CO2 FLUXES ESTIMATION FOR DIFFERENT VEGETATION COVER FRACTIONS

The Eddy Covariance (EC) technique is widely used to directly measure net CO2 land–atmosphere exchanges of an ecosystem. EC measurements are commonly used to validate CO2 fluxes simulated by models. Most of models are based on physical processes and physiology of a specific ecosystem. This study is about the development of a new methodology aimed to create a simplified model to estimate CO2 emissions and uptake processes for different vegetation cover fractions (λv) and in different ecosystem types. The model is based on empirical relations and common measurable variables, such as global radiation, soil temperature, and soil water content (SWC). Natural, agricultural, and suburban ecosystems, with vegetation cover fraction (λv) varying from 44% (suburban ecosystems) and 100% (forest ecosystems), were selected for the analysis. CO2 flux data measured with Eddy Covariance technique were obtained both from literature and original datasets. Per each site, a locally weighted non-parametric regression procedure (LOWESS) was used to determine the dominant trend of the photosynthesis light response curves during the vegetation growing season. Then, nonlinear least square (nls) model fitting was used to find the best coefficients of the related non rectangular hyperbola equation. From this analysis, a general relation to link the differences in vegetation cover fractions and CO2 flux was obtained per each coefficient. In this work, preliminary results both on the relationships between measured CO2 flux and global radiation and SWC are presented, as well as the most relevant model results. The model is able to capture the general trend of the CO2 flux at varying vegetation cover fractions. However, more datasets are needed to improve the accuracy of the model and to better understand the role of the specific model variables.