The Impact of Amazonia Droughts on Carbon Dynamics as Simulated by an Ensemble of Community Earth System Models

Amazonia biomass contains significant amount of carbon (~120Pg), and so relatively small changes in Amazon carbon dynamics have the potential to substantially affect the atmospheric CO2 concentration and thus the rate of climate change (e.g., Cox et al., 2013). Disturbance from severe climate anomalies, such as drought, can potentially result in large carbon emissions comparable to deforestation (Saatchi et al., 2013). In this study, we analyze the impact of 2005 and 2010 Amazonia droughts on changes in carbon dynamics and atmospheric CO2 simulated by the Community Earth System Model (CESM). The CESM has coupled atmosphere and land processes. The land model (CLM) includes an active biogeochemical module (CN). We assimilate raw meteorology observations with Ensemble Kalman Filter (EnKF) to create a 64-member ensemble of realistic atmospheric forcing and modeled carbon states. The mean and spread of the carbon states are analyzed in terms of their relationships to the mean and spread in atmospheric forcing and feedbacks, to see if threshold behavior is inherent in the coupled system. The simulated GPP and the CO2 responses to terrestrial biosphere fluxes are also compared with observations, such as MODIS GPP and the chlorophyll fluorescence observations from Greenhouse Gases from Satellite (GOSAT), to quantify the constraints the observations may place on the modeling.