Smart forecasting: joined-up flood forecasting (FF) infrastructure with uncertainties
Lead Research Organisation:
University of Sheffield
Department Name: Civil and Structural Engineering
Abstract
Reliable and comprehensive flood forecasting is crucial to ensure resilient cities and sustainable socio-economic development in a future faced with an unprecedented increase in atmospheric temperature and intensified precipitation. Floodwaters from the areas surrounding a city can heavily affect flood cycle behaviour across urban areas, introducing uncertainties into the forecast that are often non-negligible. However, currently the extent to which we can predict flood hazards is limited, and existing methods cannot for example deal with inter-regional dependencies (e.g. as was seen when floods affected nine different countries across Central and Eastern Europe). Presently in the UK approx. 25% of yearly flood insurance claims are from areas outside the zones forecast to be at flood risk, and annual flood damage costs are already high (approx. £1.5 billion). Also more than 20,000 houses per year continue to be built on floodplains.
The need to transform flood forecasting for a range of applications and scales has already been recognised by various parties. The UK Climate Change Risk Assessment 2017 Evidence Report prioritises flooding as the greatest direct climate change related threat for UK cities now and in the future, and urges urgent action to be taken, including the development of new solutions over the next 5 years. The hydraulic software industry and consultancy firms have expressed a desire for more reliable and sophisticated flood forecasting approaches, which can also reduce the manual labour required. In addition, mathematics and engineering research communities are still searching for forecasting models that are joined-up, reliable and efficient, as well as versatile and adaptable.
To address this need, 'Multi-Wavelets' technology will be employed in this fellowship with a view to transforming flood forecasting routines from a disparate set of activities into a unified automatic framework. The applicant's vision is to exploit the innate capability of Multi-Wavelets technology to reformulate flood forecasting methods by providing a smart modelling foundation for the delivery of timely and accurate flood maps, alongside statistically quantified uncertainties. This research presents a unique opportunity for the applicant, UK academia and UK industry, to establish a world leading capability in a nascent field while addressing Living With Environmental Change (LWEC) priorities for improved forecasting of environmental change.
The fellowship research will stimulate the creation of new software infrastructure capable of significantly improving our flood forecasting ability across length scales and under multiple uncertainties, helping us to better design infrastructure against flood risk and to plan for the consequences.
The need to transform flood forecasting for a range of applications and scales has already been recognised by various parties. The UK Climate Change Risk Assessment 2017 Evidence Report prioritises flooding as the greatest direct climate change related threat for UK cities now and in the future, and urges urgent action to be taken, including the development of new solutions over the next 5 years. The hydraulic software industry and consultancy firms have expressed a desire for more reliable and sophisticated flood forecasting approaches, which can also reduce the manual labour required. In addition, mathematics and engineering research communities are still searching for forecasting models that are joined-up, reliable and efficient, as well as versatile and adaptable.
To address this need, 'Multi-Wavelets' technology will be employed in this fellowship with a view to transforming flood forecasting routines from a disparate set of activities into a unified automatic framework. The applicant's vision is to exploit the innate capability of Multi-Wavelets technology to reformulate flood forecasting methods by providing a smart modelling foundation for the delivery of timely and accurate flood maps, alongside statistically quantified uncertainties. This research presents a unique opportunity for the applicant, UK academia and UK industry, to establish a world leading capability in a nascent field while addressing Living With Environmental Change (LWEC) priorities for improved forecasting of environmental change.
The fellowship research will stimulate the creation of new software infrastructure capable of significantly improving our flood forecasting ability across length scales and under multiple uncertainties, helping us to better design infrastructure against flood risk and to plan for the consequences.
Planned Impact
Flooding presents a major long term threat to the economy and wellbeing of people living and working in their communities. More than 5 million properties in England are already 'at risk' of flooding, with an increase in flood risk across all UK regions predicted given the current trend of global temperature increase (CCRA Evidence Report 2017). This fellowship will support the development of new mathematical approaches designed to increase the reliability and timeliness of the provision of flood forecast information (in line with ambition R5 of the 'Resilient Nation' theme of the EPSRC Delivery Plan 2016/17-2019/20). Together with the general public, those with the responsibility of forecasting, assessing and/or managing flood risk will benefit from more comprehensive quantification of flood risk at a city-wide scale and more accurate prioritisation of the highest risk zones. These beneficiaries include the Environment Agency and those working in the water consultancy and software industries, local authorities, water companies, emergency responders and the insurance industry.
Hydraulics-based software providers and engineering consultants involved in this fellowship (BMT-WBM, CH2M, XP Solutions, Innovyze and DHI) will have immediate access to the transformative approaches developed which are designed to move flood forecasting from a disparate set of activities into a unified automatic framework. Developing and extending their products as a result of the fellowship will enable them to: (i) more effectively and efficiently handle day-to-day flood modelling activities, e.g. by a reduced reliance on user iterations to build and run a model for individual flood risk assessment projects, and (ii) to become more competitive in the global market for water modelling services.
The Environment Agency (EA) will also have immediate access to the research outputs. Their 'Flood and Coastal Risk Management Research and Development' team will monitor the fellowship's research, with a view to answering key strategic questions related to flood forecasting and risk management (Collaborative Research Priorities for the Environment Agency 2016-20). This fellowship will develop a modelling framework capable of modelling the flood flow interdependencies that exist between different regions. This can be expected to benefit the EA in mitigating against severe events such as flash floods, and policy-makers in addressing key national strategic needs, such as the need to better protect and/or restrict the development of assets located on floodplains (the UK is building faster on floodplains than anywhere else, with approx. 1500 new homes built annually in 'high risk' areas (Committee on Climate Change 2015)).
