B32A-04
Effect of non-homogeneity in flux footprint on the interpretation of seasonal, annual, and interannual ecosystem carbon exchange
Wednesday, 16 December 2015: 11:05
2006 (Moscone West)
Anne Griebel1, Lauren T. Bennett2, Daniel Metzen1, James R Cleverly3, George G Burba4 and Stefan K Arndt5, (1)University of Melbourne, Parkville, VIC, Australia, (2)University of Melbourne, School of Ecosystem and Forest Sciences, Creswick, Australia, (3)University of Technology Sydney, Ultimo, Australia, (4)LI-COR Biosciences, Lincoln, NE, United States, (5)University of Melbourne, Parkville, Australia
Abstract:
Carbon flux measurements using the eddy covariance method rely on several assumptions, including reasonably uniform terrain and homogenous vegetation. These are not always possible in complex terrain, structurally variable native vegetation or in disturbed ecosystems. Consequently, an increasing number of flux sites are located over not fully homogeneous areas. This implies that observed year-to-year variations in CO2 budgets may not always be related only to changes in the key driving factors such as weather, canopy state and physiology, but may also be affected by differences in the flux footprints between years. This may bias budget estimates over many locations, since a large number of flux sites are affected by wind channelling, contrasting climatic conditions with wind direction (e.g. maritime sites) and by variations of continental-scale climate patterns that modify prevailing wind directions. We tested the effects of a non-homogeneous footprint on annual carbon estimates for an evergreen forest, where the combination of terrain, weather and anthropogenic management shaped the local forest structure. Interactions among these factors caused the key drivers regulating carbon fluxes (such as LAI, temperature, VPD and turbulence) to vary significantly with wind direction, and their combinations resulted in pronounced carbon sequestration ‘hotspots’ that impacted instantaneous fluxes. These were most distinctive during the summer months, and they varied in extent and magnitude depending on prevailing weather. Consequently, interannual variations in footprints affected up to 18.9% of seasonal estimates during the summer months, and up to 23.1% of annual carbon budget estimates. The footprint-related bias was largest at 48.7% under ‘ideal’ uptake conditions (clear sky, mid-day during summer). We further present a procedure to recognise and quantify the apparent interannual variations in carbon estimates attributable to year-to-year variations in flux footprint.
