Coastal erosion is a natural geomorphological process amplified by rising sea level and more frequent storms. The coastline receding increases the vulnerability of communities and rich ecosystems. Monitoring the coastline is necessary to assess the infrastructure risks, the ecosystem's dynamic and sediment budget and transport. We developed Earth Observation (EO) solutions for monitoring the coasts of France, Spain, United Kingdom, Ireland and Canada in the framework of the ESA-funded Coastal Erosion initiative. We will present a rigorous validation exercise of instantaneous satellite-derived waterline (SDWL) with in-situ measurements carried on along the north shore of the Gulf of St. Lawrence, Qc, Canada. SDWL is defined as the instantaneous limit between land and water and is extracted from Landsat and Sentinel 2 satellite missions. After a rigorous co-registration of the imagery, the processor computes spectral indices (Normalized Difference Water Index, etc.) and applies thresholds to detect land and water. An edge detection algorithm is then applied for the waterline delineation and an automatic quality control step is performed.
SDWL were validated using instantaneous in situ waterlines (ISWL) derived from a network of autonomous RGB video cameras at three highly dynamic sites (beaches) monitored for coastal erosion for the last 5 years that record images of the beach at very high temporal (4 Hz) and spatial (from 0.30 cm to 3 m) resolution. The video imagery was orthorectified using terrestrial LiDAR-derived topography and DGPS ground control points. Finally, the ISWL were extracted on a 5-minute frame average using an RGB ratio.
ISWL and SDWL with less than a 6-minute time difference were compared. The distance between the ISWL and SDWL was obtained for 20 matchups during summer months in 2016, 2017 and 2018. The mean distance between ISWL and SDWL is ~15 m and ~3 m for Landsat 8 and Sentinel 2, respectively. The accuracy mainly depends on the sensors and the pixel’s resolution. Further work is needed to address the gaps and limitations of the current product and improve the SDWL. Finally, we will discuss how the unique temporal and spatial capability of space-based EO products such as SDWLs prove to be a solution to support the present and future needs of the coastal stakeholders (i.e. coastal management authorities, coastal communities and research institutes) in regions where in situ data are lacking.
Coastline monitoring is a valuable tool used by the scientific community and coastal managers to assess the sustainability of their littoral and to assist them in their decision-making.
In this context, our approach is based on the most frequently used coastline indicators and covers all types of European coasts. These indicators have been extracted from high resolution satellite imagery, which remains today the best resource for a comprehensive assessment of last decades coastal evolutions at the European scale.
Various methods were experienced, some already existing in-house, others newly developed for the purpose of extracting nearly 4,000 km of coastline, on several types of beaches having micro to macrotidal regimes at mid to high latitudes. These beaches are distributed across France, Greece, Portugal, Germany, Norway, and Romania. Also, to cover the entire Danube Delta, Ukraine was included in this scope. Using optical and SAR satellite images, we monitored a variety of coastline indicators, e.g., waterlines, swash excursion limits, beach widths, dune foot, and seaward vegetation edges.
Different time periods were studied to match local dynamics, from long-term trends (up to 25 years) to the assessment of impacts induced by storm events. Considering available image frequency, we adapted the temporal resolution to the environmental dynamics of each study area, with monthly, seasonal, or annual rhythms. Coastal changes were estimated and mapped using different statistical methods, including cross-shore migration and surface changes. Our large-scale demonstration provides an overview of the hotspots of vulnerable coasts in Europe, with various diagnosis of recent acceleration or extensive coastal erosion.
In this presentation, we highlight our main results and some typical examples of highly erosive coasts. Between the examples we considers : (i) the case Danube Delta (Romania) which has retreated locally up to 330 m over the last 30 years, (ii) the Cotentin Coast (France) where the Eleanor storm induced an erosion of several tens of meters of the dune foot in a few days, (iii) the coastline of Evia Island (Greece) which suffered dramatic damage after large fires and heavy flooding during the last summer (2021), (iv) and the erosion hotspots on the highly touristic coasts of Portugal and the German island of Sylt, which has experienced coastal erosion of several tens of meters over the last ten years.
Following the successful 2-year development and production of coastal products derived from Sentinel 2 and Landsat 5/8 missions that have been delivered to IHCantabria for validation, there has now been time to thoroughly assess the results, their accuracies, their utility and equally important identify those areas that are ready for further development and improvement.
This presentation will identify the challenges and process adopted, will focus on the validation of the data set and will identify the next steps that have been enabled due to an additional year of research awarded under the ESA Contract Change Notice arrangement.
