s.surf.idw

[Chris Skelly]

Hi people,
 Following the recent (resurfacing) thread of kriging...

> > While on this subject: what is a good source of kriging programs for
> > use with GRASS? I would like to do cokriging on data that I have in
> 
> this is still on the "to do" list as far as I know. I have toyed with
> exploratory data analysis in GRASS, but have not got around to implementing
> kriging or variogram fitting.  I plan to begin implementing some of these
> things after the end of this month.
> 
> > grass site maps to produce a raster map. 
> 
> I will be working with points in vector files, so conversion of sites
> to vector will be the first step.
> 
> Darrell McCauley
> Purdue University
> 
> P.S. I am aware of one other grassu person (Chris) working toward these 
>      goals. If any other people have plans, then perhaps we should
>      coordinate our efforts.

I have been using my first attempt at writing a kriging routine for a
couple months now, and at a superficial level it seems to work quite
well.  I have been having more of a problem deciding on how to evaluate
variograms than the programming of the simple kriging routine took.

I concur with Darrell that we should collectively tackle this issue 
each to his/her interest level.  I'm doing it simply because I need it
for some research I am doing and I don't want to trust a black box
routine, such as may be found in Surfer for instance (or other GISs?
not that I'm aware of any other GISs with kriging).  Furthermore, for
the researcher, GRASS is the only platform that offers this level of
tinkering...so don't confuse my interests with those of the professionals
who actually know their way around software :-) I take my hat off to
the GRASS people and other who really seem to know what they are doing

...however, in the same breath, as I have already said, I do have a
"working" kriging routine for GRASS and it has been available on our
server at Macquarie for about a 2 and 1/2 months, with some nine 
people downloading it...until it completely disappeared last week!
I will see that it is back there tomorrow.  The only catch is that you need
to put in your own matrix manipulation algorithm, I used numerical 
recipes for my own use, but I cannot distribute it.  Also there is
a tar file containing stats and variogram routines in fortran that are
for the taking.


our ftp server is mqmet.cic.mq.edu.au
and the directory is /pub
and the files are krig_w_bugs (my own personal disclaimer ;-)
                  stats.tar
krig_w_bugs is a modified s.surf.idw so just recreate the
relavent s.surf.idw directory and rename it krig and make the
appropriate changes in make files and recompile, I really don't
want (can't...my supervisor will kill me) to support this but
general geostats/kriging/variogram discussion is always welcome
and the very odd email on I can't get this to work will 
probably (gulp) be okay.

cheers,
chris
-- 
|\/\/\/|                             Chris Skelly
|      |                             School of Earth Sciences 
|      |   High expolsives and       Macquarie University  
|  (o)(o)   school do NOT mix.       North Ryde, 2109, NSW, OZ 
C       _)                           Tel: 612-805-8433 
 |  ,___|                            Fax: 612-805-8428 
 |    /  chris at mqclimat.cic.mq.edu.au