Coastal areas contain some of the world’s most diverse and productive resources that support a variety of economic activities, including fisheries, tourism, recreation, and transportation. They are, nevertheless, extremely vulnerable to human intervention and climate change. Their recreational, ecosystem and protection services, and economic values give them high societal attention, especially given that these areas are home to more than a trillion dollars' worth of infrastructure and investments globally (Sutton-Grier et al., 2015; Holman et al., 2003). Many coastal areas have undergone extensive modification and development over the past few decades, greatly increasing their vulnerability to both the anticipated effects of global climate change and the natural coastal dynamics (He and Silliman, 2019; Nicholls et al., 2014).
Africa's coastal zone extends over 41,184 km, covering 32 continental countries. Most of the continent's economic activities are located in coastal areas, with the presence of industrial zones around the main deep-water harbours. This has led in recent decades to a demographic explosion, and major and rapidly expanding cities on the coast. Indeed, each year, 10 million people are added to the urban population of sub-Saharan Africa (State of the World’s Cities Report 2010/11). Accordingly, several cities along the African coasts have more than one million inhabitants (see Fig. 1). The African coasts are subject to other activities such as artisanal fishing (e.g., Tanner et al., 2017; Camprodon and Cuq, 2001), logging for firewood and charcoal (e.g., Sakyi et al., 2019; Adeyeye and Oyewole, 2016), sand mining (e.g., Jonah et al., 2015), tourism (e.g., Saarinen et al., 2022; Atanga and Tichaawa, 2020), as well as onshore and offshore industrial activities (e.g., processing industries, offshore oil extraction). Such activities lead to pressures and conflicts over the use of resources, often resulting in environmental degradation, which can threaten their development potential (e.g., Alves et al., 2020; Croitoru et al., 2019; Feka and Ajonina, 2011). Small-scale fishing towns and natural ecosystems are often the first victims of the induced vulnerability, as they are highly sensitive to climate variability and change due to their low adaptive capacities, which are insufficient to cope with future climate change (IPCC, 2007).
Africa is the continent with the highest fraction of sandy beaches and low-lying coastal areas (Luijendijk et al., 2018; Vousdoukas et al. (2022)) and is therefore very exposed to ocean, climate and human activities impacts. Indeed, several recent works showed that in recent decades, the coastal environment has been facing several disturbances in waves forcing and shoreline stability in West Africa (e.g. Guerrera et al., 2021; Aman et al., 2019; Dada et al., 2016; Almar et al., 2015; Nicholson, 2001), in North Africa (e.g. Hzami et al., 2021; Amrouni et al., 2019), in Central Africa (e.g. Fossi et al., 2018; Abessolo et al., 2018), in Southern Africa (e.g. Cawthra and Zyl, 2015; Theron, A.K., and Rossouw, 2008) and in East Africa (e.g. Shaghude et al., 2015). These works highlighted that as a result, the Africans coasts have become particularly vulnerable to erosion and flooding due to high-energy storm waves, extreme storm-surge water levels, sea level rise, and the high proportion of soft low-lying coastlines. In West Africa, this vulnerability is particularly aggravated by the decrease in rainfall and the presence of dams on the main rivers (Volta and Niger Rivers) thus reducing the supply of sediments to the coast (Anthony et al., 2019). According to Vousdoukas et al. (2022), 20% of the 284 reported African coastal natural and cultural heritage sites are currently at risk from an extreme coastal event, including the iconic ruins of Tipasa (Algeria) and the North Sinai Archaeological Sites area (Egypt). The vulnerable sites are expected to triple by 2050 if greenhouse gas emissions are high.
Decision-making for coastal management in Africa remains a challenge in the face of increasing risks to people and their property (Vousdoukas et al., 2022; Alves et al., 2020), as several factors impede the sustainable development of these areas. Firstly, multi-scale morphological evolution remains poorly understood on tropical coasts (Almar et al., 2014a). To ensure more integrated and efficient management of coastal areas, it is important to characterize the natural coastal systems, the dominant processes, and their integrated effects adequately. Secondly, there are few local indicators to predict the large-scale (spatial and temporal) impact of coastal infrastructure construction and development and changes in ocean forcing on coastal morphology (Hzami et al., 2021; Tano et al., 2016). Numerical modelling would undoubtedly enable the derivation of this type of indicators (Vousdoukas et al., 2022; Abessolo et al., 2021; Giardino et al, 2018). Nevertheless, calibration and improving existing models require large databases and better descriptions of nearshore processes. Thirdly, the discordance of actors and responsibilities among African countries, and sometimes even within the same country, has too often prevented the planning of effective and sustainable long-term solutions (Giardino et al, 2018). There are regional programs that synergize efforts to understand the dynamics of these environments such as the West African Coastal Areas (WACA) Management Program and the West African Coastal Observatory (ORLOA). But the results of these integrated programs and their impacts are not yet prominent.
