Simona Niculescu   Editor European Spatial Data for Coastal and Marine Remote Sensing Proceedings of International Conference EUCOMARE 2022-Saint Malo, France European Spatial Data for Coastal and Marine Remote Sensing Simona Niculescu Editor European Spatial Data for Coastal and Marine Remote Sensing Proceedings of International Conference EUCOMARE 2022-Saint Malo, France Editor Simona Niculescu Laboratory LETG-Brest, UMR 6554 CNRS University of Western Brittany Plouzané, France The European Commission’s support for the production of this publication does not constitute an endorsement of the contents, which reflect the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein. ISBN 978-3-031-16212-1 ISBN 978-3-031-16213-8 (eBook) https://doi.org/10.1007/978-3-031-16213-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Editorial for Special Issue: European Spatial Data for Coastal and Marine Remote Sensing Within the framework of the European Copernicus program, satellite data are effort- lessly accessible to the scientific community, companies, and public. At the same time, the massive availability of digital data in various forms (spatialized or not) and across different fields opens new avenues of research and applications. Recent advances in the radiometric, spatial, temporal, and spectral resolution of sensors, as well as new data processing approaches in remote sensing, provide an extremely valuable set of tools for monitoring and mapping the various characteristics of marine, coastal, and aquatic systems (macroalgae, seagrasses, temperate reefs, coral reefs, salt marshes, mangroves, peatlands, aquatic vegetation, river systems, estuar- ies, deltas, etc.). It is crucial to increase coastal monitoring efforts in light of the growing number of studies highlighting the impacts of climate change, and in par- ticular the sea-level rise that will radically redefine the physical state of the near- shore zone in the twenty-first century [2, 4]. In addition to climatic factors, there are also constraints on marine services imposed by the growth of coastal human popula- tions, whose number is expected to nearly double by mid-century [1]. In conjunc- tion with population growth, the exploitation of coastal resources has produced increasingly dramatic changes to coastal and inland aquatic habitats over the past 100 years [3]. In this special issue, 11 long papers, 6 short papers of original research and 3 abstracts of keyspeakers were published on topics covering a wide range of remote sensing applications, including marine, coastal, and aquatic ecosystems’ monitor- ing and biodiversity management, coastal wetlands, nearshore land cover changes, deep learning and artificial intelligence innovations, marine, coastal and aquatic geomorphology, and coastal risk management. This special issue also provides new methodologies for optimizing the combined use of satellite/airborne data and field measurements for a comprehensive approach. Using satellite images of different spatial and temporal scales and applying innovative methods, several different types of coastal environments (bays, estuaries, sandy, and muddy systems) and several key parameters (vegetation, wind, etc.) and morphologies (shorelines, mud banks, wetlands) were analyzed and studied. v vi Editorial for Special Issue: European Spatial Data for Coastal and Marine Remote Sensing Harid et al. identify coccolithophores blooms of Holococcolithophora sphaeroi- dea in Algiers Bay from in situ and satellite observations (Modis data) over a period of 16 years (2003–2018). For the processing and analysis of the satellite images, a remote sensing reflectance index for coccolithophores (Cocco-Index) is used. The Cocco-Index proposed in this paper is based on a remote sensing reflectance (RSR) threshold between the green (531 nm, 547 nm, and 555 nm) and blue (443 nm, 469 nm, and 488 nm) bands of the MODIS sensor. This topic would benefit from further studies on the interactions between biological, physical, chemical, and envi- ronmental factors in the coccolithophore bloom areas. Collin et al. investigate continuous mapping of salt marshes based on the normal- ized difference vegetation index (NDVI) ranges at three spatial and two temporal scales using Sentinel-2 (10 m), Dove (3 m), and unmanned aerial vehicle (UAV) (0.03 m) images. The NDVI ranges of the Sentinel-2 and Dove images were tracked with a five- and four-year lag, respectively. The NIR band of