Understanding changes to extreme rainfall

Summer Weather, June 2012 - The Great Tyneside Storm with Heavy Rain and Flooding - People brave the downpour in Newcastle. Pic Paul Norris

Rainfall data can be used to help us adapt to climate change. Hayley Fowler, School of Civil Engineering and Geosciences at Newcastle University explains

Heavy rainfall events that only last a few hours or less are associated with flash flooding, which poses a significant challenge to public safety, infrastructure, and the economy. Such events are of growing concern as greenhouse gas-induced global warming will increase the level of moisture in the atmosphere, causing heavier rainfall events. New research is, therefore, trying to understand how changes to atmospheric moisture and circulation dynamics will combine to amplify or weaken regional increases in extreme precipitation events which cause flash floods. This knowledge can be used to provide a better basis for climate change adaptation planning.

The INTENSE Project

INTENSE (INTElligent use of climate models for adaptation to non-Stationary Hydrological Extremes) is a large, 5-year research project (€2m) funded by the European Research Council. It is the first project to examine sub-daily precipitation extremes, enabling substantial advances to be made in observing current and past changes. It will also provide a physical understanding of processes relating to precipitation extremes necessary for improved regional prediction of change.

INTENSE is leading the global research effort in this area as a Global Energy and Water Exchanges (GEWEX) Hydroclimatology Panel Cross-Cutting project on sub-daily precipitation extremes, with a large number of international collaborators and partners. Begun in 2014, it is comprehensively analysing the response of precipitation extremes to global warming by:

  • Constructing a new global sub-daily precipitation data set;
  • Using this and high-resolution climate modelling to quantify the nature and drivers of global precipitation extremes across multiple time scales;
  • Examining the influence of local thermodynamics and large-scale atmospheric circulation modes on observed precipitation extremes; and
  • Using these to identify climate model deficiencies in the representation of precipitation extremes.

Data collection, quality and indices

Sub-daily rainfall extremes are particularly important as causes of urban flooding, but compared with heavy rainfall on timescales of a day or longer these events have been studied relatively little in most regions. Records of sub-daily rainfall are not as extensive, either in time or space, as those for daily rainfall totals. Such data is most commonly available from the 1990s onwards, given advances in rain gauges and electronic recording devices/telemetry. Further, sub-daily rainfall data is much harder to get access to than daily data as many organisations do not make the former freely available.

The INTENSE project is focussing on collecting gauge-based rainfall data only as it gives the most accurate representation of the amount of water reaching the ground. Many sub-daily rainfall datasets exist that cover a large proportion of the globe but are not based on gauged observations. For example, satellite datasets such as GPM record precipitation every 3 hours. Global gridded precipitation datasets also exist, such as MSWEP based on merged gauged, satellite and reanalysis data products, but these are coarse at 3hr and 0.25°. Radar and merged rainfall measurements are limited in usefulness, particularly for the aims of INTENSE which focus on extreme rainfall, as they are yet to be fully validated by observations as no global sub-daily gauge data set exists.

We have collected data from 22,644 stations globally (just under half of these are from the UK Met Office’s Integrated Surface Database). However, even where we have data, it may not be adequate for the analysis of extreme events – for example, undocumented gauge breakdowns may be recorded as periods without rain, daily totals may be erroneously recorded as hourly values or mechanical failure of the rain gauges may produce erroneous extreme events. Furthermore, changes such as gauge location, site characteristics, or equipment type may introduce inhomogeneities in climatic series. The detection of such errors is important for the evaluation of extreme events and for the assessment of longer term variability and trends. We are therefore developing an automated quality control programme for global sub-daily data that will be applied to the station records to produce a high-quality global dataset.

We are using the data collected to produce a set of sub-daily extreme rainfall indices describing monthly maxima, frequencies over thresholds and the diurnal cycle, which will be made freely available to all users and will be very useful for understanding current patterns of extreme rainfall and for the validation of climate model outputs at sub-daily scales. Additional work will use the new dataset to validate new precipitation satellite products from NASA and explore options for combining the gauge and satellite data to produce a gridded global sub-daily dataset.

The dataset collected by INTENSE is a platform for future development by the larger scientific community and policy makers. In particular, we are working to find a global organisation who will maintain and update the dataset. This will require a tremendous amount of effort as licences will need to be negotiated and secured to make the data itself available to researchers to further scientific understanding.

Scaling of intense rainfall with temperature

Climate models suggest that rainfall will intensify under global warming as the physics of a warmer atmosphere are capable of holding more moisture. We expect that rainfall intensities will increase with temperature according to the Clausius–Clapeyron (CC) relation (a rate of ~6-7% °C−1), although precipitation data for some parts of the world show larger (super-Clausius-Clapeyron) rates of change for short-duration (hourly and shorter) extremes. Historical records of UK hourly summer (June-August) rainfall show CC scaling. Our work will examine the new global dataset to see how precipitation extremes scale with temperature change and moisture availability globally.