Effective decision-making for coastal management in Africa, therefore, requires the best and most up-to-date science. And this depends on the ability to collect long-maintained and real-time data, necessary to interpret and predict nearshore processes occurring at multiple temporal scales. Pioneering field measurement campaigns have been conducted for this purpose, providing crucial baseline measurements of short-term time-scales processes (e.g., Almar et al., 2014a; Ndour et al., 2020). However, they are expensive and limited to days to weeks. The use of satellite radar and optical images enabled large spatial coverage of morphological variations. It is, however, limited to resolutions of tens of meters (10 m for Sentinel, 30 m for LandSat) with repeatability of several days (2–5 days for Sentinel 2 – Bergsma and Almar, 2020), thus limiting the long-term observation of short-term processes. Moreover, altimetry, which measures sea surface variations, is still limited to a few kilometers from the coast because of inappropriate sensors for the nearshore. While satellite earth observation represents a real opportunity to monitor African coasts (Almar et al., 2022), they face several challenges and the lack of ground truth is one of them. Buoys allow the measurement of hydrodynamic parameters (e.g., sea level, currents) closest to the coast, but they are almost non-existent or unavailable along the African coasts. Purchase and installation of buoys are, however, expensive, and their maintenance is very complex, especially for a network covering large regions. And African coasts face a lack of historical data from tide gauges (Woodworth et al., 2007).
Nearshore video camera systems are a low-cost option for high-frequency continuous long-term monitoring, with the advantage of being complementary to traditional measurement techniques (satellite, buoys, tide gauges, etc.) and particularly suited for tropical developing countries and open-wave areas (Almar et al., 2014a). Several parameters are estimated from video imagery, such as the shoreline position (e.g., Almar et al., 2012b, Pianca et al., 2015), intertidal topography (e.g., Osorio et al. 2012), nearshore bathymetry (e.g., Holman et al., 2013, Bergsma and Almar 2018, Abessolo et al., 2020c), breaking wave height (e.g., Almar et al. 2012a), sea level variations (e.g., Abessolo et al., 2019, Thuan et al., 2019), nearshore currents (e.g., Radermacher et al., 2014). Video cameras allow for the collection of hard-to-access data on a continuous, long-term basis, effectively filling the data gap and, most importantly, ensuring effective monitoring of ocean and atmospheric conditions (Holman, Stanley and Ozkan-Haller, 2003).
A pilot video camera station was installed at Grand Popo (Benin) in 2013, and the network is currently extending to West and Central Africa (Fig. 2). In this paper, we demonstrate how this pilot network can be used to build a database to document African coasts evolution, and how the network can support effective land-use planning policy. We present the main scientific advances induced by the data collected, and we discuss the benefit of building data and scientific expertise in the management and planning of the coastal environment in Africa.
New Insights Through Coastal Video Camera Monitoring
All-in-one measuring system at the coast: methodological developments
Most of the parameters needed to understand the African coastal environments are measured simultaneously in this network: wave’s characteristics, water level variations and beach morphology. Several studies have been conducted to investigate these video-based estimates and the results showed in general an overall error below 10% (see results in Table S2).
The main characteristics of waves (e.g., Fig. 3) that represent the most important forcing in open coastal environments are determined at wave breaking area where it is difficult to install conventional measuring equipment due to the dissipation of wave energy. For instance, it provides long-term local wave datasets (Abessolo et al., 2017a; 2017b) with much better accuracy than regional coarse available data (such as global coarse ERA5 reanalysis - Fig. 3). Available models are very often inaccurate when they are simplified for fast computational needs (e.g., Larson et al., 2010), or require a lot of computational power, often out of the reach of African countries, to account for complex coastal processes and wave transformation.
