An empirical model simulating diurnal and seasonal CO&amp;lt;sub&amp;gt;2&amp;lt;/sub&amp;gt; flux for diverse vegetation types and climate conditions

M Saito,S Maksyutov,A D Richardson,R Hirata

doi:10.5194/bg-6-585-2009

Abstract

Abstract. We present an empirical model for the estimation of diurnal variability in net ecosystem CO2 exchange (NEE) in various biomes. The model is based on the use of a simple saturated function for photosynthetic response of the canopy, and was constructed using the AmeriFlux network dataset that contains continuous eddy covariance CO2 flux data obtained at 24 ecosystems sites from seven biomes. The physiological parameters of maximum CO2 uptake rate by the canopy and ecosystem respiration have biome-specific responses to environmental variables. The model uses simplified empirical expression of seasonal variability in biome-specific physiological parameters based on air temperature, vapor pressure deficit, and annual precipitation. The model was validated using measurements of NEE derived from 10 AmeriFlux and four AsiaFlux ecosystem sites. The predicted NEE had reasonable magnitude and seasonal variation and gave adequate timing for the beginning and end of the growing season; the model explained 83–95% and 76–89% of the observed diurnal variations in NEE for the AmeriFlux and AsiaFlux ecosystem sites used for validation, respectively. The model however worked less satisfactorily in two deciduous broadleaf forests, a grassland, a savanna, and a tundra ecosystem sites where leaf area index changed rapidly. These results suggest that including additional plant physiological parameters may improve the model simulation performance in various areas of biomes.

Highlights

Simulation of atmospheric CO2 variability by atmospheric transport modeling depends critically on the use of terrestrial ecosystem models to accurately simulate diurnal and seasonal variations in terrestrial biospheric processes
The biomes consisted of six evergreen needle-leaf forests (ENF), two evergreen broadleaf forests (EBF), four deciduous broadleaf forests (DBF), four mixed forests (MF), three grasslands (GRS), two savannas (SVN), and three tundra ecosystems (TND) (Table 1)
We explored a simple approach to predicting diurnal variations in net ecosystem CO2 exchange (NEE) over seven biomes and proposed an empirical model based on the use of a nonrectangular hyperbola and eddy covariance flux data obtained from the AmeriFlux network

Summary

Introduction

Simulation of atmospheric CO2 variability by atmospheric transport modeling depends critically on the use of terrestrial ecosystem models to accurately simulate diurnal and seasonal variations in terrestrial biospheric processes. Models based on light-use efficiency, such as CASA and TURC (Ruimy et al, 1996), assume a linear relationship between monthly net primary production (NPP) and monthly solar radiation (Monteith, 1972) that is limited by water availability and temperature. These models appear to be successful in seasonal cycle simulation as a whole, their extension to cover diurnal cycles should be accompanied by the introduction of a more realistic, non-linear relationship between CO2 uptake by terrestrial vegetation and solar radiation at an hourly time scale.

Methods

Results

Conclusion