Modelling the biogenic CO2 exchange in urban and non-urban ecosystems through the assessment of light-response curve parameters

The biogenic CO2 surface–atmosphere exchange is investigated and linked to vegetation cover fraction for seven sites (three urban and four non-urban) in the northern hemisphere. The non-rectangular hyperbola (NRH) is used to analyse the light-response curves during period of maximum ecophysiological processes, and to develop two models to simulate biogenic vertical CO2 fluxes. First, a generalised set of NRH coefficients is calculated after linear regression analysis across urban and non-urban ecosystems. Second, site-specific NRH coefficients are calculated for a suburban area in Helsinki, Finland. The model includes a temperature driven equation to estimate ecosystem respiration, and variation of leaf area index to modulate emissions across the year. Eddy covariance measured CO2 fluxes are used to evaluate the two models at the suburban Helsinki site and the generalised model also in Mediterranean ecosystem. Both models can simulate the mean daily trend at monthly and seasonal scales. Modelled data typically fall within the range of variability of the observations (differences of the order of 10%). Additional information improves the models performance, notably the selection of the most vegetated wind direction in Helsinki. The general model performs reasonably well during daytime but it tends to underestimate CO2 emissions at night. This reflects the model capability to catch photosynthesis processes occurring during the day, and the importance of the gross primary production (GPP) in modifying the net ecosystem exchange (NEE) of urban sites with different vegetation cover fraction. Therefore, the general model does not capture the differences in ecosystem respiration that skew nocturnal fluxes. The relation between the generalised NRH plateau parameter and vegetation cover improves (R2 from 0.7 to 0.9) when only summer weekends with wind coming from the most vegetated sector in Helsinki and well-watered conditions for Mediterranean sites are included in the analysis. In the local model, the inclusion of a temperature driven equation for estimating the ecosystem respiration instead of a constant value, does not improve the long-term simulations. In conclusion, both the general and local models have significant potential and offer valid modelling options of biogenic components of carbon exchange in urban and non-urban ecosystems. © 2016 Elsevier B.V.