The future of flood hydrology in the UK

Floods have devastating effects on lives, communities and livelihoods, and the averaged economic risk of flooding in the UK is estimated (Sayers et al. 2020) to be £2.1 bn per annum. In its National Risk Register, the UK government has judged only pandemic disease and nuclear or biochemical attacks to have greater potential impacts (Cabinet Office 2020).

Rivers are the biggest source of flood risk, with £1.1 bn in expected annual damages (EAD). Surface water (pluvial) flood risk is £0.6 bn EAD, coastal risk £0.4 bn and groundwater £0.05 bn. Hydrology is therefore an essential foundation for evidence-based decisions about managing economic risks of at least an estimated £1.7 bn EAD. Alongside the economic assessment, 1.9 million people have more than an estimated 1/75 annual chance of being flooded, with the risk falling disproportionately in the most socially vulnerable neighbourhoods. Flood risk is assessed separately by the responsible authorities in each of the UK's nations. Here, and in what follows, UK-wide baseline data and future projections have been aggregated from the Third UK Climate Change Risk Assessment (CCRA3, Sayers et al. 2020).

Flood risk EAD in the UK has been projected to reach £2.7–3.0 bn annually in the 2080s under a global warming trajectory of +2 °C by 2100, or £3.5–3.9 bn under a +4 °C trajectory. Inland hydrological risks will continue to predominate in economic terms, although relative increases in risk will be largest for surface water (from £587 m to £1.2 bn in a +4 °C future) and coastal (£361 m to £1 bn) flooding. These projections assume that risk management measures, including land use planning and investment in flood and coastal defences, will continue at rates commensurate with current policy objectives. Enhanced flood management, including increased investment, appears to have the greatest potential to mitigate the projected increases in risk for fluvial flooding, which further highlights the importance of flood hydrology.

The risk-based approach to flood management in the UK (Hall 2014) demands scientifically robust analysis to prioritise resources where they will deliver best value for society (HM Treasury 2020). Economic projections (JBA and Sayers 2018; Environment Agency 2019) suggest that investments to mitigate the increasing risk are both feasible and economically justified, but with substantial residual risks that will be difficult to plan for, particularly in relation to surface water flooding, which may occur suddenly. The risk and investment projections are built on detailed modelling carried out for national flood management agencies, ultimately driven by hydrological analysis.

There is a long tradition of hydrological research and observation in the UK (Rodda & Robinson 2015), dating back at least as far as quantitative experiments by Edmund Halley in the seventeenth century (Deming 2021), and often intrinsically linked with practical problems in water management (McCulloch 2007, 2022). The water sector today reflects a complex mosaic of public, private and third sector stakeholders. Although the interconnectedness of the hydrological cycle has long been appreciated (Biswas 1970), a fragmentation of responsibilities (Pitt 2008) and the generally increasing depth of technical knowledge in science and engineering disciplines mean that methods applied in practice have become increasingly functionally specialised. Many methods routinely used in UK flood hydrology have roots in the 1960s–1990s. For example, the Flood Studies Report (NERC 1975) paved the way for the Flood Estimation Handbook (FEH, Institute of Hydrology 1999) and its derivatives in current use. The FEH methods are widely used, including in the national risk mapping that underpins the risk assessments cited in the opening paragraphs of this paper. Yet some choices made in the past, for example assumptions of spatial uniformity or temporal stationarity, could be questioned when viewed with a contemporary perceptual understanding of hydrology (Wagener et al. 2021), or when taking account of advances in observations (Beven et al. 2019) or evidence of change in UK flood data (Faulkner et al. 2020; Hannaford et al. 2021).

Regionally and nationally significant events in Britain (including in 1998, 2000, 2005, 2007, 2009, 2013–16, 2019 and 2020) have highlighted the impacts of flooding (Environment Agency 2018). Meanwhile, an increasing emphasis on whole-system thinking points to a need for integrated models to support improvements in flood resilience (Cabinet Office 2016). In recent years, transformations in our capacity to share information rapidly through digital communications, an increasing appreciation of the importance of community ownership of risk, and perhaps also an expansion of educational and training opportunities in flood risk management have stimulated a wider range of demands on hydrological data and methods.

