Historic Droughts

Historic Droughts aimed to develop a cross-disciplinary understanding of past drought episodes that have affected the UK, with a view to developing improved tools for managing droughts in future.

Our starting point was that droughts are not simply natural hazards. There are also a range of socio-economic and regulatory factors that may influence the course of droughts, such as water consumption practices and abstraction licensing regimes. Consequently, if drought and water scarcity are to be better managed, there is a need for a more detailed understanding of the links between physical (i.e. meteorological, hydrological) and social and economic systems during droughts.

With this research gap in mind, the Historic Droughts project developed an interdisciplinary understanding of drought from a range of perspectives. Based on an analysis of information from a wide range of sectors (hydrometeorological, environmental, agricultural, regulatory, social and cultural), the project has characterised and quantified the history of drought and water scarcity since the late 19th century.

The project has developed the first systematic account (the UK Drought Inventory) of past droughts in the UK, incorporating new datasets on past drought characteristics, impacts and human responses. The Inventory is the basis of a novel joint hydrometeorological and socio-economic analysis that is leading to a ‘systems-based’ understanding of drought.

The project has been applying these new datasets and methodologies to enhance drought management, principally through interfacing with the ENDOWS work with decision-makers.

Below we discuss some of the Historic Droughts outputs and activities in more detail.

Systems-based analysis and conceptual framework

We advocated a ‘systems-based’ view of drought, i.e. an understanding of the multiple and inter-connected drivers of drought, the impacts of and responses to drought, and the feedbacks between them.  A key part of this has been the development of a Conceptual Framework for the joint hydro-meteorological-social understanding of drought.

We published this framework in 2017 and illustrated its application to two past drought episodes (1976 and 2003 – 2006). We expect this systems-based understanding to improve decision-making for future drought management and planning, and to facilitate more informed and effective public discourse related to drought. We envisage the conceptual framework will also be applied in other research settings and also other environmental problems beyond drought, in the UK and internationally.

New sectoral datasets and analyses

Historic Droughts set out to improve our understanding of past droughts from a range of ‘sectoral’ perspectives, before bringing these together in the UK Drought Inventory. This has led to a wealth of new datasets on drought.

A key foundation of Historic Droughts is hydrometeorological reconstruction, aiming to improve our understanding of past drought occurrence and severity across the UK. Underlying this was a major data rescue and recovery exercise undertaken by the Met Office, which digitized hundreds of past weather records to improve the national network of rain gauge and temperature observations in the late 19th and early 20th centuries especially. This effort has led to a major strengthening of our national climate archives and extension of the datasets back to 1862.

Not only is this beneficial for Historic Droughts, however – the rescued data features in the Met Office’s recently released ‘HadUK’ archive of national climate data, available to all. The temperature dataset has then been used by the CEH team to reconstruct a national, gridded daily evaporation dataset, extending back to 1891.

River flows and groundwater

The new historical rainfall and evaporation datasets were run through hydrological and hydrogeological models to reconstruct river flows and groundwater back to 1891, for over 300 river catchments and 50 boreholes across the UK. This represents a significant advance in national hydrological datasets, as a majority of observed river and groundwater datasets start in the 1960s.

Historic Droughts published these datasets on the Environmental Information Data Centre to make them available and developed web tools to allow users to explore them. The new datasets have also allowed novel analyses that have allowed us to better characterise UK droughts back to the late 19th century.

Improved understanding of the process of drought

Historic Droughts examined the frequency and severity of past hydrological drought episodes and quantified changes over time. We have also examined the spatial coherence of drought, regionally and nationally, and explored the spatial and temporal evolution of major past drought episodes. We have also improved understanding of drought processes, especially the atmospheric / oceanic drivers of droughts in the UK. The hydrometeorological analysis allows us to characterise the occurrence of drought as a natural hazard over the last 120 years and has shed significant new light on past episodes that occurred before the start of most observed river flow and groundwater records (e.g. 1890 – 1910, 1921-22, 1933-34 and through the 1940s). But from a societal or environmental perspective, drought impacts are the most important aspect, and to help inform future management strategies we need to investigate the human responses that have been taken to mitigate these impacts.

Drought impacts on freshwater ecosystems

Historic Droughts has built up a range of datasets characterising these aspects from various sectoral perspectives. As this has used a range of disparate source datasets and methodological approaches across several disciplines, a key component of this has been a standardized approach to collecting information, including a consistent approach to capturing when and where the relevant impact / action occurred. From an environmental perspective, our focus has been on quantifying drought impacts on freshwater ecosystems using an environmental monitoring dataset collated by the Environment Agency from the early 1990s.

As the impact datasets are limited in time and space, we have developed statistical models to allow us to reconstruct the impacts of past drought episodes on river ecosystems more fully. This has demonstrated how the resilience of river ecosystems to drought varies around the country, and in particular how more heavily modified rivers are more vulnerable to drought impacts.

