Research and analysis

Review of the research and scientific understanding of drought: summary report

Published 28 November 2023

1. Chief Scientist’s Group summary report

2. Research at the Environment Agency

Scientific research and analysis underpins everything the Environment Agency does. It helps us to understand and manage the environment effectively. Our own experts work with leading scientific organisations, universities, and other parts of the Defra group to bring the best knowledge to bear on the environmental problems that we face now and in the future. Our scientific work is published as summaries and reports, freely available to all.

This report is the result of research led by the Environment Agency’s Chief Scientist’s Group.

You can find out more about our current science programmes at https://www.gov.uk/government/organisations/environment-agency/about/research

If you have any comments or questions about this report or the Environment Agency’s other scientific work, please contact [email protected].

Dr Robert Bradburne
Chief Scientist

3. Acknowledgements

The authors would like to thank Ms Stephanie Cole (Environment Agency Chief Scientist’s Group) for her help with project management, and Mr Richard Davis (Environment Agency Water Resources), Mr Richard Amos (Dŵr Cymru Welsh Water, and UK Water Industry Research), and Dr Geoff Darch (Anglian Water) for their contributions to the expert workshop.

4. Executive summary

This project reviewed the current scientific knowledge about drought in the UK, including how drought may alter due to climate change, and the implications for both the catchment environment and the management of water resources, now and in the future.

Droughts are complex events and processes that can vary in duration, timing, location, and severity. These differences can cause different impacts on the environment, farming, and water supply. Droughts are infrequent, which means there is limited data and experience of impacts and responses. This is further complicated by uncertainty in how climate change may affect drought.  This review focused on three themes: the physical processes that drive droughts; the impacts of drought; and the management of drought. Each theme was further divided into separate topics, and a review of each was undertaken by experts in that topic. We worked with over 40 academics from 13 different universities, research institutes, and consultancies.

Authors were asked to review the existing literature and comment on what is known and not known about drought for their specialist topics. A workshop was held to share information among the group, to promote discussion and develop consensus.

The review draws together and consolidates the current body of knowledge on drought and drought impacts and highlights the extent of our current understanding. It has identified gaps in our knowledge, both topic specific and recurring themes that cut-across drought management and research:

• Changing nature of drought
• Modelling, monitoring and data recovery
• Catchment processes and dynamics
• Development and ending of a drought event
• Drought as a social construct
• Impacts and interactions
• Vulnerability and resilience

We want to build on the findings of this review and identify where further research could deliver benefits for drought management and resilience, both now and in the future.

5. The need for a review

5.1 Background to the problem

The Environment Agency is responsible for safeguarding water resources in England and protecting the environment, including during drought. Droughts occur naturally when a period of low rainfall creates a shortage of water, reducing water availability, with a consequent effect on the environment and people.

Droughts are complex events and processes that can vary in duration, timing, location, and severity, with different impacts on the environment, farming, and water supply. Droughts are infrequent, which means that there is only limited experience, data and understanding. For example, the large-scale atmospheric drivers of drought remain poorly understood.  The nature of UK droughts is expected to change as the climate changes, but there is uncertainty about how, including the magnitude of any changes.

Consequently, our understanding of drought and how to manage or respond are limited. Increased knowledge would improve drought planning and management in England.

5.2 Current research into drought

There have been several recent initiatives in UK drought research, including projects undertaken as part of the Natural Environment Research Council’s (NERC) UK Drought and Water Scarcity research programme, and scientific meetings such as The Royal Society’s “Drought risk in the Anthropocene”. There have also been projects commissioned by the Environment Agency and Defra, and water companies through UK Water Industry Research (UKWIR). The research has often focused on specific questions, and the knowledge created is held across a wide range of different organisations, often in unpublished reports.

5.3 Aims and objectives

This project reviewed existing research into drought in the UK, including the drivers, impacts and management of drought, both now and in the future. The report provides insights into the state of the current understanding, including the knowledge gaps. The review brought together experts to share knowledge that is currently held across different academics, consultants, and practitioners. It provides a focus for the transfer of existing ideas into practice and an outline for further research to meet scientific and operational needs.

6. The review method

6.1 Overview of the approach

The review focused on three themes: the physical processes that drive droughts; the impacts of droughts; and the management of droughts.