Monitoring morphological variations of a beach profile is shown to be feasible (Fig. 4) on a daily scale and over years from a video camera station (Abessolo et al., 2020c). The accuracy of topo-bathymetric measurements is crucial, as they are probably the most critical variables useful for understanding and modelling coastal dynamics and variability. However benchmark data are not always available to assess the accuracy of video estimates. Bergsma and Almar (2018) and Thuan et al. (2019) defined two new error estimators based solely on video processing to assess the accuracy of video-derived bathymetry: the wavelength-related-error proxy and the tide-related-error proxy. These proxies express a reliability rate on bathymetric estimates to identify erroneous data and extract a useful data set when applying a Kalman filter, in the same way as the cBathy method (see Holman et al., 2013). Abessolo et al (2020b) tested these two proxies on data collected at Grand Popo. Their results showed that both estimators reduce errors by at least 30%.
Capturing coastal water level
There is an obvious need to extend the satellite-based sea level record toward the coast, where altimetry cannot yet reach, and where tide gauges are difficult to install. Abessolo et al. (2019) showed the potential of video stations to directly measure tide and sea level rise at the coast (Fig. 5). Therefore, video stations could help in understanding the main factors that determine the accuracy of near-shore altimeter data; and the propagation of sea level variations to the coast (see Abessolo et al., 2022). In addition, imagery from video stations are needed to refine estimates of wave contribution to sea level variations at the coast, which have so far been estimated by empirical formulations (Melet et al., 2020). Indeed, waves undergo a series of complex transformations when propagating from deep-water to shallow water, resulting in the well-known process of wave set-up (Pugh and Woodworth 2014) which adds a physical contribution to sea level variations. Using collocated dataset, tide gauge, and wave buoy, Abessolo et al. (2022) showed that this contribution could be estimated from a video camera station, thus highlighting the potential of optimizing the location of coastal observation networks with respect to satellite ground tracks (as expressed in Marti et al., 2021). Tidal harmonics were successfully extracted together with long-term sea level rise, and the signature of sea level variations induced by coastal waves, such as Kelvin waves, could well be present in the video data, but this remains to be studied and investigated through more developments and studies.
Witnessing African beaches changes
From the monitoring of topo-bathymetric variations at the storm free environment of Grand Popo (Benin), beach profile evolution (shoreline and terrace, see Fig. 4) was observed to be dominated by seasonal to interannual scales (Bonou et al., 2018; Abessolo et al., 2017a). The action of coastally-trapped waves (Kelvin waves) has also been demonstrated: a 7-cm change in sea level leads to nearly 2 m of horizontal terrace deformation (Abessolo et al., 2020c). Based on these video-derived morphological variations, Mingo et al. (2021) proposed and discussed a simple cubic parameterization based on the widely used Dean Number as the dominant control parameter for such beaches, while extending the common view of an equilibrium range profile as a power law of cross-shore distance.
The importance of intraseasonal to interranual variation of sea level on beach dynamics was also evidenced at James Town (Ghana). Angnuureng et al. (2018) demonstrated that shoreline changes are preferentially explained at daily scales by sea level anomaly (86%), waves (9%), and tide neap-spring cycles (5%) scales. These results show the role that the sea level anomaly can play in modulating the magnitude of wave action on the beach as stipulated by Abessolo et al. (2020c).
Revealing major drowning hazard
Studies carried out in Benin allowed to better understand the rip currents to prevent and reduce the numerous cases of drowning deaths due to this phenomenon on the beaches of Africa.
Wind and waves bring large amounts of water onto the beach while the previous backwash (violent wave return) is discharged laterally. This flow of water parallel to the beaches can only return to the open ocean if it finds an opening in the wall of water from the open ocean. This usually occurs at depth discontinuities in a sandbar, under a jetty or pier, which forces the discharging water into a thin corridor and accelerates the flow. The study by Castelle et al. (2014) is the first ever carried out in West Africa on the functioning of rip currents using video images. These strong currents carry the water brought by the large waves breaking on the beaches towards the open sea. This phenomenon is the main deadly danger for beach users all over Africa and the world. Floc’h et al. (2018) developed a new method based on the colour processing of video images to detect the presence of rip currents via a change in turbidity. Morphological characteristics of the flash rip were computed from a set of instantaneous images. More than 50 rip events were counted per day (from 07:00 to 17:00 local time) during the 10-day measurement campaign in Grand Popo, Benin (Floc’h et al., 2018), most occurring at low tide and migrating downdrift.