Dataset of reservoir development

From a water supply perspective, we have delivered a dataset of reservoir development in the UK from the late 19th century onwards, illustrating how reservoir storage capacity has changed in this time. This has allowed us to develop timelines of past water supply, which we can then compare with timelines of demand, assembled from proxies (e.g. population). These supply / demand balances tell us much about changing water availability, which has a bearing on how droughts unfold and how they are managed. We have also cross-referenced this with a database of water supply impacts and drought management measures (drought orders, permits etc.), which we constructed using various water company reports and other sources.

Agricultural drought inventory dating back to 1970s

From an agricultural perspective, we have developed an agricultural drought inventory of past drivers, impacts and responses to drought, gathered using the agricultural media (e.g. Farmers Weekly) back to 1976. We have used this to demonstrate improvements in agricultural resilience over time – how farmers in eastern England have moderated the impacts of droughts over this timeframe through improved planning and communication with regulatory bodies. From a legal / regulation perspective, a database was developed using Hansard, to characterise the legal / policy responses of drought since the early 1970s. We have used this to demonstrate the changing governance backcloth to drought and water resources management in the UK, and how this has had an influence on the management of drought episodes, via the regulatory tools available and their relative efficacy.

First-hand oral histories from 1960s onwards

From a social and cultural perspective, two contrasting approaches were taken. One approach has been through first-hand interviews with affected communities where we gathered over 100 Oral Histories from people who have lived through drought episodes from the 1960s onwards. These tell us something very different from the hydrometeorological / water supply focused analysis, something that has been missing from the picture – how drought is understood by communities, people’s (diverse) lived experiences in drought episodes and how impacts have been felt and responded to. This sometimes diverges markedly from the way that drought is communicated by official bodies, and the understanding of people’s behavioural responses to drought can be used to frame improved drought communications in future.

Archive of drought reports in the media going back 200 years

A second approach has been to capture how drought is represented in the media. This has been done through a ‘Corpus Linguistics’ approach which, uses computerised methods to objectively extract information from published newspaper archives back to the early 1800s. Not only does this provide a comprehensive archive of drought coverage over nearly 200 years – it also allows us to examine how the discourse around drought has changed over time. In the 19th century, for example, the word ‘drought’ was rarely used in connection with the many episodes of water shortage that occurred, whereas in recent years, ‘hosepipe bans’ have been a staple of the (often somewhat combative) narrative. Bringing it all together – accessibility and applications.

Bringing it all together – accessibility and applications

The various sectoral datasets have been combined into the UK Drought Inventory. This is a virtual collection of data, as the datasets are stored for long-term curation in different archives (the EIDC on the one hand and the UK Data Service for the social / economic data). However, all of the sectoral datasets are listed in one place and available to download (https://historicdroughts.ceh.ac.uk/ content/datasets). The hydrometeorological datasets can also be explored in the CEH Drought Portal and various web tools (Link to p.44).. For the social and economic datasets based on text entries, we have developed a UK Drought Inventory Explorer (link to (p.45)  to allow users to search for data spatially, in time, or using keywords.

The real power of these datasets is how they can be analysed together, bringing our environmental science and social / economic perspectives to bear on the same question. For example, we have linked our hydrometeorological drought indicators with the information on impacts on the agricultural sector, to demonstrate the likelihood of experiencing impacts given a certain drought severity – showing significant regional variations around the UK depending on local hydrological conditions and agricultural practices. We are also in the process of publishing an analysis linking our hydrometeorological timelines of drought (extended back to 1800s) with the timelines developed through the Corpus Linguistics analysis using the news media.

The datasets and methods assembled in Historic Droughts are finding practical application, especially through ENDOWS. For example, the indicators we have developed to quantify drought severity are being used for the UK Drought Portal and Water Resources Portals (link to (see p. 44).. The historical hydrometeorological datasets are feeding into the development of consistent Drought Libraries to support regional and national-scale water resources planning, and we have engaged very actively with water companies, the Environment Agency and Regional Planning groups to demonstrate how the datasets can be applied in future.

Water management strategy for agriculture

The agricultural inventory was updated to include information on the impacts of the 2018 drought on agriculture, which has fed into the water management strategy developed in the agricultural workstream of ENDOWS. The social and cultural learning is helping the communities workstream and feeding into new educational resources. While these datasets have had a demonstrable impact already, it will take time for the full impact of such a step change in drought data availability to be realised. These unparalleled resources will be a long-term legacy that, we hope, encourages new research avenues as well as practical applications in future.

Jamie Hannaford, Historic Droughts Project leader, Centre for Ecology & Hydrology

Posted October 2019