• carbon;
  • boreal forests;
  • InTEC model;
  • climate change;
  • remote sensing


Reconstruction of interannual variability of net ecosystem productivity (NEP) in forests provides an important approach to analyse impacts of future climate change on global carbon (C) cycling. However, lacking climate data at sufficient temporal scales constrains NEP simulation and the potential of monthly climate data in modelling annual NEP remains largely poorly understood. In this study, annual NEP at 12 Fluxnet-Canada forest sites (93 site-year) was simulated using a process-based Integrated Terrestrial Ecosystem C-budget (InTEC) model driven by forest inventory data, site-level meteorological measurements, site-specific indicators, and remote sensing observations. Our results indicate that the InTEC model can capture the first order of interannual NEP variability with coefficients of determination (R2) of 0.84 (p < 0.001) between simulated and measured NEP, providing a significant opportunity to reconstruct long-term climate change on forest C dynamics using only available monthly historical climate records. The usefulness of model simulation was further evaluated at three post-clearcut chronosequences Douglas-fir stands of British Columbia, suggesting that the Douglas-fir ecosystem would remain a C source for 15–20 years after clearcut and the maximum annual NEP may occur at the age around 50.