There is no null hypothesis
On 12 August 2020 a 125 hit a landslip near Stonehaven, killing three and injuring 6 (there were only 9 on the train; the pandemic meant that few people were travelling). Today we found out why:
A drainage system was installed in 2011 and 2012 by now-collapsed contractor Carillion – but it was not in accordance with the design. Had it been constructed correctly, it was “highly likely to have safely accommodated the flow of surface water” on the day of the crash, the report concluded.
But a memory stirred sluggishly below the surface of my mind. Wasn’t this tragedy blamed almost instantly on climate change? 12 August 2020, the date of the crash itself:
Scrolling down, we find analysis by David Shukman, Science Editor. Quoth the sage:
It’s long been recognised that landslips are one of the greatest risks to Britain’s railways and that a changing climate will make them more likely. Heatwaves and droughts can dry out the steep embankments beside the tracks, and over the years will start to weaken them. Add to that the effects of heavy rain, of the kind seen just now in Scotland, saturating and eroding the soil, and undermining its strength. And with rising global temperatures set to bring extremes of weather never anticipated by the Victorian engineers who built the lines, Network Rail has been studying how best to keep the tracks safe. But it admits that “we know we can’t rebuild every mile of railway”.
So dry is bad, and so is wet. (I’ve removed the line breaks from the above quote. Every paragraph was formed of a single sentence, including the one beginning “and.” )
After a month Network Rail released an interim report, described thusly by the BBC:
Stonehaven derailment: Report says climate change impact on railways ‘accelerating’
Not much wriggle room there. But on this occasion the BBC is accurately reporting what Network Rail said:
The report notes: “Climate change considerations are being embedded in our standards and planning. “But it is clear from the impact of severe weather events experienced in recent years that this is an area that is accelerating faster than our assumptions, and as a result it has become even more important to implement these plans.”
Scrolling down, we have analysis by Kevin Keane, Environment Correspondent. Quoth the junior sage:
Technology is being developed, like more detailed rainfall forecasting and movement sensors for the most ‘at risk’ sites. But until then it hints that more speed restrictions will be introduced during the heaviest rain along with line closures to allow for inspections. So you can add climate change to the list of potential reasons for your train being delayed.
If something bad happens, it’s because of climate change. There is no null hypothesis that it might have happened anyway. There certainly isn’t an alternate hypothesis, for example that an improperly installed drain might have had something to do with it.
A snip from the Rail Accident Investigation Branch’s animation of the derailment, to be found in today’s story from the BBC.