A Visual Aesthetic Prediction Method for Use in Benefit-Cost Analysis

Abstract

This report discusses the development of a method to predict landscape visual aesthetic changes caused by the siting of nuclear power plants. The methodology uses public perceptions as a measure of visual aesthetics. Individuals scored landscape photographs on a 0 to 50 visual aesthetic scale. The visual aesthetic scores were explained statistically by landscape characteristics, percent of the scene in clear, still water, and characteristics of the individuals scoring the photographs. Three visual aesthetic relationships were empirically estimated. The first is the relationship among group mean visual aesthetic scores and landscape characteristics. The second is the relationship among individual visual aesthetic scores, landscape characteristics, and the characteristics of the individuals who ranked the landscapes. These relationships were estimated using data from two regions in the U.S. and a diverse set of landscape photographs. The third relationship is among group mean visual aesthetic scores for landscapes with a visible nuclear power plant, landscape characteristics, and mean individual characteristics of the groups who scored the landscapes. This relationship was estimated using data from six regions in the U.S. and landscapes showing nuclear plants with a closed cycle cooling system. The statistical results are highly significant. Prediction validity test results indicate that the estimated relationships can predict visual aesthetic scores for groups of individuals outside the samples used to estimate the visual aesthetic relationships. This prediction method is not intended to determine social welfare impact. The social welfare impact of visible change is a measure of how consumers value changes in visual aesthetics relative to all other goods and services. The prediction methodology only provides an estimate of the impact of visible change on the perceived aesthetics associated with that landscape relative to other landscapes. It does not provide a measure relative to all other goods and services consumed.