The fellowship will involve personnel co-working with mathematicians, engineers and scientists, conducting research placements, and participating in industry and public engagement activities. Hence, it will generate unique skills and capabilities that can be deployed by UK organisations and private companies to maximise the social, economic and environmental impacts of the research.
In summary, the outcomes of this research will significantly benefit the UK public living in and around flood-prone areas, who will be better informed of the likely consequences of impending climatic events, and businesses who can take steps to reduce disruption of their activities, helping to ensure affordable insurance premiums and better quality of life.
Hydraulics-based software providers and engineering consultants involved in this fellowship (BMT-WBM, CH2M, XP Solutions, Innovyze and DHI) will have immediate access to the transformative approaches developed which are designed to move flood forecasting from a disparate set of activities into a unified automatic framework. Developing and extending their products as a result of the fellowship will enable them to: (i) more effectively and efficiently handle day-to-day flood modelling activities, e.g. by a reduced reliance on user iterations to build and run a model for individual flood risk assessment projects, and (ii) to become more competitive in the global market for water modelling services.
The Environment Agency (EA) will also have immediate access to the research outputs. Their 'Flood and Coastal Risk Management Research and Development' team will monitor the fellowship's research, with a view to answering key strategic questions related to flood forecasting and risk management (Collaborative Research Priorities for the Environment Agency 2016-20). This fellowship will develop a modelling framework capable of modelling the flood flow interdependencies that exist between different regions. This can be expected to benefit the EA in mitigating against severe events such as flash floods, and policy-makers in addressing key national strategic needs, such as the need to better protect and/or restrict the development of assets located on floodplains (the UK is building faster on floodplains than anywhere else, with approx. 1500 new homes built annually in 'high risk' areas (Committee on Climate Change 2015)).
The fellowship will involve personnel co-working with mathematicians, engineers and scientists, conducting research placements, and participating in industry and public engagement activities. Hence, it will generate unique skills and capabilities that can be deployed by UK organisations and private companies to maximise the social, economic and environmental impacts of the research.
In summary, the outcomes of this research will significantly benefit the UK public living in and around flood-prone areas, who will be better informed of the likely consequences of impending climatic events, and businesses who can take steps to reduce disruption of their activities, helping to ensure affordable insurance premiums and better quality of life.
Organisations
- University of Sheffield (Fellow, Lead Research Organisation)
- BMT Group (Collaboration)
- LOUGHBOROUGH UNIVERSITY (Collaboration)
- Lawrence Berkeley National Laboratory (Collaboration)
- DHI Group (Collaboration)
- CH2M HILL (Collaboration)
- University of East Anglia (Collaboration)
- Innovyze (Collaboration)
- Environment Agency (Collaboration, Project Partner)
- Democritus University of Thrace (Collaboration)
- UNIVERSITY OF LEEDS (Collaboration)
- NORCE Norwegian Research Center AS (Collaboration)
- University of Bristol (Collaboration)
- BMT Group (United Kingdom) (Project Partner)
- Danish Hydraulic Institute (Project Partner)
- Sheffield City Council (Project Partner)
- Jacobs (United Kingdom) (Project Partner)
- XP Software Solutions Ltd (Project Partner)
- University of Bristol (Project Partner)
- University of East Anglia (Project Partner)
- Innovyze Limited (Project Partner)
- University of Leeds (Project Partner)
People |
ORCID iD |
| Georges Kesserwani (Principal Investigator / Fellow) |
Publications
Sharifian M
(2018)
A discontinuous Galerkin approach for conservative modeling of fully nonlinear and weakly dispersive wave transformations
in Ocean Modelling
Kesserwani G
(2018)
Discontinuous Galerkin formulation for 2D hydrodynamic modelling: Trade-offs between theoretical complexity and practical convenience
in Computer Methods in Applied Mechanics and Engineering
Sharifian M
(2019)
Performance study of the multiwavelet discontinuous Galerkin approach for solving the Green-Naghdi equations
in International Journal for Numerical Methods in Fluids
Kesserwani G
(2019)
(Multi)wavelets increase both accuracy and efficiency of standard Godunov-type hydrodynamic models
in Advances in Water Resources
Shaw J
(2020)
Stochastic Galerkin Finite Volume Shallow Flow Model: Well-Balanced Treatment over Uncertain Topography
in Journal of Hydraulic Engineering
Shaw J
(2020)
Probabilistic Godunov-type hydrodynamic modelling under multiple uncertainties: robust wavelet-based formulations
in Advances in Water Resources
Shirvani M
(2020)
Agent-based modelling of pedestrian responses during flood emergency: mobility behavioural rules and implications for flood risk analysis
in Journal of Hydroinformatics
Kesserwani G
(2020)
(Multi)wavelets increase both accuracy and efficiency of standard Godunov-type hydrodynamic models: Robust 2D approaches
in Advances in Water Resources
Özgen-Xian I
(2020)
Wavelet-based local mesh refinement for rainfall-runoff simulations
in Journal of Hydroinformatics
Shirvani M
(2021)
Agent-based simulator of dynamic flood-people interactions
in Journal of Flood Risk Management
| Description | There is a benefit from deploying a second-order discontinuous Galerkin flood (DG2) model for flood modelling as an alternative to existing finite volume (FV) based flood models. This is useful to particularly produce more accurate and efficient flood maps over coarse terrain resolution, and more informative velocity and water level hydrographs. Multiwavelet with DG2 (MWDG2) can be made efficient and robust for real-world flood modelling applications but particualry requires exploiting the parallisation of Graphical Processing Units to overcome the overead adaptivity costs associated with the presence of large dry areas. Wavelets can made the direct stochastic Galerkin approach robust for probabilistic hydrodynamic modelling with wet-dry zones over frictional topographies under many sources of uncertainty. The direct approach has efficiency benefits over Monte Carlo when coping with a single uncertain parameter, but its integrative variants is a much better option in favor efficiency when coping with multiple uncertain parameters. More research is ongoing to find out more about the efficiency of other alterative method for quantifying uncertainties in flood modelling. Agent-based modelling is becoming trendy for flood risk management to incorporate human behaviour dynamics into the flood risk analysis. Modelling flood people interactions could be achieved via an agent based platform on Graphical Processing Units, but more science driven research is needed to produced the evidence base for integrating more realistic rules on human behaviours. |
