Land surface models LSMs examine the exchange of carbon, water and energy between Earth’s surface and the atmosphere, and are a cornerstone of climate models. However, they are also simplistic, and rely on fixed values to estimate the amount of water transpiring from forests.
The EU-funded M-TRAIT project set out to improve the quality and predictive power of LSMs. “This type of model is calibrated heavily to current conditions, so is not credible outside of the current climate,” explains Aude Valade, M-TRAIT project researcher. As climate change is expected to give rise to more frequent and severe droughts, it’s important to be able to model forests under these new conditions, she adds. LSMs use a large number of parameters, some of which are related to the physiology of plant cover. But plants can adapt their behaviour in response to a changing climate. “Typically, the calibration is set under current climate conditions, and tuned so the model matches observations,” notes Valade. “The problem is that the future climate we’re expecting is not something that exists today.” In order to improve the LSM, more accurate data about trees’ physiological response to climate change needed to be included. This included parameters such as p50 and the hydrological safety margin, two measures of how much water stress a tree can endure.
“We want to know how forests will react to drought, so we need to know how the behaviour of a single tree will change,” adds Valade. By understanding how each trait changes under drought stress, researchers could predict the change in global behaviour of the forest. To find this data, Valade and her colleagues turned to published data on traits measured around the globe, as well as forest plots across Catalonia inventoried by the Centre for Ecological Research and Forestry Applications (CREAF). They gathered over 4 500 datapoints across 24 different traits to determine the plasticity of the trees’ behaviour. The results are still being analysed, but Valade says early evidence shows the models are more robust. “Currently we use one p50 value per tree species for all of Europe. Instead, I had a value for each plot, so the effect of climate conditions was taken into account,” she says. “I could already see the measured value was very different from the pre-calibrated one. You could really see the model was better.”
The research was undertaken with the support of the Marie Skłodowska-Curie Actions programme. “This work wouldn’t have been possible without that funding,” explains Valade. “It gave me the ability to jump into a big group, join my objective with the one going on, and take the time to explore the model’s behaviour.” The support also gave Valade the freedom to explore dissemination efforts, such as developing a board game to introduce climate concepts such as mitigation and adaptation to high school and university students. “This was translated into Catalan and Spanish and is now being used in schools there,” notes Valade. Having completed her project, Valade has since joined the French Agricultural Research Centre for International Development (CIRAD) in Montpellier to apply her research to tropical forests.
M-TRAIT, tree, model, water, forest, drought, p50, Catalonia, plant