Periodic Reporting for period 2 - 4S (Satellite Seafloor Survey Suite)
Reporting period: 2022-05-01 to 2023-10-31
Knowledge on shallow water seabed habitats and morphology is key for marine and coastal management, such as EC reporting on maritime directive and coastal and marine engineering and planning activities which are part of the blue economy. Digital information on seabed characteristics is required for amongst others shallow-water dredging, sand reclamation and navigation. Furthermore, the shallow water euphotic zone is of particular importance, because it is home to critical habitats which impact biodiversity, carbon storage and coastal protection.
Thus, a strong demand exists to map and monitor shallow water zones on small and large scale. Currently this data demand is filled with labour and equipment intensive survey and sampling methods, such as boat or airborne surveys, diving transects or drop-down cameras, and – increasingly, but still on a low technological level – manual satellite or airborne interpretation. None of these methods have the capability to allow a continuous and objective monitoring within standard budgets and time constrains.
We believe, that modern satellite capabilities and aquatic Earth Observation (EO) analysis can be brought to a level which significantly improve the current state of the art of data gathering of shallow water morphology, habitats and its trends over time. The Copernicus Sentinel-2 data, recent breakthroughs in the physics-based modelling of aquatic EO and machine learning procedures together with new EO archiving and processing centre form the solution of the propose concept. EOMAP and team have already shown the potential by mapping parts of the German Baltic seafloor habitats and the Great Barrier Reef. However, work has to be done to improve algorithms to increase overall accuracy, reduce manual interpretation and interaction, generate operational software solution and integrate the big data from current EO missions, including the Copernicus fleet. Especially the Sentinel-2s with its high revisit time and high spatial resolution will be key to map seafloor habitats, and more than that, have the potential to generate new information, such as the seasonal growth of seagrass, which will allow to better map and understand this environment and respond to EC reporting and shallow water data demand more effective and precise.
The 4S solution provides a cloud-based software solution which empowers the users to map and monitor seafloor habitats, morphology and depth from the comfort of his office. The 4S software is accessed by an intuitive user interface (webapp) and through programming interfaces, which allow to smoothly integrate data and analytical process into own workflows, such as demonstrated by QPS, which developed a 4S extension for their widely known hydrographic software suite.
Objective 1 (O1): Easy access and generation of satellite-derived data and tools to map and monitor seabed with Copernicus data being a core dataset
Objective 2 (O2): A solution which allows an easy integration into clients’ workflows.
Objective 3 (O3): Global evaluation and multi-purpose use
Objective 4 (O4): Substantial increase in the market enabled by integrating Copernicus data in the coastal and marine sector with state-of-the-art innovative technologies
Objective 5 (O5): Increase impact and dissemination through dedicated capacity building
Thus, a strong demand exists to map and monitor shallow water zones on small and large scale. Currently this data demand is filled with labour and equipment intensive survey and sampling methods, such as boat or airborne surveys, diving transects or drop-down cameras, and – increasingly, but still on a low technological level – manual satellite or airborne interpretation. None of these methods have the capability to allow a continuous and objective monitoring within standard budgets and time constrains.
We believe, that modern satellite capabilities and aquatic Earth Observation (EO) analysis can be brought to a level which significantly improve the current state of the art of data gathering of shallow water morphology, habitats and its trends over time. The Copernicus Sentinel-2 data, recent breakthroughs in the physics-based modelling of aquatic EO and machine learning procedures together with new EO archiving and processing centre form the solution of the propose concept. EOMAP and team have already shown the potential by mapping parts of the German Baltic seafloor habitats and the Great Barrier Reef. However, work has to be done to improve algorithms to increase overall accuracy, reduce manual interpretation and interaction, generate operational software solution and integrate the big data from current EO missions, including the Copernicus fleet. Especially the Sentinel-2s with its high revisit time and high spatial resolution will be key to map seafloor habitats, and more than that, have the potential to generate new information, such as the seasonal growth of seagrass, which will allow to better map and understand this environment and respond to EC reporting and shallow water data demand more effective and precise.
The 4S solution provides a cloud-based software solution which empowers the users to map and monitor seafloor habitats, morphology and depth from the comfort of his office. The 4S software is accessed by an intuitive user interface (webapp) and through programming interfaces, which allow to smoothly integrate data and analytical process into own workflows, such as demonstrated by QPS, which developed a 4S extension for their widely known hydrographic software suite.
Objective 1 (O1): Easy access and generation of satellite-derived data and tools to map and monitor seabed with Copernicus data being a core dataset
Objective 2 (O2): A solution which allows an easy integration into clients’ workflows.
Objective 3 (O3): Global evaluation and multi-purpose use
Objective 4 (O4): Substantial increase in the market enabled by integrating Copernicus data in the coastal and marine sector with state-of-the-art innovative technologies
Objective 5 (O5): Increase impact and dissemination through dedicated capacity building
The project progressed well, especially during the second part of the project, only some use cases were delayed. During the second part of the project, we focused on further market analyses and optimization of the business plan. This included the business model canvas as well as a thorough competitor analysis.
The 4S project has been able to achieve major technological developments which include automatic routines to generate bathymetric point data from the active green Lidar on board ICESat-2 satellite and the webapp to generate dense shallow water bathymetric grids using physics-based SDB (SDB-Online). This webapp comes with a cloud backend where API interfaces allow to integrate the solution into the application of the 4S project partners and third parties. In June 2023, the web-app was re-launched to incorporate multiple new features such a calibration/validation option with own field data. Further activities were related to the use cases where drone based multispectral imagery were recorded which feed into the innovation and research activities.