Those demands stem from many areas of flood management, including: the design and maintenance of flood defences, flood risk mapping, risk assessments for investment or development planning, the design and operation of forecasting and warning systems, reservoir safety, sustainable drainage systems, the evaluation of nature-based flood management and understanding of the impact of environmental change on flood risk.

This paper describes how the evolving demands for hydrological analysis and advances in scientific knowledge have prompted a comprehensive reappraisal of research and innovation in UK flood hydrology, culminating in the publication of a roadmap for the next 25 years (Environment Agency 2022), available from https://www.gov.uk/flood-and-coastal-erosion-risk-management-research-reports/flood-hydrology-roadmap.

The roadmap is a community initiative, building on a strongly collaborative approach that will be introduced in the following section. We then describe the objectives of the roadmap project, and how it fits with related initiatives both in the UK and internationally. The following sections set out the methodology for development of the roadmap and summarise its contents. We then discuss how the roadmap is being implemented, what is needed to ensure its long-term success, and aspects of its development that may be transferable to other settings. We conclude with an overview of the roadmap project's key outcomes.

Flooding and coastal risk management (FCRM) are considered together in UK policy. No recent estimate is available of total spend on research or innovation in flood hydrology, although analysis in 2011 found that over the preceding decade there had been annual public spend of between approximately £7 and £14 m (£12.5 and £25 m at 2021 prices) in FCRM research (Moores & Rees 2011). The private sector also invests in FCRM research and innovation, but this has not been quantified.

An important feature of flood risk management in the UK is the history of cooperation between universities, the public sector and private industry. Collaborative research and development programmes (including those in footnotes^1,2,3,4) have helped to promote exchanges of knowledge about the scientific and practical drivers for continuing developments in flood hydrology. Research programmes with a focus on flood risk have included Flood Risk from Extreme Events (FREE, Hardaker & Collier 2013), the Flood Risk Management Research Consortium (FRMRC, Environment Agency 2021) and Flooding From Intense Rainfall (FFIR).⁵ This coordination helps support communities of practice, enhanced through knowledge exchange networks including professional societies, notably the British Hydrological Society (BHS), a volunteer organisation with more than 900 members, the British Geomorphological Society and chartership institutions in water management,⁶ civil engineering,⁷ meteorology,⁸ geology⁹ and geography.¹⁰

Recognising the broad drivers for change discussed above and the long-term (∼10 to ∼100 years) influence of flood hydrology on infrastructure and land-use plans, the flood hydrology roadmap was initiated in 2018 through a research and development programme¹ run jointly by public risk management authorities and government departments in England and Wales. First, a project board was established to take overall responsibility. The board was supported by, and worked closely with, a steering group drawn from the regulatory, academic and non-profit organisations. For brevity, the board and steering group will be referred to in this paper as the ‘project team’.

Broad initial objectives were set by the project team, framed as ambitions that the roadmap should:

• take a 25-year view,

• be inclusive and community-owned,

• combine scientific credibility and practical utility,

• consider inland flood hydrology (flood risk from rivers, surface water, groundwater and reservoirs),

• consider both forecasting the near future and longer-term risk,

• enable and drive change (e.g. in research, guidance, data or organisations).

Recently, there have been several important consensus statements about challenges in hydrology. A landmark study is the international ‘Twenty-three unsolved problems in hydrology’ (23 UPH) initiative (Blöschl et al. 2019) to identify major scientific problems, motivated by a need for stronger harmonisation of research efforts. The 23 UPH project did not have a specific focus on flooding, but, like the roadmap, it involved iterative co-creation through a blend of digital channels, in-person meetings and working groups.

Three of the UPH relate specifically to hydrological extremes, either droughts or floods, and one relates exclusively to exceptional runoff produced by rain-on-snow events. The 23 UPH in general reflect fundamental questions about variability, scaling, process interactions, empiricism, modelling and the role of water in society. Hence, there are connections with flood hydrology embedded within many of the remaining 20 UPH. Blöschl et al. (2019) concluded that hydrological applications and fundamental research reinforce each other. They viewed the UPH as proof of concept that a broad consultation process was feasible and welcomed by the hydrology community. Both findings also characterise the roadmap. The 23 UPH coincided with the International Association of Hydrological Sciences (IAHS) scientific decade 2013–2022 denoted ‘Panta Rhei – Everything Flows’ (Montanari et al. 2013), which has been dedicated to research about changes in hydrology systems and their relationships with a rapidly changing society.

