University of Edinburgh logo British Antarctic Survey logo National Oceanographic Centre logo University of Leeds logo


How much did human influences contribute to the 2020 Stonehaven derailment?

Derailment at Stonehaven in 2020. B y Environment Agency – Oblique Aerial Photography – 20151227

Scientific Background

Extreme precipitation events have significant impacts on human society as shown by the tragic derailment near Stonehaven in August 2020.  The causal chain for this tragic incident was extreme precipitation that likely generated a landslip which then subsequently caused the derailment.  Some media reported that the heavy rain, and so the derailment, was due to climate change.

After a damaging weather and climate event science and society would like to know to what extent, if at all, human influences have changed the risk of such an event. One of the expected impacts of human driven climate change is an increase in extreme precipitation, of about 7%/K local warming (Allen & Ingram, 2002). Mean precipitation is expected to increase at a smaller rate of only 2-3%/K local warming due to energy balance requirements. However, both local and global scale atmospheric circulation changes could considerably modify both the extreme and mean increases.  Atmospheric processes that lead to extreme precipitation are generally small scale and rare. In many climate models these are heavily parameterised and so are highly uncertain. Recent developments in high resolution modelling mean it is now  possible to study individual extreme events (Pall et al 2017) though for this event defining the relevant extreme requires a good understanding of the relevant weather conditions that lead to land slips over parts of Scotland.

Research Questions

  • What are the relevant weather and precipitation thresholds for landslips in eastern Scotland
  • What was the human contribution to the relevant heavy precipitation event of 2020
  • Have the characteristics of convective events changed  due to anthropogenic climate change.
  • If so, what are the mechanisms for this change?


The planned approach is for the student to use remote sensed data and radar rainfall data to establish a landslide database from which  precipitation and weather thresholds will be estimated that generate landslips. The student would then run two ensembles of global scale models nudged to observations to drive high-resolution models  over eastern Scotland.  The ensembles differ in that one runs with current boundary conditions while another runs with boundary conditions estimated for a world without human influences. Differences between the ensembles allow estimation of the changing probability of breaching relevant extreme precipitation thresholds.  Any changes in extreme precipitation can then be traced to either local warming, circulation changes or changes in precipitation efficiency (Pall et al 2017).

The student would set up a computational system focussed on Scotland using the Met Office’s high resolution weather model (UKV)  and test these ideas. One issue is to make the free-running component sufficiently dispersive that the simulations span a broad range of possibilities.


Allen, M. R., and W. J. Ingram (2002), Constraints on future changes in climate and the hydrologic cycle, Nature , 419 (6903), 224.

  1. Pall et al (2017) Diagnosing Anthropogenic Contributions to Heavy Colorado Rainfall in September 2013. Weather and Climate Extremes 17, 1-6. 10.1016/j.wace.2017.03.004