Research interest

The development and application of  land surface models with a focus on the biogeochemical cycles of carbon, nitrogen and phosphorus. My overall aim is to advance the understanding of the biogeochemical cycles and their interactions with the climate system by introducing ecological theory into physically-based models.

Feel free to contact me in case you are interested in any of the topics below: email address


Nutrient limitation of land carbon uptake

Currently, about one third of the fossil fuel CO2 emissions is taken up by land ecosystems, but there is no guarantee that this ecosystem service will continue into the future (Global Carbon Project). In particular, the limited availability of plant nutrients could pose a thread to the continuation of land carbon uptake in the near future (for example Norby et al. (2010)Goll et al., 2012).

Simulations with the Max Planck Earth System Model (MPI-ESM) indicate a dominant role of phosphorus availability in constraining land carbon uptake – not only for tropical regions, but also for higher latitudes (Goll et al., 2012). A recent analyses of foliar nutrition data of European ecosystems supports this finding, showing a decline of leaf P concentrations during the last decades (Jonard et al. , 2015Talkner et al., 2015).

Unfortunately, the interactions between the cycles of carbon, nitrogen and phosphorus are poorly understood. which hampers projections of nutrient availability. A better process understanding is needed which is the aim of my current work in the IMBALANCE-P project.


Figure: The reduction of carbon uptake during the 21st century due to nitrogen (top) and phosphorus (bottom) limitation. Simulations with MPI-ESM under SRES A1B scenario. From Goll et al., 2012.

Uncertainties in CO2 emission from land use and land cover change 

The quantification of sources and sinks of carbon from land use and land cover changes (LULCC) is uncertain (Houghton et al., 2012). We investigated the uncertainty in  the historical and future carbon fluxes in the MPI Earth System Model (MPI-ESM).

The net LULCC flux strongly depends on the assumed fraction of biomass which enters the atmosphere directly due to burning or use in short-lived products. Globally, historical emissions vary up to 25% depending on this fraction which is insufficiently constrained from data. This effect strongly depends on the parametrization of decomposition. Overall, we found a uncertainty in the decadal LULCC fluxes of the recent past due to the parametrization of decomposition and direct emissions of 0.6 Pg C yr−1, which is three times larger than the uncertainty previously attributed to model and method in general (Houghton et al., 2012).

This finding indicates that there is a imbalance between the evaluation and representation  of fundamental processes, like decomposition, and implementing second order processes, like for example dynamical changes in natural land cover, nutrient cycles, or shifting cultivation.

Ecosystem adjustments to increasing temperatures: adaptation and acclimation

Changes in temperature affect fundamental processes and their interactions in ecosystems. Models which incorporate our best understanding of all the modes of action of the individual factors will also capture many of the major interactions regarding temperature.

In model projections for the 21st century, the response of tropical ecosystems to warming dominates the dynamics of the terrestrial sink (Raddatz et al., 2007). It is the tropical ecosystems where temperatures are most likely to exceed the optimum for photosynthesis in a warming climate. There is considerable controversy about the vulnerability of tropical trees to a future warming (Clark, 2004; Feeley et al., 2007; Lloyd & Farquhar, 2008). In particular, the limited understanding of the extent to which ecosystems adjust to gradual changes in temperature (Medlyn et al., 2002; June et al., 2004; Kattge et al., 2007Lin et al., 2012) hampers our ability to constrain ecosystem responses to global warming (Lloyd & Farquhar, 2008; Booth et al., 2012).