Future socioeconomic development along the West African coast forms a larger hazard than sea level rise

The study addresses how sea level rise (SLR) and socioeconomic development will affect future coastal flooding risk across West Africa (WA). The context is that global flood and storm damages have increased in the 21st century, largely due to socioeconomic change and evolving climate hazards. Even small differences in global mean sea level by 2100 could translate into trillions of dollars in additional annual flood losses without adaptation. Vulnerability is disproportionately borne by poorer populations in developing countries, notably in Africa, where adaptive capacity is limited. Relative SLR experienced by populations is often dominated by land subsidence, much of it anthropogenic and tied to coastal development and resource extraction. In WA, low-lying coastal nations (Mauritania to Nigeria) already host about one-third of their populations and generate 56% of GDP below 10 m elevation, making them highly exposed. Rapid population growth, coastward migration, urbanization, and unregulated development are increasing exposure and may exacerbate subsidence, as seen in places like Saint-Louis (Senegal). The purpose of the study is to quantify present and future exposure of population and assets to normal (median) and extreme (98th percentile) coastal water levels (CWLs) along WA under IPCC AR6 sea-level projections and Shared Socioeconomic Pathways (SSPs), identify regional and national hotspots, and assess the relative roles of climate-induced SLR versus socioeconomic development in driving future risk.

The paper situates its analysis within literature showing rising global coastal flood damages driven by both climate and socioeconomic changes, and highlights studies quantifying the sensitivity of damages to small SLR differences. It reviews evidence that low adaptive capacity and anthropogenic subsidence (from groundwater, oil and gas extraction, and heavy infrastructure) contribute substantially to relative SLR and flood risk in coastal zones. Prior global and regional assessments project continued increases in flood exposure in West and Central Africa, where many coastal areas are low-lying and undergoing rapid urban growth. The paper also references work demonstrating that under certain SSPs exposed population can be highest even when asset exposure is not, and that socioeconomic change can dominate flood risk increases absent adaptation. Case examples (e.g., Saint-Louis, Senegal; Lagos, Nigeria; Greater Accra, Ghana) from prior studies illustrate how urban expansion degrades wetlands, increases flooding, and how groundwater abstraction can drive subsidence, compounding flood hazards.

• Study domain and CWL estimation: Coastal water levels (CWL) were estimated at 24 locations approximately every 50 km along the West African coastline, prioritizing even spatial coverage and socio-environmentally relevant coastal cities. CWL is decomposed as CWL(u) = T(u) + W(u) + S(u) + SLA(u), where T is tide, W wind setup, S storm surge (including tide anomalies), and SLA regional sea level anomaly, all as a function of time, date, and location. Normal CWL (NCWL) refers to median (50th percentile) and extreme CWL (ECWL) to the 98th percentile.
• Baseline and projections: Historical CWLs were reconstructed for 1993–2015 to provide a baseline. Future CWLs to 2050 and 2100 were projected by coupling NCWL/ECWL with IPCC AR6 sea-level projections under SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5.
• Inundation modeling: A static “bathtub” (bathymetric inundation) approach was applied to delineate areas flooded by NCWL and ECWL under present and future scenarios, assuming no coastal defenses. The method identifies connected areas below projected water levels using coastal topography and drainage.
• Topographic data: The MERIT DEM (3 arc-second global product) was used to represent coastal elevations due to its reduced noise and corrected regional biases compared to other global DEMs. AW3D30 was used for deriving coastal slopes and drainage where appropriate, acknowledging DEM-related uncertainties.
• Exposure estimation: Population exposure was computed by overlaying inundation extents on gridded population rasters consistent with SSPs. Asset exposure was estimated by translating exposed population to exposed assets using country-specific GDP per capita and an asset-to-GDP ratio of 2, i.e., A = P × G, where A is exposed assets (US$), P exposed population, and G GDP per capita. The analysis distinguishes contributions from climate-induced SLR (holding socioeconomic factors static) and from socioeconomic development (population and GDP growth) to quantify their relative roles in changing exposure.
• Tools and data sources: Analyses were conducted in a GIS environment (QGIS), using IPCC AR6 SLR projections, MERIT DEM, gridded population projections consistent with SSPs, and GDP data. Figures were generated using QGIS and satellite basemaps; additional dynamic products and auxiliary datasets are referenced in the Data Availability.