Each theme was further divided into separate topics and each topic was the subject of an individual review undertaken by experts providing answers to questions about drought in the UK, including:

• What is the observed experience of drought?
• What is the current understanding about the causes and variability of droughts?
• What is the current approach to drought management?
• How do we currently model drought and how might droughts change in the future?
• What is known about drought impacts now and expected in the future?

The review group included more than 40 academics from 13 different universities, research institutes, and consultancies. Authors were encouraged to review the existing literature on a specified topic and to comment on what is known and where there are knowledge gaps and uncertainties. To gain the widest range of ideas, authors were given freedom to form their own views without editorial control from the Environment Agency.

6.2 In-person workshop

In addition to their individual reviews, authors also attended an in-person workshop in early February 2023. They were joined by practitioners from the Environment Agency and representatives of UK water companies.

The workshop provided an opportunity for the authors to share their work and to ask questions of each other. In so doing, the group was able to debate both what is known about drought, and the gaps in understanding, both current and future. This report summarises the main themes and findings.

6.3 Findings of the review

The following sections are a summary of the discussions and views expressed by the authors both individually and collectively.

For details of the review topics and authors, see Appendix A. For the reviews see the Annex to this report published on GOV.UK.

7. Extent of current knowledge

The review identified the extent of the current scientific understanding of drought. The main findings are summarised here by theme and individual topic.

7.1 Physical processes of drought

Rainfall and atmospheric conditions

• Climate drivers of low rainfall. Low rainfall is often associated with anticyclonic blocking, which has different impacts in summer and winter. Blocking is linked to the strength of the North Atlantic jet stream (either being deflected northwards towards Iceland or southwards towards the Azores) and North Atlantic Oscillation (NAO). Different patterns of variability in atmospheric circulation govern variation in regional rainfall in the UK.
• Historical droughts. Historic droughts (e.g., long droughts of 1890-1910) can differ from those experienced in the past few decades
• Sub seasonal forecasts. Skilful dynamical seasonal forecasts of the wintertime North Atlantic Oscillation are now possible.
• Climate model projections and biases. Climate model projections indicate warmer, wetter winters and hotter, drier summers for the UK. The World Climate Research Programmes Fifth and Sixth generation Coupled Model Intercomparison projects (CMIP5 and CMIP6) models do not always simulate patterns of atmospheric circulation accurately, which has implications for assessing the impacts of climate change on low rainfall and thus droughts in the UK.

Surface water

• Historical droughts. There is a good understanding of the most severe hydrological droughts that are likely for given drought types (short or multiannual) and how this varies around the country.
• Indicators. There have been advances in the development of indicators and indices to characterise hydrological drought, including threshold methods and standardised indices. There is a reasonable understanding of how to apply these indicators in a UK setting, and generally when and where they work and for what purposes.
• Drought propagation. There is a good general understanding of how the propagation of meteorological to hydrological deficits varies around the UK and some understanding of how this is influenced by catchment characteristics.
• Trends in river flows. In the last 50 years, winter and autumn river flows have increased, and spring flows have decreased. Summer shows more mixed trends. These trends are in line with changes in rainfall and evapotranspiration. Low flows have generally increased in the north and west of the UK and decreased in the south and east – but trends are statistically significant for only a few catchments.
• Climate drivers. The NAO is an important driver of streamflow variability with a strong positive association between the NAO Index and river flow in the winter months, especially in northern and western areas.  Atmosphere-ocean drivers in both the Pacific (El Niño-Southern Oscillation (ENSO), and Pacific Decadal Oscillation (PDO)) and in the Atlantic (Atlantic Multi-decadal Oscillation (AMO) and Atlantic sea surface temperatures (SSTs)) can influence hydrological droughts.
• Anthropogenic drivers. Human-made influences, such as abstraction and land use change, can exacerbate or mitigate hydrological drought. However, it remains challenging to quantify the impact of a given influence on hydrological drought – there have been many studies aimed at detecting the effect of various interventions, but conclusive proof of the effect is often elusive – in part due to a lack of readily available datasets of artificial influences.

