Calibrating INPUFF-2 model by resident-panelists for long-distance odor dispersion from animal production sites

An air dispersion model, INPUFF-2, was calibrated for long-distance odor estimation. Odor events were monitored from June to November 1999 by 19 trained resident-panelists living on a 4.8 × 4.8 km (3 × 3 miles) grid of farmland containing 20 livestock/poultry farms. The panelists used a numerical 0 to 3 scale to assess the odor intensity. To convert odor intensity to concentration, the correlation between odor intensity (on a 0 to 5 n-butanol water solution scale) and concentration (in OU/m³) for swine and cattle odors were developed. Odor emissions from the farms were measured and weather conditions were monitored over the same period. A total of 296 odor events were detected among which 170 odor events were simulated by the model. One hundred twenty three of these 170 odor events, i.e. 72.3%, were classified as faint odor (intensity 1). The comparison between the modeled and the measured odor intensity indicated that the model successfully estimated odor intensity 1 (faint odor) traveling up to 3.2 km under stable atmospheric conditions (P > 0.05). However, the model underestimated moderate to strong or very strong odors and odors that occurred during neutral or unstable weather as compared with the field measured data (P < 0.05). The overall percentage of agreement was 81.8%. The results indicate that the model could serve as an effective tool for agriculture odor dispersion estimation from animal production sources.