In the UK, a working group was formed in 2018, under the auspices of the BHS, to debate and make recommendations about the future of UK hydrology, leading to two journal papers (Beven et al. 2019; Wagener et al. 2019). Most individuals who participated in the BHS working group also contributed to the roadmap. Additionally, the roadmap team examined the working group's outputs to understand areas of strong or weak alignment with the emerging roadmap, a process that informed the final action plan.

Another source of advice that informed the roadmap was a UK flood resilience review (Cabinet Office 2016), which made the case for integrated modelling, encompassing both physics-based and statistical approaches, regular updating of risk assessments and tests of resilience based on extreme event scenarios. An earlier report commissioned by the Government Office for Science (Royal Society 2015) highlighted the need for improved observations of natural hazards, including flooding, to increase the UK's resilience. This prompted a study (from 2020 to 2022) to establish the requirements for national Floods and Droughts Research Infrastructure (FDRI). The aim of the FDRI is to transform research capability to improve flood and drought forecasting, planning, incident response and management. The FDRI study engaged with a broad range of stakeholders from public, private and non-profit sectors, using similar methods to the flood hydrology roadmap, and producing several proposed investment options, which are being taken forward into a business case for funding at the time of writing. The Reservoir Safety Research Strategy (Environment Agency 2016) also highlighted needs for research on extreme rainfall and runoff, which have informed the roadmap.

The initial consultation questionnaire represented a form of purposive sampling. Subsequent stages of engagement were designed to achieve greater depth and breadth, with a larger pool of participants being encouraged through promotion in professional newsletters, email lists, meetings and webinars. No fixed target was set for the number of respondents; instead, the aim was to ensure that anyone with an interest in hydrology and flood risk management in the UK had the opportunity to contribute through at least one of the engagement processes.

An initial questionnaire was formulated by the project team. Its structure is shown in Table 2, which details the motivation for each section. The questionnaire was sent to 52 individuals or groups, selected in line with the Environment Agency's internal guidance on engagement. Responses, totalling more than 50,000 words, were submitted in free-text format. To avoid constraining the solution space too soon, the questionnaire was intentionally directed towards mapping the stakeholder community and identifying problems, needs or opportunities, rather than identifying specific actions or solutions. It was designed to inform a wider discussion at the subsequent workshop, enabling the refinement of problem areas and the co-creation of proposed solutions and actions.

The project team parsed the raw questionnaire responses, collating them into four themes: ‘ways of working’, ‘data’, ‘methods’ and ‘scientific understanding’, which provided a foundation for the workshop. Alongside this subjective process, a machine learning approach was applied to the questionnaire responses to seek out topics with distinctive meanings as groups of key words, in this case, the 10 most frequent words associated with each of four putative topics. The word groups discovered using machine learning aligned well with the themes chosen by the project team (Environment Agency 2022, Appendix I), giving some assurance that the choices were evidence-based and not strongly biased by the backgrounds of the project team.

The workshop aimed to build ownership of the roadmap among influential stakeholders, and to start creating its content in terms of a vision for the future, analysis of perceived needs, and the actions required to meet those needs. Over 2 days, 50 individuals from 27 organisations generated draft vision statements and 73 potential actions grouped into 16 clusters. For example, one cluster of actions was ‘Improve access to flood hydrology data’; see Environment Agency (2022, Appendix C) for a comprehensive account. The 73 actions were further consolidated by the project team into 38 potential areas for future work, which formed the basis of the next step, an online survey.

Four task groups were established in March 2019 to summarise the current state of knowledge with respect to reservoirs, groundwater, surface water and fluvial flood risk, considering both forecasting and flood risk management planning perspectives. The groups comprised volunteers with expertise in each topic from academic, government and non-government organisations. Their reports (Environment Agency 2022, Appendix E) helped to consolidate the evidence gathered during the questionnaire and workshop and were used by the project team as baselines in developing proposals for future improvements.

