Assessing Model Treatment of Drought Legacy Effects in the Amazon

Abstract:

Extreme climate events play an important and potentially lasting role in terrestrial carbon cycling and storage. In particular, satellite and in-situ measurements have shown that forest recovery time following severe drought can extend several years beyond the return to normal climate conditions. However, terrestrial ecosystem models generally do not account for the physiological mechanisms that cause these legacy effects and, instead, assume complete and rapid vegetation recovery from drought. Using a suite of fifteen land surface models from the Multi-scale Synthesis and Terrestrial Model Intercomparison Project (MsTMIP), we assess models’ ability to capture legacy effects by analyzing the spatial and temporal extent of modeled vegetation response to the 2005 Amazon drought. We compare the simulated primary production and ecosystem exchange (GPP, NPP, NEE) to previous recovery-focused analysis of satellite microwave observations of canopy backscatter. Further, we evaluate the specific model characteristics that control the timescale and magnitude of simulated vegetation recovery from drought. Since climate change is expected to increase the frequency and severity of extreme climate events, improving models’ ability to simulate the legacy effects of these events will likely refine estimates of the land carbon sink and its interannual variability.