Analysis
Preprocessing
The preprocessing stage went
pretty smoothly. All the data layers were available at the same location.
They were all available online free of charge and downloaded fairly
quickly considering the size of some of the files (e.g. SSURGO Soils). After
obtaining the data, layers they were converted from vector to raster format.
Once this was complete, we determined what erosion risk factors would
be considered in the analysis. A search of the data tables within the
each layer identified the location of these elements. The flow chart
for the analysis was then developed.
Analysis
The process of analysis began
with the watershed boundaries. Specifically, the Harlow Creek watershed was
identified within the Marquette County hydrology coverage. Because the hydrology
layer was stored as lines, we converted first to grid, then to a polygon
raster file so that subsequent grid coverages could be clipped using this
polygon. Percent slope was derived for the 30M DEM model. Next,
we reclassified this coverage based on the range of slope values identified
from the erosion risk factors (Figure 2). In
order to determine where the cover types that are vulnerable to possible
logging occur, the 1992 NLCD Landuse data layer was reclassified. Deciduous
forest (Figure 1), identified within the data
layer as class 41, was given a point value of one while all other cover
types were given a point value of zero. The resulting reclassification
was then intersected with the slope reclassification (Figure 3) to obtain a coverage showing slope values
for areas of deciduous forest. The SSURGO Soils layer was first converted
from a line coverage into polygons then converted to GRID format. In
order to reclassify the soils according to K Factor two tables were joined.
Once complete the soils were reclassified according to K factor as
identified in our erosion risk factors. This resulting reclassification
was then intersected with the land cover reclassification to obtain a coverage
showing the K factor associated with each soil of the areas of deciduous
forest. Finally, slope class/forest and soil class/forest were unioned
together to derive the final index of susceptibility with values ranging from
2 to 6 (Figure 4). However, because of the rather
flat topography of the Harlow Creek watershed, values ranged from 2 (least
susceptible to erosion) to 4 (most susceptible to erosion).