With input from the prioritisation task group about the relative importance and scheduling of the 11 initiatives, the project team developed a draft action plan during 2020.

Two further task groups were established in parallel to help finalise the roadmap; a ‘delivery partners’ group advised on the timing of actions and who could fund them, while an ‘enablers’ group advised on how contributions could be made in other ways. Advice from the two groups was gathered during four workshops in early 2021 and supported the project team in drafting a final roadmap action plan including estimated costs.

The development process and draft action plan were previewed through a public webinar hosted by the BHS in March 2021 and attended by nearly 500 people. Following this webinar, the draft plan was made available on request and eight sets of comments were received to feed into the final roadmap.

The roadmap has highlighted the breadth and depth of the stakeholder community involved with flood hydrology in the UK. Figure 6 shows the distribution by sector of organisations represented throughout the entire co-creation process. Nearly half (45%) of engagements were with public sector organisations, reflecting the regulatory and policy landscape. Private industry and academia were the next largest groups, representing 28 and 23% of engagements, respectively, and providing reassurance about the representation of both the research community and practitioner stakeholders.

Figure 7 shows the distribution of technical interests among those engaged during creation of the roadmap. Both long-term analysis of river flood flows (‘flood estimation’) and near-future prediction (‘forecasting’) were important areas of technical interest. While fluvial hydrology was the most common primary interest, surface water, reservoirs and groundwater were all strongly represented.

The community's overall vision for the future of UK flood hydrology is that:

• during the next 25 years society will have improved hydrological information and understanding to manage flood hazard in a changing world;

• flood hydrology and whole-system process understanding will be underpinned by excellent evidence with quantified uncertainty;

• leadership and collaboration are crucial to achieving this vision.

The vision embodies the different dimensions of the roadmap, encompassing the importance both of science and applications. Specific references are made to uncertainty and holistic systems-based thinking, which is reflected throughout the action plan at the interfaces between flood hydrology and related environmental, physical and social systems.

The final part of the vision highlights the importance of leadership and partnerships, which reflects both the community ownership of the roadmap, achieved through the co-creation process, and the development of the ways of working theme. Early priorities within this theme include the establishment of a governance structure and a scientific and technical advisory group to identify funding opportunities, to promote the delivery of projects and to review progress against the roadmap action plan.

The roadmap has an estimated budget requirement of £110 m (present value, PV, of £74 m) over 25 years. Adjustments to account for optimism bias in cost and timing have been calculated using methods detailed in Environment Agency (2022, Section 4.2), giving upper estimates of £165 m (PV £111) costs and a 37.5-year programme. The best estimate average annual funding requirement is £4.4 m, with a peak in 2032 in the range £9.9 to £14.9 m (after optimism bias adjustment), and will need to be met by, and coordinated across, multiple funding bodies. Already, the Environment Agency in England has preliminarily allocated £6.9 million of funding over 6 years, from April 2021, to begin implementing the roadmap. Additional funding for science, and its translation into practice and policy, will be required to deliver the roadmap in full.

There are few estimates of the economic value of flood hydrology, or, specifically, research and innovation in flood hydrology. The annual economic benefits of the FEH (Institute of Hydrology 1999) were estimated in 2006 (CEH 2020) to be between £8.3 million and £31.3 million (£11.2 and £42.2 m at 2021 prices), while the reduction in damages from flood early warning systems, including both meteorological and hydrological components, was estimated by NERC (2015) at between £76 million and £127 million (£86 m and £145 m at 2021 prices). The above assessments are not comprehensive evaluations of all the benefits of flood hydrology, which may be larger. Even so, the figures demonstrate the potential for the roadmap action plan to deliver benefits that considerably exceed its costs. An action within the roadmap (‘W4: Identify ways to improve investment in flood hydrology’) will address in detail the business case for further investment in flood hydrology by quantifying the contributions of hydrology and hydrometric data to flood risk management outcomes, while other actions will explore the value of the (flood) hydrology skills base and provide for evaluation of the benefits realised through the roadmap.

