The Quest for Data
To perform the analysis I wanted to do, I only required three layers of information for a particular watershed: a digital elevation model, a soil composition layer and a layer with landuse data.  However, it was difficult finding all three of these layers for any single area with the resolution I needed.  I spent many hours searching through many data websites, especially the U.S.G.S, State of Washington, State of Oregon and State of Montana sites.  A good DEM was the most critical layer and also proved to be the most problematic in finding.  Most of the DEMs had resolutions that were too coarse for this project.  They would have been useful if I were modelling a substantially larger watershed, but I needed a higher resolution DEM so I could model a smaller watershed.  The Rational Method of calculating watershed discharge (see Methodology) is best suited to smaller watersheds.  Of the higher resolution DEMs that I found at research sites, most of them had errors that impeded translation.  Others simply did not cover an area condusive to the analysis I wished to conduct - usually this was because the data layer did not cover a full watershed of any decent size.

Foundation Data Layers

DEM
Eventually, I went back to examining the DEMs available on the SIS server.  I settled upon using the MCEelev.rst DEM from our lab exercises.  This appeared to be the best choice for two reasons.  Firstly, it has the highest resolution at 20m/cell, compared with 30-120m/cell for the other DEMs.  Secondly, it contained four watersheds that were composed of more than 20000 cells, three of which consisted of more than 50000 cells.  This is much greater than the other DEMs that had similar cell sizes.

Landuse
The MCELanduse.rst layer was used as a starting point from which to build a landuse layer suitable for this analysis.

Soils
There was no soil layer covering the same region as the other two layers.  Since the goal of the project was only to model a hypothetical watershed, I created my own soil class attribute layer.  To do this, I imported the MCElanduse.rst file into ARCView and used it as a reference layer over which to create my own new layer of soil zones.  I created a layer with 37 hypothietical soil polygons.