Stream 1

Day 1

Duncan Faulkner, JBA Consulting

Welcome to the future: living and working in a climate-changed world

Duncan Faulkner Duncan is Head of Hydrology at JBA Consulting. After five years of research on rainfall and flood frequency at CEH in Wallingford, he joined JBA and moved north just in time to see in the new millennium. Since then he has worked on a large portfolio of projects across the UK, Ireland, Australia, Canada and south-east Asia. Current topics of Duncan’s consultancy and research work include application of non-stationary flood frequency estimation and improving methods of flood estimation for dam safety.

Abstract

Once upon a time, climate change was something that we expected to impact our future. Now we are living in that future. How does it compare with our expectations?
I will present results from a new trend analysis of flood flow data from UK rivers, asking whether trends are more likely on some catchment types and how we can know if the trends are caused by climate change.
I will compare estimates of design flows for 2020 with those derived using pre-1990 data. How do the differences compare with the climate change uplifts we have been applying in the UK? Are the latest uplifts fit for purpose?
To close, some thoughts on whether Tesco offers a suitable slogan when it comes to climate change mitigation.

Will McBain

Keeping our eyes on the prize

Will McBain is Arup’s flood resilience leader in the UK. He is also Arup’s framework manager for the Environment Agency’s Collaborative Delivery Framework across the Midlands and North East/Yorkshire. He has 26 years’ wide-ranging experience of the appraisal, planning and design of projects to improve flood resilience and enhance the water environment. He also has extensive experience of assessing and mitigating the impacts of major infrastructure projects (eg roads, rail, nuclear and hydropower) on the water environment.

Abstract

As the impacts of climate change start to bite, client organisations are asking a great many questions of modellers that lie well beyond the realms of “how high do I need to build my flood defences?” There is a hunger for answers to questions about future hazard characterisation, natural flood management, long term morphological change, habitat evolution, carbon sequestration potential, design exceedance, residual risk…the list goes on. But we are effectively on a war footing tackling the environmental crisis. There has never been a greater need for us to deploy our available skills and resources wisely to make sure we’re asking the right questions about the right problems. And this means keeping our eyes on the prize. My talk will elaborate on what we might usefully consider that prize to be. I hope to stimulate some thinking over the course of the conference about the crucial role of the modeller in formulating solutions that decarbonise infrastructure and improve the long-term resilience of both society and the environment.

Bill Syme, TUFLOW

Changing Nature of Hydraulic Modelling

Bill Syme Bill has over 30 years’ experience working on riverine, estuarine and coastal studies, of which most have been in the flood hydraulics field. During this time, he has successfully managed and led a wide range of flood and flood risk management studies in Australia and overseas. The widely used TUFLOW hydrodynamic modelling software was developed by Bill starting in 1989.
Today, Bill is BMT’s TUFLOW Software Business Lead, managing TUFLOW Products and associated services, which won BMT a place in the prestigious 2020 Australian Financial Review’s Most Innovative Companies Awards. He also continues to provide specialist flood modelling and flood risk management advice and was the Project Manager for the award-winning Brisbane River Flood Study Hydraulic Assessment. 

Abstract

Whilst the equations that govern the physics of fluid flow have not changed, how our industry goes about applying those equations for hydraulic modelling is in a flux of change.
Firstly, the advancements in computational hardware have allowed us to increasingly break down or discretise the real world at finer and finer resolutions. This has brought great change and greater accuracy just like high resolution monitors have enhanced our visual experience of watching a movie at home.
Secondly, the other less clear change is how we go about solving the equations. For the flood modelling industry, we’ve seen progression from using the highly simplified 1D equations to the more realistic 2D equations, with often a combination of 1D and 2D being optimal. Whilst this was inevitable with the advancement of computers, what is not so transparent is how the 2D equations are solved.
Behind the easily produced and highly visual flood mapping images, there is a multitude of different approaches to solving the hydraulic equations, especially for the more complex 2D form. Many of these approaches, unbeknown to practitioners and decision makers, are often not fit-for-purpose.
However, for coastal and estuarine hydraulic modelling the proliferation or use of inappropriate 2D and 3D solvers is much less of an issue as accurate representation of the physics is paramount. The low friction, deep water environment requires good representation of inertia and turbulence, and therefore the solvers mostly used for coastal and estuarine hydraulics tend to be of high quality and accuracy.
When solving the equations for flood flows, where the flow is benign (i.e. slow moving), the more complex physical terms (e.g. inertia and turbulence) are often inconsequential, especially where the ground or surface over which the water is moving is rough and the bed resistance term (e.g. Manning’s equation) dominates. In these situations, nearly all hydraulic solvers can produce a reasonable representation of flooding given accurate input data.
Where flood flows are not benign (i.e. fast moving; forced to change speed or direction; or over smooth surfaces), the accuracy of 2D solvers varies greatly. This is for two primary reasons:
The numerical order of the solver’s mathematics. For example, where the hydraulics are not benign, a 1st order scheme will be numerically diffusive generating a compounding error resulting in inaccurate results compared with a 2nd order scheme.
The omission or poor representation of key physical terms such as inertia and turbulence.
A third reason exists that is related to model resolution, in particular the ability to utilise terrain and surface roughness data sub-cell. Whilst model resolution issues can be resolved by reducing the cell size smaller and smaller, this is rarely practical due to increased simulation times. The recent advent of sub-grid sampling (SGS) by some 2D solvers has greatly assisted with resolving this constraint.
The future of the flood hydraulic modelling industry is dependent on the industry using 2D solvers that are accurate under all flood conditions, not just when the flows are benign. The presentation backgrounds the evolution of hydraulic modelling and highlights current day innovations that are changing the flood modelling industry for the better.