Future surface water

• Climate projections. Climate projections show nationwide increases in temperature, increases in summer potential evapotranspiration, and a reduction in summer rainfall, all factors that could result in declining river flows. However, projections also indicate increased winter precipitation, heavier summer downpours and an increased chance of high intensity rainfall events, leading to a complex picture for changing river flows and hydrological droughts.
• Hydrological projections. Hydrological projections tend to show reductions in median flows, summer flows and low flows across the UK, although with large uncertainties and local variations. Most projections indicate an increase in the severity of future drought events, although changes to other drought characteristics such as intensity and duration vary regionally. Southeast England may emerge as a hot spot for future multi-year droughts, with increases in both drought intensity and duration.
• Top-down approach. Most hydrological climate change impact studies follow a ‘top-down’ approach using a climate-hydrological modelling cascade. This can cause a growth in uncertainty through the model chain and unlike bottom-up or vulnerability-led approaches, focuses the results only on the outcomes possible from the original climate projections.

Groundwater observed and projected

• Drought in groundwater. Groundwater droughts are defined as a below normal groundwater level that can either be characterised using a threshold or standardised time series.
• Groundwater drought development. In natural settings, groundwater droughts develop as a function of catchment setting and aquifer properties that determine the responsiveness and memory of an aquifer. This translates into the drought duration and intensity following from a deficit in recharge, resulting in different spatio-temporal footprints for groundwater droughts.
• Long-term changes. Groundwater droughts are influenced by several decadal-scale changes, as evidence shows the influence of atmospheric circulation and changes in water resource management practices.
• Monitoring. Near-real time drought conditions for groundwater for 40 locations are published on the UKCEH water resources portal. Drought conditions are commented on in the Hydrological Summaries and Hydrological Outlooks that are currently available as static monthly PDF documents.

7.2 Impacts of drought

Water quality

• Complex effects. The physical, chemical, and biological impacts of drought are strongly interlinked and vary by waterbody and across catchments depending on the hydrology, morphology, catchment land-use and pollution sources.
• Diffuse inputs. Drought reduces diffuse, precipitation-related inputs of nutrients, suspended sediments, pesticides, herbicides, and road runoff.
• Low flow velocities. The slower speed of water at low flows will increase the rate of in-channel deposition and reduce dissolved oxygen levels and water quality.
• Groundwater. Groundwater nutrient input (which can be a major source of nitrate) can reduce as the water table falls, but the relative proportion of groundwater in surface water may increase.
• Sewage effluent. Rivers dominated by nutrient inputs from sewage treatment works effluent have increased nutrient concentrations and highest solutes, pollutant, and metal concentrations during drought low-flow periods.
• Metal concentrations. Droughts usually result in increased metal concentrations in both rivers and lakes. But pollution from mines will vary depending on groundwater levels.
• Algal blooms. The long flow residence times and concurrent increased water temperatures and nutrient concentrations can result in high algal and bacterial biomass and rapid biogeochemical cycling in rivers and lakes, reducing water quality and dissolved oxygen concentrations.
• Community changes. Low flow conditions in rivers can shift phytoplankton community towards smaller algae such as cyanobacteria, which often dominate when river water temperatures are high. Increased phytoplankton growth in rivers and lakes can cause a shift in plant community composition or loss of plants and associated adverse impacts on the invertebrate and fish communities.
• Lakes. Reduced lake flushing rates and higher water temperatures associated with droughts can increase the incidence and duration of lake thermal stratification, and the related release of pollutants and nutrients from bed sediments.
• Heavy rainfall during drought. Rainfall events following periods of dry weather can rapidly mobilise deposited sediments, nutrient, metals, and chemicals from the catchment and in river sediments.

