The Role of Vegetation in Past and Future Global Hydroclimatic Change
- Lead PI: Jason E Smerdon , Dr. Richard Seager , Park Williams, Kate Marvel, Ben Cook
-
Unit Affiliation: Ocean and Climate Physics, Lamont-Doherty Earth Observatory (LDEO)
- August 2021 - August 2024
- Inactive
- Project Type: Research
DESCRIPTION: Vegetation plays an important role in hydroclimate through its influence on water and energy fluxes at the land surface. Vegetation will respond in potentially competing ways to changes in temperature, precipitation, and CO2 itself, creating uncertainty in projections of future hydroclimate. Accurately simulating and projecting hydroclimate in Earth System Models (ESMs) requires that their representations of vegetation are constrained and assessed. Our project therefore aims to understand the processes that have coupled vegetation and hydroclimate in the past, assess how state-of-the-art land-surface models represent these processes, and evaluate the implications of these first two insights for future projections of hydroclimate risks. We are pursuing a model-by-model evaluation of the CMIP6-generation of ESMs (with a heightened focus on the E3SM model) to determine how vegetation processes within state-of-the-art land-surface models influence simulated hydroclimate characteristics globally, with a specific focus on droughts defined by soil moisture and runoff. Three questions are driving our research: (1) how do biases in modeled vegetation impose biases in simulations of past droughts? (2) how do these biases affect assessments of future drought risks as CO2 continues to rise? and (3) where should model development focus to improve representations of vegetation and reduce uncertainties in hydroclimate projections? The answers to these questions will allow us to translate modeled ecohydroclimate projections and their uncertainties into ecological and hydrological risk assessments relevant for stakeholders and decision makers. We additionally are working to identify model features and parameterizations tied to systematic biases, therefore identifying clear pathways for model improvements.