Impacts of several years consecutive summer droughts on forest-climate feedbacks - an ecosystem modelling approach

There is an urgent need to assess the resilience to extreme weather events such as repeated and long-lasting summer droughts of common tree species in Sweden. On the global scale, vapor pressure deficit is considered a predictor for climate-vegetation feedbacks. To facilitate development of regional and global models with more robust climate-vegetation feedbacks, we need to improve our understanding of water-carbon relationships at the local scale.

Norway spruce stands for 40% of trees in managed forests in Sweden, and forests are the most important national carbon sink in Sweden. It has relatively shallow roots and may grow on shallow soils, making it extra vulnerable to drought. Drought can cause a chain of vegetation responses, which may trigger several negative impacts on vegetation-climate feedbacks. The forest in SW Sweden:
1) The immediate response of trees to drought is stomatal closure to reduce water loss,
which may reduce plant CO2 uptake and cooling effects of latent heat flux.

2) Drought reduces the abundance and biodiversity of root symbionts within a year and thereby nitrogen (N) and phosphorous (P) supply to the tree host and, hence lowers photosynthesis. We developed a model that explicitly describes the climate-vegetation-symbiont feedback.

3) When drought continues, trees allocate a higher proportion of photosynthetic assimilates to roots and stem at the cost of needles, which further reduces C storage by trees. Higher stem
density reduces stem hydraulic conductivity, leading to many years reduced growth.

4) After long-lasting drought, trees may prematurely defoliate, aggravating further the
reduction of C uptake by trees.

Time periods to trigger these responses to drought as well the time needed for recovery vary from hours to years. Hence, a dynamic process-based model, that integrates dynamical cycling of water, heat, C, and macronutrients N and P within an ecosystem at a high temporal resolution (e.g., Coup-CNP), is needed to investigate impacts of drought on ecosystems and ecosystem resilience. There is growing experimental evidence that water storage (WS) in stems can alleviate drought stress and facilitate transport of nutrients in the trees. Many Earth system models ignore WS. With the development towards applications of Earth System models with shorter time-steps, there is an increasing need to explicitly describe diurnal patterns of climate-vegetation feedbacks that can be of importance for responses to extreme weather at the seasonal and annual scale.

The novelty of the proposed modelling approach is that we simulate the dynamics of water,
heat, C, N and P cycles with an explicit description of the internal tree WS. The overall aim is to analyze to what degree a state-of-the-art model can explain the dynamics of a forest ecosystem C sequestration in response to several years of consecutive summer droughts. Hereby, we disentangle the coupling and hierarchy of vegetation-climate feedbacks at the ecosystem scale and identify possible overseen feedbacks of importance. We hypothesize that including the tree WS will improve the predictive power of the model for estimating the ecosystem C cycle.

Modelling approach
We will use the version of CoupModel v6.0 called Coup-CNP. This version explicitly integrates the P cycle and root-mycorrhiza fungi symbiosis, besides the C and N cycles and its development was co-financed by MERGE. The CoupModel dynamically calculates water, heat, C, N and P fluxes in plant ecosystems in concert at an hourly or daily resolution and at the local scale, making it particularly suitable for applications related to forest resilience, recovery, and management. The Coup-CNP was evaluated against observations of managed forests in different climatic regions in Sweden. The CoupModel is commonly applied by assuming no WS in plants, but it is possible to explicitly calculate the dynamics of WS with the SPAC module as a function of water potentials in plants and soil.

Earlier versions of the CoupModel have been extensively evaluated against observational data, including data gathered at the Skogaby experimental site and managed forests or used to estimate climate change impacts. In this project, we will evaluate the model’s predictive power of multiple components of the water, energy, and C cycles by common measures such as model efficiency, normalized average error, and variance ratio.

This project is financed by the strategic research area MERGE.