Contact details
Links
Scientific classifications
- 1. Natural sciences
- 1.5 Earth and related Environmental sciences
- Environmental sciences (social aspects to be 5.7)
- Meteorology and atmospheric sciences
- 1.5 Earth and related Environmental sciences
Main research areas
My research focuses on the urban atmospheric environment, in particular the high-resolution in situ investigation of the street-level spatial and temporal variability of aerosol particles (PM10 and PM2.5). Conventional fixed monitoring stations cannot capture the heterogeneity caused by traffic, urban morphology and local microclimatic processes. Therefore, we investigate the number and mass concentrations of different aerosol size fractions in Budapest using a bicycle-mounted mobile measurement platform that combines calibrated reference instruments and low-cost sensors. The aims of the measurements are to establish a quality-assured, georeferenced reference database, to identify air-pollution hotspots, to develop field validation methods for mobile and low-cost instruments, and to quantify personal aerosol exposure. The dataset, comprising more than two hundred measurement routes, enables the analysis of local air-quality zones and of traffic- and morphology-related effects. Closely related work includes the comparison of indoor and outdoor aerosol concentrations and the development of measurement QA/QC and data-processing algorithms.
My second main research direction is the dynamic downscaling of global machine-learning weather prediction (MLWP) models — primarily GraphCast and ECMWF AIFS — with WRF and WRF-Chem for urban and regional air-quality forecasting. MLWP forecasts are comparable to, and often better than, conventional physical forecasts, but their spatial (0.25°) and temporal (6 h) resolution, the limited number of vertical levels and the absence of several surface, soil and turbulence variables prevent their direct use in urban-scale atmospheric chemistry. My task is to nest WRF/WRF-Chem in MLWP fields, to model the missing input variables, to tune mesoscale parametrizations and to validate the simulations. Chemistry calculations use CAMS emission inventories and boundary conditions; the target species are mainly PM2.5, PM10 and NOx. The aim is to produce fine-resolution air-quality forecasts for Hungary and to assess the suitability of AI-generated atmospheric fields — which do not contain explicit physics — for coupled chemistry-transport modelling. Street-level mobile aerosol measurements from our own campaigns support model verification and subsequent statistical downscaling.