[GRASSLIST:4599] Re: Interpolating DEMs with r.surf.spt

[Gordon Keith]

On Sat, 28 Sep 2002 14:32, Bruce Saunders wrote:
> I want to analyze landforms and watersheds in the southeastern US
> using 7.5 min USGS quads.  10 m DEMs are available for most but
> not all of the analysis area, so I am thinking about using GRASS's
> s.surf.rst module to create 10m DEMs from 30m DEMs of the missing
> quads.  The missing quads cover about 20-25% of 1 section of the
> analysis area; very few quads are missing from the other 9 sections.
>
> My first question is, does this sound like the best strategy
> to produce a consistent DEM using standard GRASS modules?

I've been trying to join datasets of different resolutions for 
visualisation purposes (eg flythroughs in nviz) and have found that 
s.surf.rst gives a pronounced patchwork effect with the various segment 
areas being very clearly visible. I doubt the patchwork effect is 
significant in terms of landforms, but might affect watershed analysis.

What I wound up doing was:
r.mapcalc temp='if(isnull(hires),lowres,hires)'
r.neighbours in=temp out=pretty method=average size=3
r.mapcalc pretty='if(isnull(hires),pretty,hires)'

I chose the size in the r.neighbours command so that for each cell of 
the low resolution DEM there would be (as close as possible to) exactly 
one cell of high resolution DEM that was averaged over the values of 
just that cell.

The results were the best I managed to get from a whole series of 
attempts. 

If anyone has a way of getting rid of the patchwork effect from 
s.surf.rst I'd like to hear about it. All my attempts failed, but I 
don't make any claim to understanding what all the parameters mean.

My requirements are different from yours so YMMV.

Regards
Gordon

-- 

Gordon Keith
Programmer/Data Analyst
Marine Acoustics
CSIRO Marine Research
http://www.marine.csiro.au

    Machines running Windows are by far the most popular,  
    with about 90% machines in use worldwide. Linux users, on  
    the other hand, may note that cockroaches are far more  
    numerous than humans, and that numbers alone do not  
    denote a higher life form.