Search for a command to run...
This study presents the results of smoke simulations from wildfires that occurred in September 2024 in northern Ukraine. The case study focuses on peatland fires near Loshakova Huta (Chernihiv region), specifically during 19–23 September, when smoke was transported toward the capital city of Kyiv under northerly wind direction. Burnt areas were identified using multiple data sources, including Sentinel-2 and Planet satellite imagery (false-colour analysis), EFFIS (European Forest Fire Information System) polygons, land surface temperature anomalies obtained from MODIS and VIIRS satellite instruments (NASA FIRMS (Fire Information for Resource Management System) platform) and EUMETSAT LSA SAF (Land Surface Analysis Satellite Application Facility), and Sentinel-3 fire radiative power data. Emissions of major air pollutants were estimated using the Fuel Fire Tools application, which integrates the FCCS (Fuel Characteristic Classification System), the CONSUME fuel consumption model, and FEPS (Fire Emission Prediction Simulator). The FCCS module provided fire behaviour parameters and fuel loading for the identified burnt areas based on a 30-m fuel map developed for Ukraine, which had previously been modified by introducing peat fuel classes using available soil data. Total emissions and their temporal dynamics were then calculated. Smoke transport and dispersion were simulated by means of the CALPUFF Lagrangian puff model and HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory Model) at 1 km spatial resolution. Both models were applied to simulate the transport of carbon monoxide and particulate matter (PM2.5 and PM10) using meteorological fields simulated by the WRF (Weather Research & Forecasting) model driven by ERA5 reanalysis. Model performance was evaluated by comparing background-corrected modelled pollutant concentrations with measurements from eight air-quality monitoring stations in Kyiv, using several statistical metrics. Data from seven stations were provided by the Department of Environmental Protection and Climate Change Adaptation of the Kyiv City State Administration, and data from one station were provided by the SI “Marzieiev Institute for Public Health of the National Academy of Medical Sciences of Ukraine”. The results showed acceptable accuracy for both models, despite the relatively large distance from the active fires (~60-70 km) and uncertainties related to other emission sources and surrounding conditions near the monitoring sites. Periods of substantial concentration increases were also well reproduced by the models.Peatland wildfires pose a significant public health hazard due to their enormous emissions of air pollutants, particularly the PM2.5 fraction. This case study is valuable due to the availability of qualitive satellite-based information, on-site documentation, and reliable air-quality measurements in Kyiv. The results indicate that the CALPUFF and HYSPLIT models can adequately reproduce smoke plume transport and dispersion when peat-fire source terms are properly parameterized, especially with respect to emission rates, and when accurate meteorological input is used.