B031-0015
Global controls of environmental and land surface conditions on forest carbon stock dynamics

Wednesday, 9 December 2020
Poster
Simon Besnard1,2, Maurizio Santoro3, Oliver Cartus3, Naixin Fan4, Sujan Koirala1, Nora Linscheid1, Ulrich Weber1 and Nuno Carvalhais1,5, (1)Max Planck Institute for Biogeochemistry, Department of Biogeochemical Integration, Jena, Germany, (2)Wageningen University and Research Center, Laboratory of Geo-Information Science and Remote Sensing, Wageningen, Netherlands, (3)Gamma Remote Sensing, Guemligen, Switzerland, (4)Max Planck Institute for Biogeochemistry, Jena, Germany, (5)Universidade Nova de Lisboa, Departamento de Ciências e Engenharia do Ambiente, Lisbon, Portugal
Abstract:
Quantifying the response of forest dynamics and the changes in forest carbon pools to changes in climatic and land surface conditions at global scales has been rather challenging, leading to large uncertainties in characterizing the terrestrial carbon cycle. It is expected that favorable environmental conditions (e.g. via water supply or temperature) result in a positive balance between biomass inputs through growth and outputs through mortality, while opposite responses would be observed during adverse conditions (e.g. limited water supply or nutrient availability). Yet, it has been difficult to corroborate such assumptions at large scales across different ecosystems and climate gradients due to a lack of observations describing long-term changes in above-ground biomass (AGB) over-time globally. Based on a global time series of annual AGB estimates from C-band scatterometer data, we map the changes in AGB in space and time as well as provide insights on the responses of forest dynamics to fluctuations in forest cover changes and environmental conditions from local to regional scales for the period 1992-2018. Globally, changes in forest carbon stocks correlate positively with changes in forest cover fraction (r = 0.62), and to a lesser extent with environmental conditions (r=0.29 and 0.43 for air temperature and total precipitation, respectively). However, the controls on forest changes from environmental conditions, productivity and vegetation properties differ at local scales, conditional on climate and ecological space. A variable importance analysis, performed using a Random Forest algorithm, reveals that among productivity, forest cover change and climatic conditions, the latter is overall the main control of changes in forest dynamics at local scales, albeit productivity also exerts a control on forest dynamics regionally (e.g. dry tropics). Finally, we find regional evidence that carbon gains from forest growth are generally coupled with a carbon sink at the ecosystem level. These findings provide new insights into quantifying the contributions from management and disturbance regimes to vegetation growth, apart from the direct links from climate and productivity, providing insights into the global importance of the underlying mechanisms of forest dynamics on the global carbon cycle.