Watershed analysis uses Washington State's Standard Methodology for Conducting Watershed Analysis to determine the cumulative effects of forest practices on public resources in a watershed—primarily water quality, fish habitat, and capital improvements—and establish guidelines for future management and regulation. The process is typically carried out by a large forest landowner within a period of 4-6 months on mixed-ownership land comprising an entire watershed of 5,000-50,000 acres. It makes use of available data from a variety of sources along with a limited amount of data collected through field surveys. Watershed analysis is part of Washington's forest practices regulatory process, but has also been used voluntarily by forest landowners in Oregon and Idaho.
GIS is an important tool in watershed analysis. Off-the-shelf digital data—such as USGS digital line graphs and digital elevation models, Forest Service cartographic feature files, and digitized soil surveys—provide a foundation for the analysis. The forest landowner can provide harvest history, land classification and other useful data. Current vegetative cover and logging roads can be mapped from aerial photos. Field surveys collect data on fish populations and habitat, hydrologic and riparian condition, erosion hazards, and mass wasting sites.
The product of a watershed analysis is a resource assessment report divided into several sections, or modules, including: hydrology, mass wasting, surface erosion, stream channel, riparian function, fish habitat, causal mechanisms, and prescriptions. The document includes approximately 25 maps as well as statistics derived from the GIS database. ARC/INFO has a wealth of analytical and display tools that are well-suited for watershed analysis. In particular, the multi-layer modeling capabilities of GRID are ideally suited for this application. An example is a rain-on-snow model we implemented on two watershed analysis projects that integrated five GIS layers: slope, aspect, elevation, precipitation, and forest canopy closure.