The coupling between boreal arctic land and the near-surface atmosphere is influenced by a combination of soil conditions, vegetation dynamics, and atmospheric properties. Understanding how the land-atmosphere couplings respond to warming and enhanced CO2 concentrations is crucial for gaining insights into the carbon, energy, and water cycles. In this study, we analyzed these coupled systems using both observational data and fully coupled simulations from the Energy Exascale Earth System Model (E3SM), focusing on boreal arctic forest ecosystems. We utilized a causal inference approach along with network metrics to uncover patterns and the strength of land-atmosphere coupling network under both historical conditions and a future high-emission scenario (SSP585). Our findings indicate that, based on observations, the current extratropical forest coupling network exhibits high connectivity (72%–88% of the targeted processes are coupled). E3SM effectively simulated such coupled network and projected a significant 28% (±3%) increase in coupling strength under conditions of warming and elevated CO2. Additionally, E3SM factorial coupled experiments suggested that the enhancement in future land-atmosphere coupling is driven by warming-enhanced soil nitrogen mineralization, which promotes nitrogen uptake by plants and vegetation growth.