Complex Network Dynamics of Climate Interactions Driving Carbon Source–Sink Transitions

Climate change alters the regulation of carbon fluxes in terrestrial ecosystems, raising fundamental questions about the collective dynamics of climate drivers. Here, we develop a stochastic network-based framework to investigate how complex interactions among climatic variables control carbon source–sink dynamics. Using data from the FLUXNET (global flux network) database, we construct seasonal interaction networks based on statistically significant correlations among ecosystem and atmospheric variables in two contrasting biomes. A spectral ranking algorithm inspired by PageRank is applied to estimate the steady-state contribution of each variable within the network dynamics. Results reveal distinct emergent behaviors: the grassland system exhibits strong seasonal reorganization, while the forest maintains a more stable network structure. Air temperature acts as a dominant driver promoting carbon source dynamics, whereas relative humidity reinforces sink behavior. The cumulative annual contributions highlight a structural asymmetry in climate regulation, with source-driving mechanisms outweighing sink-enhancing effects. These findings demonstrate that accounting for full network topology uncovers emergent regulatory patterns and provides a transferable framework for analyzing carbon dynamics in complex ecological systems.