Using Airborne Microwave Remotely Sensed Root-Zone Soil Moisture and Flux Measurements to Improve Regional Predictions of Carbon Fluxes in a Terrestrial Biosphere Model
Abstract
North American ecosystems are critical components of the global carbon cycle, exchanging large amounts of carbon dioxide and other gases with the atmosphere. Net ecosystem exchange (NEE) of CO2 between atmosphere and ecosystems quantifies these carbon fluxes, but current continental-scale estimates contain high levels of uncertainty. Root-zone soil moisture (RZSM) and its spatial and temporal heterogeneity influences NEE and improved estimates can help reduce uncertainty in NEE estimates. We used the RZSM measurements from the Airborne Microwave Observatory of Subcanopy and Subsurface (AirMOSS) mission, and the carbon, water and energy fluxes observed by the eddy-covariance flux towers to constrain the Ecosystem Demography Model 2.2 (ED2.2) to improve its predictions of carbon fluxes. The parameters of the ED2.2 model were first optimized at seven flux tower sites in North America, which represent six different biomes, by constraining the model against a suite of flux measurements and forest inventory measurements through a Bayesian Markov-Chain Monte Carlo framework. We further applied the AirMOSS RZSM products to constrain the ED2.2 model to achieve better estimates of regional NEE. Evaluation against flux tower measurements and forest dynamics measurements shows that the constrained ED2.2 model produces improved predictions of monthly to annual carbon fluxes. The remote sensing based RZSM can further help improve the spatial patterns and temporal variations of model NEE. The results demonstrate that model-data fusion can substantially improve model performance and highlight the important role of RZSM in regulating the spatial and temporal heterogeneities of carbon fluxes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.B33B0480Z
- Keywords:
-
- 0414 BIOGEOSCIENCES Biogeochemical cycles;
- processes;
- and modeling;
- 0430 BIOGEOSCIENCES Computational methods and data processing;
- 0480 BIOGEOSCIENCES Remote sensing;
- 0428 BIOGEOSCIENCES Carbon cycling