In an exhaustive new global assessment, researchers reveal startling discrepancies in coastal elevation measurements that suggest sea levels are significantly higher than commonly assumed in the vast majority of coastal hazard analyses. This revelation challenges prevailing understandings and poses critical implications for flood risk evaluations and climate adaptation strategies worldwide.
The investigation was underpinned by a meticulous and systematic literature review adhering to rigorous standards akin to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). By interrogating a dataset of over 35,000 publications sourced primarily from the Scopus database across a fifteen-year period, the team honed in on 385 meticulously selected studies—each integrating satellite-borne elevation data within coastal hazard contexts. This comprehensive approach targeted the most up-to-date research coinciding with the availability of global digital elevation models (DEMs) and mean dynamic topography (MDT) datasets, ensuring a consistent data horizon for meaningful comparison.
Central to the analysis was the often-overlooked aspect of vertical datum conversion, a process critical to aligning elevation data with mean sea level (MSL). The team found that many studies failed to properly document or apply the necessary conversion procedures, leading to significant underestimations of coastal elevations and, by extension, flood susceptibility. Vertical datum conversion entails not only recognizing the original vertical reference system of a DEM but also integrating auxiliary datasets such as geoid models and local sea-level indicators to calibrate elevation data precisely.
To empirically quantify these discrepancies, the researchers processed four state-of-the-art global DEMs—CoastalDEM v2.1, FABDEM v1.0, GLL-DTM v.2, and DeltaDTM v1—each originally referenced to different geoid models (EGM96 or EGM2008). By rigorously translating these DEMs into a common vertical framework using the latest hybrid MDT data (HYBRID-CNES-CLS2022), they reconciled variations in geoid heights and sea surface levels derived from satellite altimetry and gravitational field measurements. This conversion employed advanced spatial interpolation and smoothing techniques to extrapolate sea-level data over land, a necessary step given the coastal focus of the study.
The implications of aligning elevation datasets vertically to MSL were profound. When comparing unconverted and converted DEMs, stark differences emerged in estimates of land areas and populations at risk from a hypothetical one-meter relative sea-level rise (RSLR). The uncorrected models consistently underestimated exposure, sometimes by substantial margins, due to misaligned vertical references. The research extended beyond area metrics, incorporating an ensemble of three global population datasets—WorldPop, LandScan 2020, and LandScan 2023—to robustly quantify how many people currently reside below or near local MSL thresholds.
Further scrutiny uncovered a widespread and systematic issue: many influential scientific publications, some referenced extensively in the latest IPCC Assessment Reports, failed to incorporate precise sea-level datum referencing. Out of studies incorporated into the IPCC AR6 cycle, less than 15% demonstrated proper vertical datum integration. This revelation suggests that even well-recognized climate assessments may underrepresent coastal vulnerability globally, potentially skewing policy guidance and resource allocation.
Methodologically, the researchers employed the ArcGIS Pro platform to standardize transformations, employing bilinear resampling and inverse distance weighting algorithms to meticulously adjust DEMs for datum biases. The use of robust geoid models from the GFZ Helmholtz Centre and hybrid MDT products enhanced the geographic fidelity and temporal relevance of the analyses. This methodological rigor establishes a replicable framework that future coastal hazard assessments can adopt to improve accuracy.
A critical aspect of this work was the recognition that accurate vertical referencing transcends mere technical detail; it is foundational for correctly estimating low-elevation coastal zones (LECZ), which harbor some of the world’s densest human populations and most economically vital infrastructure. By recalibrating elevation data consistently with sea-level datasets, risk assessments can better capture subtle but consequential degrees of inundation potential, especially in vulnerable deltaic and estuarine environments.
Beyond the technical elevation adjustments, the researchers highlight the limitations inherent to current approaches to exposure modeling. Notably, their exposure statistics are conservative, focusing only on relative elevation without applying hydrodynamic inundation models or factoring in dynamic population growth in coastal zones. Recognizing these constraints, the study nonetheless represents a critical correction to the baseline data inputs that underlie numerous impact projections.
This breakthrough recalibration of coastal elevation data aligns with emerging calls for more transparent, data-driven, and reproducible methodologies within the climate science community. It provides an indispensable corrective lens to reconcile disparate elevation datasets and sea-level assumptions, setting a higher standard for environmental impact assessments and infrastructure planning.
The findings underscore an urgent need for widespread adoption of consistent vertical datum conversions and up-to-date sea-level referencing in coastal research. Given global population trends concentrating billions in vulnerable coastal regions, ignoring these methodological imperatives risks underestimating future hazards and compromising adaptive responses.
Moreover, this paradigm shift offers a clarion call for the IPCC and other scientific consortia to revisit foundational datasets informing policy models. The systematic review uncovered that nearly 90% of coastal hazard studies neglected adequate sea-level reference integration, suggesting that global hazard aggregation may be fundamentally flawed. Rectifying this will improve the precision of projections guiding climate resilience investments.
This research thus represents a milestone in coastal risk science, marrying satellite geodesy, geospatial analysis, and rigorous bibliometric assessment to expose a critical blind spot in existing hazard appraisals. It charts a path forward for harmonizing global elevation data with real-world sea-level dynamics—an essential step in safeguarding vulnerable coastal populations in the Anthropocene.
As coastal impacts accelerate worldwide, the precision of elevation and sea-level datasets emerges as a linchpin in effective climate adaptation. By driving awareness of vertical datum discrepancies and establishing best-practice conversion protocols, this study empowers scientists, policymakers, and planners to craft more accurate and actionable hazard maps.
The broader implications extend to urban planning, disaster risk reduction, and insurance modeling, where marginal differences in elevation measurement can translate into huge shifts in exposure and vulnerability profiles. The refined datasets and conversion workflows pioneered here are poised to become standard tools for next-generation coastal assessments.
Ultimately, this comprehensive reevaluation challenges prevailing narratives of coastal safety and vulnerability, advocating a recalibration of risk perspectives grounded in rigorous geospatial fidelity. As the world grapples with rising seas, such heightened accuracy in elevation referencing represents a critical advancement in our collective capacity to anticipate and mitigate coastal hazards.
Subject of Research: Coastal elevation accuracy, vertical datum conversion, sea-level reference frames, global digital elevation models, coastal hazard assessment
Article Title: Sea level much higher than assumed in most coastal hazard assessments
Article References:
Seeger, K., Minderhoud, P.S.J. Sea level much higher than assumed in most coastal hazard assessments. Nature (2026). https://doi.org/10.1038/s41586-026-10196-1
