This study explores the role of numerical modelling in understanding the fate of injected CO2 through the lens of noble gas partitioning. Uncontrolled releases of CO2 can lead to atmospheric re-emission and threaten groundwater, and surface water resources. Furthermore, naturally occurring CO2 can complicate efforts to monitor storage security and identify potential leaks; as such, differentiating anthropogenic and natural CO2 in the shallow subsurface is crucial. Noble gas tracers are ideal for this purpose due to their stability and predictable partitioning behaviour. Our research employs a lab-validated model capable of simulating realistic gas fingering behaviour coupled to groundwater flow via dissolution, exsolution and multicomponent chemical partitioning. We present the results of simulations of shallow CO2 injections with realistic noble gas mixtures under varying conditions of groundwater flow and subsurface heterogeneity. These results reveal how factors such as soil structure and groundwater flow affect the vertical migration of CO2, specifically through the impact on dissolution rates. The study also uncovers how trapped gas influences noble gas ratios to aid interpretation , as less soluble gases like helium gravitate towards the gaseous phase, affecting both noble gas ratios and surface gas flux. These insights underscore the effectiveness of noble gases in monitoring while highlighting the need to account for compositional changes during dissolution.