Thomas Kjeldsen, University of Bath

Detecting change in UK flood data and implications for design flood estimation

Dr Thomas Kjeldsen is trained as a civil engineer (MSc, PhD) and has more than 20 years research experience focusing on mathematical and statistical modelling of hydrological and environmental systems, with particular emphasis on extreme events and sustainable use of water resources. He led the scientific development of current industry standard methods for risk-based hydrological design in the United Kingdom. Recent research includes studies of: estimating the probable maximum flood, the effect of urbanisation on catchment hydrology, and the development of a statistical extreme value procedure for regional and non-stationary analysis of flood events.

Abstract

The impact of past and future change in flood characteristics is of major importance for design flood estimation. In particular, the existence of change will potentially invalidate parts of existing methods and calls for new tools to support flood management. This talk will present the outcome of ongoing investigations into detection, attribution and implications for design flood estimation of trend and change in UK flood data. A new regional non-stationary statistical model based on the FEH pooling-group concept is presented. The new model is shown to provide more credible and robust estimates of design floods under change than possible using the corresponding single site analysis.

Dr Chris Whitlow, Edenvale Young

Modelling Natural Flood Risk Management in both the Urban and Rural environment

Dr Chris Whitlow is an expert in hydrometeorology, water quality, sediment transport and hydraulic modelling with more than 35 years postgraduate experience. Arguably his primary claim to fame is associated with the development of the ISIS (now Flood Modeller) software which is still the main software used to simulate river channels in the UK.
Chris has also published more than 30 technical papers and acted as a reviewer for the ASCE Hydraulics Division, the International Association of Hydraulic Research and the ICE Journal. He has also worked as a research fellow with the Universities of Birmingham and Nottingham and supervised or examined postgraduate level research at the University of Leeds, the Open University and the Institute of Hydraulic Engineering in Delft. Historically he was appointed to be a member of the Engineering and Physical Sciences Research Council College where he reviewed research application and reports.
Chris is also an experienced expert witness, having completed the Cardiff University Bond Solon course, and appeared in the High Court in November 2017 on behalf of the National Farmers Union.
He developed the first Environment Agency (EA) modelling strategy to draft level and was the technical advisor during the process to create the first National Flood Forecasting System, also for the EA. He also authored or co-authored EA documents such as the real-time modelling guidance and the extension of rating curves at gauging stations.
He retains a very strong capability in flood forecasting and led Edenvale Young in the development more than 50 forecasting models to support river and coastal flood warning in the cities of Manchester, Carlisle, Edinburgh, Perth, Leicester, Gloucester, York, Bristol and London among others.
Overseas Chris has worked on projects in the US, South Africa and Australia, he was the expert modeller for the World Bank funded CRIP project in Sri Lanka between 2016 and 2018. Work on the latter project was a key reason that led to Edenvale Young winning the 2018 NCE 100 Award for Innovation in Design.
Chris is also proud to be a director of the Bristol Avon Rivers Trust consultancy and has developed close relationships with various rivers trusts in recent years which have included working on projects to quantify the impact of Natural Flood Risk management (NFM) interventions as well as small storage areas, constructed wetland design, weir removals, river restoration and reconnection of watercourses.
.

Abstract

This talk focusses on the potential role of Natural Flood Risk Management (NFM) interventions to reduce flood risk in four catchments in Greater London ,namely the Salmons Brook, Pymmes Brook, Edgware Brook (Harrow Headwater) and the Rise Park Stream. The project was undertaken jointly with Thames21 with significant input from staff from the London Borough of Enfield.
Four new direct rainfall models were developed and calibrated to available hydrometric data or flood extent data. The NFM measures considered were different amounts of woodland planting, river restoration, multiple rural SUDS measures, multiple leaky dams and constructed wetlands.
The hydraulic modelling results showed the relatively high level of effectiveness in terms of flood peak reduction over a range of return periods of extensive woodland planting in the upper reaches of the catchments as well as rural SUDS and constructed wetlands.

