Skip to content

gstat for interpolation

3 messages · Ashton Shortridge, Mintewab Bezabih

#
Dear the R-sig-geo users,


I have been trying to interpolate rainfall and temperature data for 8
stations and that needs to be interpolated at a farm level so that I
have farm level temperature and rainfall observations.

I am trying to use R for that. First off is kriging in gstat something
that I can use for that purpose? I have latitude and longitude
information. Can somebody give me some tips on how I can go about
that?

many thanks in advance
minti
#
Dear Minti,

with only 8 stations there is not enough data to construct reliable covariance 
(variogram) models. Eight stations simply isn't very many for reliably 
modeling spatial structure. You would be much better off just using inverse 
distance weighting (IDW), which gstat can also do, or even nearest-neighbor 
interpolation. 

Second, it would be best to project the latitudes and longitudes to a planar 
coordinate system appropriate for your study region. I think gstat may be able 
to deal with spherical distances, but I haven't any experience with that. The 
rgdal package is one of several that can do this in R; it can also be done in 
GIS or with other geospatial tools.

Yours,

Ashton
On 09/07/12, Mintewab Bezabih, wrote:
-----
Ashton Shortridge
Associate Professor			ashton at msu.edu
Dept of Geography			http://www.msu.edu/~ashton
235 Geography Building		ph (517) 432-3561
Michigan State University		fx (517) 432-1671
4 days later
#
Dear Ashton,

I am fine with using the Inverse Distance Weighting or the Nearest
Neighbour and I can also use your suggestion of the two coordinate
plotting. The purpose of my exercise is to get interpolated figures
that I will use in further statistical analysis (I am an economist and
my understanding of GIS issues is very limited).

I would be delighted if you could give me more tips on how I could go
about this. I had previously attempted doing interpolation with R
(which is perhaps totally wrong but I can post it here if you think I
can incorporate your suggestions on there

Many thanks.
Regards,
Minti
On Fri, Sep 7, 2012 at 2:45 PM, Ashton Shortridge <ashton at msu.edu> wrote: