Engineering with Nature: combining Artificial intelligence, Remote sensing and computer Models for the optimum design of coastal protection schemes
Lead Research Organisation:
University of Liverpool
Department Name: Geography and Planning
Abstract
Currently, 41% of power stations, 17.9% of railway track, 14.3% of railway stations, 33% of wastewater treatment and half a million of properties are at risk of coastal flooding. The average damage to properties is over £260million each year. Hard engineering solutions are becoming economically unviable due to the high costs of construction, maintenance and adaptation to changes in sea level and storms.
For this reason, there is a growing interest in engineering with nature (including the creation of salt marshes, seagrass beds, beach nourishment and mega-nourishment) which offers a more economically viable alternative and also support net Zero-Carbon emissions and local amenities value as highlighted into the 25 years Government plan to improve the environment, FCERM strategies for England, Scotland and Wales. However, despite the growing recognition about the necessity to move towards this greener alternative for coastal protection, there is little to no guidance on the implementation on engineering with nature. There are no quantitative and process-based decision-making tools and guidelines to aid engineers, planners, and governments to select coastal management strategies fit for their unique local environment. There are still many uncertainties in relation to conditions maximizing the establishment and longevity of engineering with nature and uncertainties in relation to their effectiveness.
This fellowship will develop novel understanding necessary to protect coastal infrastructures and coastal communities through widespread adoption of engineering with nature. The fellowship will use a novel combination of remote sensing, artificial intelligence and computer models to provide -for the first-time- design criteria for coastal protection using engineering with nature and knowledge necessary for the choice of the most durable and efficient coastal management type and location. Results will be summarized into an interactive decision support tool which will be distributed to stakeholders and government agencies for a consistent evaluation of pros- and cons of different coastal management interventions including uncertainties in relation to their effectiveness under different sea level rise and storms scenarios.
For this reason, there is a growing interest in engineering with nature (including the creation of salt marshes, seagrass beds, beach nourishment and mega-nourishment) which offers a more economically viable alternative and also support net Zero-Carbon emissions and local amenities value as highlighted into the 25 years Government plan to improve the environment, FCERM strategies for England, Scotland and Wales. However, despite the growing recognition about the necessity to move towards this greener alternative for coastal protection, there is little to no guidance on the implementation on engineering with nature. There are no quantitative and process-based decision-making tools and guidelines to aid engineers, planners, and governments to select coastal management strategies fit for their unique local environment. There are still many uncertainties in relation to conditions maximizing the establishment and longevity of engineering with nature and uncertainties in relation to their effectiveness.
This fellowship will develop novel understanding necessary to protect coastal infrastructures and coastal communities through widespread adoption of engineering with nature. The fellowship will use a novel combination of remote sensing, artificial intelligence and computer models to provide -for the first-time- design criteria for coastal protection using engineering with nature and knowledge necessary for the choice of the most durable and efficient coastal management type and location. Results will be summarized into an interactive decision support tool which will be distributed to stakeholders and government agencies for a consistent evaluation of pros- and cons of different coastal management interventions including uncertainties in relation to their effectiveness under different sea level rise and storms scenarios.
