Interpolation for Precipitation Grid

[ConradoDe_León]

Hi, I´m trying to create a precipitation grid to use as an input data for N-SPECT and I would like to know how to proceed. How can I create it if I only have the data from the meteorologic stations and I need to interpolate the values?
Thanks a lot for your time.

[JeffLee]

I would try ordinary Kriging.  If you have station elevations, I would use co-kriging with elevation as co-variate.

[MariaRosales_Rueda]

I would try ordinary Kriging.  If you have station elevations, I would use co-kriging with elevation as co-variate.

Dear Jeff, I also need to interpolate precipitaion data from stations (point data) to counties in ecuador (shapefile). I do have elevation of the stations. I'm new at GIS. But think it is important to account for altitude.  Can you please tell me what tool should I use to use co-kriging? and also, do I need an elevation measure for the counties?
I really appreciate your help

Thanks a lot
Maria

[MarkBoucher]

There is a program called PRISM, developed at Oregon State Univ that does this in a very complex way. I think it does the spatial interpolation from rain gauge points using 12 or so variables such as elevation, aspect, etc. I do not believe it is an ArcGIS tool in any way, though I'm certain it is raster based. My personal conversation with some working with it at OSU convinced me that it is not meant to be run by the "average" hydrologist. It is very tedious to use and an expensive tool to use to set up for a state or region. Few people are trained on it. It is being used by NOAA in the latest NOAA Atlas 2 development.

Having said all this, I suspect there are parts of the PRISM logic that would be useful to the "average" hydrologist or researcher that could be emulated in ArcGIS using Spatial Analyst. Might be worth looking into.