Data assimilation

Data assimilation is a data–fusion method that is commonly used to combine observations and models, for example to estimate the state of a dynamical system. Our group uses data assimilation in numerical weather prediction to improve the initial conditions and thereby the weather forecasts. Additionally, we develop new data assimilation methods to estimate carbon emissions and uptake at the Earth’s surface based on atmospheric measurements, also known as “inverse modeling” or “the top-down approach”. This application has two main objectives: (1) to support greenhouse gas emissions monitoring, and (2) to advance our understanding of the carbon cycle.

People working on this topic


Hans Chen

Assistant Professor

Hans Chen was trained in data assimilation with an emphasis on ensemble-based approaches at The Pennsylvania State University under the guidance of Fuqing Zhang. During his PhD, Hans took the PSU WRF EnKF system, which is well-known for hurricane forecasting, and extended it to perform atmospheric CO2 inversion. The new CO2 data assimilation system is now known as the TRACE Regional Atmosphere–Carbon Ensemble system (TRACE; Chen et al., 2023). One of the special features of TRACE is that it uses a full-physics high-resolution mesoscale atmospheric model, which makes it possible to explicitly simulate atmospheric transport errors (Chen et al., 2019) and reduce such errors by assimilating meteorological observations. This is promising, because errors in the simulated atmospheric transport can alter the inference of carbon emissions and uptake from the atmospheric observations.

Additionally, the ensemble-based data assimilation framework implemented in TRACE is computationally efficient at assimilating satellite observations. One of our current research focuses is to extend the TRACE system to take advantage of upcoming measurements from next-generation CO2 satellites, including ESA’s CO2M.

Flow chart of the information flow in the TRACE data assimilation system. TRACE includes both an atmospheric and a carbon cycle data assimilation component and can optimize both to provide optimal estimates of surface emissions and uptake of CO2. From Chen et al. (2023).

We also work together with Marko Scholze at Lund University, as well as Thomas Kaminski and colleagues at the Inversion Lab, on the Carbon Cycle Fossil Fuel Data Assimilation System (CCFFDAS; e.g., Kaminski et al., 2022). This system can simultaneously estimate millions of parameters related to both fossil fuel emissions and the natural carbon cycle.

Other examples of data assimilation applications from our research include paleoclimate data assimilation to estimate Arctic amplification over the past millennium (Fang et al., 2022).

Related publications

Chen, H. W., F. Zhang, T. Lauvaux, M. Scholze, K. J. Davis, and R. B. Alley, 2023: Regional CO2 inversion through ensemble-based simultaneous state and parameter estimation: TRACE framework and controlled experiments. Journal of Advances in Modeling Earth Systems, 15, e2022MS003208,

Kaminski, T., M. Scholze, P. Rayner, S. Houweling, M. Voßbeck, J. Silver, S. Lama, M. Buchwitz, M. Reuter, W. Knorr, H. W. Chen, G. Kuhlmann, D. Brunner, S. Dellaert, H. A. C. Denier van der Gon, I. Super, A. Löscher, and Y. Meijer, 2022: Assessing the constraint of atmospheric CO2 and NO2 measurements from space on city-scale fossil fuel CO2 emissions in a data assimilation system. Frontiers in Remote Sensing, 3, 887456,

Kaminski, T., M. Scholze, P. Rayner, M. Voßbeck, M. Buchwitz, M. Reuter, W. Knorr, H. Chen, A. Agusti-Panareda, A. Löscher, and Y. Meijer, 2022: Assimilation of atmospheric CO2 observations from space can support national CO2 emission inventories. Environmental Research Letters, 17, 014015,

Scholze, M., H. Chen, T. Kaminski, and M. Voßbeck, 2020: Inversion strategy based on joint QND assessments. CHE Consortium.

Chen, H. W., F. Zhang, T. Lauvaux, K. J. Davis, S. Feng, M. P. Butler, and R. B. Alley, 2019: Characterization of regional-scale CO2 transport uncertainties in an ensemble with flow-dependent transport errors. Geophysical Research Letters, 46, 4049–4058,

Ying, Y., X. Chen, Y. Zhang, M. Minamide, R. Nystrom, H. Chen, J. Poterjoy, C. Melhauser, Y. Weng, Z. Meng, A. Aksoy, F. Zhang, 2018: PSU WRF EnKF/4DVar hybrid regional data assimilation system: Technical notes. Department of Meteorology and Atmospheric Science, and Center for Advanced Data Assimilation and Predictability Techniques, The Pennsylvania State University, University Park, Pennsylvania.