[R-sig-Geo] Point pattern analysis on a sphere?

[Terhi Tormanen]

Dear List,

This is a bit more general question but I'd be very happy about any  
advice.

I have two data sets of geographic (lon, lat) locations covering a  
large area (large continent scale; and also global sets). I'd like to  
perform some basic spatial analyses of the data, i.e. are they  
clustered/dispersed compared to CSR, are there differences between  
the data sets, are there any clusters.

I've tried to look at various spatial packages but so far I've not  
found anything that would have been impelemented on a sphere (i.e.  
with true spherical distances and surface areas).

I know that someone somewhere must have done this before, so I  
started looking at R packages before I try to programme these myself.
(I've done basic reading about the spatial statistcis so (I think) I  
know which statistical methods to try to use). (Also, if someone  
knows about any other available software to use, I'd be very grateful.)

Thank you!


Terhi Tormanen
PhD candidate
Planetary Sciences, Dept. of Physics
Univ. of Oulu, Finland