Ecosystem collapse or recovery? A model-data fusion framework for assessing drought-induced tipping points in a Californian forest.

More frequent and severe droughts are driving increased forest mortality around the globe. We urgently need to describe and predict how drought affects carbon uptake, storage and cycling within ecosystems, and identify thresholds of water stress that can trigger ecosystem collapse. However, quantifying the effects of drought at an ecosystem level is complex because dynamic climate-plant relationships can cause rapid and/or prolonged shifts in carbon balance. The CARbon DAta MOdel fraMework (CARDAMOM), a model-data fusion approach, uses meteorological observations to constrain parameters that assess above and below ground carbon and water ecological processes, while also tracking the historical patterns within the ecosystem (i.e., a memory function). This can provide an observation-informed, long-term, mechanistic link between drought and forest mortality. We employed CARDAMOM to investigate legacy effects of drought on forest carbon pools and fluxes. Our model was constrained by eddy covariance observations and we investigated the cascading effects of insufficient precipitation on plant available water and its consequences on foliar, labile, root and litter carbon pools. We will present our findings on the carbon and water balance of a forest in the Southern Sierras that was severely impacted by the exceptional 2011-2015 California drought. We will show how slight shifts in mean daily precipitation (~10-20% or 0.12-0.24mm increase for 2014 only) can tip an ecosystem from a state of collapse to recovery. Finally, we will highlight the importance of assessing historical fluxes and pools in ecosystems and describe how failure to capture this will increase the uncertainties of the terrestrial carbon cycle.