| Exploitation Route | Regular meetings are taking place with consulting companies supplying hydraulic software. The DG2 model has been being integrated into LISFLOOD-FP (http://www.bristol.ac.uk/geography/research/hydrology/models/lisflood/) for wider community access. |
| Sectors | Digital/Communication/Information Technologies (including Software) |
| URL | https://www.seamlesswave.com |
| Description | Development of new training/educational materials for postgraduates/research users |
| Geographic Reach | Multiple continents/international |
| Policy Influence Type | Influenced training of practitioners or researchers |
| Impact | Updated the materials of a course taught to postgraduate students to include latest development on flooding modelling technologies (course name: "Computational methods for water engineering", University of Sheffield). Uploaded research software and data for the published papers into an open-source platform (https://zenodo.org/record/4073011, more than 1000 downloads to date), and a detailed user guide (https://www.seamlesswave.com/LISFLOOD8.0) to train many community to learn the new developments and to apply them in their research projects. The changes have been noted from the many quieries received from our students and research scientists worldwide, asking questions inspired from the educational materials placed on the website. |
| URL | https://www.seamlesswave.com/LISFLOOD8.0 |
| Title | An agent-based flood-pedestrian simulator ported into a UK national facility |
| Description | The simulator tool allows microscopic modelling of individual people's responses as they interact with floodwater. It is useful to support flood risk management research and applications. For example, it can be used to evaluate the spatial and temporal changes in the flood risk state of people, as well as their mobility and stability during an urban flood incident. It can also be used to evaluate options for an emergency intervention just within hours before a severe flooding. As a flexible agent-based simulator, it allows to realistically characterise individual people behaviours in floodwater, e.g. to integrate different physical body characteristics, walking/running speeds in water depending on the age and gender, and to even insert new attributes, features and vulnerability metrics to facilitate multidimensional flood-related studies. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | The most notable imact from this development has been a series of diaglogues with the developers of the world-leading Life Safety Model (https://lifesafetymodel.net/), onto how it can be supported with a functionality for micro-simulation of the uncertain interaction of people with urban floodwaters. These dialogues resutled in a joint grant proposal submission that is pending a decision. |
| URL | https://dafni.ac.uk/project/flood-people-simulator/ |
| Title | Augmenting the LISFLOOD-FP software with new solvers |
| Description | Many advanced numerical solvers and efficiency enhancement tools have been integrated into LISFLOOD-FP (http://www.bristol.ac.uk/geography/research/hydrology/models/lisflood/). LISFLOOD-FP was initally developed in the UK Food Risk Managment Research Consortium. It is a two-dimensional hydrodynamic model specifically designed to simulate floodplain inundation in a computationally efficient manner over complex topography. LISFLOOD-FP has already been used by a variety of research group worldwide, and the added components have been documented including video tutorials to further promote the usability of the software worldwide (https://www.seamlesswave.com/LISFLOOD8.0). We are still collaborating with colleagues at the University of Bristol (Prof. Bates and Dr Neal) to augment the capabilities of LISFLOOD-FP with more advanced solvers for multiscale modelling and futher update the open-source version. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | Since the release of our updated version 8.0 of LISFLOOD, we have received an great deal of enquiries from researcher worldwide for using it and sharing best practice with us. So far, the sofware has been dowloaded more than 1000 times with great international interest as evidenced by the high number of quieries we have received on its usability. LISFLOOD is already used by many research group and businesses and the additions we have made to it are imacting many users intersted to support more detailed and efficient flood modelling activites. |
| URL | https://www.seamlesswave.com/LISFLOOD8.0 |
| Title | One-dimensional Fortran 2003 finite volume and discontinuous Galerkin hydrodynamic models with wavelet-based adaptivity |
| Description | Fortran 2003 models for solving the one-dimensional shallow water equations with topography and friction using Godunov-type finite volume and discontinuous Galerkin methods. Models can run with wavelet-based adaptivity enabled, or with adaptivity disabled on uniform meshes. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2019 |
| Provided To Others? | Yes |
| Impact | The proposed methods has attracted a collaboration with US researcher to add wavelet-based grid resolution adaptivity into a hydrological model for rainfall-runoff simulation (https://iwaponline.com/jh/article/22/5/1059/73853/Wavelet-based-local-mesh-refinement-for-rainfall). So far the code received 32 downloads and the paper reporting it has received more than 1000 reads on research gate, suggesting a great level of interest. |
| URL | https://www.sciencedirect.com/science/article/pii/S0309170819301770 |
| Title | Sampling-based uncertainty quantification methods |
| Description | Sampling based uncertainty propagation approaches has been integrated into a framework for probabilistic flood maping from any existing deterministic models. The framework considers a range of modern uncertainty quantification methods and is aimed to accelerate probabilistic flood modelling by using alternatives to the Standard Monte Carlo method (SMC). The methods include Latin Hypercube Sampling (LHS), Adaptive Stratified Sampling (ASS), Quasi Monte Carlo (QMC) and Haar Wavelet Expansion (HWE) as they can capture all types of histogram distributions. The code to test and run these alternatives for three realistic case studies with many uncertain inputs can be found on https://zenodo.org/record/7050213, and the paper reporting the resutls is still under reviewer for publication in 2023. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2023 |
| Provided To Others? | Yes |
| Impact | The key impact of this proposed research framework is new evidenced-based guidance on the most suited uncertainty quantification method for probablistic flood modelling featured with multiple uncertain input paramters. |
| URL | https://zenodo.org/record/7050213 |
| Title | Two-dimensional Fortran 2003 finite volume and discontinuous Galerkin hydrodynamic models with (multi)wavelet-based adaptivity |
| Description | Fortran 2003 models for solving the two-dimensional shallow water equations with topography and friction using Godunov-type finite volume and discontinuous Galerkin methods. Models can run with (multi)wavelet-based adaptivity enabled, or with adaptivity disabled on uniform meshes. The code can be compiled using Intel Fortran Compiler in both Windows and Linux. Other Fortran compilers have not been tested. Windows users can simply add the source files to the project created for Microsoft Visual Studio or any other IDE. Linux users can use the included makefile to compile the codes. The user can configure the simulations by modifying config.dat input file. A suite of two-dimensional test cases are preconfigured. |
| Type Of Material | Improvements to research infrastructure |
| Year Produced | 2023 |
| Provided To Others? | Yes |
| Impact | Will be released in July 2023 from the Zenodo repository. Meanwhile, the key impact of this research has been to collaborate with colleagues from the University of Bristol to augment LISFLOOD-FP with the wavelet-based grid resolution adaptivity alongside GPU parallelisation for faster flood simulations (https://gmd.copernicus.org/preprints/gmd-2022-259/; https://www.sciencedirect.com/science/article/pii/S0309170822002202). |
| URL | https://zenodo.org/record/394172 |
| Title | Benchmarking (multi)wavelet-based dynamic and static non-uniform grid solvers for flood inundation modelling (Simulation results) |
| Description | Simulation result data for Environment Agency benchmark test 5, Thamesmead hypothetical flood, and Carlisle 2005 case studies, using uniform DG2, adaptive MWDG2, adaptive HWFV1, non-uniform DG2, non-uniform FV1 and non-uniform ACC solvers. Model results are archived in 3 zip files: EA5.zip contains results of Environment Agency test 5 (Néelz and Pender, 2013) Thamesmead.zip contains results of Thamesmead hypothetical flood (Liang et al., 2008) Carlisle.zip contains results of Carlisle 2005 flooding (Neal et al., 2009) The results are stored with the following file extensions: ".wd" for 2D flood inundation maps in ESRI ASCII format ".stage" for water depth or water level time-series at staging points in tabulated text format ".velocity" for velocity time-series at staging points in tabulated text format Model outputs are stored under directories named for each solver. References Néelz, S., & Pender, G. (2013). Benchmarking the latest generation of 2D hydraulic modelling packages. Environment Agency: Bristol, UK. Liang, Q., Du, G., Hall, J. W., & Borthwick, A. G. (2008). Flood Inundation Modeling with an Adaptive Quadtree Grid Shallow Water Equation Solver. Journal of Hydraulic Engineering, 134(11), 1603-1610. https://doi.org/10.1061/(ASCE)0733-9429(2008)134:11(1603) Neal, J. C., Bates, P. D., Fewtrell, T. J., Hunter, N. M., Wilson, M. D., & Horritt, M. S. (2009). Distributed whole city water level measurements from the Carlisle 2005 urban flood event and comparison with hydraulic model simulations. Journal of Hydrology, 368(1-4), 42-55. https://doi.org/10.1016/j.jhydrol.2009.01.026 |
| Type Of Material | Database/Collection of data |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | They have been downloaded by other groups for use in alternative research projects |
| URL | https://zenodo.org/record/5047565 |
| Title | Efficient uncertainty propagation approaches (stochastic Galerkin approaches with and without local wavelet-based refinement) |
| Description | The models and results are to support understanding the methods reported in https://doi.org/10.1061/(ASCE)HY.1943-7900.0001705 and in https://www.sciencedirect.com/science/article/pii/S0309170819306281 |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | Opening a potential direction for future research in shallow flow modelling to model uncertainty more efficiently than the conventional Monte-Carlo methods to run many simulations (https://www.sciencedirect.com/science/article/pii/S0309170821000221). This method has started to receive popularity in the computational hydraulics community (number of views and download counts can be found on Zenodo). |
| URL | https://zenodo.org/record/2592000 |
| Title | Flood-pedestrian simulator user guide |
| Description | This document provides: a brief overview of the flood-pedestrian simulator's algorithmic structure (accessible at: https://doi.org/10.5281/zenodo.4564288) step-by-step guidance for users to run the simulator on their own machine for two test cases reported in the Shirvani et al. (2020), Shirvani et al. (2021) and Shirvani & Kesserwani (2021) papers (see also the demo videos); explanation of the outputs from the simulator with guidance on how to modify the simulations by changing the environment parameters (outputs from the test cases reported in Shirvani & Kesserwani (2021) are also available at: https://doi.org/10.5281/zenodo.4576906); and guidance on how to apply the simulator to new test cases using the available tools and models designed for this purpose. |
| Type Of Material | Data analysis technique |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | Many researchers in the UK has been refering to this document to apply the flood-pedestrian simulation to other UK case studies of flood risk mitigation and alleviation. |
| URL | https://nhess.copernicus.org/articles/21/3175/2021/nhess-21-3175-2021-discussion.html |
| Title | LISFLOOD-DG2 simulation set-up files and additional datasets |
| Description | The set-up files (including initial & boundary conditions and digital elevation model) for LISFLOOD-DG2 have been provided for the following test cases: Test 1: Vortex shedding past a conical island Test 3: Flow diversion at a T-junction, and Test 4: Flooding in an urban residential area The data for the initial condition, boundary condition and topography data for Test 2: Recirculation flow in sharp building cavities are also included, as .txt and .csv format. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | The paper describing the utility of this data sets has just come out (https://ascelibrary.org/doi/abs/10.1061/JHEND8.HYENG-13244). Therefore, it is expected that impact of this data sets will appear in the next year or so. |
