Monday, August 3, 2009 - 3:40 PM

OOS 1-7: Assimilating multi-scale observations into ecological models to reduce the uncertainty of carbon flux estimates in northern Fennoscandia

Paul C. Stoy1, M. Williams1, T Hill1, J. Evans2, B. Fletcher3, J. Gornall4, I. Hartley5, P. Ineson6, V. Sloan3, R. Poyatos7, A. Prieto-Blanco8, L. Street1, T Wade1, J Moncrieff1, and J Subke6. (1) University of Edinburgh, (2) Centre for Ecology and Hydrology, (3) University of Sheffield, (4) Met Office, (5) Stirling University, (6) York University, (7) Durham University, (8) UCL

Background/Question/Methods
The structure and function of Arctic ecosystems is responding rapidly to observed climate change. Quantifying the magnitude of these changes, and their implications for the climate system, requires extrapolation and modelling, i.e. 'upscaling' across time and space. Upscaling is a general challenge for ecological science, and we develop a simple framework based on formal information conservation while desceibing the major results of the ABACUS project. ABACUS is a multi-scale investigation of plant and soil process studies, isotope analyses, flux and micrometeorological measurements, process modelling, and aircraft and satellite observations that is designed to improve predictions of the response of the arctic terrestrial biosphere to global change. Results from multiple research sites in Abisko, Sweden and Kevo, Finland are discussed and implications for pan-Arctic ecosystem ecology are addressed.

Results/Conclusions
We demonstrate how bias in multi-scale estimates of CO2 flux in Arctic ecosystems can be reduced by preserving the information content of high spatial and spectral resolution aircraft and satellite imagery. Assimilating spatial observations alongside chamber, tower and aircraft-measured biosphere-atmosphere fluxes using the ensemble Kalman filter further reduced errors for improved spatial prediction. A number of novel factors that are critical for quantifying ecosystem response to global changes were identified, including: (i) reduced photosynthesis in transition zones between vegetation types, (ii) ‘priming’ of older soil C stock respiration during the period of vegetative activity, (iii) the sensitivity of this priming to N and P additions, and (iv) the lack of soil microbial acclimation to temperature. Potential ‘shortcuts’ for quantifying multi-scale Arctic ecosystem state and function include: (i) the strong relationship between leaf area index and fine root biomass, (ii) the short residence time and respiration of recently assimilated C, (iii) the relationship between canopy N and albedo in some ecosystem types, and (iv) predictable changes in leaf area index as a function of micro-scale topography. Continuing challenges include quantifying the extent and functional role of sub-canopy vascular plant and moss cover, and quantifying the effects of land cover change on radiation balance partitioning and surface temperature change.