Simulation of greenhouse gases in the atmosphere

Our group uses and improves atmospheric transport models to enhance our current understanding of the carbon cycle and provide informed support for mitigation efforts. We use the transport models to simulate the dispersion of greenhouse gases (especially carbon dioxide and methane) to compare between reported emissions and measured concentration. The comparison allows conclusions about the location, timing, and magnitude of sources and sinks.

Greenhouse gases on different scales

Greenhouse gases are dynamically distributed in the global atmosphere.  Therefore, in addition to sources and sinks, transport processes influence the actual concentration in the atmosphere. These transport processes act on scales ranging from global over regional to local. We, therefore, use different models to map the relevant processes on the respective scales:

From global...

Enhancements in Australian XCO2 caused by transported fire CO2 emissions from different regions.

Enhancements in Australian XCO2 caused by transported fire CO2 emissions from different regions. Source: Eva Schoemann

Global atmospheric inverse model transport prior surface fluxes forward and assimilate measurements to generate global modeled atmospheric concentrations.  By varying the surface fluxes, the differences between measured and modeled concentrations are minimized. In collaboration with Sourish Basu (Earth System Science Interdisciplinary Center, University of Maryland), we use the TM5-4DVAR transport model with input from meteorological data to determine the effects of transport and emissions on a continental scale. TM5-4DVar employs a 4-D variational method and optimizes fluxes on a 6° × 4° longitude–latitude grid. regional...

Comparison of total column XCO2 from OCO-2 overpass and WRF simulation.
Comparison of total column XCO2 from OCO-2
overpass and WRF simulation. Source: Lukas Pilz


To analyze mesoscale phenonoma, we use the Weather Research and Forecasting (WRF) Model to simulate the transport of greenhouse gases in the atmosphere. WRF is a modular Eulerian mesoscale model, which simulates atmospheric dynamics based on actual atmospheric conditions or idealized conditions. It allows for nesting a higher-resolved "child" domain in a coarser "parent" domain to zoom into specific regions. WRF-Chem simulates CO2 concentrations by coupling the atmospheric dynamics to CO2 surface fluxes. A Langrangian particle model is used to determine the areas influencing specific measurements.  This knowledge is used in an inversion scheme to quantitatively compare different observation strategies. This is part of the BMBF-funded project "Intergriertes Treibhausgas Monitoring System (ITMS)". local...


Dispersion of a source in Heidelberg as simulated by GRAMM/GRAL
Dispersion of a source in Heidelberg as simulated
by GRAMM/GRAL, Source: Sanam Vardag

To support political decision-makers in their mitigation efforts, inner-city-resolved CO2 information is needed. The high-resolution transport model GRAMM/GRAL is composed of a mesoscale model (GRAMM) coupled with computational fluid dynamics (CFD) modeling (GRAL). It is a Reynolds Averaged Navier Stokes (RANS) model that can run at 10m x10m resolution over long periods, ranging from months to years. The high temporal and spatial resolution makes the model suitable for monitoring suburban greenhouse gas dispersion. To achieve the required high spatial and temporal resolution with the available computational capacity, the GRAMM/GRAL model uses a so-called "catalog approach": the model simulates a catalog of stationary wind situations for different stability classes, wind speeds ,and wind directions. These 1000+ catalog entries represent a hypothetical simulated wind field library. We choose the wind field by minimizing the differences, between observed and simulated wind speed and direction. The group is collaborating with the group of Prof. Dr. Dominik Brunner (EMPA).



...from natural to social sciences and with stakeholders!

Diurnal variation of NO2 concentration at Berliner Straße in Heidelberg (2017-2019).

Diurnal variation of NO2 concentration at Berliner Straße in Heidelberg (2017-2019).
Source: Simone Wald

The work in the group benefits from the close scientific cooperation of researchers at Heidelberg University. Our overarching goal is to advance climate research in an action-oriented manner. The Heidelberg Center for the Environment (HCE) bundles the activities from different disciplines. Also, the group benefits from a lively exchange with decision-makers e.g. the city of Heidelberg. In this context, we apply the research results in a needs-oriented manner.

The group "Simulation of greenhouse gases in the atmosphere" is part of the group "Remote Sensing of the atmosphere"  (Prof. André Butz). Collaboration with other groups and institutes is welcome. 


Dr. Sanam Vardag
Universität Heidelberg
Institut für Umweltphysik
Im Neuenheimer Feld 229 (3rd floor, room 330)
69120 Heidelberg

Phone: +49 6221 54 6511
Email: svardag ( at )


Dr. Sanam Vardag (group leader)

Simon Cello (Master candidate)

Sonja Gabriel (Master candidate)

Miao Huang (Master candidate)

Marcia Kroker (Master candidate)

Pernilla Kühn (Master candidate)

Christopher Lüken-Winkels (PhD candidate)

Robert Maiwald (PhD candidate)

Maximilian May (PhD candidate)

Eva-Marie Metz (PhD candidate)

Lukas Pilz (PhD candidate)

Claus Sarnighausen (Master candidate)

Anna Sommani (PhD candidate)



Lukas Artelt (M.Sc., 2024), Ines Dillerup (M.Sc., 2023), Leonie Kemeter (M.Sc., 2023),  Robert Maiwald (M.Sc., 2023), Christopher Lüken-Winkels (M.Sc., 2022), Lukas Pilz (M.Sc., 2021), Lena Schreiner (M.Sc., 2021), Simone Wald (M.Sc., 2022), Lea Lilli zur Lage (B.Sc, 2023)