A scenario-based digital twin framework for assessing urban exposure to sea level rise: a case study in Abu Dhabi, United Arab Emirates

Climate change is increasing the exposure of many coastal cities to sea level rise (SLR), including those in the United Arab Emirates (UAE), where detailed scenario-based simulations remain limited by 30 m or coarser spatial-resolution elevation data. Traditional cartographic and Geographic Information System (GIS) mapping approaches, such as the bathtub method, also limit impact interpretation and constrain asset-level exposure statistics. This study addresses this gap by implementing GIS data in the Unity game engine to develop a 3D digital twin that generates quantitative statistics across three key urban components: buildings, road networks, and vegetation. A game engine provides a real-time physics solver, collision detection between water and urban objects, and free navigation inside a single interactive environment. These capabilities enable simultaneous measurement and immersive visualization of inundation extent, interactive navigation, hydrological connectivity, and object-level exposure. Abu Dhabi was selected because the extreme April 2024 rainfall events revealed vulnerabilities in its coastal infrastructure. The Corniche and Great Mosque areas are modeled at 11 m spatial resolution and assessed under four SLR scenarios of 1, 2, 3, and 4 m. These custom levels follow earlier research on potential SLR impacts on mangroves in the same city, under the Intergovernmental Panel on Climate Change (IPCC) Shared Socio-economic Pathway (SSP) projections. The 3 and 4 m levels correspond to low-likelihood, high-impact scenarios and are considered as stress tests. Simulation behavior is verified against the current shoreline under the null scenario and compared with independent inundation mapping of Abu Dhabi. Results reveal a non-linear exposure trend and two opposite regimes. At 1 m, only the shoreline is impacted. In the Corniche, exposure occurs early, with 36% of buildings affected at 2 m, after which the marginal rate decreases. In the Great Mosque area, low margins absorb the first 2 m, and the exposure accelerates thereafter, making this area more affected at 4 m, with 94.7% of buildings against 88.8%. A vertical sensitivity test shows that the difference between both areas is the same, but not the absolute percentages. The elevation at which exposure accelerates differs depending on the area and should guide adaptation planning.