Project highlights

  • First identification of risk hotspots for this catastrophic compound hazard
  • Innovative mix of methods to quantify evolving risk in a warming climate
  • Results of immediate interest to broad spectrum of society (from climate scientists to decision makers)


How might the severe risk posed by convective storms become worse? As global temperatures rise, individual thunderstorms might become more intense or frequent. But, perhaps our concern should focus on entirely new or unrecognized ‘compound’ risks caused by combinations of two or more hazards or events (Zscheischler et al., 2018; Matthews, Wilby and Murphy, 2019), which may prove more catastrophic. However, these possibilities are largely unexplored.

“Black-swan” events are perceived as unpredictable, but with suitable physical understanding of the climate system it should be possible to identify likely devastating natural hazard combinations before they first impact. There is a strong incentive to develop plans for minimizing the impact of these emergent threats. This PhD will address the as yet unseen hazards associated with severe convective storms (SCS), driven by impact-inspired questions about SCS extremity, and focussed by pairing a case-study site with excellent modelling (UK) with an area of higher hazard (Himalaya).

The PhD’s first question is: How severe might individual SCS events get? In the Hindu Kush Himalaya (HKH) region, monsoon-related ‘cloudbursts’ are of concern (Thayyen et al., 2013), especially in glacierised catchments, where extreme rainfall may increasingly coincide with snow/ice melt to drive catastrophic flooding. Similarly, in the UK SCS hazards (Hillier, 2017; p209) of wind, flooding, lightning and landslide might compound within events, or with the potentially deadly heat that drives SCS (Zhang and Villarini, 2020). And, interactions may intensify as climate changes.

The second question is: How might SCS events combine? After a heatwave this August (9-12th), storms caused €10s millions of damage across Europe (UK, Greece, Spain, Germany). But, limited attention has been paid to how SCS phenomena co-occur at this larger (inter-)national scale, and how damage might build across a season to be socially and/or economically catastrophic (Hillier et al., 2015; Hillier and Dixon, 2020).  In the limit, as boundary layer heat builds with climate change (Diffenbaugh, Scherer and Trapp, 2013) are widespread tornadoes akin to the USA possible?

Through co-supervision by the Met Office, links to India’s National Institute of Hydrology, and strong connections to the insurance sector, this PhD is well placed to feed into impactful decisions.


Loughborough University


  • Climate and Environmental Sustainability


Project investigator

John Hillier



Freya Garry

Tom Matthews

Nick Dunstone

How to apply


Case studies will be blended with the UNSEEN method (Thompson, 2017), which uses atmospheric physics encoded in climate/weather models to seek unobserved yet plausible future extremes that might (co-)occur.

(1) Case studies: Historic UK storms and Himalayan (HKH) cloudbursts will be selected to compare/contrast them, understand ‘ingredients’ that drive SCS activity (e.g. moisture, heat, wind shear), and how/where damage arises.

(2) UNSEEN driving conditions and impacts: Examine environments (25-60 km resolution) identified in (1) as conducive to the most severe SCS, and potential impacts (Hillier and Dixon, 2020).

(3) UNSEEN dynamics: After fidelity tests, high-resolution (2 km, convective permitting) custom model runs made available to this project by the Met Office will be used to explore projected SCS events in detail.

Throughout, there will be an assessment of large-scale modes of variability (e.g. NAO, ENSO) that may enhance the SCS risk at seasonal timescales.

Training and skills

To explore the most severe SCS extremes, the student will be provided with training that

  • develops a deep meteorological understanding of convective storms
  • guides them towards programming skills required for the analysis of large climate data sets (e.g. Linux/Python)
  • includes training in R, which is the statistical gateway to interpreting such data
  • includes support on using the NERC JASMIN computing facility (“using a super computer”)
  • builds to the ability to manipulate and examine custom numerical weather forecasting runs
  • allows them to show leadership (e.g. co-convening a session at an international conference)

Partners and collaboration

In this project Loughborough University is partnered formally with the Met Office, and also to India’s National Institute of Hydrology through Dr Renoj Thayen.  The Met Office has co-designed the project, will co-supervise the project, and is making available the opportunity for two 2-month secondments (physical/virtual) to their offices in the UK. The Met Office bring world-leading forecasting and modelling skills to the project. Dr Thayen brings a global, impact-based perspective with his knowledge of the meteorology and societal challenges of SCS in India. A research visit to Dr Thayen might be possible, conditions allowing.

