Mon - Sat 8:00 - 6:30, Sunday - CLOSED
Heritage sites have important cultural, ecological, historical, social and economic value1,2. Yet climate change hazards such as river floods, heatwaves and wildfires threaten heritage globally3,4,5. Multiple heritage sites, including World Heritage Sites of ‘Outstanding Universal Value’, are located in the low-lying coastal zone and therefore also face threats from coastal hazards due to rising sea levels. Sea levels have been rising at a faster rate over the past three decades compared with the twentieth century6,7, a process that is expected to gather pace through the twenty-first century8,9. Together with changing weather patterns10,11, this is expected to intensify coastal flooding12,13 and coastal erosion14, exacerbating damages to coastal zone assets15. However, in contrast to other continents16, few studies have assessed climate change risks along the 300,000 km African coastline that spans 39 countries3,4,17. Even sparser is information about the future of the continent’s cultural and natural heritage sites, many of which are found in the coastal zone.
We assess exposure of African heritage sites (AHS) to coastal flooding and erosion along the entire African coastline. We create the first continent-wide, digitized, geospatial database of 284 coastal AHS, combining 71 cultural World Heritage Sites and 213 natural World Heritage Sites that are either already recognized, or currently under consideration by the United Nations Educational, Scientific and Cultural Organization (UNESCO) World Heritage Centre and the Ramsar Convention on Wetlands of International Importance18,19,20,21. Our analysis focuses on the coastal area exposed to a 1-in-100-year (that is, once in a century) coastal flooding and coastal erosion event. We estimate the exposed area at each site for the baseline (reference year 2010), as well as through the twenty-first century, under moderate (Representative Concentration Pathway (RCP) 4.5) and high (RCP 8.5) greenhouse gas emission scenarios. To assess exposure to current and future coastal floods, we derive inundation maps using a hydrodynamic model forced by extreme sea levels (ESLs; combination of sea level, waves, tides and storm surges)22. To assess coastal erosion, we post-process recent shoreline change projections14, together with site-specific geological information that can limit shoreline retreat. At each site we derive exposed area for flooding and erosion separately and then calculate the total exposed area as the union of these two, so that reported values express the combined effect. We present median values of exposed area, as well as the very likely range; expressed as the 5th and 95th percentiles. We provide information for each AHS and present our findings at country, regional and continental levels (see Methods for more details on the different steps of the analysis, including a detailed discussion of any limitations in a dedicated section).
Fifty-six (20%) of the 284 identified AHS are currently exposed to a 1-in-100-year coastal extreme event. Thirty-five of the 213 natural heritage sites (16%) and 21 of the 71 cultural heritage sites (30%) are exposed to a 100-year coastal extreme event (Table 1), corresponding to 1,719 km2 and 419 km2 of exposed natural and cultural heritage area, respectively. On average, each site has 4.5% of its area exposed. Fifty sites have <50% of their area exposed, 3 sites have >50% area exposed and 3 sites have >75% of their area exposed (Fig. 1).
North Africa has the largest number of exposed sites (23 of a total of 109; Extended Data Fig. 1). West Africa has 18 sites exposed, Southern Africa has 7, while East Africa and Small Island Developing States (SIDS) each have 4 exposed sites. Tunisia contains the most heritage sites (34), 7 of which are exposed to a 100-year event, with 2 of them being highly exposed (>75% exposed area). Morocco and Senegal have 7 exposed sites each and Egypt has 4. No Central African sites are currently exposed.
The number of sites threatened by a 100-year coastal extreme event is projected to more than triple under moderate emissions, reaching 191 (very likely range: 191–196) by 2050 (Table 1). Considering the median estimates, 68, 47, 24, 23, 16 and 13 exposed sites are found in North, West, Southern, SIDS, Central, and East Africa, respectively. High emissions will increase the total number of exposed sites by 2050 to 198 (very likely range: 198–210), with 4 of the 7 additional sites (for the median estimate) found in North Africa and 3 distributed among the SIDS, Southern and West Africa.
