Constraining Modeled Biospheric CO2 Fluxes in the Rocky Mountains with Atmospheric Observations: a WRF-STILT-CLM Framework

High-elevation forests in the Rocky Mountains represent a major carbon sink in the Western United States. These forests are particularly vulnerable to climate change, which is expected to increase the frequency, severity, and duration of summer droughts and also the frequency and severity of wildfires and bark-beetle outbreaks, leading to a weakening of their carbon sequestration capacity or even to their transition from a carbon sink to a source. Despite the relevance of these ecosystems, direct carbon budget measurements with flux towers are sparse over the Rockies, especially due to the complex topography. Land-surface models, constrained by observed atmospheric CO₂ concentrations, emerge as an alternative for quantifying and projecting carbon dynamics into the future.

In this study we used CLM4.5 (Community Land Model version 4.5) as the land-surface model. Meteorological data from a high-resolution WRF (Weather Research and Forecasting Model) simulation for the summer of 2012 were used to drive CLM4.5 and also a transport model (STILT; Stochastic Time-Inverted Lagrangian Transport Model). WRF-STILT was used to model the footprints of atmospheric CO₂ observations at three sites in the Regional Atmospheric Continuous CO₂ Network in the Rocky Mountains (Rocky RACCOON) (Hidden Peak, Niwot Ridge, and Storm Peak Laboratory). The footprints were combined with the modeled surface fluxes from CLM4.5 in order to simulate CO₂ concentrations at each site. Surface CO₂ emissions associated with anthropogenic emissions and wildfires (both minor components in comparison with the biospheric fluxes in the study region) were taken from EDGAR (Emission Database for Global Atmospheric Research) and WFEI (Wildland Fire Emission Inventory) inventories, respectively. Modeled CO₂ concentrations were then compared against observations at each site, and the differences were used to guide parameter adjustments in CLM4.5. The initial results of this modeling effort will be presented.