Individual actions and their outcomes are detailed comprehensively within the published roadmap. Rather than duplicate the details, we will discuss the key characteristics below.

Flood hydrology is often described in terms of physical domains (e.g. surface water, rivers, urban), scales (e.g. long-term risk, real-time forecasting, small catchments) or sectors (e.g. regulatory, academic, private). The four themes that emerged through the roadmap's co-creation process (ways of working, data, methods and scientific understanding) instead provided a useful and parsimonious architecture of flood hydrology, both as a technical discipline and profession, that helped enable the roadmap to develop along multi-disciplinary and multi-sectoral lines.

As expected, the roadmap includes strong emphases on observations, data and modelling. Data from past events cannot be changed (unless systematic measurement errors are discovered and corrected), but it may be possible to assess uncertainties better to help decide on priorities for future observations, either to improve current estimates or to commission new types of measurement. It is notable that uncertainty features in the roadmap as a fundamental topic, along with an open philosophy with respect to data, methods and models. Future environmental change is a primary example of how knowledge uncertainties may influence decisions that shape flood resilience.

Funding is inevitably a critical challenge for any programme as ambitious as the roadmap. The allocation of funds by the EA (see above) marks a significant early outcome. Further sources in future could include other UK regulatory or public sector bodies, UKRI, European and other international programmes, private industry and charities. By combining scientific research and applied needs in a coherent package, the roadmap will help to inform and substantiate the pathways to impact through which future research funding will deliver wider benefits. A critical element of this is likely to be continued efforts to understand, quantify and reduce uncertainties.

The roadmap recognises the importance of sustaining the flood hydrology community if the 25-year vision is to be realised. It represents a coalition of interests from across hydrology, flood management and allied disciplines. The roadmap offers a focal point for this community; its future evolution is likely to depend in part on funding, combined with continued voluntary activities of societies such as the British Hydrological Society, which has agreed to support the implementation of the roadmap through engagement activities. The vision is not static; rather, continuing review and opportunities for updating of the roadmap are foreseen through the ‘ways of working’ theme.

Returning to comparisons with international perspectives, in Supplementary Information S1, we present a text-mining analysis to map between the 23 UPH and the UK flood hydrology roadmap actions. This shows strong alignment across many topics. The analysis indicates that the international perspective perhaps places more explicit emphasis on links and feedbacks between social and physical systems (see, e.g., Di Baldassarre et al. 2015). This may in part reflect the fact that the roadmap was from the start embedded within the wider practice of flood risk management within the UK, which is, fundamentally, concerned with interfaces between physical and social systems. Two of the three key pillars for progress in hydrology identified in the 23 UPH were ‘generalisation and open data/models’ and ‘activities organised around integrated questions’; the UK flood hydrology roadmap is aligned with both. Blöschl et al. (2019) gave equal importance to the substance of the 23 unsolved problems and to the process of community-level learning involved in their development, and they advocated for similar consultations to be carried out in future. This finding is perhaps echoed in the roadmap's ‘ways of working’ theme, which aims to sustain a community of science and practice in flood hydrology. However, it is interesting to reflect on differences in framing of this concept; the roadmap process is perhaps more oriented towards outputs and applications, rooted in scientific progress, whereas the 23 UPH is framed in terms of collective learning. In taking the roadmap forward, it may be beneficial to identify explicitly the opportunities for community-level learning.

The third key pillar of the 23 UPH initiative may offer a useful challenge in the future implementation of the UK roadmap. It relates to risk and reward. The roadmap has been developed with explicit consideration of opportunity costs, delivery risks and optimism bias. Perhaps reflecting similar considerations, the 23 UPH project recognised that progress is often incremental and most of the UPH might not be ‘solved conclusively but can likely be realistically advanced in the next couple of decades’ (Blöschl et al., 2019 Section 4.2.1). However, the authors also gave explicit consideration of high-risk/high-gain activities, noting that apparently ‘outrageous’ scientific hypotheses (difficult as they are to define) can turn out to be true. A hydrological case in point may be the existence of preferential flows in soils (Beven 2018), perhaps as an element of a fundamental reconsideration of the natural hysteresis in hydrological systems (Beven 2019). The 23 UPH could prompt a useful debate about where the balance should lie between such apparently riskier ideas, and a professional culture that tends to favour ‘solid, proven methodologies’.

