Future Shoreline Projections

This dataset provides global projections of future shoreline positions for the years 2030, 2050, and 2100 under three Shared Socioeconomic Pathway (SSP) scenarios: SSP1-2.6, SSP2-4.5, and SSP5-8.5. The data is structured around the Global Coastal Transect System, providing probabilistic estimates of shoreline change for 88,710 km (9.3%) of the world's coastline where the underlying physical model is considered valid (sandy, gravel or shingle beaches on sediment plains and dune coasts). The projections are delivered as GeoParquet files, with each row representing a single transect for a specific climate scenario and time horizon. The dataset includes the final projected shoreline geometry, the probabilistic distribution of total change, and the individual contributions from ambient (historical) trends and sea-level rise. This dataset is packaged as a multi-part ZIP archive (shoreline-projections.zip together with .z01–.z12). To extract the data, place all parts in the same folder and unpack the archive starting from the shoreline-projections.zip master file. MethodologyFuture shoreline positions were projected using a probabilistic model that decomposes shoreline change into two independent components: Ambient Change (AC): The background, long-term erosion or accretion trend. This was estimated by projecting historical shoreline change rates derived from satellite data (from the GCTR sds:change_rate field). Sea-Level Rise (SLR) Retreat: The shoreline retreat induced by sea-level rise. This was calculated using the Bruun rule, applied to the full ensemble of local relative sea-level rise projections from the IPCC AR6. A multi-stage Monte Carlo simulation (n=1,000) was performed for each transect to robustly propagate and combine uncertainties from both components. The total shoreline change was derived through a numerical convolution of the AC and SLR distributions, yielding a final empirical probability distribution. From this distribution, the median (total_change_p50) and the 90% confidence interval (total_change_p5 and total_change_p95) were extracted. The model was applied only where its underlying physical assumptions are valid. This includes unconsolidated coasts (sandy, gravel, or shingle beaches) on erodible profiles (sediment plains and dune coasts) with high-quality historical shoreline data and valid nearshore slope information. Data Structure and ContentThe dataset is organized by transect, with key columns including: Identifiers and Location: transect_id, lon, lat, continent, country, and quadkey (copied from Global Coastal Transect System) Input Data: Key source data for the projection: historical shoreline change rates (sds:change_rate), nearshore slope (gcc:ns), and the full AR6 sea-level rise projections for each SSP-year combination (e.g., ar6:ssp2-45_2050). Scenario Information: The specific scenario for each row, defined by ssp (e.g., 'ssp2-45') and datetime (e.g., '2050-01-01'). Projection Results: Probabilistic estimates for ambient change (ambient_change_p5, p50, p95). Probabilistic estimates for SLR-induced retreat (slr_retreat_p5, p50, p95). The final combined total shoreline change (total_change_p5, p50, p95). Geometry: transect_geometry: The original WKT LineString of the analysis transect. geometry: The final projected shoreline position as a WKT Point. When using this dataset, please cite the following paper: Calkoen, F. R., Luijendijk, A. P., Barli, P., Hemmes, J., Ranasinghe, R., & Nicholls, R. J. (2025). Global Coastal Erosion: Present-Day Exposure and Future Risk. Unpublished manuscript. And please consider citing this paper as well (underlying transects/metadata that enable the analytics). For the foundational transect system (GCTS): Calkoen, F. R., Luijendijk, A. P., Vos, K., Kras, E., & Baart, F. (2025). Enabling coastal analytics at planetary scale. Environmental Modelling & Software, 183, 106257. https://doi.org/10.1016/j.envsoft.2024.106257