Page path:

Biogeochemical Prediction Research Group

Group Leader

Biogeochemical Prediction Research Group

Dr. Hongmei Li

MPI for Marine Microbiology
Celsiusstr. 1
D-28359 Bremen
Germany

Room: 

2125

Phone: 

+49 421 2028-7010

Dr. Hongmei Li

The Biogeochemical Prediction Group investigates how the global carbon cycle and ecosystem-relevant ocean stressors and extremes evolve across timescale - from near-term variability to long-term climate change. The carbon cycle is central to the Earth system as it connects physical climate dynamics with biogeochemical processes, shapes the accumulation of CO₂ in the atmosphere, and influences climate risks and mitigation pathways.

A key scientific challenge is understanding how natural climate variability and human-driven change interact with ocean biogeochemistry and marine ecosystems. These interactions influence the ocean’s capacity to absorb carbon and the combined pressures of ocean warming, acidification, and oxygen loss.

With Earth System modelling and AI/ML approaches combining observational evidence, we address how marine biogeochemical change translates into ecosystem stress and climate feedbacks. Our vision is to build an interdisciplinary research program that links process understanding with a special focus on the pathways from microbial mechanisms to large-scale Earth system variability, prediction, and projection.

Illustration
The illustrated time series of the atmospheric CO₂ growth rate shows annual means from model simulations plotted together with observations. We conduct three sets of simulations (from left to right in sequential order): (a) uninitialized “free” simulations, which are the same as the freely evolving Coupled Model Intercomparison Project (CMIP) historical type simulations, (b) an assimilation simulation to reconstruct the evolution of the climate and carbon cycle towards the real world by nudging the model towards observation and reanalysis data during its integration, and (c) initialized predictions started from reconstruction states produced by the assimilation simulation and integrated freely (i.e., no nudging towards observations) for 5 years. The time series in (a) shows that the uninitialized simulations capture the long-term trend well, but the year-to-year variations are out of phase with the observations. The time series in (b) shows that the assimilation simulation forces the variations in the uninitialized freely run simulation towards the real world and results in a reconstruction that is closer to the observations. Panel (c) shows the reconstruction together with the 5-year-long initialized retrospective predictions.
Source: Li et al. (2023), Reconstructions and predictions of the global carbon budget with an emission-driven Earth system model, CC BY 4.0
Back to Top