| URL | https://ascelibrary.org/doi/abs/10.1061/JHEND8.HYENG-13244 |
| Title | Modelling results for the analytical assessment (Sec. 2.3) and EA benchmark tests (Sec. 3) |
| Description | These datasets contains the flow profiles and time histories produced by the FV1, MUSCL-FV2, DG2-NL and DG2-LL for the following tests: Sec. 2.3: Analytical assessments of model conservation properties: Time histories of normalised total energy and mass error from FV1, MUSCL-FV2, DG2-NL and DG2-LL at 20m and 40m grid resolution. Water level profiles at 7tau/2 and 8tau from FV1, MUSCL-FV2, DG2-NL and DG2-LL at 40m grid resolution, along with the analytical solutions. Sec. 3: Comparison against industrial flood model outputs Sec. 3.1.1: Flooding and drying cycle over a sloping topography: Time histories of water level from DG2-LL and DG2-NL at 10m resolution, DG2-NL and MUSCL-FV2 at 20m resolution, and the reference solution. Sec. 3.1.2: Symmetrical flow propagation over a flat bed: Time histories of water level and velocity from from DG2-LL and DG2-NL at 5m resolution, DG2-NL and MUSCL-FV2 at 10m resolution, and the reference solution. Sec. 3.1.3: Slow filling of multiple ponds: Time histories of water level from from DG2-LL and DG2-NL at 20m resolution, DG2-NL and MUSCL-FV2 at 40m resolution, and the reference solution. Sec. 3.2.1: Momentum conservation over an obstruction: Time histories of water level and velocity from from DG2-LL and DG2-NL at 5m resolution, DG2-NL and MUSCL-FV2 at 10m resolution, and the reference solution. Sec. 3.2.2: Torrential flooding over a rural catchment: Time histories of water level and velocity from from DG2-LL and DG2-NL at 50m resolution, DG2-NL and MUSCL-FV2 at 100m resolution, and the reference solution. Sec. 3.2.3: Dam-break over an oblique building: Time histories of water level and velocity from from DG2-LL and DG2-NL at 0.1m resolution, and DG2-LL and MUSCL-FV2 at 0.2m resolution. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | Very useful for benchmarking other flood models. |
| URL | https://zenodo.org/record/3760628 |
| Title | Reconstructed DEM of the Morgan-Selwyn floodway with added building piers |
| Description | These datasets contain the DEM of Morgan-Selwyn floodway in Merewether, Australia, which has been generated using the rescaled bathymetry data in Smith et al., (2017), with small blocks representing the building piers added. Two different resolutions are considered, namely: Merewether DEM with 0.175m resolution (i.e. 'merewether-0p175m.asc') that is used to produce simulation results in Ayog et al. (2021), and Merewether DEM with 0.010m resolution (i.e. 'merewether-0p01m.zip') for the start-up training of LISFLOOD 8.0 (https://www.seamlesswave.com/Merewether.html) |
| Type Of Material | Database/Collection of data |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | A complex data set that was produced as part of our research to analyse "Shallow-Flow Velocity Predictions Using Discontinuous Galerkin Solutions" (https://ascelibrary.org/doi/abs/10.1061/JHEND8.HYENG-13244). This particualr data set has received a high number of downloads, more than 135, suggesting that it is useful to other researcher working on bencharking the ability of flood models to predict the velocity fields. |
| URL | https://zenodo.org/record/5069224 |
| Title | Smart and flexible multiresolution wavelet-based flood models |
| Description | In support of the methods described in https://www.sciencedirect.com/science/article/pii/S0309170819301770 and in https://www.sciencedirect.com/science/article/pii/S0309170820303079. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2019 |
| Provided To Others? | Yes |
| Impact | One direct impact was a collaboration with US researchers in Lawrence Berkeley National Laboratory, to develop and validate Wavelet-based local mesh refinement (wLMR) for rainfall-runoff simulations for real-word catchments. The wLMR approach is being integrated into the hydrogeological tool 'TINerator' to augment its mesh generation capability. The general impact of these new methods can also be evidenced by the number of downloads and views on Zenodo. |
| URL | https://zenodo.org/record/1745471 |
| Title | Two-dimensional Fortran 2003 finite volume and discontinuous Galerkin hydrodynamic models with (multi)wavelet-based adaptivity |
| Description | Fortran 2003 models for solving the two-dimensional shallow water equations with topography and friction using Godunov-type finite volume and discontinuous Galerkin methods. Models can run with (multi)wavelet-based adaptivity enabled, or with adaptivity disabled on uniform meshes. The code can be compiled using Intel Fortran Compiler in both Windows and Linux. Other Fortran compilers have not been tested. Windows users can simply add the source files to the project created for Microsoft Visual Studio or any other IDE. Linux users can use the included makefile to compile the codes. The user can configure the simulations by modifying config.dat input file. A suite of two-dimensional test cases are preconfigured. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | The models are embargoed until the end of 2023. |
| URL | https://zenodo.org/record/3941723 |
| Description | Academic and industrial advisory network |
| Organisation | BMT Group |
| Country | United Kingdom |
| Sector | Charity/Non Profit |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | CH2M HILL |
| Country | United States |
| Sector | Private |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | DHI Group |
| Country | Global |
| Sector | Private |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | Environment Agency |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | Innovyze |
| Country | United Kingdom |
| Sector | Private |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | Loughborough University |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | University of Bristol |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | University of East Anglia |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Academic and industrial advisory network |
| Organisation | University of Leeds |
| Country | United Kingdom |
| Sector | Academic/University |
