I'm attempting to create a fairly large raster (Hydrologic Unit Code 8 [HUC8) from a polygon HUC8. Basically I want to illustrate the HUC 12 values within that HUC 8 to run a weighted overlay analysis on three datasets. All of these datasets are the same HUC 8 with multiple HUC 12s within, they just contain different content. For instance, population, stream miles, road miles. When specifying the cell size, I assumed I could simply choose another raster dataset that had the cell size I desired, and Arc would just chose the cell size. I chose a raster dataset with a 3 arc second (90 meters), but this raster's extents were not in the same location as the polygon I wished to convert. Do I need to specify a raster dataset that is already contained in my study area? Is the sotware specifically looking for the raster's corners to mimic or a random cell size?
On a similar note, my goal is rather simple. I am overlaying three datasets that align perfectly, and assigning weights. It seems this type of analysis would be a simple calculation using vector data. Is there a method to perform weighted overlay analysis within a vector analysis.
The raster defining the cell size of the output raster does not have to be in the same extent. Do keep in mind that the output raster will use the same coordinate system of the polygon layer. If your input cell size raster uses 3 arc seconds then the tool will take the number (3) and apply it to the coordinate system of the polygon layer. If it uses feet then your output cell size will be 3 feet. I would recommend manually defining the cell size based upon the polygon coordinate system units. Also, you can define the processing extent in the environment settings so each conversion is consistent. Keep in mind that if the polygon layers are not exact you will need to use the Snap Raster environment setting to get the cells to line up properly.