The results of a thorough validation analysis have been published in the peer-reviewed International Hydrographic Review. Further analysis for the 4S use cases have been conducted by each use case partner with own data and SDB-Online results. The overall perception is well, with issues arising in very turbid areas where a manual scene selection is crucial for a successful derivation of bathymetry.
Several workshops and presentations, including those at EOMAP-hosted SDB Days 2021 and 2022 were conducted. Notably, the SDB Best Practice Working Group (SDB-PT) was established with EOMAPs guidance and became part of the International Hydrographic Office’s (IHO) permanent Hydrographic Survey Working Group (HSWG). Marketing efforts included the creation of tutorial videos and testimonials, strengthening LinkedIn presence, and winning the "Bavarian Innovation Prize" for SDB-Online, while publications, such as the validation paper and the anticipated IHO Publication B-13, contributed to the project's impact. Initial sales of SDB-Online affirm its success, indicating a significant and promising trajectory for the project members and 4S.
The 4S project has been able to achieve major technological developments which include automatic routines to generate bathymetric point data from the active green Lidar on board ICESat-2 satellite and the webapp to generate dense shallow water bathymetric grids using physics-based SDB (SDB-Online). This webapp comes with a cloud backend where API interfaces allow to integrate the solution into the application of the 4S project partners and third parties. In June 2023, the web-app was re-launched to incorporate multiple new features such a calibration/validation option with own field data. Further activities were related to the use cases where drone based multispectral imagery were recorded which feed into the innovation and research activities.
The results of a thorough validation analysis have been published in the peer-reviewed International Hydrographic Review. Further analysis for the 4S use cases have been conducted by each use case partner with own data and SDB-Online results. The overall perception is well, with issues arising in very turbid areas where a manual scene selection is crucial for a successful derivation of bathymetry.
Several workshops and presentations, including those at EOMAP-hosted SDB Days 2021 and 2022 were conducted. Notably, the SDB Best Practice Working Group (SDB-PT) was established with EOMAPs guidance and became part of the International Hydrographic Office’s (IHO) permanent Hydrographic Survey Working Group (HSWG). Marketing efforts included the creation of tutorial videos and testimonials, strengthening LinkedIn presence, and winning the "Bavarian Innovation Prize" for SDB-Online, while publications, such as the validation paper and the anticipated IHO Publication B-13, contributed to the project's impact. Initial sales of SDB-Online affirm its success, indicating a significant and promising trajectory for the project members and 4S.
4S aims to bring a more automatic and standardized mapping concept to map seafloor habitats. We will do this by introducing innovations in different processing steps, which are
(1) the image selection process, a crucial step right at the beginning of the mapping. Satellite data needs to be cloud, sunglint and turbid (!) free in order to map seabed characteristics. We developed an AI processor which is trained by EOMAP’s seafloor analysts to allow an automatic prediction/classification of the feasibility for satellite image data.
(2) the satellite image processing. Our progress beyond the state of the art is to run physics-based correction modules in order to minimize numerous effects on the signal through the water and atmosphere. The final data represents the seafloor albedo and is thus, a standardized reflectance product which can be used for automatic (and trained) classification processors.
(3) introducing machine learning procedures, trained by harmonized aquatic EO information on seafloor reflectance, seafloor texture, shape and known/surveyed seafloor information accessed in the use cases. With this harmonized advanced aquatic EO raster data we believe, that we can train a procedure which allows to map and monitor seabed in most of the shallow waters worldwide.
(4) Bringing ICESat-2 Atlas data into practice and integration with Copernicus data. We will make use of ICESat-2 Atlas Lidar bathymetry and integrate those with multispectral data derived from Copernicus data. This concept will provide reduced product uncertainties.
(5) Sensor fusion of very high-resolution drone imagery with satellite data. The aim is to combine the very high spatial resolution of drones, typically few centimetres recorded by RGB cameras, with the high quality and multispectral band information of satellite data.
(1) the image selection process, a crucial step right at the beginning of the mapping. Satellite data needs to be cloud, sunglint and turbid (!) free in order to map seabed characteristics. We developed an AI processor which is trained by EOMAP’s seafloor analysts to allow an automatic prediction/classification of the feasibility for satellite image data.
(2) the satellite image processing. Our progress beyond the state of the art is to run physics-based correction modules in order to minimize numerous effects on the signal through the water and atmosphere. The final data represents the seafloor albedo and is thus, a standardized reflectance product which can be used for automatic (and trained) classification processors.
(3) introducing machine learning procedures, trained by harmonized aquatic EO information on seafloor reflectance, seafloor texture, shape and known/surveyed seafloor information accessed in the use cases. With this harmonized advanced aquatic EO raster data we believe, that we can train a procedure which allows to map and monitor seabed in most of the shallow waters worldwide.
(4) Bringing ICESat-2 Atlas data into practice and integration with Copernicus data. We will make use of ICESat-2 Atlas Lidar bathymetry and integrate those with multispectral data derived from Copernicus data. This concept will provide reduced product uncertainties.
(5) Sensor fusion of very high-resolution drone imagery with satellite data. The aim is to combine the very high spatial resolution of drones, typically few centimetres recorded by RGB cameras, with the high quality and multispectral band information of satellite data.