Simulation of Greenhouse Gases in the Atmosphere (Vardag - Group) Research methods
In our group, we use atmospheric transport models to study the dispersion of trace gases in the atmosphere. Our work spans multiple scales—from urban to global—which requires the use of different models with varying resolutions.
Urban
To simulate the dispersion of trace gases within cities, it is necessary to account for airflow around buildings and within street canyons. While city-wide modeling at such high resolutions is typically computationally demanding, we use a Reynolds-Averaged Navier-Stokes (RANS) model capable of computing high-resolution (10 m) wind fields over extended periods. GRAMM/GRAL simulates a wide range of hourly stationary wind fields that cover all realistic meteorological conditions. Each hour, one of these precomputed wind fields is selected based on a matching algorithm that compares simulated and observed wind speed and direction. By linking the best-fitting wind fields hour by hour, the model generates an artificial dynamic. This so-called “catalog approach” further reduces computational demand. GRAMM/GRAL therefore offers a promising way to simulate urban concentration enhancements with sufficient accuracy and manageable computing costs. The model thus has the potential to provide information on timescales relevant for policy-making.
Regional
To investigate regional and mesoscale effects, we use the Weather Research and Forecasting (WRF) model. WRF simulates atmospheric dynamics based on actual or idealized atmospheric conditions. A nested modeling approach—embedding a higher-resolution domain within a coarser one—allows us to focus on specific regions. WRF-Chem simulates CO₂ concentrations by coupling atmospheric dynamics with CO₂ emissions. We currently use WRF-Chem at a 1 km resolution in selected focus regions.
The meteorological fields calculated in WRF can also serve as input for Lagrangian models to better determine the source regions influencing specific measurements. For this, we use the model FLEXPART-WRF. This approach is particularly advantageous when high resolution is needed, such as for measurements near large emission sources. The information about the influence of surrounding areas on potential observation sites ("footprints") is used in an inversion to quantitatively compare different observation strategies. These activities are part of a project funded by the Federal Ministry of Research, Technology and Space (BMFTR).
Global
To study continental-scale biogenic fluxes, we use coarser-resolution global models, specifically TM5-4DVAR, which is operated by Sourish Basu (Earth System Science Interdisciplinary Center, University of Maryland). TM5-4DVAR uses observed meteorological data as input and applies a four-dimensional variational (4D-Var) method to optimize fluxes on a 3° × 2° longitude-latitude grid. These so-called “top-down” fluxes can be compared with process-based (“bottom-up”) models to gain a better understanding of biogenic processes of continental importance.
To better interpret total column measurements, we use the FLEXPART model, which is driven by reanalyzed meteorological fields and allows us to infer large-scale source and sink patterns.