People |
ORCID iD |
| Nicoletta Leonardi (Principal Investigator / Fellow) |
Publications
Forrester J
(2024)
Seagrass as a nature-based solution for coastal protection
in Ecological Engineering
Forrester J
(2024)
Seagrass as a Nature-Based Solution for Coastal Protection
Kumar P
(2023)
Coastal forecast through coupling of Artificial Intelligence and hydro-morphodynamical modelling
in Coastal Engineering Journal
Kumar P
(2024)
Exploring Mega-Nourishment Interventions Using Long Short-Term Memory (LSTM) Models and the Sand Engine Surface MATLAB Framework
in Geophysical Research Letters
| Description | The increasing threat of coastal flooding and erosion necessitates the development of nature-based solutions as viable alternatives to traditional hard engineering approaches. This research has advanced our understanding of the eco-geomorphic processes underpinning the effectiveness and resilience of wetlands and seagrass systems in coastal protection. Additionally, it has pioneered the development of AI-based predictive models, representing a significant advancement in coastal forecasting capabilities. Seagrass as a Coastal Protection Mechanism: This project has demonstrated that while seagrass alone may not replace conventional hard engineering solutions, its integration with salt marsh restoration and hybrid approaches can enhance coastal resilience. Delft3D modelling was employed to assess the potential of seagrass as a wave-attenuating intervention. Findings indicate that seagrass meadows significantly reduce wave energy, with maximum wave height reductions potentially exceeding 50% within seagrass patches during extreme conditions. Notably, lower-density seagrass configurations can aslo be effective in wave dissipation. Salt Marsh Resilience to Climate Change and Human Intervention: The research has highlighted the importance of considering both natural and anthropogenic factors in designing effective engineering-with-nature interventions. A key focus was the role of storm events in salt marsh sediment dynamics, revealing that storm-driven sedimentation substantially benefits marsh interiors by promoting vertical accretion. However, long-term resilience depends on sediment availability, which can be influenced by human activities such as embankment construction. AI-Driven Innovations in Coastal Geomorphology: Despite the increasing role of Artificial Intelligence (AI) in scientific research, its application in coastal geomorphology remains relatively limited. This research has pioneered the development of AI-based predictive models, providing a step-change in coastal forecasting capabilities. As part of this research, we have demonstrated the potential of AI models to predict suspended sediment transport (SSST) and morphological changes in coastal environments based on hydrodynamic forcing. For instance, extensive numerical simulations in Delft3D were conducted, analysing over five hundred scenarios of sand engine interventions to train ensemble Artificial Neural Network (ANN) models. These models accurately predict hydrodynamic and sediment transport responses under varying forcing conditions. Additionally, we developed state-of-the-art Long Short-Term Memory (LSTM) models for time-series predictions related to the morphological evolution of mega-nourishment interventions. To our knowledge, this represents the first application of such models for this coastal engineering topic. AI-Powered Decision Support Tools: Two innovative software applications-the Sand Engine App and the Sand Engine Surface App-have been developed, integrating ANN-based predictive modelling into intuitive, user-friendly interfaces. These open-source tools are freely available, complete with clear instructions and usage videos, showcasing the potential of AI models to deliver fast, comprehensive results with minimal computational effort. These research outcomes address important national and global priorities in coastal resilience, aligning with the UK Government's 25-Year Environment Plan and Flood and Coastal Erosion Risk Management (FCERM) strategies. The integration of eco-geomorphic insights with AI-driven predictive capabilities represents an viable approach to engineering with nature, providing sustainable and scalable solutions for coastal challenges. |
| Exploitation Route | The outcomes of this funding have been deliberately structured to ensure accessibility and usability by external stakeholders. A specific effort has been made to make publication resources widely available and easy to use. Specifically, two innovative software applications have been developed-the Sand Engine App and the Sand Engine Surface App-which incorporate ANN-based predictive modelling into user-friendly interfaces. These open-source tools are freely available, complete with clear instructions and usage videos, showcasing the potential of AI models to deliver fast, comprehensive results with minimal computational effort. These tools might serve as a first-level approximation for engineers, planners, and policymakers, enabling them to dynamically assess the impact of sand engine designs on coastal morphology. Beyond academic dissemination, steps have been taken to promote accessibility of the outcomes, and easy-to-read leaflets and updates have been shared with the Environment Agency and their Coastal Research newsletter. |
| Sectors | Environment Other |
| Title | SandEngine App |
| Description | SandEngineApp, a MATLAB based Application, part of the manuscript "A novel framework for the evaluation of coastal protection schemes through Integration of numerical modelling and Artificial Intelligence into the Sand Engine App". The App allows users to utilize an ensamble of Artificial Neural Networks models to evaluate the efficacy of different Sand Engines configurations. Specifically, more than five hundred numerical simulations with different sand-engine designs and different locations along Morecambe Bay were conducted with the hydromorphodynamic model Delft3D. Twelve Artificial Neural Networking ensemble models structures were trained on the simulated data to predict the influence of different sand engines on water depth, wave height and sediment transports with good performance. The ensemble models were then packed into the Sand Engine App developed in MATLAB and designed to calculate the impact of different sand engine features on the above variables based on users' inputs of sand engine designs. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | The article emphasizes synergies and complementarity between hydro-morphodynamic modelling and artificial intelligence techniques. The computational efficiency of ANN allows optimizing the evaluation of coastal protection schemes. |
