Meteorological drivers of carbon flux variations on Xinglong Mountain in the transition zone between the Qinghai–Tibet and Loess Plateaus

The transition zone between the Qinghai–Tibet and Loess Plateaus is an important ecological functional area and carbon (C) reservoir in China. However, long-term monitoring data in this area are lacking, and the mechanisms underlying the carbon sink remain unclear. Using 6 years (2018–2023) of eddy covariance data from Xinglong Mountain, this study analyzed the response mechanisms of forest carbon fluxes to meteorological factors using structural equation modeling (SEM). The results showed that meteorological factors, including monthly average daily maximum radiation (Rg), monthly precipitation (PP), monthly average daily maximum vapor pressure deficit (VPD), monthly effective accumulated temperature (Tac), monthly average atmospheric temperature (Tair), and monthly average soil temperature (Tsoil), exhibited significant seasonal variations, peaking from June to August. Net ecosystem productivity (NEP), gross primary productivity (GPP), and ecosystem respiration (Reco) also exhibited significant seasonal variations, peaking from May to August. The forest ecosystem exhibited stable carbon sequestration (NEP: 299.64–448.62 g C m −2 a −1 ) from 2018 to 2023. SEM revealed that Rg and Tair were significant direct drivers of NEP; Reco was influenced directly by Tsoil and indirectly by Rg; and GPP was mainly affected directly and indirectly by Rg, and was also directly affected by PP and VPD. Overall, the forest ecosystem carbon fluxes at Xinglong Mountain were primarily subjected to direct and significant positive influences from Rg, PP, Tair, Tsoil, and VPD. This study provides mechanistic insights into the response of forest carbon cycling to climate change in semi-arid regions, which is highly significant for evaluating the ecological barrier function of the transition zone between the Qinghai–Tibet and Loess Plateaus.