| PI Contribution | The fellowship team has been committed to demonstrate the relevance of new or research-led theoretical developments to improve real-world modelling of flood inundations. Contributions made to the partners and the wider scientific community includes: (1) Proposing a robust and efficient design for a second-order discontinuous Galerkin (DG2) flood model, benchmarking it with respect to the finite volume industry standard flood models based on test cases that are approved by the UK Environment Agency, and integrating and releasing an open source DG2 flood model within the widely-used LISFLOOD-FP software alongside a detailed user guide with video tutorials; (2) Proposing a radical redesign for multiwavelet-based adaptive DG2 flood models that adapts multiwavelets theory to fit the setting whereby 2D solvers are efficient and robust for real-world applications. The results are significant in demonstrating how multiwavelet-based 2D solvers can be made operational for 2D flood inundation modelling, and offers an approach to augment Godunov-type solvers to deliver the accuracy of yet unaffordable research-based models, while simultaneously boosting their efficiency; (3) Demonstrating a direct and/or reduced-cost alternative approach to the expensive Monte Carlo uncertainty sampling that is commonly used in flood inundation modelling; and (4) Designing new flood-pedestrian simulator that dynamically couples and updates flooding and people in a single agent-based modelling platform. Visit https://www.seamlesswave.com for more elaborate details on the contributions made by the fellowship team to this collaboration. |
| Collaborator Contribution | Many of the industrial partners contributed by sharing their software results and data, and their expertise on the modelling needs for real-world applications. Many of the academic partners have contributed to provide feedback over publication drafts and to steering the direction of the research. The Environment Agency provided information of relevance to the computerization of the flood people simulator. The academic partner from Bristol provided access to the LISFLOOD-FP software for the team to integrate the DG2 model and many other enhancement including GPU parallelisation and wavelet-based grid adaptivity. |
| Impact | LISFLOOD-FP version 8.0 that integrates a new DG2 hydrodynamic solvers and a new parallel platform on GPU (https://doi.org/10.5194/gmd-14-3577-2021) and grid adaptivity for fast compulation at the catchement scale (https://doi.org/10.5194/gmd-2022-259); and, the flood-people simulator for dynamic modelling of flood-people interactions that has been ported to DAFNI, https://www.dafni.ac.uk, (https://onlinelibrary.wiley.com/doi/10.1111/jfr3.12695). All the members of this Network, among other collaborators, agreed to support a workshop in Sheffield: "Flood modelling and forecasting challenges in industry workshop", via contributing with a talk. The workshop was postponed twince (June 2020 to June 2021) in response to the COVID19 pandemic, and finally took place remotely (https://www.seamlesswave.com/Workshop_16Sep2021). One key oputput of the workshop has been to conduct a compartive study for a range of alternative uncertainty quantification method to the standard Monte Carlo method to more efficiently reporduce flood-related probability distributions, which practionners and decision makers are interested to capture for flood frequency analysis. |
| Start Year | 2018 |
| Description | Efficient uncertainty propagation in flood modelling |
| Organisation | Democritus University of Thrace |
| Country | Greece |
| Sector | Academic/University |
| PI Contribution | A collaboration with an applied mathematician, Dr. Per Pettersson from the Norwegian Research Centre, who is expert on advanced uncertainty propagation methods for a wide range of engineering problems, and civil engineer, Dr Vassils Bellos, who is expert on uncertainties in the field of flood modelling. Our main contribution has been to transfer and explore many new mathematical concepts for improving the efficiency of uncertainty propagation methods in the context of flood inundation modelling. |
| Collaborator Contribution | Dr. Pettersson has provided hands-on support and software access to relatively new uncertainty quantification methods for applying them in the context of flood modelling; whereas Dr. Bellos has provided expertise on how to randomise the input uncertain parameter for flood modelling. |
| Impact | The first output of this collaboration has led to new method highlighting a potential direction for future research in shallow flow modelling (https://www.sciencedirect.com/science/article/abs/pii/S0309170819306281), as part of a special issue on "Innovations towards the next generation of shallow flow models (https://doi.org/10.1016/j.advwatres.2021.103867). The method has the potential to enable much more efficient quantification of uncertainty in various applications such as probabilistic flood forecasting and risk assessment, and has been made accessible (https://zenodo.org/record/3600494; and https://zenodo.org/record/3600480). The second output has been a comparative study of this method and four other methods for efficient capturing of flood frequency for the relevant quantities of interest to help practionners choose better altenatives the standard Monte Carlo Method for probabilsitic flood modelling (Hajihassanpour et al., 2023, under review). The codes for reusing these methods have been made accessible on Zenodo with guidance on how to run them (https://zenodo.org/record/7050213, and https://www.seamlesswave.com/Uncertainty_Propagation). |
| Start Year | 2020 |
| Description | Efficient uncertainty propagation in flood modelling |
| Organisation | NORCE Norwegian Research Center AS |
| Country | Norway |
| Sector | Academic/University |
| PI Contribution | A collaboration with an applied mathematician, Dr. Per Pettersson from the Norwegian Research Centre, who is expert on advanced uncertainty propagation methods for a wide range of engineering problems, and civil engineer, Dr Vassils Bellos, who is expert on uncertainties in the field of flood modelling. Our main contribution has been to transfer and explore many new mathematical concepts for improving the efficiency of uncertainty propagation methods in the context of flood inundation modelling. |
| Collaborator Contribution | Dr. Pettersson has provided hands-on support and software access to relatively new uncertainty quantification methods for applying them in the context of flood modelling; whereas Dr. Bellos has provided expertise on how to randomise the input uncertain parameter for flood modelling. |
| Impact | The first output of this collaboration has led to new method highlighting a potential direction for future research in shallow flow modelling (https://www.sciencedirect.com/science/article/abs/pii/S0309170819306281), as part of a special issue on "Innovations towards the next generation of shallow flow models (https://doi.org/10.1016/j.advwatres.2021.103867). The method has the potential to enable much more efficient quantification of uncertainty in various applications such as probabilistic flood forecasting and risk assessment, and has been made accessible (https://zenodo.org/record/3600494; and https://zenodo.org/record/3600480). The second output has been a comparative study of this method and four other methods for efficient capturing of flood frequency for the relevant quantities of interest to help practionners choose better altenatives the standard Monte Carlo Method for probabilsitic flood modelling (Hajihassanpour et al., 2023, under review). The codes for reusing these methods have been made accessible on Zenodo with guidance on how to run them (https://zenodo.org/record/7050213, and https://www.seamlesswave.com/Uncertainty_Propagation). |
