Surrogate modelling for the forecast of Seveso-type atmospheric pollutant dispersion

Abstract This paper presents a framework for the development of computationally-efficient surrogate model air pollution dispersion. Numerical simulation dispersion is fundamental importance mitigation in Seveso-type accidents, and, extreme cases, design evacuation scenarios which long-range forecasting necessary. Due to high computational load, sophisticated programs are not always useful prompt studies and experimentation real time. Surrogate models data-driven that mimic behaviour more accurate complex limited conditions. These computationally fast enable efficient computer with them. We propose two methods. The first method develops grid independent dynamic second reduced interpolation outputs. Both demonstrated an example realistic, controlled experiment complexity based on approximately 7 km radius around thermal power plant Šoštanj, Slovenia. results show acceptable matching between original noticeable improvement load. makes obtained appropriate further confirms feasibility proposed method.