Sharla McGavock & Quy Vu, Mott MacDonald

Machine learning for real-time decision support in Bangkok

Sharla McGovock ” is a principal hydrologist Mott MacDonald in Singapore, with a background in flood modelling, water resources management and hydropower. She is currently leading the flood studies for two urban flood resilience projects in Bangkok and Ho Chi Minh City within the FCDO Global Future Cities Programme, providing targeted technical support which supports sustainable, inclusive and equitable urban development.”

Quy Vu
“is a Data Scientist within Mott MacDonald’s Moata Division. He specialises in developing and productionising data science and machine learning products to save costs and increase efficiency in asset management and predict natural disasters. He has a proven track record of developing machine learning frameworks to support decision-making through data engineering and building data-driven models. For the Bangkok Flood Management Decision Support System pilot, he has developed a machine-learning based surrogate model based on a TUFLOW-ESTRY model to deliver real-time surface water flood forecasts which are available to Bangkok Metropolitan Administration through Moata, Mott MacDonald’ digital twin platform.”

Abstract

Mott MacDonald has delivered a pilot Flood Management Decision Support System (DSS) in Bangkok as part of the FCDO Global Future Cities Programme. The DSS, delivered in Moata Smart Water, provide a central, easily accessible and visual single source of truth to enable better informed operational and maintenance decisions before and during severe rainfall events. Rainfall observations from existing radars in Bangkok were improved using rain gauge and Vertically Profiling Radar data, facilitating automatic quick generation of accurate near real-time rainfall maps and operational Nowcasting. A real-time surface water flood forecasting system was developed by applying machine learning to a TUFLOW-ESTRY hydraulic model and using rainfall Nowcasts as an input. These observations and predictions are integrated in the DSS, with a spatial interface and dashboard for interrogation. Alarms provide early warnings to Bangkok Metropolitan Administration (BMA) to inform operational response, improving BMA’s ability to prepare for and respond to surface water flooding.
This presentation will provide an overview of the project including the radar improvements, hydraulic modelling and machine learning undertaken to deliver real-time flood forecasts.

Ellis Symons, TUFLOW

New TUFLOW Input/Output File Formats

Ellis Symons “Ellis is a senior engineer at BMT and has a background in hydrologic and hydraulic modelling for a number of studies within Australia using a range of software. Ellis is responsible within the TUFLOW team for GIS tool developments and was instrumental in the development of the TUFLOW Viewer QGIS plugin.  As well as GIS, Ellis also takes the lead developing and maintaining the TUFLOW utilities.  Ellis is a python expert and has developed a range of eLearning training courses for TUFLOW modellers from introductory courses to advanced python training.” 

Abstract

TUFLOW use of open GIS formats such as MapInfo Interchange Format (MIF) files and ESRI Shape files has allowed easy and rapid development of hydraulic models from commonly used GIS mapping formats.  In the last few years there have been developments of additional file formats which are increasingly being used as the go to format for mapping and Digital Elevation Model (DEM) data sources.  TUFLOW 2021 introduces direct reading/writing of popular open format vector and raster GIS data: GeoPackage and GeoTiff. This presentation discusses the development and testing of these formats, how TUFLOW interacts with them and the available options, and looks at the exciting potential that the database format, GeoPackage, offers to TUFLOW modellers in terms of speed, interoperability, versioning, and file management.

Ellis Symons, TUFLOW

Python Scripting – Improve Your Workflow Efficiency

Abstract

Python is one of the most popular programming languages currently available. It’s free, there’s plenty of free online training material, and a lot of key software such as QGIS, ArcGIS, Mapinfo, and cloud compute providers offer Python Application Programming Interfaces (APIs).  This makes Python perfect for use in combination with TUFLOW and it’s open input/output file formats. This presentation talks about how Python can be incorporated into your TUFLOW workflow to improve efficiency and quality control, as well as the new resources TUFLOW is producing for new and experienced Python programmers alike.

Ellis Symons, TUFLOW

QGIS TUFLOW Viewer and other Tools

Abstract

The availability of QGIS and the development of the QGIS TUFLOW plugin has provided TUFLOW with a graphical interface to build, run and analyse results for 1D, 2D and coupled models.  The QGIS TUFLOW Plugin is a very popular tool with over 50,000 downloads and a 5-star rating in the plugin repository. Along with tools such as the asc_to_asc.exe utility, it has become a part of most TUFLOW modellers workflow. Learn about the latest and future development within TUFLOW Viewer, the QGIS Plugin, as well the utilities and add-ins to other software including some exciting developments particularly relevant to UK users. 

