Coastal flooding is one of the main concerns worldwide associated with climate change. Historically, coastal cities have regularly experienced hazardous flooding across the globe, contributing to coastal population vulnerability. In Aotearoa New Zealand, an island nation, the outlook of rising sea levels is especially concerning due to the number of people that live in the coastal zone. Predicting extreme water levels is challenging and depends on properly identifying the main physical processes. Swells, wind waves, storm surges, and astronomical tides are transformed when propagating inside shallow coastal lagoons, bays and estuaries, mainly because of their mutual interaction (e.g., phase alteration) but also because of the effects of constrictions and friction imposed by the estuarine morphology (e.g., amplification and asymmetry). Furthermore, the compound action of different drivers and extreme events in close succession can exacerbate flooding. The understanding of non-linear interactions between flooding drivers can substantially improve the power of prediction of flooding in coastal areas with complex morphology such as estuaries. For that, proper bathymetry and topography data are fundamental for assessing flooding risk, which can be challenging because some of the current acquision methods are economically expensive or not applicable to remote and shallow areas exposed during low tide. Satellite-derived bathymetry and topography are a recent approach which are increasingly used to tackle this issue. Although flooding drivers have already been identified and studied around the coast of Aotearoa New Zealand, few studies have focused on their interaction inside estuaries. Similarly, very little has been undertaken on the applicability of satellite-derived elevation data for hydrodynamic modelling. This thesis regionally assesses the non-linear interactions between storm surges, waves, astronomical tides and estuarine morphology and the suitability of satellite-derived data to represent these complex interactions in hydrodynamic models. The main findings are that tide-surge interactions are a major contributor to the water level inside estuaries and cannot be neglected when simulating flooding events. Tide-surge interactions can represent variations in the water level ranging from -16 cm (below mean sea level) to +27 cm (above mean sea level). In addition, non-linear interactions induced by the estuarine morphology — especially the coverage of the intertidal zone related to the total surface area of the estuary — modulate the co-occurrence rate between events outside and inside an estuary. It was found that co-occurrence rates can vary largely according to the estuary (20–86%). Furthermore, storm-surge and wave interactions have shown strong potential for compound and clustering effects in estuaries on the North Island, linked to severe south and southwesterly winds generated by local weather patterns. Wave height and storm surges corresponding to extreme water levels are significantly correlated, especially inside estuaries in locations close to the estuarine entrance. Wind waves are more strongly correlated to storm surges on the west coast while swell waves show a stronger correlation on the east coast of the North Island. However, when wind and swell waves are combined (i.e., integrated wave hindcast data) the correlation between wave heights and storm surges is stronger than when analysed separately. Furthermore, independent and clustered extreme events of waves and storm surges follow similar seasonality, can co-occur and can be composed of up to 12 extreme events occurring within 15 days. Finally, based on existing techniques, a framework was applied to obtain satellite-derived topo-bathymetric data. Although the applied method is sensitive to several factors (e.g., the complexity of the relief, number of processed images, bottom reflection), the derived topographic data showed good approximation to LiDAR data validation — showing similar accuracy of ~20 cm, in an elevation range of approximately 2m. The elevation estimates were used in hydrodynamic modelling to assess extreme water levels in a complex estuarine system, considering non-linear interactions. Simulation scenarios using only satellite-derived elevations could represent extreme water levels inside an estuary with complex morphology with comparable accuracy of the scenarios using topo-bathymetric data acquired using in-situ methods (< 0.1 m difference). The results show that satellite-derived data can be a good replacement when data originating from traditional methods are scarce or unavailable. In conclusion, the thesis shows the regional patterns of non-linear interactions between different flooding drivers around Aotearoa New Zealand, showing where and when these interactions are important. Furthermore, it shows how satellite-derived techniques can improve the power of predicting flooding in remote areas with data scarcity. These findings are crucial for enhancing the prediction of flooding in morphologically complex areas such as estuaries, where a great part of the coastal community and civil infrastructure are based and have major social-economic importance. A better understanding of estuarine flooding would make coastal adaptation policies more effective by making the current models more accurate.