Water temperature

• High river temperature extremes were noted during nationally iconic droughts (e.g., 1976, 1995, 2018). However, wider drought-river temperature associations have been inconsistent, reflecting the influence of environmental controls beyond low-flows experienced during droughts on river water temperature.
• River water temperature is influenced by:
  • energy flux dynamics
  • reach-scale habitat conditions, including hydraulic (e.g., residence time) and physical conditions (e.g., riparian vegetation coverages, wetted perimeters)
  • water source contributions (surface water and groundwater)
  • human influences (pressures and management activities)
• Riparian vegetation shading effects are a complex interaction between channel width, gradient, orientation, aspect, tree height, vegetation density and functional properties and solar geometry.
• Residence times. River systems with greater residence times, including those with shallow gradients (e.g., lowland environments, plateaus) or notably wide or deep systems (e.g., online ponds, unconfined reaches), are more susceptible warming during drought.
• Human influences. Four main human influences that affect river water temperature during drought are riparian vegetation modifications, flow regulation, water abstraction and channelisation (physical modifications to river channels). These often occur in combination at the same location.
• Afforesting slower flowing headwater systems can be most effective in reducing high river temperature extremes during low-flow conditions.
• Key changes to thermal drivers during droughts based on a process-based understanding can be conceptualised.

Riverine ecosystems

• Drought effects. The biodiversity and biomass of aquatic communities including plants, invertebrates, and fish decrease during drought. Responses vary within and among rivers and communities and are exacerbated by human pressures.
• Drying. Drought-driven streambed drying events cause severe declines in biodiversity.
• Species loss. Drought increases the risk of local or larger-scale species extinctions, with droughts and heatwaves interacting to exacerbate losses.
• Priority species. Species of conservation concern, including habitat-forming plants such as water crowfoot (Ranunculus) and salmonid fish, are among those at risk of population decline and loss.
• Recovery. Post-drought recovery of biological communities can be rapid—but some species take years to return. Moreover, reports of rapid recovery are often snapshots that lack pre-drought data, fail to track recovery to completion, and characterise communities already shaped by disturbances.
• Resilience. Refuges are key to supporting post-drought recovery. Communities in relatively natural rivers with diverse refuges are more resilient to drought than those exposed to greater human impacts.

Soils

• Impacts of droughts on soil. The main processes associated with water movement and processes in soils are understood, as are some of the impacts of extreme drying and re-wetting on soil physical, biological, and chemical properties.
• Soil Types. Different soil types will respond differently to drought conditions and the spatial distribution of soils within a catchment leads to complex patterns of recharge and discharge when drought conditions end.
• Interactions. There are many interactions between soil physical, chemical, and biological properties that determine response to drought. Soil physical structure (essential for the infiltration of precipitation) is strongly linked to interactions between the soil matrix and soil biology. Changes in soil biogeochemical cycles are often linked to the changes in soil physical structure, such as aggregate breakdown.
• Soil bacterial and fungal communities, and soil fauna have been shown to be susceptible to drought.

Agriculture

• Drought impact. Meteorological, soil moisture and hydrological droughts affect different agricultural sub-sectors in different ways. Impacts include changes in crop, grass and livestock productivity and quality, livestock welfare, business profitability and farmer wellbeing.
• Timing. Drought impacts depend on both the timing and severity of the drought.  Some impacts can be rapid, but others are longer term or lagged. For example, drought can reduce crop yield and/or quality of the harvest, as well as limit the effective winter refill of on-farm reservoirs for irrigation of subsequent years. On a livestock farm, drought can lead to a shortage of feed and fodder which can cause a reduction in both livestock production and fertility, affecting production in future years.
• Farmers’ response. Most drought responses implemented by farmers are reactionary short-term coping strategies. Drought responses of livestock farmers range from management of fodder and grazing fields, to reducing livestock numbers and buying additional feed. Longer term measures, including diversification of fodder crops and the introduction of drought tolerant species, are not currently widely used by farmers. Other short-term measures that can be available to farmers include temporary relaxation of regulations by the government.
• Drought resilience. Investing in increased resilience to drought is a decision for individual farms based on their own perception of drought risk as it carries costs which do not deliver financial benefits in most years. Deciding to change or diversify the crops being grown, or to implement new approaches to increase the retention of soil moisture, or to build a new on-farm reservoir, are very expensive decisions to make in a highly competitive financial environment. This individual approach persists despite evidence of the benefits of collaborative approaches such as local abstractor groups.
• Alternative water supply. Due to the pressure on water supplies, there is an increased interest in water harvesting techniques, on-farm reservoirs, and collaborative water management.
• Irrigation demands. Irrigation for outdoor crops increases yield and improve quality of the harvested produce. Where water is available, farmers may be granted an abstraction licence for purposes of irrigation. There can be financial penalties for farmers when yield and quality of crops grown under contract to supermarkets is lower than expected, for example under conditions of drought when restrictions on abstraction licences may apply.

