I'm trying to calculate the average slope and standard deviation of the slope for each county in the USA. As the measures don't have to be very precise, I'm using DEMs from the GMTED2010 to save on disk space.
I think I might be making a mistake in terms of correctly projecting the DEMs and was wondering if anyone could provide advice.
My procedure is as follows:
1) From the GMTED2010 dataset, I'm using the DEMs that are aggregated using mean elevation.
2) Project each DEM from "WGS_1984" to "North_America_Equidistant_Conic" using the "WGS_1984(ITRF00)_To_NAD_1983" transformation.
3) Create a mosaic of the projected DEMs
4) Calculate the slope of the Mosaic
5) Run zonal statistics by county on the slope raster
Is it sensible to calculate slope using the North_America_Equidistant_Conic projection? Does it make more sense to use ASTER data and finer projections? As I'm only looking at county level averages, the results don't have to be precise, but they shouldn't be systematically biased.
