Poleward moisture transport into the Arctic occurs via short-lived, episodic intrusions of warm, moist air masses. These extreme transport events are associated with an enhanced greenhouse effect that is expected to slow sea-ice growth in winter and hasten the start of spring melt season. Yet, their impact on the Arctic energy balance averaged over longer time scales has been difficult to quantify, in part due to the challenge in separating local and remote Arctic moisture sources. Additionally, though much of the poleward moisture transport into the Arctic occurs via moist intrusions, popular diagnostic frameworks emphasize only a modest contribution of atmospheric energy transport to Arctic amplification. To reconcile these contradictory findings, this project will advance a process-oriented understanding of the role of moist intrusions in driving Arctic warming via pre-conditioning sea ice for rapid ablation during the warm season and therefore initiating positive local feedbacks. Numerical water tracers in the Energy Exascale Earth System Model (E3SM) will be used to quantify how moist intrusions, both at present and in a warmer world, draw moisture from source regions outside the polar cap and inject it over the polar cap. Source-aware radiative locking experiments in E3SM will reveal how moist intrusions and their vertical structure help sustain water vapor and cloud feedbacks over the Arctic, and thereby elicit sea ice retreat and its attendant feedbacks. A multi-model analysis of the Coupled Model Intercomparison Project Phase 6 (CMIP6) will enable characterization of how jet stream characteristics, such as position and sinuosity, control present-day and future moist-intrusion climatologies. Lastly, biases in model representation of moist intrusions and their impacts will be assessed through a like-for-like application of causal inference techniques, which can establish causality in the absence of targeted model experiments, to E3SM experiments, the CMIP6 ensemble, and reanalysis products. Through the coordinated activities of E3SM experiments and statistical approaches rooted in AI techniques, this project will systematically uncover how the large-scale environment governs the climatology of moist intrusions and how, in aggregate, these extreme transport events drive Arctic summertime sea ice loss and warming across an ensemble of coupled Earth system models.