved with JPEG
quality factor of 50%.  If we stored 160 000 images like this, it would
amount to just over 1 Gb of data -- fairly modest by today's standards.
The same image at full resolution (~30 m pixels, or about 1200 x 900), but
still saved as a 50% JPEG file, is about 120 kb.  The resulting 20 Gb
volume if all images were stored that way would still probably not be a
problem.

It might be good to have the capability to store multiple images per
glacier, in case we have interesting images or photos from various
sources, or a short time series (ground-based photos showing marked
recession, for example).

Other thoughts/comments from anyone on the list?

Bruce

-- 
Bruce Raup
National Snow and Ice Data Center                     Phone:  303-492-8814
University of Colorado, 449 UCB                       Fax:    303-492-2468
Boulder, CO  80309-0449                            Bruce.Raup at colorado.edu


-------------- next part --------------
The following attachment was sent,
but NOT saved in the Fcc copy:
    A Image/JPEG segment of about 7,727 bytes.