For both moderate and high emissions scenarios, the number of exposed sites remains stable in the second half of the century, but there is a sharp increase in the level of exposure. Under RCP 4.5, the number of very highly exposed sites (that is, fraction of site’s exposed area >75%) increases fivefold from the present-day value, to reach 15 sites (14–20) by 2100; while under high emissions this estimate increases more than sixfold to 20 (17–30; Fig. 1). The latter is the result of the fact that the increase in exposed heritage area accelerates as sea-level rise gathers pace (Fig. 2). By 2050 and under high emissions, the median additional exposed area is limited to about 25% of the baseline value (1,719 km2 versus 2,171 km2), while under moderate emissions the increase is less than 2%. However, by the end of the century the median additional exposed area increases by 9.5 times its present-day value under moderate emissions, reaching a total exposed area of 16,638 km2 (13,192–22,596; Table 1). The median exposed area under high emissions is 20,969 km2 (15,168–28,051), about 12 times the baseline value (Table 1). These findings underline the benefits of reducing greenhouse gas emissions, as mitigation from high to moderate emissions would result in a 21% reduction of the median exposed area, as well as 25% fewer sites that would be highly exposed by the end of the century (Fig. 1).
On average, AHS will have 11.2% (9.4–14.2) and 13.7% (10.7–18.3) of their area exposed under moderate and high emissions, respectively (Table 1). These percentages are higher for natural sites than for cultural sites; for example, under high emissions, 15% (11.6–20.4) versus 9.7% (7.7–12.0), respectively, by the end of the century (Fig. 2c,d). Note that the mean exposed area is used as a proxy of the overall effect of coastal hazards on AHS, but it doesn’t imply that sites are interchangeable, nor that the characteristics of each site are homogeneous. Each site contains unique outstanding value, can be made up of multiple components, and potential loss even across one site would probably be unequal. Projections show that at least 151 natural and 40 cultural sites will be exposed to the 100-year event from 2050 onwards, regardless of the scenario (median values, Table 1). As natural sites occupy almost 10 times more area than cultural ones, most of the exposed area belongs also to the former. Under moderate emissions and by the end of the century, the exposed natural and cultural area will be equal to 15,053 km2 (11,906–20,545) and 1,585 km2 (1,287–2,051), respectively (Table 1 and Fig. 2a,b). Under high emissions, the same values rise to 18,930 km2 (13,652–25,545) and 2,039 km2 (1,516–2,507), respectively.
At country level and in terms of median estimates, there are several countries that are projected to have all their coastal heritage sites exposed to the 100-year coastal extreme event by the end of the century, regardless of the scenario: Cameroon, Republic of the Congo, Djibouti, Western Sahara, Libya, Mozambique, Mauritania and Namibia (Fig. 3 and Table 2). Under high emissions and for the worst-case scenario (that is, 95th percentile), four more countries are added to this list: Côte d’Ivoire, Cabo Verde, Sudan and Tanzania. Morocco and Tunisia have the highest number of sites exposed by 2100, at least 13 more than at present (and at least 20, regardless of the scenario). With respect to the heritage area exposed, Mozambique is the most exposed country (median value exceeding 5,683 km2 under moderate mitigation; Fig. 3), followed by Senegal (>2,291 km2), Mauritania (>1,764 km2) and Kenya (>822 km2). Tanzania, Mozambique, Côte d’Ivoire, Benin, Togo and South Africa are countries that by the end of the century will have at least 100 times more exposed heritage area than at present. Ghana, Sierra Leone, Libya, Mozambique and Seychelles are projected to have 51%, 30%, 25%, 21% and 20% of their heritage area exposed by the end of the century, respectively, under high emissions (Fig. 4).
SIDS heritage sites are especially at risk. For example, Curral Velho (1,575, Ramsar site in Cabo Verde) is an important wetland that will be exposed to coastal hazards by 2050. Under high emissions and by the end of the century, 44% of the site’s total area will be exposed. Aldabra Atoll (1,887), the world’s second-largest coral atoll, and Kunta Kinteh Island (The Gambia) could both see up to 17% and 46% area exposed by 2100 under high emissions, respectively.
Notable increases of percentage site exposure are also projected for some archaeological and cultural sites, such as the North Sinai Archaeological Sites Zone (189; 91%; RCP 8.5, 2100), Agglomération Aného-Glidji (1,505; 37%; RCP 8.5, 2100), the ancient Punic and Roman trading post, Tipasa (193; 11%; RCP 8.5, 2100), the archaeological site, Sabratha (184; 7.7%; RCP 8.5, 2100) and the archaeological site, Carthage (37; 5.9%; RCP 8.5, 2100). Although Qaitbay Citadel, the ancient site of the Lighthouse of Alexandria (1,822) and one of the Seven Wonders of the Ancient World23, is projected here to experience minimal exposure due to existing protections, it has already experienced severe flooding in 2019, leading to the construction of coastal defences24.
