SCENIC ASSESSMENT OF THE ST. CROIX NATIONAL SCENIC RIVERWAY
Bart Richardson, Minnesota Department of Natural Resources David Pitt, University of Minnesota, College of Design, Department of Landscape Architecture Andrea Wedul, Washington Conservation District
Sponsored by the Midwest Regional Office of the National Park Service (NPS), this project sought identification and mapping of scenic values as perceived by stakeholders who live, work, or play in the St. Croix River valley extending five miles on either side of the river from Prescott, WI to Danbury, WI.
Sixty-four photographs representing landscape views in the valley were rated for visual attractiveness by 209 study participants on an 8-point scale. A perceived attractiveness value for each view was created by calculating the photograph's mean attractiveness value. The landform and land cover content of each photograph was catalogued on 25 variables. A multiple regression model of the attractiveness scores as a function of the combined effects of five view content variables explains 72% of the variability in the attractiveness scores. The standardized form of the model is:
Y^ = 0.50 (X1) – 0.55 (X2) – 0.23 (X3) + 0.24 (X4) + 0.17 (X5)
where:
Y^= predicted visual attractiveness value of a landscape view and
X1= presence of water, wetland or meadow
X2= presence of any form of development
X3= presence of billboards or towers
X4= presence of older development
X5 = presence of topographic diversity
Key landscape dimensions identified in the analyses were rated in a grid-based GIS database on a scale of 1 to 100 and weighted based on the standardized regression coefficients. The following equation was applied to the viewshed analysis grids of the database:
total scenic value = ([naturalism_score] * 0.5) + ([topo_score] * 0.17) + ([historic_score] * 0.24) + ([forest_score] * 0.1) + ([water_score] * 0.1) - ([highway_score] * 0.1) - ([powerline_score] * 0.23) - ([road_score] * 0.1)
The spatial model was validated by correlating scenic values for selected views from the survey analysis with predicted values derived from the GIS analysis.
