Modeling CO2 And Water Vapor Exchange Of A Temperate Broadleaved Forest Across Hourly To Decadal Time Scales

  • Sites: US-WBW
  • Publication Type: JOUR
  • Authors: Baldocchi, D. D.; Wilson, K. B.

  • Fluxes of carbon dioxide, water and energy between a temperate deciduous forest and the atmosphere were quantified across time scales of hours, days, seasons, years and decades. This exercise was performed using stand-level eddy covariance flux measurements and a biophysical model, CANOAK. The CANOAK model was tested with measurements of carbon dioxide, water vapor and energy flux densities we have been collecting since October 1994. Model calculations reproduced 80% of CO2 and water vapor flux variance that are contained in a year-long time series, when the model was forced with hourly weather data and seasonal information on plant structure and physiological capacity. Spectral analysis of measured and computed time series revealed that peak time scales of flux variance have periods of a day, half-week, season and year. We examined questions relating to inter-annual variability of mass and energy exchange by forcing the validated model with a decade-long meteorological record. Theoretical estimates of year-to-year variability of net ecosystem CO2 exchange were on the order of ±200 gC m−2 year. We also deduced that significant variance of water vapor and CO2 exchange occurs on the time scale of 5–6 years, the time scale associated with El Nino phenomena. Sensitivity tests performed with the model examined issues associated with model complex and parameterization issues. Of particular importance were the effects of leaf clumping and length of the growing season on canopy photosynthesis and net ecosystem CO2 exchange. Ignoring the effects of leaf clumping caused an error as large as 50% in the estimation of annual biosphere–atmosphere net carbon exchange. Each incremental day change in the length of the growing season altered the net ecosystem CO2 exchange by 5.9 gC m−2.


  • Journal: Ecological Modelling
  • Funding Agency: —
  • Citation Information:
  • Volume: 142
  • No: 1-2
  • Pages: 155-184
  • Publication Year: 2001/08
  • DOI: 10.1016/s0304-3800(01)00287-3