Chat Tboul (1,044) and Parc National du Diawling (666) in Mauritania are examples of sites already exposed to extreme flood events despite efforts to ecologically restore the floodplain25,26. Similarly, the vulnerability of the socio-ecological system of the Densu River Basin (Ghana) might be accentuated by the predicted total flooding of the Densu Delta Ramsar site (564) by 2100 under both emission scenarios, compared with a present-day value of only 47% area flooded27. Relict Guinean coastal forests have already largely disappeared due to coastal erosion28,29. For Parc National du Diawling (666), future flooding and erosion will affect the entire site, compared with a baseline of only 45% area exposed. Such a substantial increase may affect the ecological equilibrium of the site’s ecosystem.
Africa is home to some of the most diverse cultural and bio-cultural heritage in the world, internationally recognized for its uniqueness and ‘Outstanding Universal Value’30. Heritage sites have continuously served as ‘living’ heritage31 and therefore are deeply interwoven with the people’s identity and tradition, are essential for social wellbeing, safeguarding traditional knowledge and livelihoods, and constituting a prerequisite for sustainable development32. Yet, we find that 1 out of 5 coastal AHS are already at risk from a 1-in-100-year ESL event, a number that is projected to almost quadruple by the end of the century.
More heritage area is exposed to flooding compared with erosion, but as the impact mechanism of the two hazards is different, their relative importance is site-specific. Erosion would have a stronger effect compared with inundation, the effect of which depends on the interplay between the flood depth, the flow velocity and the element inundated. Cultural sites, which tend to be either archaeological or historical built heritage, will be affected by both erosion and flooding, while bio-cultural and natural areas are more likely to recover from episodic flooding. A partially flooded or eroded natural area may accommodate these disruptions and maintain ecological equilibrium, either by migrating landwards, or even while shrinking, in the frequent case that retreat is constrained by coastal development.
How much area a heritage site can lose to flooding and erosion and still maintain its value (for example, cultural, ecological, Indigenous and economic) is a question of growing importance for all protected areas and World Heritage Sites, and demands site-specific local studies. The same applies to the capacity of natural coastal systems to adapt and absorb other external shocks, such as changes in salinity, which remains unknown17.
Anthropogenic modification of coastal processes will also affect natural systems’ responses to shoreline change. For example, in the Bight of Benin, West Africa, the construction of dams on the Volta River, combined with lower rainfall, contributed to a decrease in sediments on the coast, thereby increasing the effects of coastal flooding and erosion33. Several sandy beaches of the continent are naturally protected by ecological elements, such as coral reefs, seagrass and mangroves34,35,36. However, the fate of coral reefs depends on future marine heatwaves37 and ocean acidification trends38—both of which are expected to increase all around the continent—while mangroves are also threatened by rising seas. For example, five species of mangrove are listed among biota likely to become locally extinct in Ghana, if rising seas outpace the rate of forest migration39. In Central Africa (Cameroon) as well as in other regions, mangrove logging and anthropization also accelerates these effects40. The eastern African coast, considered a region of high diversity of seagrass41, is subject to frequent anthropogenic disturbance, resulting in the loss of about 21% of Kenya’s seagrass cover between 1986 and 201642. Such transitions could have further indirect effects and weaken natural coastal protection, further exacerbating flood risk; with substantial social consequences43.
Our findings help with prioritizing sites at risk and highlight the need for immediate protective action for AHS; the design of which requires in-depth local-scale assessments of vulnerability and adaptation options. For many cultural and natural sites, relocation or managed retreat might be the least favourable option due to intrinsic values of site locations and potential impacts on local communities3. Coastal protection through the construction of breakwaters, groins and beach nourishment may be effective, where they are technologically and financially feasible. For example, protections to Qaitbay Citadel (1,822), Egypt, have recently been reinforced44. However, such ‘hard’ protection strategies need to consider future sea levels and are known to distort the site’s natural ecological and morphological equilibrium45. Hybrid protections that include ecological infrastructure, such as rock sills combined with saltmarshes, seagrasses or restored mangroves, may prove more effective17,45. Improving local and Indigenous governance, monitoring and evaluation, and broader land management actions, such as expanded buffer zones, can further provide enabling conditions for site protection that address existing vulnerabilities2,46,47.
As understanding of climate risk to heritage grows, there is potential for these exposure findings to raise public concern and mobilize rapid and ambitious greenhouse gas mitigation to reduce overall risk and potential loss and damage3. Future research needs to quantify climate risk to heritage more broadly, including risks to inland heritage across Africa. Knowledge is also needed on risks from a broader range of climate hazards, and particularly potential impacts from compound climate extremes. Finally, better understanding is needed of risks from responses to climate change that will also affect heritage48, such as migration, managed retreat, ecosystem-based adaptation and relocation.