Near-surface soil gas monitoring is widely used at carbon storage sites (e.g., geological storage and enhanced oil recovery) to detect leaking CO2. Reliable detection is critical as CO2 migration into the shallow subsurface can reduce groundwater quality and create greenhouse gas emissions. Current diagnostic techniques for soil gas samples rely on stoichiometric ratios between CO2, N2, and O2, derived from biochemical reactions in the unsaturated zone, to distinguish leakage signals from natural CO2 variability. However, these techniques do not currently consider the effect of gas transport (e.g., leak type, transport mechanisms) on soil gas ratios. This study investigates how gas transport mechanisms (e.g., multicomponent diffusion, leakage rate) effect the spatial and temporal reliability of near-surface detection techniques. CO2 was injected at varying rates into a 2D transparent flow cell filled with a variably saturated, quasi-homogeneous sand pack under abiotic conditions, isolating gas transport effects from natural variability. Light transmission techniques were used in the saturated portion of the cell, enabling real-time tracking of gas flow behaviour. Unsaturated zone gas samples were taken at varying depths and surface flux was monitored continuously. Removing natural variability, CO2 and atmospheric gases (O2, N2) produced strong linear trends over time, providing a useful baseline for comparison with soil gas signals at field sites. Results highlighted the importance of sampling depth, as multicomponent diffusion caused decreasing linear correlations between CO2, N2, and O2 at shallower depths. These findings will help inform site sampling recommendations and better characterize environmental signals indicative of leakage.