Areas of interest along the Mediterranean Spanish coast were specifically selected as test sites to demonstrate the value that Earth observation brings based on an extensive local knowledge and in situ data collection. Coastal Erosion is a global phenomenon and has been around since the early seas formed, however the effect of climate change bringing increased sea levels, more frequent and more severe storm surges has certainly exacerbated the effects.
Within the Mediterranean coastal regime human intervention is almost everywhere. Populations, industry, infrastructure to support them are all heavily invested in and there is a considerable risk attached. Tourism within Spain accounts for 11.8% of GDP with 13.5% of the national workforce being employed within this sector, so beach management is exceptionally valuable. Within Spain there are over 2000km of public beaches, many of them require constant and costly maintenance and renourishment. Over the past 10 years approximately €130m has been spent in over 500 sediment management projects (renourishment, redistribution and by-pass) and in over 200 beach rehabilitation projects.
Recent developments in satellite processing tools under this ESA funded Coastal Erosion from Space project have enabled cost effective, fast and automatic processing of a large amount of information from Earth observation, enhancing the capability of detecting coastal changes from space at different temporal scales.
IHCantabria, who is the Spanish partner in the consortium, provided the user requirement and local knowledge and have evaluated the capability to monitor changes in coastal morphology using satellite Earth observations, namely, 1D (coastlines) and 3D (satellite-derived bathymetry) data products using the Argans Ltd developed processors. The accuracy and ability of satellite derived products were assessed at several strategic sites in Spain (based on local knowledge, socio-economic value, and varying geomorphologies). The results indicated high horizontal accuracy, with errors on the order of half of the pixel size. Time-series analysis using satellite-derived shorelines showed that coastal change processes can be detected at several temporal and spatial scales, such as short-term erosion and accretion events on a small beach, seasonal beach rotation, and long-term trends at local and regional scales.
The high horizontal accuracy and skill of satellite derived data can be attributed to site-specific information that were included in the Argans Ltd processors in several phases of the development of the satellite products: i) VHR images from each pilot site were used in the co-registration process to guarantee high geolocation accuracy in images from different missions, ii) different spectral indices were tested at each pilot site to guarantee reliable detection of the waterlines at all sites and iii) auxiliary data and measured topo-bathymetry data were used to obtain datum-based satellite shorelines and bathymetry.
However, the results from satellite-derived bathymetry indicated that the quantitative assessment of the coastal morphology with 3D products is still limited. Some in situ measurements are necessary to obtain satellite data that represent site-specific conditions. Although the quantity of required data measured in situ is significantly lower than the quantity required by traditional monitoring methods.
So, what next?
A clear advantage of using Earth observed evidence is that this technology enables a frequent re-visit and therefore can detect changes based on actual events, whether man-made or storm induced. A requirement that has been placed upon the development and production team has been to increase the number of observations available by optimising the selection of cloud free coastal strips and improving the continuity of the lines derived. Although greater resolution is always requested this cannot be delivered, however an inter pixel best fit assessment can be achieved via a marching squares approach. This approach coupled to the adaptive threshold and variable index selection has also improved the reliability of the product set.
With these improvements there is also a requirement to further migrate the product set up the value chain and so a vulnerability approach is being designed that will harness the change rates observed over time and join this data with coastal fragility based on geomorphology and coastal make up and storm surge input.
At Living Planet Symposium 2022 the approach and results will be demonstrated.
Coastal erosion strongly depends, among others, on the distribution of different substrates, on morphology, on marine conditions, and mostly on cross-shore and longshore sediment imbalance. When coastal erosion occurs, changes in the coastline position is the most obvious and measurable phenomenon. However, the result of the coastline retreat is often the manifestation of a larger sediment imbalance, which generally begins in the submerged part of the beach system as it usually constitutes a larger sediment catching area than onshore. Knowing the morphological dynamics of the foreshore area and quantifying the erosion that occurs are key elements in sizing the weakness of the littoral system, its instability, and the ways to adapt or protect.
Within the ESA Coastal Erosion Project, the Space for Shore consortium has developed several products that focus on submerged and intertidal areas. As a first step, structures that are temporarily or permanently covered by water are identified. Then, their temporal and spatial changes are monitored, and finally, these changes are related to the various processes that happen on shore. This workflow provides tools to assess changes and has potential to predict risk areas of increased coastal erosion.
