Measurement of CO2 exchange between Boreal forest and the atmosphere.

The Boreal forest is the world's second largest forested biome occupying the circumpolar region between 50 degrees N and 70 degrees N. This heterogeneous biome stores about 25% of all terrestrial carbon. We have reviewed EC measurements of CO2 exchange between the atmosphere and Boreal forests, and assessed progress in understanding the controlling processes. We have assessed net ecosystem productivity, the net balance between net primary productivity and heterotrophic respiration, measured using the EC method, for 38 Boreal forest sites. Gross ecosystem productivity has been estimated by adding day-time EC-measured CO2 fluxes to respiration estimated from night-time relationships between respiration and temperature. Maximum midday values of gross ecosystem productivity vary from 33 pmol m(-2) s(-1) for aspen to 6 micromol m(-2) s(-1) for larch stands. Long-term EC flux measurements, ongoing at nine Boreal sites, have shown the strong impact of spring weather and growing season water balance on annual net ecosystem productivity. Estimation of net biome production, incorporating the effects of disturbance resulting from forest fires and logging, has progressed significantly in recent years. After disturbance, summer measurements in Boreal chronosequences suggest that it takes about 10 years before growing season carbon uptake offsets the decomposition emissions. Small-scale exchange rate measurements using chambers and manipulative experiments such as stem girdling and soil heating help to understand the processes and mechanisms playing major roles in the carbon balance of terrestrial ecosystems. Aircraft EC flux measurements, convective boundary layer carbon budgets, and (13)C/12C changes in the atmosphere play an important role in validating estimates of regional carbon exchange based on scaled up EC measurements. Atmospheric inverse models are an important approach to studying regional and global carbon balance but need further improvement to yield reliable quantitative results.