Vegetation and wildfire

• Detection and prediction of vegetation drought. Earth observation and remote sensing-based approaches have been successfully adopted for detecting vegetation and agricultural drought.
• Resilience to drought. Some UK plant species exhibit a high degree of adaptability, resistance, and resilience to drought, compounded by additional variation resulting from differing environments.
• Response to drought. The collation of a consistent and comprehensive database of plant responses to drought (including agricultural and invasive) would support further investigation of vegetation response to future climate change scenarios.
• Impact of drought on plants. Extended drought conditions lead to a reduction in the uptake and movements of water through a plant, which leads to a decrease in photosynthesis, evapotranspiration, and carbon fixation. The timing of a drought event can interrupt the plant life cycle in different ways, producing different impacts including whole body damage, limited growth and productivity, bud development or seed development/germination.
• Impact of drought on vegetation. Droughts impacts species differently and affects species competition so can change species community composition. The extent of change depends upon the local environment and the intensity and length of the drought period.
• Wildfire in the UK: There has been limited research into wildfire, its impacts on different vegetation types and post fire recovery in the UK. This is mainly because predicting when and where fires will occur is very difficult. Further research is needed as wildfires are expected to increase in frequency and intensity with climate change and drought.
• Drought and wildfire. Drought events typically increase the amount of dry biomass available to fuel wildfires whilst also providing increased general temperatures that lend themselves to increased combustion rates, fire intensity and the potential for vegetation recovery post fire. Vegetation fires can damage soil and reduce the local water quality.
• Wildfire impact timescales. The impacts of wildfires demonstrate impacts over varying timescales:
  • Immediate loss of vegetation, wildlife, and infrastructure
  • Longer term impacts on the environment (recovery or total loss)

7.3 Water and demand management

Social and behavioural scientific evidence for public messaging

• Public messaging. There is no single understanding of droughts nor a single homogeneous ‘public’ experiencing or contributing to drought equally. The term messaging itself is inherently diverse. The evidence base about messaging effects is patchy in terms of quantitative studies but is considerable in terms of examples of strategies. Top-down messaging can assume that what is sustainable is known. Individual behaviours are equally the primary causes of unsustainability, and such behaviours exist in all parts of a society or community.
• Message framing. Messaging needs to move beyond framing drought as simply a hydrological or climatic event and recognise that different social groups contribute to and will be affected differently by drought impacts and related possible water restrictions.
• Behaviours. Behaviours cannot be assumed to be irrational and therefore ‘corrected’ with more or better information. The context in which messages are received, and the judgement made about the author of the message (ethical resonance and level of trust) are perhaps the most important factors in determining behaviour responses and practices.
• Drought warnings. Early interventions such as warnings need to be followed up and reinforced with later messaging, both because behaviour ‘fade’ and because short-term actions must be contextualised in long-term actions.
• Enabling positive action. Messaging must be meaningful and relevant for the public being targeted, and enable positive action i.e., individuals can choose to enact some kind of change based on the message.
• Collective messaging. Messaging from a collective platform of stakeholders (rather than an individual company or organisation) is more likely to be trusted and acted upon.