In this contribution, we will present three sets of submerged and intertidal coastal indicators derived for various time scales and in several coastal regions of Germany, Romania, and France, using hundreds of optical, infrared, and SAR satellite images. These indicators highlight the migration of submerged sandbars in foreshore areas, the spatio-temporal changes in bathymetry, and the dynamics of intertidal flats. We will demonstrate how our automated processing schemes lead to an assessment of fore-shore changes, and how they can help coastal managers in theirs decision-making, without the common delays for similar surveys during field campaigns.
Within the consortium, we are also investing a lot of effort in providing this type of information, which is not always familiar to local authorities and decision makers. Correlations were noticed between the dynamics of these submerged features and the shoreline changes over time. However, the interpretation of the achieved results often remains complex, hence our goal in this project is to make them accessible and understandable to users, and to provide them with additional information and tools to further analyze their coastal systems.
The aim of this study is to identify the full potential of satellite remote sensing to capture sub-annual to multi-decadal shoreline variability by investigating the differing temporal scales of potential shoreline change (i.e., single storm events, beach rotation, cyclic-seasonal variability, long-term erosion/accretion trends, and engineering interventions) that can be resolved by shorelines extracted from satellite imagery. In this study, the capability to extract shoreline change data from Multi Spectral Imagery (MSI) and Synthetic Aperture Radar (SAR) publicly available satellite imagery at timescales from individual storm events, to months, seasons and several years and decades is assessed at three study sites along the England coastline; macro-tidal estuarine, macro-tidal soft cliff environment and meso-tidal gravel barrier beach. To better understand the uncertainty of satellite derived shoreline changes from MSI, we have assessed the accuracy of one near operational algorithm [1] for all three coastal environments. To better understand the uncertainty of satellite derived shoreline changes from SAR imagery, we have compared the shoreline changes derived from waterlines extracted using the proposed algorithm by [2] and differentiating between ascending and descending orbits. To ensure that the analysis is transparent and can be reproduced by others, we have used the new Open Digital Shoreline Analysis System (ODSAS) which enables the user to explicitly include the resolution of the study when using the transect and baseline approach [3] to measure coastline changes. We argue that by explicitly requesting the user to define a minimum resolution is important to reduce the subjectivity of the transect and baseline method. Figure 1 illustrates the different erosion rate EPR and LRR values (EPR: End Point Rate & LRR: Linear Regression Rate) obtained for the meso-tidal gravel barrier beach.
Arctic coastal areas can experience higher erosion rates than temperate regions due to the combined influence of seasonal permafrost melt and extreme temperatures. In addition to these ordinary dynamics, high latitude coastal areas are even more affected by climate-induced changes such as increased weather hazards, rising temperatures or changes in river discharges and sediment supply.
The Svalbard region consists of an archipelago of Arctic islands and a rocky and sandy coastline chiseled by numerous fjords connected to glaciers and a complex hydrographic network. Highly sensitive and exposed to the impacts of climate change, this coastal area is a perfect witness to the environmental changes of our century.
The Svalbard Archipelago has recently become a key hotspot with an increasing number of studies, mainly focusing on glacier melt, temperature change or soil destabilization. The environmental, geographical, and geomorphological conditions of Svalbard make it extremely difficult to monitor coastal change on a large-scale. However, several studies, including Lim et al. (2020), Jaskolski et al. (2018), and Sisneros-Kidd et al. (2019), have highlighted the strong pressure of climate change, population, and human activities on the Svalbard coastal area.
This littoral is fully in line with our approach to apprehend the past, present, and expected consequences of climate change on the environment and populations.
In close collaboration with local researcher Maria Jensen from the University of Svalbard (UNIS) and other experts in France (Agnès Baltzer and Franck Garestier), we have explored the potential of the Copernicus satellite images to produce key information on the past and recent dynamics of nearly 300 km of coastline in the Svalbard archipelago. This is a major challenge, given the complexity of the environment and the meteorological and climatic conditions of the region, which limit the volume of usable spatial data (cloud cover and seasonal ice on the monitored sediments).
After a first phase of adaption of our in-house algorithms to this new typology of coastal areas, we extracted the coastline over 25 years, computed associated coastline changes as well as the evolution of the banks and the extent of the hydrographic network along several major fjords in Svalbard. Particular attention was given to Advenfjorden to improve our effort, due to the greater availability of in situ data to validate the satellite products. We experienced satellite-derived bathymetry into the fjord. This information was one of the most complex challenge in terms of methodology and algorithm development regarding the environmental context, but it is also a crucial insight to consider the full climate change impact on coastal sediment dynamics. Finally, we extracted another coastal indicator to focus on changes in the deltaic areas, namely the pioneer vegetation coverage, which reveals the impact of warming on these highly dynamic regions.
For the first time, we present our new results issued from these investigations.