PARENT SESSION
Symposium 11: Global land science: Dynamics of the coupled human-environment system
Organized by: DS Ojima and EF Moran
Wednesday, August 10, 8:00 AM - 11:30 AM, Meeting Room 517 B, Level 5, Palais des congrès de Montréal

Estimation of regional carbon dynamics in complex landscapes by model-data fusion using ground, airborne and satellite observations.

Schimel, David^*,1, Ojima, Dennis², Running, Steve³, Stephens, Britt¹, Monson, Russ ⁴, Sacks, Bill ¹, ¹ NCAR, Boulder, CO, USA² Colorado State University, Ft Collins, CO, USA³ University of Montana, Missoula, MT, USA⁴ University of Colorado, Boulder, CO, USA

ABSTRACT- Ecologists and meteorologists have often shied away from focusing on biogeochemical and micrometerological processes in mountainous landscapes, because of the complexity, heterogeneity and practical challenges of working there. However, mountainous landscapes play a crucial role in the biological, biogeochemical and hydrological function of, especially, mid-latitude regions where they are also experiencing intense land use. While mountain landscapes are complex, as identified in pioneering watershed biogeochemistry studies, they also have highly organized gravitational flows of water and air that provides an organizing principle absent in simpler landscapes. In 2004, we conducted a research program aimed at quantifying carbon fluxes in the Colorado Rockies that combined airborne, ground-based and satellite data using novel modeling techniques. The results reveal the carbon cycle of a large forested region at multiple scale. We link detailed process information from plot and eddy covariance measurements to spatially extensive and explicit information from remote sensing and spatially extensive (but not explicit) constraints from aircraft measurements. The approach provides quantification of regional fluxes from aircraft and explanation from su-scaled process studies. We integrate the top-down and bttom-up approaches using "data assimilation" models that optimally transfer parameters between scales of measurement. In this approach, local time-series and experimental data provides information on dynamic processes, while aircraft and satellite data provide "snapshots" of information needed to address scaling to entire regions. The project involves diverse techniques, linked by their common dependance on well-defined orographic air flow as an integrating technique. Because of their historic landuse, the Rockies are a highly active carbon flux region with peak flux rates similar to mesic temperate forests.

Key words: carbon cycle, mountains, assimilation modeling