Modeling human judgments of urban visual air quality

Abstract

The overall approach to establishing a complete predictive model link between pollutant emissions and human judgments of urban visual air quality (UVAQ) is presented. The field study design and data analysis procedures developed for analyzing the human components of visual air quality assessment are outlined. The air quality simulation model which relates pollutant emissions to human judgments of visual cues which comprise visual air quality judgments is described. Measured and modeled cues are compared for five typical visual air quality days in the winter of 1981 for Denver, Colorado. The comparisons suggest that the perceptual cue model, based on dispersion and radiative transfer theory, does not adequately predict human judgments of UVAQ cues. Analysis of the limits of predictability of the human judgments and the predictive capability of the model components indicates that the greatest improvements toward achieving a predictive UVAQ model lie in a reformulation of the theoretical descriptions of visual cues.