Almar et al. (2016) described a method based on the application of the Radon transform on longshore time-stack images to directly estimate longshore currents and tested the method with field measurements at Grand Popo. In addition, transient circulation was detected. Flash rips and surf eddies are transient horizontal structures of the order of 10 to 100 m, which can be generated in the surf zone in the absence of bathymetric irregularities. Using video-derived current fields, Marchesiello et al. (2021; 2016) revisited the processes of surf eddy generation with a new three-dimensional wave resolution model (CROCO) and provided a plausible demonstration of new 3D non-hydrostatic instability and turbulent cascade. Instant fields of surface current were retrieved from shore-based and unmanned aerial vehicle videos by an optical flow (OF) method named "Typhoon" (Derian and Almar, 2017). Such computer vision algorithm estimates dense two-dimensional 2-component velocity fields from the observable motion of foam patterns in the surf zone.
Coastal Zone Management
Beach volume monitoring
The video camera station installed at the enclosed Elmina beach is considered to first test multi-platform monitoring of coastal erosion and secondly, evaluate the effectiveness of beach defense structures (revetments and jetties). The multi-platform data collection strategy (Angnuureng et al., 2022) was deployed for a year at Elmina, through the use of the installed video camera station, drones (unmanned aerial vehicles), Sentinel satellite images, and a dumpy level. Angnuureng et al. (2022) revealed that data from local video cameras and drones are more effective for operational monitoring of shoreline changes at all-time scales compared to satellite imagery. They also clearly identified the potential causes and areas of erosion at Elmina, associated with breaking wave conditions. The obtained results indicated that the presence of the defense systems adequately protected the section of the beach, and a larger unprotected portion of the beach was out of balance with high erosion rates (Angnuureng et al., 2022). This is an example of the application of video data: to verify and evaluate the effectiveness of protective structures that are built along the coasts of Africa, and how they can induce upstream or downstream morphological imbalances. This type of multi-platform monitoring allows coastal management to gauge the potential hazards associated with the siting of housing and coastal infrastructure.
Early warning system
The rapid development of video cameras (Fig. 6) now offers many applications that can provide quantitative information for swimmers, naval engineering, safety, and research. A vulnerability evaluation of the coasts in Africa using data from video camera stations is feasible nowadays in a sustainable way. The video camera stations could lead to the establishment of early warning systems for extreme events such as storms or floods on a regional scale. For example, in Senegal, the installed video camera can measure local conditions and serve as an alert system to fishermen in case of very dangerous swell conditions. In this case, the real-time processing of the data collected by the video camera station will facilitate the development of now-casting modules. In the locality of Cap Cameroon (Cameroon), which is inhabited mainly by fishermen, the recession of the coastline has led to the immersion of more than 300 hectares of coastal land per year (Fossi et al., 2018). Similarly, the Langue de Barbarie, near Saint Louis, has undergone significant morphological changes following the opening of an artificial breach (Taveneau et al., 2021). The coastal African camera network offers the potential for more effective and controlled management of coastal development and protection. For example, the relocation and construction of houses on stilts can be better coordinated, rather than leaving them to the vagaries of weather and flooding. Indeed, the regular monitoring of oceanic forcing allows one to foresee with reasonable accuracy the associated morphological variations and thus to define zones at risk. The coastal African camera network represents an excellent risk assessment and early warning tool for the security of people and goods.
Toward more science-based coastal zone management
According to Dada et al. (2021), coastal geophysical processes (particularly coastal erosion) and public policies are parts of the greatest challenges to be addressed and should be seen as a priority for research in Africa. They, therefore, recommended an integrative conceptual framework that encompasses the study of phenomena ranging from geophysics to society, and the development of coastal management policies. Other studies (e.g., Giardino et al., 2018, Abessolo et al., 2021) have highlighted the interdependence between different coastal developments, whether along rivers or on the coast, on the overall sediment budget of the entire region, which is not limited to the geographic boundaries of countries. These studies thus recommend the development of a much more integrated large-scale coastal management plan.
The video camera network described in this paper is one of the guarantees for the implementation of a more integrated management policy along the African coasts. Video camera stations serve as a multi-variable monitoring platform, with the collection over long periods of real-time estimates of morphological and hydrodynamic variations around the wave breaking zone, where in-situ measurement techniques and other types of remote sensing (e.g., satellite, drone, Lidar) are either too costly or inadequate. The multiplicity of video camera station sites contributes to the strengthening of the database necessary to study the associated physical phenomena. Moreover, the low financial cost of establishing a video camera station guarantees the spatial development of this network whose maintenance depends almost only on human resources. The generated database could serve as a basis for the study of natural processes, the prediction of associated risks, and therefore for the development of a common regional coastal management policy. This includes the identification of objectives, priorities, strategies, and timetables common to several countries to achieve them, even if compliance with legal and regulatory frameworks is a challenge that falls under the sovereignty of the States, with of course the consideration of United Nations guidelines at the level of the African Union. In addition, this camera network can be the basis for the development of monitoring systems specific to the sub-region concerning the management of, for example, the registration of the departure and arrival of artisanal fishing boats, which are essential in the region. Other local challenges, such as Sargassum-related problems, could be addressed through the data collected by the video camera station network.
