PROGNOSTIC VS MEASURED MET DATA FOR ODOR DISPERSION MODELLING: A DUAL-SITE CASE STUDY

Olfactory nuisance is a parameter that is increasingly growing in importance within environmental impact assessments. The technical problem of odor exposure assessment is not trivial. Despite this fact, the most widespread technique, at a regulatory level which is prescribed to be used for odor impact assessment, is atmospheric dispersion modelling. Although some criteria for the choice of model type are widely accepted, or at least prescribed, this is not the case for the choice of weather data source. In the present work, a simulation of a real-case odor emission source is considered, and different kinds of meteo datasets have been considered: WRF prognostic data, surface and upper air measured data, and a composition of both of them. The simulation of the wind field has been conducted with the CALMET diagnostic meteo preprocessor considering the mentioned different input data; the odor dispersion simulation has though been conducted with the Gaussian Lagrangian CALPUFF model. Two different geographic areas have been considered: one in a tropical american island and one in a central european site. Odor impact is itself a peak and not an average exposure phenomenon: the regulatory levels are currently expressed as yearly peaks or different levels of yearly percentiles. In the present study, the Italian regulatory guideline has been considered valid for both the geographical sites: so percentiles of 98th order have been considered as representative of odor impact. The outcome of the study is that, despite the choice of the kind of the meteo input dataset, the outcomes of the odor impact assessment arise largely comparable.