Public water supply

• Characteristics of historical droughts. Most if not all English water supply systems have experienced multi-seasons drought events (more than 18 months and up to 3-years). These longer droughts tended to be more spatially coherent than single-season events and to end in the autumn. Groundwater-dominated systems are also vulnerable to single-season droughts ending in March (dry winters).
• Causes of historical droughts: The severity of multi-season droughts is generally affected by large scale circulation patterns, although many overlapping climatological drivers play a role. Little is known about how domestic water demand responds under drought conditions, but we know that spatial patterns of demand can greatly affect drought resilience and they vary from one event to another. Industrial and energy demand is distributed unequally around the UK and has varied significantly over time, which has both increased and reduced water deficits. Other factors related to the supply network operation, such as outage or water quality issues, can exacerbate drought conditions.
• Vulnerability to drought. The vulnerability of water supply systems to drought changes with time. As we experience more droughts, we expose new vulnerabilities, but problems can also arise from different responses to droughts like those we have seen previously. For example, changing levels of demand for water can make drought impacts more severe and increase the risk of failure.
• Future water supply droughts. Projections based on “change factors” (the approach recommended in current planning guidelines) indicate a likely increased risk to water supply from single-season droughts and decrease to multi-season, because of warmer and wetter winters and hotter and drier summers. However, the change factors approach misses important aspects simulated within climate models, such as the projected lengthening of summer and shortening of winter and increased spatial coherence of drought events. Alternative approaches to incorporating climate change may capture future drought events better. There are also large uncertainties around future demand, but household peak demand may increase because of increasing dryness and rising temperatures.
• Present and future drought impacts. There is large uncertainty in present and future impacts, and significant spatial differences in impacts across different scales. Impact assessments are hampered by uncertainties that are introduced and can grow through the modelling chain, including bias correction and downscaling of climate model outputs. Reconciling secure water supplies and a better water environment will be increasingly difficult in the future.
• Planning approaches. Multiple metrics are used to evaluate supply system performance but there is no agreement on which one is preferable. Current planning approach is mainly top-down, although there are attempts at including elements of bottom-up approaches to better dealing with ‘deep uncertainty’ about future demand and policy, as well as the plausibility of future drought projections.

Models and data

• Drought Assessments. Over the last decade, drought assessments in the UK water sector have become increasingly complex with a significant increase in the models and data available. Most of these datasets and models are published in the scientific literature and freely accessible.
• Datasets. Available datasets include historical observations of temperature and precipitation at high spatial (1 km) and temporal (hourly/daily) resolution, and a range of future projections from climate models and weather generators at variable resolution (from 2.2 km to 60 km, daily/monthly).
• Models. A range of hydrological and water resource models can be used to quantify the effects of present and future meteorological droughts on river flows and water resource systems (e.g., reservoir levels, water transfers and abstractions) at various scales (national, catchment, water resource zone).
• Capability. There is currently a strong capability for producing long-term, spatially coherent, national-scale climate, hydrological and water resource projections for drought management, planning and assessment.

Novel drought indicators

• Indicators. There have been significant advances in the development of indicators and indices to characterise hydrological drought.

8. Gaps in understanding

The scarcity of droughts in the historical record means that there are many gaps in the understanding of droughts in England. Some of these are specific to sectors or receptors, but others are system wide. Here these are grouped under four headings:

• Physical, chemical, and biological processes
• Recognising and monitoring drought
• Drought impacts
• Effective drought actions, responses, and recovery

8.1 Physical, chemical, and biological processes

Droughts have traditionally been defined as the result of an extended period of below average rainfall. Dry periods are common in the UK, but they are usually punctuated by wetter weather. There is limited understanding of the atmospheric drivers that lead to prolonged periods of drier weather and how they interact to cause the most extreme droughts, and the relationship between decadal climatic variability and long droughts remains unclear. This uncertainty has two consequences.  While seasonal forecasts are becoming increasingly accurate, forecasts are still highly uncertain, so the onset, extent and duration of droughts can rarely be predicted with confidence. Perhaps more importantly, this uncertainty means that it is not possible to know if global circulation models (GCMs) used for climate projections reproduce current and future drought risk accurately, meaning that there is considerable uncertainty around plausible drought intensity and duration in current and future climates. Under global warming past droughts will remain useful but may not always provide a reliable guide to the future. Most serious droughts in the 20th and 21st centuries were the result of two consecutive dry winters; a third dry winter could cause impacts that have not previously been experienced. Statistical and stochastic approaches to generating extreme droughts are very sensitive to input data and assumptions, very difficult to test, and perhaps should be best treated as generating scenarios for sensitivity analysis.

8.2 Recognising and monitoring drought

Every drought starts differently, is particular to the local conditions, and it usually takes some months before there is agreement that a drought has begun. Many different drought metrics have been proposed, often based on rainfall deficits or river flow deficits. These provide a consistent approach to identifying and understanding drought onset and intensity, although they are not always easy to reconcile with people’s or the landscape experience of drought. There is no consistent way to characterise groundwater droughts, and there has been little work to link drought indicators with drought impacts on people and the environment. 