We noted earlier that a stakeholder analysis was maintained throughout the roadmap's development to help reduce the scope for bias. This was done to ensure representation within the roadmap across different technical disciplines and from different categories of stakeholder, including by sector (e.g. public, private), function (e.g. service user, service provider, regulator), geography (representation across UK regions, international) and impact or influence (e.g. directly affected, indirectly affected, able to affect the work).

Equality, diversity and inclusion (EDI) emerged among the principles to be embedded within the roadmap. This is reflected in the emphasis given to EDI within roadmap actions, such as in the establishment of a technical advisory group for UK flood hydrology (Action W2). Future stakeholder engagement could include an explicit EDI plan, and usefully gather information about the characteristics of the community.

It is a challenge to understand the professional community around flood hydrology because it is varied and does not necessarily speak with one voice (as evidenced by the many different priorities identified in the roadmap). The inter-disciplinary nature of hydrology means that the boundaries between flood hydrology and related disciplines are difficult to draw. The roadmap engagement was framed as a practical exercise rather than as a research project. Resources were not available to fund research about the engagement process itself, but this perhaps missed an opportunity to gain additional insights. An initial review of the representation of different disciplines and groups within similar national or international initiatives (not only in hydrology) might help steer the initial consultation in future scoping exercises.

Stakeholders were not asked to identify their backgrounds with fixed disciplinary categories precisely because of an awareness of the inter-disciplinary nature of flood risk management and the consequent risk of reductionism. However, their backgrounds spanned multiple disciplines including, in alphabetical order: climate science, ecology, economics, engineering, geology, geomorphology, meteorology, policy and social sciences. Input from the meteorology and climate science communities was strong with participation from the Met Office (UK), incorporating perspectives on integrated environmental modelling from the wider Unified Model Partnership¹¹, from the Bureau of Meteorology (Australia) and from the European Centre for Medium-Range Weather Forecasts (ECMWF). Better integration across different sectors and disciplines was ranked 17th most important out of 38 potential work areas in the online survey, and inter-disciplinary issues feature explicitly in Actions W5 (Build hydrological skill, esteem and value), D3 (Assess the potential of new observational technologies to improve flood hydrology), S7 (Develop proposals for research programmes) and S8 (Potential research programmes).

The flood hydrology roadmap is a significant step in the evolution of hydrology within the UK, both as a scientific and an applied profession. It has brought together different sectors and disciplines to produce a costed, long-term plan with shared ownership and a clear vision. Its origins reflect a diverse need for applications of hydrology in flood risk management and related activities, and an engaged research community.

The roadmap captures the energy and ambitions of a large, representative coalition of individuals and public-, private-, academic- and third-sector organisations. Although its original stimulus came in large part from an applied perspective, it recognises the importance both of scientific progress and of practical drivers. Comparisons with other national and international hydrological scoping initiatives show much commonality with the roadmap, while suggesting it may be useful to explore further the interfaces between flood hydrology and water quality, health and social science. The roadmap includes mechanisms to enable its action plan to adapt and evolve to address any scientific gaps or changes in context that become apparent.

In total, the roadmap embodies a substantial, coordinated intellectual effort by more than 270 people over 47 months. It has been developed using multiple modes of engagement, an approach that has succeeded in galvanising and bringing together a community of differing interests. The multi-modal approach to engagement was particularly important in enabling the roadmap to continue its development despite disruptions caused by the COVID-19 pandemic. The roadmap presents an ambitious, 25-year programme that has already helped to enable significant new investment into the improvement of flood hydrology within the UK. By publishing this commentary in Hydrology Research, an international journal, we hope to have placed the roadmap in context, to have highlighted insights from the headline report and its appendices and to encourage further exchange and learning between the UK flood hydrology community and international communities of practice.

We are grateful to Lucy Barker for help with British Hydrological Society membership data. The published roadmap contains acknowledgements, too numerous to be reproduced here, detailing individual contributions to the co-creation process described in this paper. We are grateful to the Editor and reviewers for their guidance and comments.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.