| URL | https://github.com/pavitra979/SandEngine |
| Title | Sand Engine App |
| Description | SandEngineApp is an innovative MATLAB-based tool for evaluating coastal protection schemes through the integration of numerical modeling and Artificial Intelligence. Published in Scientific Reports, it leverages over 500 simulations and AI models to predict the impact of sand engine features on water depth, wave height, and sediment transport. The app is designed to assist users in optimizing sand engine designs for coastal protection. Detailed technical specifications and the creation process are provided in the associated publication, with an easy-to-follow installation guide available in the video. |
| Type Of Technology | Software |
| Year Produced | 2024 |
| Impact | As of 11/03/25, the article describing the software has received over 2,500 accesses. It is novel as one of the few incorporating Artificial Neural Networks (ANN) into a downloadable MATLAB framework, and to our knowledge, it is the only MATLAB framework and ANN-based software developed specifically for sand engine studies. |
| URL | https://github.com/pavitra979/SandEngine/tree/main |
| Title | Sand_Engine_Surface |
| Description | This repository contains a MATLAB based application SandEngineSurface, which is a part of a research work titled as "Exploring Mega-Nourishment Interventions Using Long Short-Term Memory (LSTM) Models and the Sand Engine Surface MATLAB Framework". This research work is published in Geophysical Research Letter Journal. The technical specifications, usage and component description is present in "Description" file. Installation and usage are demonstrated in "Installation" and "Usage" video files. |
| Type Of Technology | Software |
| Year Produced | 2024 |
| Impact | This MATLAB-based SandEngineSurface repository provides a powerful tool for predicting sand engine evolution using LSTM models. Published in Geophysical Research Letters, it combines AI with hydro-morphodynamic modeling to optimize coastal protection solutions. The repository includes clear installation and usage instructions, making it accessible for researchers. It helps decision-makers design effective sand engine configurations, enhancing climate change adaptation. |
| URL | https://zenodo.org/doi/10.5281/zenodo.10654361 |
| Title | Sand_Engine_Surface |
| Description | This repository contains a MATLAB based application SandEngineSurface, which is a part of a research work titled as "Exploring Mega-Nourishment Interventions Using Long Short-Term Memory (LSTM) Models and the Sand Engine Surface MATLAB Framework". This research work is published in Geophysical Research Letter Journal. The technical specifications, usage and component description is present in "Description" file. Installation and usage are demonstrated in "Installation" and "Usage" video files. |
| Type Of Technology | Software |
| Year Produced | 2024 |
| Impact | This MATLAB-based SandEngineSurface repository provides a powerful tool for predicting sand engine evolution using LSTM models. Published in Geophysical Research Letters, it combines AI with hydro-morphodynamic modeling to optimize coastal protection solutions. The repository includes clear installation and usage instructions, making it accessible for researchers. It helps decision-makers design effective sand engine configurations, enhancing climate change adaptation. |
| URL | https://zenodo.org/doi/10.5281/zenodo.10654362 |
| Description | Engagement and presentation of the preliminary results to the project partner, external industry partners and agencies. |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Third sector organisations |
| Results and Impact | Presentation of the project goals and preliminary findings to the project partner, external industry partners, and agencies. Leaflets with summaries of the main project goals and outcomes were distributed, followed by online meetings. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Meeting with Local industry |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Industry/Business |
| Results and Impact | Discussions with Arup about the research outcomes and possible collaborations within the context of Arup Ventures. We discussed a lot of what Arup Venture does and is looking for and the PI shared her research on the use of neural networks and AI to solve the data challenge and train models on costal erosion and use of nature based solutions to mitigate it. |
| Year(s) Of Engagement Activity | 2025 |
| Description | Open workshop |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Public/other audiences |
| Results and Impact | The research was discussed during an engagement event organized around the theme of Nature Based Solutions and sustainability. This workshop was attended by members of the general public and included discussions on the theme of coastal resilience and environmental factors that contribute to the sustainability of communities. Thanks to networking, the event served as a springboard for further discussions and connections with local authorities interested in coastal change. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Workshop on costal system. |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Public/other audiences |
| Results and Impact | The workshop, focused on the theme of coastal resilience, provided a platform to showcase the outcomes and findings from coastal research associated with this award. It also highlighted collaborative efforts with colleagues working on coastal-related themes within other UK Research and Innovation (UKRI) funded projects. The event successfully facilitated interdisciplinary exchange and contributed to a deeper, shared understanding of coastal resilience, drawing on a diverse range of expertise and insights from leading researchers in the field. |
| Year(s) Of Engagement Activity | 2024 |