[GRASSLIST:2148] Re: Aerial photos and noxious weeds

[Rich Shepard]

On Tue, 24 Jul 2001, Roy Sanderson wrote:

> I've tried using GRASS with to identify Juncus infestations in semi-natural
> vegetation in upland grazing areas in UK with variable levels of success.
> These were based on true-colour aerial photos, scanned in on a flatbed
> scanner, imported into GRASS and split into RGB components with r.mapcalc
> before a supervised classification.  The availability of other wavebands
> e.g. nr-IR would improve the classification, and obviously the lower the
> over-flight altitude the better.  However, the key factor determining the
> relative success is the contrast in spectral signature between your noxious
> weeds and the other vegetation.  If the contrast is good then I see no
> reason why the technique shouldn't work well.

Roy,

  Thank you for sharing your experience. I knew that it should work -- in
theory -- and I was curious whether anyone had tried it. I think that I can
get my client to agree to low-level flights as they have the site flown on a
regular basis in any case.

  It would be much easier to survey approximately 100 square miles for
noxious weeds using GRASS than driving/walking over the area for several
days. High contrast should not be a problem because this is a high
environment. Most of the vegetation is sage brush, bunch grasses, junipers
on the hillsides (north facing, of course), and very light ground because
most of it is diatomaceous earth. If we go this route, we can probably
classify all vegetation while we're at it.

Again, thanks,

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
            2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
 + 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) | rshepard at appl-ecosys.com
                 Making environmentally-responsible mining happen.