| Start Year | 2020 |
| Description | Integrating a wavelet-based grid generator for rainfall runoff modelling at catchement scale |
| Organisation | Lawrence Berkeley National Laboratory |
| Country | United States |
| Sector | Public |
| PI Contribution | Collaboration with US researchers in Lawrence Berkeley National Laboratory to develop and validate Wavelet-based local mesh refinement (wLMR)for rainfall-runoff simulations for real-world catchments. The wLMR apporach is being integrated into the hydrogeological tool 'TINerator' to augment its mesh generation capability (I. Özgen-Xian and G. Kesserwani). |
| Collaborator Contribution | The partners provided software for rainfall runoff modelling to explore how much it can be made efficient by integrating a multiscale grid generator support by the wavelets algorihtm devleopped by our university. |
| Impact | The outomes of this collaboration has been published in the Journal of Hydroinformatics and the software that has been improved as a resut of this collaboration is openly accessible on GitHub: https://github.com/lanl/LaGriT/tree/tinerator |
| Start Year | 2019 |
| Title | Flood-pedestrian simulator |
| Description | The flood-pedestrian simulator is developed within the Flexible Agent-based Modelling Environment for the GPU FLAMEGPU framework for modelling multiple agent interactions on CUDA Cores for parallel processing on Graphical Processing Units (GPUs). |
| Type Of Technology | Software |
| Year Produced | 2021 |
| Open Source License? | Yes |
| Impact | To be continued |
| URL | https://www.seamlesswave.com/Flood_Human_ABM.html |
| Title | LISFLOOD-FP 8.0 hydrodynamic model |
| Description | LISFLOOD-FP is a two-dimensional hydrodynamic model specifically designed to simulate floodplain inundation in a computationally efficient manner over complex topography. |
| Type Of Technology | Software |
| Year Produced | 2020 |
| Open Source License? | Yes |
| Impact | To be continued |
| Title | Two-dimensional Fortran 2003 finite volume and discontinuous Galerkin hydrodynamic models with (multi)wavelet-based adaptivity |
| Description | Fortran 2003 models for solving the two-dimensional shallow water equations with topography and friction using Godunov-type finite volume and discontinuous Galerkin methods. Models can run with (multi)wavelet-based adaptivity enabled, or with adaptivity disabled on uniform meshes. The code can be compiled using Intel Fortran Compiler in both Windows and Linux. Other Fortran compilers have not been tested. Windows users can simply add the source files to the project created for Microsoft Visual Studio or any other IDE. Linux users can use the included makefile to compile the codes. The user can configure the simulations by modifying config.dat input file. A suite of two-dimensional test cases are preconfigured. |
| Type Of Technology | Software |
| Year Produced | 2020 |
| Open Source License? | Yes |
| URL | https://zenodo.org/record/3941722 |
| Description | Engagement with software developers and end-users |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Six industrial delegates and 5 academic delegate attended a focused industrial meeting at the University of Sheffield which sparked critics and discussion afterwards, and the industry delegates reported increased interest and support in terms of provision of data and existing software results to help better validate the new flood modelling tools being developed. In this meeting, the panel was particularly in favor of conducting a comparative study of uncertainty quantfication mehtods for practical flood inundation modelling, and such a study is currently ongoing. |
| Year(s) Of Engagement Activity | 2021,2022 |
| Description | Engagement with software developers and end-users |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Six industrial delegates and 5 academic delegate attended a focused industrial meeting at the University of Sheffield which sparked critics and discussion afterwards, and the industry delegates reported increased interest and support in terms of provision of data and existing software results to help better validate the new flood modelling tools being developed. |
| Year(s) Of Engagement Activity | 2018,2019,2020 |
| URL | https://www.seamlesswave.com/Activities.html |
| Description | Engagement with the public at Sheffield to make them understand: How do computers forecast floods? |
| 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 | Public engagement presentation, "How do computers forecast floods?" It was held as part of the "Pint of Science, Sidney & Matilda, Sheffield (M. K. Sharifian). Mohammad talked about the importance of flood forecasting systems, how they work and what are the challenges in improving the flood forecasts. The presentation followed with Q&A from many members of the public, asking clarification on the role of computer modelling in the understanding of flood risk to properties. |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://docs.google.com/presentation/d/1-r7LbF9SkBfxxsSLyGZ2axyEm2iWGHnJ/edit#slide=id.p1 |
| Description | Exhibition at Flood Expo 2019 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Participation in the Flood Expo 2019 (http://www.thefloodexpo.co.uk) to showcase new research developments to a range of stakeholders, through exhibition panels (J. Ayog, M. Shirvani & G. Kesserwani). In particular, our 'agent-based flood-people simulator' attracted great interest from flood defense designers and companies in the UK and the EU, leading to them to providing us with the data needed to computerise alternative to sandbagging into our flood-people simulator. |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://www.thefloodexpo.co.uk/ |
| Description | Festival of the Mind 2020 |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Public/other audiences |
| Results and Impact | The event was to showcase the real-time visualization capability of a 'food-pedestrian simulator' applied to a flood-prone site around Hillsborough football stadium, which sparked questions and discussions from members of the general public, in particular those who experienced the 2007 Sheffield flood who provided us with valuable data to improve the simulator. |
| Year(s) Of Engagement Activity | 2020 |