Stream 1

Day Two

Dr Eric Gillies, CBEC Eco Engineering

Hydraulic, morphological and habitat modelling of a restored, dynamic, gravel bed river using TUFLOW FV

Dr Eric Gillies. For the last 8 years I have been technical director for CBEC eco engineering, a UK consultancy specializing in river restoration, leading both our field surveying and hydraulic modelling teams. We undertake flood modelling, natural flood management modelling, modelling of river restoration designs, bank and channel stability modelling, and sediment transport/morphodynamic modelling, using a range of tools including TUFLOW FV. Prior to working with CBEC, I had 16 years experience in computational and experimental aerodynamics/fluid dynamics, modelling a diverse range of phenomena from unsteady flows of rotary and fixed wing flight vehicles, bluff bodies, avalanche/debris flows, to bio-medical flows.

Abstract

We present the geomorphic and hydraulic evolution of a restored gravel bed river in the Scottish Highlands, using both annual 2D survey data, and TUFLOW FV hydraulic and morphodynamic/sediment models. The morphodynamic models are shown to reliably predict a growth in hydromorphic complexity, accurate thalweg evolution, sediment sorting and formation of gravel bars seen at the site. We further use TUFLOW sediment modelling together with modelled hydraulic variables (depth, velocity and Froude), to infer Atlantic salmon spawning habitat growth in the restored reaches during the winter season; and we use particle tracking of turbulent, invertebrate drift to infer that the spawning habitat is closely located to good salmon fry habitat in the spring and summer seasons. The presentation highlights the utility of TUFLOW morphodynamic modelling in reducing risk while restoring dynamic gravel river environments, and also demonstrates that it can be combined with additional post-processing to infer improvements to aquatic habitat, the end-goal of most river restoration.

Antoine Lebrun, Environment Agency & Kevin Haseldine, JBA Consulting

Environment Agency Updated Flood Modelling Standards

Antoine Lebrun
Antoine works as a Senior Technical Advisor in the Evidence & Risk (E&R), Skill & Quality Assurance team at the Environment Agency (EA).
He has worked in many different sectors of the water industry including water and sewage treatment, water transportation and flooding for over 5 years. At the EA Antoine main work streams are standards for fluvial and coastal modelling, creation and maintenance of modelling quality assurance processes and training of E&R colleagues. He also is a member of the “Large Project Review Group”, where he sits as an expert in modelling, assuring projects over 10million pounds.
Antoine is a Red Cross, Water Hygiene & Sanitation, engineer and have been deployed during the Greece refugee crisis and latest Ebola Virus Disease outbreak in Democratic Republic of Congo.
Antoine was the EA project manager on the Fluvial Modelling Standards project. He lead and shaped the project from scoping to completion as well as created and coordinated reviews of the documents. After completion, Antoine worked closely with the DEFRA publication team to create a more accessible version of the standards before their publication on gov.uk.

Kevin Haseldine
Kevin has worked at JBA for over 12 years and is a Principal Analyst in hydrology and hydraulic modelling.  He has managed many Environment Agency projects over the last seven years including the Vale of Pickering (150km of open watercourse), Model Package 19 (including five modelling projects) and the Fluvial Modelling Standards. He runs JBA’s Introduction to Flood Modeller training course and has been a mentor to many new graduates. Kevin became chartered with CIWEM in 2016.
Kevin was JBA’s lead author of the Fluvial Modelling Standards and developed the Excel based model review tool.  He organised the testing of the tool on 12 existing models and facilitated workshops with other NGSA consultants to contribute to and review the documentation and tool.

Abstract

To follow

Dr Gerald Morgan, Edenvale Young Associates Ltd

Representing upper catchment restoration schemes

Dr Gerald Morgan
Gerald is the Royal Academy of Engineering Research Fellow in Modelling for Natural Flood Risk Management and is also a director of Edenvale Young Associates. He was one of the pioneers in the UK of the distributed modelling approach using TUFLOW and has extensive experience of hydrological and hydraulic modelling and development across the UK.

Abstract

Afforestation and moorland restoration schemes are becoming more and more common across the country, but the accurate representation of these schemes in models continues to vary widely. In this workshop, Gerald will present the key hydrological effects of these restoration schemes, with an overview of the evidence for how these schemes affect flood risk. This will then be followed by a wider discussion with the workshop participants on how our modelling techniques could be improved.

Copyright © 2021 Edenvale Young Associates Ltd. Company No. 5910755

Registered Address: 30 Queen Charlotte Street, Bristol BS1 4HJ, UK