Potential for the management of the African coast
Some of the problems facing coastal managers in Africa are the availability of optimally spatial-temporal resolution data and insufficient understanding of the coastal system. Coastal video cameras can provide valuable information with adequate spatial/temporal resolution and coverage for efficient management of the African coast (Davidson et al., 2007). The traditionally in situ measurements approach for monitoring coastal systems lacks the spatial resolution and coverage required for effective coastal zone management. This limitation is further exacerbated by the cost and logistical difficulties associated with a deployment. Video camera systems exhibit the ability to at least partially address these issues by providing relatively continuous, inexpensive daylight data for periods stretching to decades when compared to other remote sensing methods (Davidson et al., 2007).
Video imagery can document coastal processes without experiencing several issues associated with in situ instrumentation, such as flow disruption, biofouling, and sensor degradation under unfavourable wave conditions (Holland et al., 1997). It can provide valuable scientific information to understand and study the complex coastal processes, particularly the nearshore dynamics. The logistics and expense of using a video camera for coastal monitoring are typically less than those of more conventional approaches, which involve placing a vast array of instruments at a limited number of locations. Unlike satellite imagery which has a low temporal resolution for detecting daily or event-based beach changes and low spatial resolution for coastal studies and management decisions, video cameras provide a pixel resolution and cover timescales from seconds to years (Angnuureng et al., 2022). Although more suitable for smaller-scale applications, video camera system networks can be used for long-term coastal monitoring and to track the impacts of human and natural processes, especially climate change, on coastal ecosystems. Thus, the ongoing data acquisition by video camera could be a crucial backbone for coastal monitoring and management along the African coast.
Unfortunately, there is significant insufficient information about sea level rise in the region. The lack of timely, accurate, and trustworthy data has consistently been a problem in Africa. It still poses a significant obstacle to the efficient monitoring and assessment of development and intervention projects in the region. But data remains a critical tool for decision-making, development, and raw material for the quality of information. The coastal African camera network has the potential to bridge the problem of non-data availability for coastal studies. The video data from the network can provide data and information that will help coastal scientists, investors, managers, and policymakers in improving the spatial planning and management of the African coastal areas. The network can be used to gather data that will be used for studies on flooding, erosion, coastal structures, infrastructure monitoring, oil pollution assessment, overfishing, and ecosystem and ecosystem services. It can provide new data and information useful for understanding and managing African coastal environments to the Coastal Observatories such as the World Bank-funded WACA and ORLOA.
Future perspective
The coastal camera network is rapidly growing, and it is set to become an African data hub for coastal video imagery. Such a platform could be linked to the World Bank-funded WACA Management Programme; or the ORLOA, an observatory established to develop the Central and West African coastal territories in a sustainable manner, through the promotion, production, and sharing of reliable and homogeneous data using harmonized data collection protocols for decision-making. The ORLOA could be used as a regional centralized repository accessible to all actors (researchers and decision makers). The repository must be made up of generated data and processing algorithms, to promote sharing and integration of all actors. The network is poised to make significant progress on coastal environmental issues that have significant societal impacts. It is, therefore, essential to expand the video camera system coverage to other locations, especially the coastal sectors that have been characterized and identified based on their level of hazards and issues.
In the long-term, the data acquired has considerable potential for use in other coastal environmental issues such as in the management of ecosystem protection, including competing coastal land uses (e.g., conservation vs. development), fisheries, and resource management, recreation, plastic and oil pollution monitoring, navigation, and coastal protection. Similarly, it has the potential of providing key information that can be used in basic input to the engineering design of the coastal zone, and support for integrated numerical model validation. This should be strongly encouraged. However, strengthening the existing stations and creating new ones requires more stable funding. The African and international coastal community, including academics, governments, regional (WACA/ORLOA), and local and international development partners are welcome to be part of the network. If this vision is achieved, coastal video camera imagery will transform environmental monitoring along the African coasts. It is the only way forward if we want sustainable African coasts.