ols and raster calculator

Discussion created by mar1p0sa on May 30, 2012

I am building a habitat model and don't have enough snow collection data sites in one area for interpolation to work well.  I ran a bilinear regression (Stata and with ols tool) on my snow depth and elevation data from NCRS, resulting in a statistically significant Pearson???s correlation at the 0.05 level (0.576).  I wanted to use this in combination with elevation DEMS from the USGS to produce a more reasonable snow depth layer by elevation than I currently have through interpolation.

I downloaded four DEMs (30m) covering the area and mosaicked together.   I thought that I could connect the raster calculator with the OLS tool (but don't know exactly how to do this since the input doesn't seem to be allowed) or alternatively, put the OLS equation results into the raster calculator with the DEM:   snow depth = (10.757 * DEM + 3996.5).  However, it doesn???t work well.  I get a raster with what seem to be hugely unreasonable numbers or the raster calculator crashes.

I obviously don't know what I'm doing wrong, although I've experimented considerably.  This is a class project, so please provide a neophyte answer as opposed to a long-term user answer.

Any help is appreciated.