As a drought progresses, soils become dry and groundwater levels and river flows fall. There is little monitoring data to help understand non-linearities in hydrological and biogeochemical processes and the response of ecosystems, particularly in long or extreme droughts. Routine monitoring is often not in the right place or at an appropriate temporal frequency to inform an understanding of thresholds and system responses in extreme droughts, and operational organisations are often too engaged in drought response to gather additional data. This contributes to the continued limited understanding of the effect of drought on catchment processes.

8.3 Drought impacts

Drought can push ecosystems past tipping points, causing shifts to alternative states that are hard to reverse. It is important to be able to identify systems on the approach to a tipping point, to inform timely and appropriate management actions. For example, greater understanding of which river and lake typologies are most vulnerable to drought, and which are the most resilient could inform actions to increase resilience in vulnerable catchments through land-use change, water quality improvements and river restoration. There is limited understanding of how the environment responded to past droughts, which makes it difficult to predict how increases in drought severity will interact with other climatic and human pressures to shape ecological responses to future events.

There is little information about the spatial extent and rates of change of soil water repellency across the UK, and how this affects infiltration and recharge. Soil bacterial and fungal communities and soil fauna have been shown to be susceptible to drought, but it is not clear if there are tipping points where soil biological functions do not recover. The vulnerability of soil carbon stores to drought remains uncertain.

There is poor quantitative and empirical evidence of the impact of drought on UK farming. Different farming systems (arable, horticultural, livestock) are affected in different ways. Drought can also affect farmers indirectly through economic costs of reduced productivity and yield, and potentially through penalties for reduced crop quality.

Drought impacts on public water supply are generally lagged. Storage in reservoirs and groundwater smooths the shocks of extreme weather, but chronic or acute conditions may eventually lead to low storage levels which could ultimately lead to interruptions in supply. When heatwaves occur during droughts the demand for water may increase greatly, sometimes to the point where part of the supply infrastructure fails. The way that water demand changes during drought remains poorly understood, partly because each drought is different but also because people’s response to drought changes through time.

8.4 Effective drought interventions and recovery

The factors that influence people’s response to drought remain unclear. There is:

• Limited understanding of the effectiveness of specific savings by households arising from different types of messaging;
• A question over the extent to which public trust and willingness to change behaviour may be eroded by increasing scrutiny of regulators, water companies and water governance systems;
• A lack of agreement between focusing on individual change against a backdrop of wider and institutional questions of water governance;
• Little understanding of the potential for citizen science projects and two-way and multi-nodal messaging to address the limitations of top-down messaging.

Management actions during droughts are relatively limited: reductions in abstraction may support environmental flows, but at financial cost to the abstractor. Other actions include fish rescues and oxygenation of water; more work is needed to understand the effectiveness of these interventions and how they can be deployed most effectively.

All droughts end differently; there is some information on the termination of historical droughts, but little knowledge of the effect of management interventions on reducing the long-term impact of droughts.

9. Summary and conclusions

By reviewing the current research into drought, this review has shown the breadth and complexity of the issue and the extent of current understanding. The review has identified both the state of existing knowledge and the remaining gaps.

It is evident from the review that as well as topic-specific gaps that there are recurring themes that cut across the topics.

9.1 Changing nature of drought

In recent decades, the droughts that have been experienced have been similar, and consequently management practices and engineering have developed to cope with them. But how severe or frequent could droughts be? Could England be at risk of different types of droughts?  Could summers become extreme? Could multi-annual events become longer? How does the variability of the climate between seasons and years, and transitions from wet to dry, affect the hydrology? And how might this alter due to climate change? What will the droughts of the future be like?

9.2 Modelling, monitoring and data recovery

Existing models perform poorly at the extremes of calibration, and the quantity, quality, and type of data available is insufficient. Recovery of old data and targeted monitoring of new will extend records and fill gaps. But how can model capability be improved? Can models and data be better integrated? Do the right models and data exist, and if not, what would they look like? Could better decisions be made with existing models and data?

9.3 Catchment processes and dynamics

There is still only limited understanding of how physical, chemical, and biological processes, operate in a catchment and the dynamic relationships between them. This may limit the ability to manage catchments effectively. How can these processes be investigated? How do any impacts propagate through the system? Do groundwater storage and recharge processes change? How do ecosystem interactions change with physical and chemical changes? Are some catchments more sensitive to drought? Do processes alter with spatial scale? Do wildfires change catchment hydrology?

