A novel time-lapse modelling scheme for airborne electromagnetic (AEM) monitoring data sets is presented, using data from multiple surveys applied to study the hydrorelated evolution of the Bookpurnong floodplain in South Australia. Additionally, it introduces a new wide-ranging approach for this type of study, incorporating new processing, validation and interpretation tools. Time-lapse studies are widespread in the literature but are not commonly applied to model electromagnetic (EM) data, particularly AEM data. This is linked to the challenges of performing overlapping data acquisition with inductive systems. The key features of the present time-lapse scheme include the definition of independent forward and model meshes, essential for considering discrepancies in the location of soundings which arise in multitemporal AEM data acquisition. Moreover, the incorporation of system flight height in the inversion revealed important for achieving satisfactory data fitting and limiting artifact propagation in the time-lapse models. A novel processing workflow for AEM multitemporal data sets is also presented. This has proven important for effectively processing the multitemporal data sets, which presents new challenges in identifying noise coupling arising from the use of different systems across vintages of data, possible variations in acquisition settings operated by different field crews, and changes in subsurface resistivity in the survey area. Results generated from the time-lapse modelling are evaluated with an independent hydrogeological validation (IHV), designed to support the geophysical models validation and interpretation by providing a first-step hydrogeological evaluation. At Bookpurnong, along a sector of the Murray River floodplain, multitemporal AEM surveys were collected in 2015, 2022 and 2024, to study natural and engineered changes in the groundwater system over time. The time-lapse models show significantly smaller variations compared to those determined with individually modelled survey data sets, while delineating sharply bounded changes in resistivity across the floodplain. This highlights the effectiveness of the new time-lapse scheme in minimizing inversion variations typically encountered with independently modelled results affected by larger equivalence issues. Here, AEM models are first compared with resistivity borehole measurements, revealing a close match between the two methodologies and spatial variations in resistivity consistent with a meandering river across the floodplain. These variations are further validated and interpreted using the IHV approach, which revealed a direct correlation between the hydrological stress of the Murray River and the response of shallow aquifers. Additionally, time-lapse geophysical models, combined with a hydrostratigraphic analysis, allow for a direct correlation between shallow and deep hydrogeological responses. We believe that the time-lapse methodology described here can be widely applied to multitemporal studies using AEM data sets, enabling the study of a broad range of natural processes with great accuracy and at the basin scale.