the UAV imagery was, therefore, modeled (R2NIR = 0.98) using a three-layer neutral network (NN) pre- diction based on red, green, and blue reflectance imagery, which in turn was cali- brated/validated/tested by the bands of the Dove imagery. The 100-fold increase in pixel size allowed for the detection of decimeter-scale salt marsh and tidal flat objects. Mury et al. investigate the contribution of near-infrared (NIR) and mid-infrared (MIR) bands from multispectral UAV and WorldView-3 (WV-3) super-spectral (SS) satellite imagery for mapping coastal ecomorphological features using two super- vised classification algorithms, Maximum likelihood (ML) and support vector machine (SVM). They propose various combinations of spectral bands, visible + NIR and visible + MIR, evaluated by overall accuracy (OA) scores, for the classifi- cation of ecomorphological features. Their results show the significant contribution of the NIR and MIR bands for mapping coastal ecogeomorphological features. They also show that adding the NIR bands to the RGB band combination signifi- cantly increases the OA scores of the classifications (by +4.99% and +6.54%, with the ML and SVM algorithms, respectively). Adding the MIR bands to the combina- tion of these bands provides classifications with even higher OA (up to 99.1% and 98.4%), demonstrating the suitability of the MIR bands for mapping coastal ecogeo- morphological features. Gadal et al. perform an analysis of coastal changes based on the use of Landsat remote sensing data sets from the 1980s to the 2020s, combined with EU geo- graphic databases and land use plans in the coastal areas of Kaliningrad Oblast, Lithuania, and Latvia. Territorial dynamics are modeled using the GEOBIA pack- age with object-oriented classification and machine learning algorithms (maximum likelihood, minimum distance to means, and parallelepiped classifiers) applied to multispectral images from the Landsat 5™ and Landsat 8 OLI satellites. The results were compared with the CCI Land Cover databases from 1995 and 2015. Zaabar et al. examine LULC changes in the Allala watershed, including the city of Ténès (Algeria), using satellite images from 1999 (Landsat image), 2009 (Landsat image), and 2020 (Sentinel-2 image). In this study, the image processing and analy- sis is based on object-oriented classification (OBIA) using random forest (RF) and Editorial for Special Issue: European Spatial Data for Coastal and Marine Remote Sensing vii support vector machine (SVM) machine learning algorithms. Subsequently, the LULC changes were evaluated using GIS (analysis of LULC transition matrix). Their results show that RF outperforms SVM on all three input data periods, with an overall accuracy of 95.6% obtained with the Sentinel-2 image. Modoosoodun Nicolas et al. investigate bathymetry predictions using Pleiades multispectral satellite data from around Europa Island, which is a coral island in the Mozambique Channel. The authors explore the possibility of using very high- resolution multispectral satellite data with a neural network architecture inspired by U-Net deep learning to infer bathymetry estimates. A 1 m resolution airborne LiDAR data set is used as ground truth to train the model. The predicted bathymetry values could potentially be incorporated into electronic navigational charts. Accurate results are obtained for estimating bathymetry in shallow waters, with the average absolute error not exceeding 1.5 m in this case. The authors show that this U-Net inspired deep learning technique is capable of predicting water depth from very high-resolution satellite data. Boussetta et al. perform an assessment of the vulnerability distribution of the NE-SE coast of Jerba (Tunisia) and highlight the areas that are likely to be most affected by future sea-level rise by analyzing geophysical and socioeconomic vari- ables processed by remote sensing approaches. They integrate remote sensing and GIS methods through supervised classification using an object-oriented approach combined with the random forest machine-learning algorithm (OBIA-RF). Their results provide baseline data on coastal zone physical processes and land use. They have found that 63% of the coastline of the northeast coast of the island has a high to very high degree of vulnerability. On the other hand, 37% of the coastline of the southeast of the island has a low to medium vulnerability. Lam et al. propose an approach using random forest methodology