This research presents a comprehensive tsunami inundation mapping and risk assessment for the Hawaiian Islands using a combination of machine learning and geospatial techniques. For the assessment of land-use and land cover changes for the years (U.S. 2018) through 2025, RF classification was required to measure the critical LULC mappings for understanding the spatial changes and their impacts regarding tsunami vulnerability. The analysis utilized high-resolution geospatial datasets including digital elevation models, which are valuable in estimating the impacts of former tsunami flooding events and evaluating potential impacts of future tsunami events. Tsunami inundation modeling was performed from a 1 m scenario up to a 60 m scenario, with 60 m being the worst-case hazard scenario. As a result, an inundation area differs between islands since they are characterized by diverse topography, coastal morphology, and LULC distributions. Proportional impact under the 60 m scenario to Nihau, which had the maximum value of 64.63%, followed by Oahu with 29.09% (or 458.03 sq k. m), Molokai (16.69%), and Kauai (15.38%). As such, even if Maui, Lanai, and Hawaii Island have lower amounts, the localized effects on key infrastructure and ecological habitats are breathtakingly enormous, whereas small area and height contributed to the least inundation of Kohoolawe. The measure of built land area was used in making exposure quantification with respect to vulnerability on the islands. Analyses showed highly differing areas between potential tsunami impact sites, with Oahu recording the most (249.58 km² of built land found within inundation zones), followed by Hawaii (74.37), Maui (52.24), and Kauai (37.81). Compared to the size of these areas, they were tiny with regard to affected urban extents: Molokai (9.02 km²), Lanai (0.52 km²), Niihau (0.017 km²), and Kahoolawe (0.015 km²). According to the tsunami inundation impact assessment, Oahu had the highest population (681,473) and household (253,939) exposure under the 60 m inundation model, followed by Hawaii and Maui, suggesting that these islands are the most at risk from tsunamis throughout the Hawaiian archipelago. The findings suggest that the larger developed islands are facing a comparable risk while smaller islands remain at risk due to limited evacuation processes and localized infrastructure. The geospatial risk model provides a critical engagement in the overall disaster readiedness to assist in evacuation planning and strengthening infrastructure in considering potential damage and loss of life through identification of main hotspots. This kind of machine learning and geospatial analysis for a large-scale tsunami risk assessment in support of resilience and sustainable coastal management in at-risk areas will lead to a path to scalable solutions.
