Meteorology-driven screening of moisture swing sorbents for passive direct air capture

Scaling passive direct air capture (DAC) with moisture swing sorbents requires an understanding of how ambient meteorological conditions constrain achievable performance. While laboratory and device-scale models of moisture swing adsorption exist, there is currently a lack of parsimonious frameworks that translate long-term weather variability into site-level productivity and water-loss metrics for screening and deployment decisions. In this study, we develop a reduced-order, meteorology-driven process model for a moisture swing passive DAC system that maps ambient temperature, relative humidity, and wind speed directly to CO₂ productivity and net water loss. The model employs decoupled proxy equilibria, a modified Langmuir isotherm for CO₂ loading and a Flory–Huggins model for H₂O loading, closed with linear-driving-force kinetics parameterized by ambient wind speed and temperature. Using 17 years of NOAA weather data for St. Johns, Arizona, we quantify climatological seasonal variability and the impact of interannual weather fluctuations. Results show that relative humidity is the dominant driver of both CO₂ productivity and water loss. Low-humidity conditions enable higher productivity at the cost of increased net water loss, whereas periods of sustained high ambient humidity can preclude capture entirely. Notably, we identify distinct wintertime intervals during which the equilibrium CO_₂ loading under ambient conditions falls below the regenerated state, resulting in zero net capture despite continued cycling. This finding highlights a minimum required swing capacity and establishes a fundamental climatic constraint on passive DAC operability. By explicitly linking meteorology to CO₂ productivity and water loss, this work provides a screening-level framework for geospatial site triage and operational planning of moisture-swing passive DAC systems, with results interpreted as bounded performance estimates rather than device-specific predictions.