[GRASSLIST:2334] Re: s.to.rast problems

[Rich Shepard]

On Thu, 9 Aug 2001, Kevin Slover wrote:

> thanks, will give it a shot, I read yesterday in an email that s.to.rast was the
> best to use, but maybe not.   Have you had any real problems using this
> command??  Or had any success with the other site interperlators (i.e.
> s.surf.rst or s.surf.krig)??

Kevin,

  Not yet. On my current, priority project I'm bringing in DEMs from USGS
SDTS format and vector files from MapInfo format.

  What you need to do is carefully read the man pages and associated
references for each of those modules. Each interpolation method is best
suited for different types of data, and different placement of data. For
example, the r.surf.idw man page mentions that it's for gridded data while
the r.surf.idw2 man page notes that it is for randomly-spaced data. Big
difference in results. Kriging is a whole 'nuther story, too. :-) So much
for data placement affecting the method used.

  For types of data, you need to understand what each point actually
represents and how the interpolation algorithm fits these points. With
elevation data we can assume that the highest value(s) and lowest value(s)
do not represent the actual highest and lowest points on the landscape, but
local maxima and minima. Therefore, it's most appropriate to use an
interpolating algorithm that fits a surface slightly above local maxima and
slightly below local minima.

  On the other hand, if you are buiding a surface of chemical concentrations
at specific sites, then you want an algorithm (such as Kreiging that takes
into account spatial dependencies among close samples) that fits the surface
_through_ each point.

  Of course, you can take the easy way out and ignore all this messy detail
and just use whatever works the quickest for you. :-)

HTH,

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.