Further details

For further information about this project, please contact Dr John Hillier ( or Dr Tom Matthews ( For enquiries about the application process, please contact the School of Social Sciences & Humanities ( Please quote CENTA when completing the application form:



Possible timeline

Year 1

The core of this work is low risk, but scope exists for a student to excel (e.g. deliver a global-scale assessment of SCS as a compound risk in a changing climate). The two questions that guide this PhD are: How severe might individual SCS events get? And, how might they act together? The latter one is lower-resolution, making the UNSEEN work technically easier, and so in our suggested timeline this will be tackled first (Years 1-2), building to higher-resolutions (Years 2-3) and increasingly in depth insights.

Beginning with case studies, you will examine ~10 historic UK storms and a similar number of Himalayan (HKH) cloudbursts to compare and contrast the events and their synoptic driving conditions (i.e. the weather ingredients that promote SCS activity). A link to damage from these environmental ‘ingredients’ (e.g. in reanalysis data such as ERA5) will be forged via novel, impact-based proxies created by calibration to loss data (e.g. rail delays, road accidents) (Hillier and Dixon, 2020).  A first virtual/physical secondment 1 Met Office (2 months) is envisaged, and familiarization with the UNSEEN method.

Year 2

Mine existing, open-access data archives to explore the changing SCS hazard. A practical introduction to the UNSEEN approach (Thompson, 2017), exploring 25-60km resolution datasets (e.g. SEAS5, GLOSEA5, UKCP18) for synoptic conditions (e.g. CAPE, moisture, heat, wind shear) conducive to the most rare and extreme SCS formation past, present and future. To continue a focus on the extremeness of individual SCS at the Indian (HKH) study site, a visit is potentially possible.

Year 3

Examine the threatening extremes and their dynamics more directly. This PhD follows a joint Environment Agency/Met Office collaboration project, from which high-resolution (2 km, convective permitting) custom model runs of a numerical weather prediction model will be available to this PhD. So, after fidelity tests, SCS within this will be examined. Illustratively, a ‘storyline’ extrapolation from Europe/UK’s 2020 summer episode of SCS could be done. Namely, how bad might it have been with similar underlying conditions? How bad might a future analogue be?  Second Met Office secondment.

Each year will deliver valuable, publishable understanding in its own right. But, the yearly blocks also add to each other, designed to yield understanding is greater than the sum of the individual parts.

Further reading

Cecil, D. J. (2014) ‘Gridded lightning climatology from TRMM-LIS and OTD: Dataset description’, Atmospheric Research, 135–136, pp. 404–414. doi: 10.1016/j.atmosres.2012.06.028.

Diffenbaugh, N. S., Scherer, M. and Trapp, R. J. (2013) ‘Robust increases in severe thunderstorm environments in response to greenhouse forcing’, Proceedings of the National Academy of Sciences, 110(41), pp. 16361–16366.

Hillier, J. K. et al. (2015) ‘Interactions between apparently primary weather-driven hazards and their cost’, Env. Res. Lett., 10, p. 104003.

Hillier, J. K. (2017) ‘The Perils in Brief’, in Natural Catastrophe Risk Management and Modelling: A Practitioner’s Guide. Oxford, UK: Wiley-Blackwell, p. pp 536.

Hillier, J. K. and Dixon, R. (2020) ‘Seasonal impact-based mapping of compound hazards’, Env. Res. Lett.

Matthews, T., Wilby, R. L. and Murphy, C. (2019) ‘An emerging tropical cyclone–deadly heat compound hazard’, Nature Climate Change, 9, pp. 602–606. doi: 10.1038/s41558-019-0525-6.

Thayyen, R. J. et al. (2013) ‘Study of cloudburst and flash floods around Leh, India, during August 4–6, 2010’, Natural Hazards, 65, pp. 2175–2204. doi: 10.1007/s11069-012-0464-2.

Thompson, V. (2017) ‘High risk of unprecedented UK rainfall in the current climate’, Nature Communications, 8, p. 107. doi: 10.1038/s41467-017-00275-3.

Zhang, W. and Villarini, G. (2020) ‘Deadly Compound Heat Stress‐Flooding Hazard Across the Central United States’, Geophys. Res. Lett., 47, p. e2020GL089185. doi: 10.1029/2020GL089185.

Zscheischler, J. et al. (2018) ‘Future climate risk from compound events’, Nature Climate Change, 8, pp. 469–477. doi: 10.1038/s41558-018-0156-3.


This project is underpinned by novel, computer-based analyses and so is highly resilient to the possible impacts of COVID-19. It is intended that the project experience is uplifted by in-person secondments (UK Met Office, India), with a fallback to conduct secondments virtually if necessary.  However, all core analyses and supervision can be conducted remotely if required with no loss of scientific integrity. Project robustness to illness is ensured by pairing supervisors in both location (Loughborough/Met Office) and speciality (meteorology/multi-hazards). All supervisors have extensive experience of working effectively in teams in remote working environments.