Temporal Dynamic of the Net Ecosystem Exchange of Carbon Dioxide and Responses to Meteorological Drivers in Mountain Forest Ecosystem of Southeastern China

Abstract

This study reports the temporal dynamics of the net ecosystem exchange of carbon dioxide (NEE), the importance of various meteorological factors influencing NEE, and the response characteristics of NEE to meteorological changes at different temporal scales (half-hourly, daily and monthly) on the basis of eddy covariance and meteorological data collected from forest ecosystems in southeastern China between June 2022 and May 2024, thereby employing the random forest regression algorithm. The results show that the forest ecosystem in the study area provides a clear carbon sink function, with an estimated daily NEE period of approximately 11–12 h. Notably, this ecosystem functions as a carbon sink almost year-round, with the lowest NEE in summer, followed by spring and autumn, whereas winter shows minimal NEE, resulting in an annual total NEE of −2983.34 g CO₂·m⁻²·a⁻¹. Furthermore, the NEE varies temporally and is influenced by several meteorological factors with varying degrees of impact. Specifically, sub-daily fluctuations are driven primarily by the light intensity, whereas the effect of light decreases at longer time scales, allowing other meteorological factors to exert increased influence. The complex and nonlinear responses of NEE to these meteorological factors often feature thresholds and optimal values. This study contributes to our comprehensive understanding of NEE behavior and provides reference data for improving forest carbon cycle models and developing regional forest management measures.