on sentinel images from between 2019 and 2022 to assess the changes in the forest, planted to prevent saltwater intrusion and coastal erosion, outside the dike surrounding the coast of the Vietnamese Mekong Delta. The authors apply a method to overcome the shortcomings of optical data in identifying areas of forest loss and forest restoration. First, the satellite image data was converted into two-dimensional row and column matrix data. Each row is an object to be analyzed (pixel, object), and each column is the value of features (spectral value, vegetation value, backscatter value, observa- tion data, etc.). Through pre-processing, optical and radar data were integrated into one dataset per year, leading to four datasets corresponding to the four observation years of 2019 to 2022. Each dataset has nine columns corresponding to three visible color bands, one near infrared band, altitude data, two polarization (VV and VH), the vegetation index (NDVI), and the observation column (class). The random for- est method was found to be effective in identifying forested and non-forested areas with accuracies greater than 90%. Rabehi et al. propose the spatial monitoring of the dynamics of Palm Beach in Algeria using the Algerian Alsat2 satellite archive (high spatial resolution (2.5 m)). They performed an evaluation “before,” “during,” and “after” coastal protective measures, using several spectral detection algorithms of the coastline (supervised spectral angle matching index and morphological mean filter, and unsupervised viii Editorial for Special Issue: European Spatial Data for Coastal and Marine Remote Sensing using principal/independent component analysis and NDWI normalized difference water index). The quality of this shoreline evaluation is compared to a DGPS refer- ence shoreline in order to obtain the overall accuracy. Monitoring enables the quan- titative observation of the behavior of the beach and thus allows the qualification of the degrees of effectiveness of these breakwaters. However, the performed monitor- ing does not take into account exceptional erosion episodes, such as storms or sea- sonal erosion. Ngan et al. investigate the possibilities of assessing the current state of the shore- line and the effects of erosion on the shoreline in Vinh Chau City – Soc Trang Province, Vietnam. Using Landsat images from the years 2005 to 2022, they com- pare the shorelines obtained to the ones seen on Google™ Earth Pro. Google™ Earth images from the years 2006 to 2020 are then used to digitize the specific beach part of the Vinh Chau shoreline. The shoreline positions were used to deter- mine the variation in the location of the shoreline and calculate possible erosion and accretion zones and rates of the spatial extent of the sandy beach using the digital shoreline analysis system (DSAS). The results are based on wave measurements over a short period of time, so it is not possible to come to a conclusion on the wave reduction effectiveness of current measures over the long term. Continuous moni- toring during different wind seasons is needed to have a more accurate assessment of the effectiveness of wave reduction. Bengoufa et al. propose an efficient method for the detection and extraction of coastal indicators. They use a satellite image classification approach, based on a deep learning algorithm, optimized and adapted to the extraction of a hydrodynamic and a biological coastline indicator. They use a very high-resolution Pleiades image, cut into sub-regions, and analyzed by a convolution kernel of size 3*3. Following the image processing and analysis by classification, they obtained a very high global accuracy of 92%. These results were validated by comparing them to the field sur- veys (reference) acquired on the same day as the satellite image acquisition. With a run-up (horizontal wave excursion) of 0.6 m, the confidence interval for the deep learning method was estimated at ±0.42 m. In the Category of Short Articles, There Are Six Articles James et al. propose to evaluate the evolution of a polder (Beaussais’ polder located at the bottom of Beaussais’ Bay on the French Emerald Coast) in the context of coastal management by drone. Using multi-spectral spectral bands (RGB, red-edge, RE, and near infrared, NIR) at very high resolution, they examine the importance of landscape cover variations by calculating several indicators, the normalized differ- ence vegetation index (NDVI), the normalized