9.4 Development and ending of a drought event

The type, magnitude and duration of drought impacts is dependent on the conditions before and after the event, as well as during. How quickly can a drought start?  How far can the quality of forecasting be improved? What drought and future scenarios should we use? What is meant by recovery and how much rainfall is needed? Are there other factors that can affect the speed of onset or recovery from drought?

9.5 Drought as a social construct

The understanding of drought is in part determined by public discourse. The way drought is presented affects public expectations and behaviours.  A better understanding of the social context of drought management could change the actions of regulators, water suppliers and users.  How do people respond to water use restrictions? Why is drought discussed separately from flooding? What is a water supply system? What is the role of public trust in the governance of water? What is the national understanding of water?

9.6 Impacts and interactions

Drought is a concern because of the impacts it causes and is often described in those high-level terms: environmental, agricultural, or public water supply. But drought impacts result from a combination of variables and to limit impacts it is important to understand how they interact together. What are the water resource, water quality, agricultural and environmental impacts of drought? How do human activities and human-made changes to catchments interact with natural impacts? Are they changing with climate? Can lessons be learnt from past events?

9.7 Vulnerability and resilience

Drought management is difficult because it requires decisions to be made in circumstances where there is much uncertainty. By exploring what a system (natural or human-made) can cope with and what it is vulnerable to, it may be possible to identify measures to increase resilience. But what is a resilient system? Are there thresholds or tipping points beyond which a system moves into a new state, and can we define them? What is resilience in natural and human-made systems? How do we use the concept of catchment resilience? Can stress testing create management plans that help to build resilience?

10. Concluding comments

The review has drawn together and consolidated the current body of knowledge on drought and its impacts and has highlighted where there are gaps in understanding. We will build on the findings of this review and identify where further research could deliver the most benefit for drought management and resilience, both now and in the future.

11. Appendix A - Review topics and authors

The following is list of the authors and institutions who contributed to the review, organised by theme and essay topic:

11.1 Physical processes of drought

1. Rainfall and atmospheric conditions: Prof Len Shaffrey - University of Reading and the National Centre for Atmospheric Science.

2. Surface water: Mr Jamie Hannaford, Dr Lucy Barker, Dr Stephen Turner, Dr Simon Parry, Dr Maliko Tanguy, Dr Amulya Chevuturi - UK Centre for Ecology & Hydrology.

3. Future surface water: Dr Rosanna Lane, Dr Alison Kay - UK Centre for Ecology & Hydrology.

4. Groundwater observed and projected: Dr John Bloomfield, Dr Chris Jackson - British Geological Survey.

11.2 Impacts of drought

1. Water quality: Dr Mike Bowes, Ms Alex O’Brien, Dr Michael Hutchins, Dr Cedric Laize - UK Centre for Ecology & Hydrology.

2. Water temperature: Prof David Hannah, Dr James White, Dr Kieran Khamis - University of Birmingham.

3. Riverine ecosystems: Prof Rachel Stubbington - Nottingham Trent University.

4. Soils: Dr Andrew Tye - British Geological Survey and Dr David Robinson - UK Centre for Ecology & Hydrology.

5. Agriculture: Prof Ian Holman - Cranfield University.

6. Vegetation and wildfire: Dr Jill Thompson, Dr France Gerard, Dr Douglas Kelley, Dr Rebecca Oliver, Dr Amy Pickard, Dr Maliko Tanguy, Dr Maud van Soest - UK Centre for Ecology & Hydrology.

11.3 Water and demand management

1. Social and behavioural scientific evidence for public messaging: Dr Kevin Collins - The Open University.

2. Public water supply: Mr Mason Durant, Dr Chris Counsell - HR Wallingford.

3. Models and data: Dr Gemma Coxon, Dr Francesca Pianosi, Dr Doris Wendt - University of Bristol.

4. Novel drought indicators: Prof Robert Wilby, Dr Josh Thompson - Loughborough University, Prof Conor Murphy, Dr Paul O’Connor - Maynooth University and Dr Tom Matthews - Kings College London.