• Coastal water level projections: Mean NCWL and ECWL along WA increase from ~0.83 m and 1.97 m (1993–2015 baseline) to 1.05–1.08 m and 2.25–2.28 m by 2050 across scenarios, and to 1.31/2.62 m (SSP1-2.6), 1.32/2.73 m (SSP2-4.5), 1.56/2.95 m (SSP3-7.0), and 2.62/3.17 m (SSP5-8.5) by 2100.
• Country variability in 2100 (SSP5-8.5): Nigeria shows the highest projected NCWL/ECWL (up to ~2.4/3.5 m). Other countries exceeding the WA mean include Benin (~2.92/3.25 m), Ghana (~2.91/3.42 m), Liberia (~2.61/3.2 m), and Guinea Bissau (~2.86/3.16 m).
• Exposed population: Potentially exposed WA population increases from ~0.7–1.1 million (2015) to multi-million ranges by 2100 across scenarios, corresponding to a roughly 9–16-fold increase compared to 2015.
• Exposed assets: Exposed assets increase from ~US$7.8–11.5 billion (2015) to on the order of hundreds of billions by 2100: ~US$464–580B (SSP1-2.6), ~US$303–620B (SSP2-4.5), ~US$153–302B (SSP3-7.0), and up to the order of ~US$655–953B (SSP5-8.5), implying roughly 100–300× growth compared to 2015.
• Dominant driver: At the regional scale, socioeconomic development is the dominant factor driving increases in exposure to coastal water flooding (CWF) across all climate scenarios. At the country level, dominance varies: SLR dominates exposure in Guinea Bissau, Mauritania, Guinea, Cameroon, and Togo, while socioeconomic development dominates in The Gambia, Benin, Senegal, Côte d’Ivoire, Nigeria, Sierra Leone, Liberia, and Ghana.
• Concentration of risk: By 2100 (SSP5-8.5), Nigeria, Senegal, Côte d’Ivoire, Benin, and Ghana together account for ~80.7% of exposed population and assets to ECWL region-wide, with Nigeria alone exceeding 50% of the exposed population.
• Scenario contrasts: Exposed population is highest under SSP3-7.0 and lowest under SSP1-2.6, whereas exposed assets are highest under SSP5-8.5 and lowest under SSP3-7.0, reflecting differing socioeconomic and climatic trajectories.

The analysis shows that while climate-induced SLR will increase coastal water levels and flood hazard, the rapid growth of population and economic activity in low-lying WA coastal zones will be the principal driver of future exposure at the regional scale. This finding reframes the risk management challenge: even if SLR progresses as projected, unchecked urbanization, coastward migration, and expansion into hazardous areas will amplify exposure beyond what SLR alone would cause. Country-level differences highlight where SLR controls risk (e.g., Guinea Bissau, Mauritania, Guinea, Cameroon, Togo) versus where socioeconomic dynamics dominate (e.g., Nigeria, Côte d’Ivoire, Benin, Senegal, Ghana). The concentration of risk in a handful of countries, especially Nigeria, underscores the need for targeted adaptation and planning. The scenario contrast—higher exposed population under SSP3-7.0 but higher exposed assets under SSP5-8.5—aligns with prior studies and reflects interactions between demographic and economic growth pathways and climatic forcing. These results support prioritizing integrated strategies: managing coastal development patterns (zoning, limiting expansion into wetlands), strengthening defenses and nature-based solutions, and addressing drivers of subsidence (e.g., regulating groundwater and subsurface resource extraction).

Under IPCC AR6 SLR scenarios and SSP-consistent socioeconomic pathways, West Africa’s exposure to coastal flooding will rise sharply this century. Although SLR will dominate hazard changes in some countries, at the regional level future coastal water flooding is likely to be dominated by socioeconomic development, given ongoing urban expansion and economic transformation along the coast. Policymakers should implement comprehensive adaptation strategies that include planned relocation, risk-informed development, restrictions in high-risk zones, protection and restoration of coastal ecosystems, and measures to curb subsidence. Future work should refine local-scale assessments, integrate dynamic defenses and adaptation, and evaluate governance mechanisms to manage rapid socioeconomic development while safeguarding coastal ecosystems and communities.

• DEM and elevation uncertainties: Flood extents are sensitive to elevation data. While MERIT DEM reduces noise and biases compared to some global DEMs, uncertainties remain of similar magnitude across available global DEMs. AW3D30 was used for slopes/drainage but is less suitable than MERIT for regional flood mapping. High-resolution LiDAR-quality DEMs exist locally but lack regional coverage; new missions (e.g., COS3D) are needed for consistent, repeatable monitoring.
• Static inundation (bathtub) approach: The method does not simulate dynamic hydrodynamics (e.g., wave setup, river discharge interactions, drainage infrastructure), potentially misestimating flood extents in complex settings.
• Exposure modeling assumptions: Assets are inferred from exposed population using GDP per capita and an assumed asset-to-GDP ratio (2), introducing uncertainty because assets are unevenly distributed spatially and socioeconomically.
• Socioeconomic projections: Population and GDP projections entail scenario and model uncertainties; adaptation measures are not explicitly modeled and could reduce future exposure.
• Data and temporal sampling: CWL estimates are based on components and percentiles from historical (1993–2015) baselines and may not capture all extremes or local processes (e.g., compound flooding, rapid morphology change).