Trends in extreme rainfall

We have produced the first analyses of trends in sub-daily rainfall records over the US. These show that hourly and daily seasonal maxima have significantly increased over the last 6 decades. The percentage of stations showing significant increasing annual maximum precipitation trends was generally higher for daily compared to hourly extremes. However, strong evidence points to more widespread increases in the magnitude and frequency of hourly extremes during winter compared to daily extremes. This work is now being extended to examine trends globally and how these relate to global mean temperature change. Our work has indicated, though, that one of the challenges of assessing changes in historical rainfall extremes is their relatively large natural variability, in part due to their dependence on large-scale atmospheric patterns such as El Niño.

Climate modelling

Climate models are the main tools used by scientists to obtain projections of future climate from which estimates of future impacts on society are subsequently derived. Although they have a high degree of skill in simulating many features of our climate they also have a number of known weaknesses and so there are large uncertainties in regional patterns of change, making the development of efficient adaptation strategies for flooding more difficult. One major weakness is that the grids they operate on are not fine enough to incorporate surface features that influence local climate or to represent the convective storms that produce intense summer rainfall.

Working with our partners at the UK Met Office Hadley Centre we adapted a weather-forecast resolution model to study potential changes in hourly rainfall associated with climate change. This model, unlike coarser resolution climate models, is able to realistically represent hourly rainfall, allowing us to make future projections with some confidence. In Nature Climate Change in 2014, together with the UK Met Office Hadley Centre, we published the first evidence that summer downpours could become heavier over the UK with climate change. While we expect summers to become drier overall by 2100, our results indicate heavier summer downpours in the future, with almost 5 times more events exceeding 28mm in 1 hour (associated with flash flooding) by 2100 than there are currently. The model also shows increases in hourly rainfall intensities in winter, consistent with projections from a coarser 12km-resolution model and previous studies at the daily timescale. The results represent the first step towards building a complete picture of how rainfall may change as our climate warms.

This work has led to the inclusion of several very-high resolution model simulations in the UK’s next set of official climate change projections (UKCP18). We have now also extended this work to compare our results with the results from other international modelling centres to build up a picture of how the results from these very high-resolution models differ from those coarser resolution models generally used to create national climate scenarios used for climate adaptation. We have also extended our work within INTENSE to examine larger model domains – building towards very high-resolution global model runs.

Putting science into practice

Changes to future rainfall intensities will impact sewer flooding and so our initial work has been applied in the UK water industry. Using our high-resolution observations and model simulations and working with UK Water Industry Research, we have developed better guidance on estimates of change in rainfall intensities (uplifts) affecting water and sewerage companies, offering the potential for their application in both flooding and pollution analyses and in investment planning. INTENSE will be able to use its high-quality observations from around the globe to provide relevant stakeholders with the most reliable return period estimates for different rainfall intensities, and so contribute to a better understanding of current flood risk. Our high-quality data could also be used to better understand how river catchments respond to intense rainfall events – for example, we have produced a high resolution gridded data set of hourly rainfall for the UK which will be freely available to hydrologists and other practitioners. The project’s advances in data provision and scientific understanding will, therefore, facilitate knowledge to help make people, business and infrastructure more resilient to flooding from intense rainfall in the future.

We have also worked hard to communicate our research with the public more generally. We recently developed a museum exhibition, FLOOD! In which we showcased our research into extreme rainfall. We aimed to help visitors navigate their way through the arguments made in the public sphere around climate change and extreme rainfall and provided a more detailed understanding of why flooding happens and what can be done to prevent it. Throughout the exhibition, visitors experienced what it’s like during a flood with an immersive virtual reality headset, understood the role of landscapes and rainfall through an augmented reality sandbox and learnt about flood management, using a giant version of the childhood game Kerplunk. The exhibition also used puppets, real field kit that we use to obtain hydrological data and a peg board game to understand how the probability of extreme rainfall events will change in the future. The walls were covered in information boards and videos explaining the causes and types of floods, how we measure and model flooding, what climate change is, how climate change will affect flooding and what we can do to reduce flood risk, and also displayed research posters from the INTENSE project. The exhibition ran from 3rd-9th July 2017 and attracted over 1000 visitors.


Research within the INTENSE project is providing a greater understanding of the nature and drivers of change in heavy rainfall events worldwide, which will help us to better understand the mechanisms associated with flash flooding. In turn, this should lead to improved prediction of likely future changes and improvements in the use of climate models for adaptation.

Publications referred to in this report can be found on the INTENSE website.



Professor Hayley Fowler

Professor of Climate Change Impacts

School of Engineering

Newcastle University

Tel: +44 (0)191 208 7113




Please note: this is a commercial profile



Please enter your comment!
Please enter your name here