| URL | https://festivalofthemind.sheffield.ac.uk/2020/ |
| Description | Festival of the Mind: How do computers forecast floods? |
| Form Of Engagement Activity | Participation in an open day or visit at my research institution |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Public/other audiences |
| Results and Impact | M. K. Sharifian was also awarded "Partnerships & Regional Engagement" funds to design a toolkit over a 6-month project that concluded with him demonstating futher to the general public "How do computers forecast floods?" supported with live visualisation. The real-time visulation toolkit was for people to use and explore to better understand the impact and concequences of flooding events. It was presented over the 10-day " Festival of the Mind 2022" during Sept. 2022. The toolkit was explored by many people, in the presence of a delegate to answer emerging questions. |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://festivalofthemind.sheffield.ac.uk/2022/futurecade/interactive-visualization-software-for-flo... |
| Description | Flood Risk Workshop assembling people in academia, industry (R&D), and government agencies |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Supporters |
| Results and Impact | An invitation-only NASA-supported 'Flood risk workshop' (Oct 2018, Colorado, USA) in which invited talks were delivered by charitable organisation or flood related businesses. The event aimed to identify minimal requirements for mapping global flood risk that meets user needs, to promoting citizen science and start new collaborative partnerships. |
| Year(s) Of Engagement Activity | 2018 |
| URL | https://sites.google.com/view/flood-risk-ws |
| Description | Flood modelling and forecasting challenges in industry virtual workshop |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | The workshop was supported by TWENTY65. It brought together practitioners and scholars, in an attempt to identify and discuss timely issues in operational-scale flood modelling, forecasting and management. The workshop aimed to increase our understanding of industrial and practical needs that require further addressing in academic research. The focus was on the themes: Overview of flooding challenges, Modelling, Resilience. The workshop was facilitated by a series of invited talks by experts from academia, industry and UK organisations. Title of the talks and speakers' information are listed in the agenda (see the appendix). During the lunch break, there was a virtual networking event on Gather Town allowing participants to meet each other online. At the end of the workshop, participates were directed into Miro's whiteboards letting them enter their thoughts, comments and ideas. A direct outcome of the workshop is an online post-workshop report written based on compiling and analysing the participants' inputs. |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.seamlesswave.com/Workshop_16Sep2021.html |
| Description | Invited talk at the University of Machester |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Postgraduate students |
| Results and Impact | Invited seminar talk for the "Water, Ocean, Coastal and Environmental Engineering with Geotechnics" (WOCEE-G) at the University of Manchester (G. Kesserwani). Title: "Benchmarking uncertainty propagation methods for flood hazard frequency analysis in response to multiple uncertain inputs". The talk addressed a wide range of audience members (15-25) from experimental to computational researchers all focused on water-related research, and stimulated questions on how to best tackle uncertainties in modelling activities with minimum computational costs. |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://drive.google.com/file/d/1_nV5uinfD4GAHJZRGTP6yHmA3muhhAMx/view |
| Description | Promoting and facilitating the utility of our new DG2 flood model and the GPU platform added to LISFLOOD-FP |
| Form Of Engagement Activity | Engagement focused website, blog or social media channel |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Postgraduate students |
| Results and Impact | This activity is part of a knowledge exchange project award from the University of Sheffield (2021, Jan.-Mar.). The project aims to document our new contribution to LISLFOOD-FP (Shaw et al. 2021), a software widely used within the hydraulic and hydrologic modelling communities (https://gmd.copernicus.org/preprints/gmd-2020-340/). |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.seamlesswave.com/LISFLOOD8.0 |
| Description | Showcase of GPU-accelerated adaptive flood modelling with the developers of TUFLOW sofware |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Two from the software development team of the TUFLOW flood modelling software attended the meeting where our team showcased the benefit from parallelising adaptive finite volume flood modelling on the Graphical Porcessing Units (GPU). The meeting sparked questions on the extent to which dynamic modelling in time can be faster and this has lead to provided further assessment of speed-up with respect to operation flood modelling tools and a report on the subject afterwards. |
| Year(s) Of Engagement Activity | 2021 |
| Description | Showcase of the flood-pedestrian simulator at the UKCRIC Digital theme workshop |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Participation in the UKCRIC Digital theme workshop, University of Cambridge. Aimed to explore the UK research activities focussing on digitalisation in the built environment across the UKCRIC institutions. The flood-pedesrtian simulator was presented, with a particular focus on how it computerise people's risk perception to floodwater in modelling a mass evacuation scenarios. |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.ukcric.com/news/ukcric-workshop-explores-the-challenges-and-opportunities-of-digitalisat... |
| Description | Talk to engage with the public in the Millenium Gallery, Sheffield (20 Sept. 2019) |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Public/other audiences |
| Results and Impact | Public engagement presentation: "What happens the day after tomorrow: how do computers forecast floods?" delivered in the Pop-Up University talks, Millenium Gallery, Sheffield (J. Shaw). James' talk explained how the Environment Agency is collecting high-resolution terrain data across the UK in order to produce national flood hazard maps, and he showed how his research, developing uncertainty quantification methods, can help flood risk managers to assess their confidence in computer flood simulations. 20 civilians attended the talk, which sparked questions and discussion afterwards. |
| Year(s) Of Engagement Activity | 2019 |
| URL | https://www.museums-sheffield.org.uk/whats-on/events/2019/9/pop-up-university |