difference water index (NDWI), and the soil adjusted vegetation index (SAVI). Boussetta et al. study the possibility of improving the monitoring of sandy coasts by satellite observation (Ten Landsat images and two Sentinel-2 images) by Editorial for Special Issue: European Spatial Data for Coastal and Marine Remote Sensing ix applying CNN (convolutional neural network) and digital shoreline assessment sys- tem (DSAS) algorithms. Caline et al. propose a coastal characterization of complex but representative morpho-sedimentary patterns of the seabed in Saint Malo Bay using very high- resolution topo-bathymetric LiDAR data, their morphometric derivatives, sonar, and sediment samples. Bulot et al. present a method for mapping using a Shannon index of benthic geo- biodiversity as a whole, using continuous morphometric predictors derived from a combination of VHR topo-bathymetric LiDAR data, discrete underwater imagery responses, and 10-neural network prediction. The study took place in Saint-Malo Bay, along the coast from Saint-Briac-sur-Mer to La Pointe du Grouin. Ternon et al. investigate the effect of local (~100 m2) habitat components such as habitat architecture, substrate composition, and benthic community, as well as the value of photogrammetry versus visual observation on the fish community. The study was conducted in Saint-Malo Bay (Brittany, France). Collin et al. propose an approach to spatially assess and augment the potential of natural coastal defenses using a combination of very high resolution terrain classi- fication and network modeling. The bathymetric terrain, derived from a state-of-the- art LiDAR survey, was enhanced by a topographic position index (TPI) based classification. The study took place in Saint-Malo’s lagoon, core of the French Emerald Coast, located on the Brittany coast of the English Channel, west of Mont- Saint- Michel Bay and south of Jersey Island. Plouzané-Brest, France Simona Niculescu References 1. Neumann B, Vafeidis AT, Zimmermann J, Nicholis RJ (2015) Future coastal population growth and exposure to sea-level rise and coastal flooding – a global assessment. PLoS ONE 10:e0118571. doi: https://doi.org/10.1371/journal.pone.0118571 2. Taherkhani M, Vitousek S, Barnard PL, Frazer N, Anderson TR, Fletcher CH (2020) Sea-level rise exponentially increases coastal flood frequency. Sci Rep 10: 6466 3. Turpie KR, Ackleson S, Bell T, Dierssen HM, Goodman J, Green R et al (2017) Spectroscopy for global observation of coastal and inland aquatic habitats. In: 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) (Fort Worth, TX). pp 3051–3054. doi: https://doi.org/10.1109/IGARSS.2017.8127642 4. Vousdoukas MI, Ranasinghe R, Mentaschi L, Plomaritis TA, Athanasiou P, Luijendijk A, Feyen L (2020) Sandy coastlines under threat of erosion. Nat Clim Chang 10:260–263 Contents Detection of Coccolithophore Bloom Episodes in Algiers Bay Using Satellite and In Situ Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Romaissa Harid, Hervé Demarcq, Shara Amanouche, Malik Ait-Kaci, Nour-El-Islam Bachari, and Fouzia Houma Multiscale Spatiotemporal NDVI Mapping of Salt Marshes Using Sentinel-2, Dove, and UAV Imagery in the Bay of Mont- Saint- Michel, France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Antoine Collin, Dorothée James, Antoine Mury, Mathilde Letard, Thomas Houet, Hélène Gloria, and Eric Feunteun Contribution of Near- and Mid-Infrared Wavebands to Mapping Fine-Scale Coastal Ecogeomorphological Features . . . . . . . . . . . . . . . . . . . 39 Antoine Mury, Antoine Collin, Dorothée James, and Mathilde Letard Monitoring Land Cover Change in the Southeastern Baltic Sea Since the 1980s by Remote Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Sébastien Gadal and Thomas Gloaguen Assessment of Land Cover Changes in the Allala Watershed Based on Object Based Image Analysis Using Landsat and Sentinel-2 Images . . . . 81 Narimane Zaabar, Simona Niculescu, and Mustapha Kamel Mihoubi Deep Learning–Based Bathymetry Mapping from Multispectral Satellite Data Around Europa Island . . . . . . . . . . . . . . . . . . 97 Khishma Modoosoodun Nicolas, Lucas Drumetz, Sébastien Lefèvre, Dirk Tiede, Touria Bajjouk, and Jean-Christophe Burnel Assessment of Coastal Vulnerability to Erosion Risk Using Geospatial and Remote Sensing Methods (Case of Jerba Island, Tunisia) . . . . . . . . . . 113 Amina Boussetta, Simona Niculescu, Soumia Bengoufa, Hajer Mejri, and Mohamed Faouzi Zagrarni xi