An Empirical Model of Gross Primary Productivity (GPP) and Relations between GPP and Its Driving Factors, Biogenic Volatile Organic Compounds in a Subtropical Conifer Plantation in China

Abstract

Measurements of net ecosystem exchange (NEE), solar global radiation, photosynthetically active radiation (PAR) and meteorological parameters were carried out on a subtropical conifer plantation in China from 2013 to 2016. These observations were used to develop and evaluate an empirical model of gross primary production (GPP) (EMGPP) with 3-factor and 2-factor models. Using a 3-factor model, the simulated hourly GPP values were consistent with observations with a relative bias of 9.96% and normalized mean square error values of 0.07 mg CO2 m−2 s−1 for the scattering factor S/Q (S and Q are diffuse and global solar radiation) < 0.5 and 15.52% and 0.15 mg CO2 m−2 s−1 for S/Q ≥ 0.5. Validations of the EMGPP for hourly, daily, monthly, and annual GPP values were carried out and showed that both 3-factor and 2-factor EMGPP models can accurately capture diurnal, seasonal and interannual variations in GPP, but most simulated GPP overestimated the observed value. When the scattering factor is not available, the 2-factor EMGPP can be used. The EMGPP using 3-factor and 2-factor models was applied to simulate GPP under all sky conditions from 2013–2016, and the estimated GPP were in reasonable agreement with the measured values and showed systematic overestimations of 31% and 29% for mean hourly GPP and 41% and 29% for annual amounts, respectively. The sensitivity test demonstrated that GPP values were more sensitive to changes in PAR than to changes in water vapor and scattering factor at low S/Q, but were more sensitive to changes in water vapor than to PAR and S/Q at high S/Q. The sensitivity test revealed some mechanisms of GPP and its related processes, including the relationships between GPP and scattering of PAR, GPP and water vapor, which were in good agreement with other observations and model studies. An empirical model based on PAR energy balance can better describe the multiple interactions between GPP and its driving factors (PAR, water vapor, S/Q). The ratio of the emissions of biogenic volatile organic compounds (BVOCs) to net ecosystem exchange clearly varied between forests in different climate zones.