Atmospheric Dispersion Modeling

A dispersion model is only as good as the question it answers, the data it is fed, and the honesty of its limitations section. This advanced five-day course takes environmental engineers, EIA consultants, permitting specialists, HSE professionals and regulators from the physics of the atmosphere to a modelling study that a reviewer can follow and reproduce. It is built around the Arabian Gulf setting: intense summer convection, sea and land breezes, shamal winds and high natural dust all challenge the assumptions behind standard models. The programme starts with the boundary layer, turbulence, Pasquill-Gifford and Monin-Obukhov stability, and mixing height. It then shows how to characterise sources correctly: emission rates, true exit velocity, plume rise, building downwash, GEP stack height, flares and fugitives. Learners master the Gaussian plume equation by hand, including dispersion coefficients, maximum concentrations, averaging-time conversion and line sources. They then move to the tiered model hierarchy: AERSCREEN and AERMOD or ADMS for permitting, NO2 conversion methods, and matching the model statistic to the way a standard is written. Next come advanced puff, particle, Eulerian and CFD models, and the preparation and quality control of meteorological data, including WRF. Special cases cover dense gas, toxic endpoints, odour, deposition, chemistry, coastal fumigation and complex terrain. Applications include impact assessment, ambient standards, process contribution and background, health risk, stack design, emission limits and emergency planning. The final technical modules cover model evaluation statistics, uncertainty, sensitivity testing and reporting with QA/QC. Eight hands-on workshops use calculators, spreadsheets, AERSCREEN, WRPLOT and ALOHA, each with a model answer. The course ends with a complete applied case on a 600 MW combined-cycle plant on the UAE coast and a review clinic in which learners critique a flawed submission. Throughout, the course stresses doing a hand check before trusting any model run, and knowing that the highest predicted values are typically accurate to within a factor of two. Assessment is an online 20-question exam at a 70% pass mark, leading to a verifiable SKEDMIA certificate.