Volcanoes and Society – Part Two
Risk and probabilities
In my last post, I wrote about two aspects of culture and natural hazards – about cultural knowledge of hazards, and about the culture surrounding the way we deal with hazards. I mentioned how we may perceive what is ‘acceptable risk’ differently for individuals and groups. In this post, I want to focus on the idea of communicating and determining acceptable risks and probabilities.
Probabilistic risk maps
One of the recurring themes of the Research Day was that scientists are reluctant to give probabilities; to me, Jonty Rougier‘s talk was particularly eye-opening. He proposed an approach of ‘time integrated risk maps’, rather than hazard maps. He argued that risk managers don’t want to know whether a village is in an area that could be affected by pyroclastic flows – they want to know what the probability is that the village will be inundated by a flow within the next five years, the next ten years, and the next thirty years. They can then consider what they could do to mitigate the risk, and see how those actions in turn affect the probability of risk.
Our ‘risk memory’
While talking about various natural disasters, I started to wonder: how long does our ‘risk memory’ last? Someone pointed out that following the 1906 San Francisco earthquake, building construction was kept to rigorous standards for about twenty years – after which it was dropped due to the expense. I wonder how long the new building standards in Christchurch – and those reviewed around New Zealand following the Christchurch earthquakes – will last.
What is the largest risk we can plan for? Can we take largest historical event we know of as the largest possible future event, and how feasible is this? And what if such an event is so large that we don’t know how to plan for it? One speaker told us how, prior to the Tohoku/Sendai earthquake and tsunami last year, despite some evidence of previous tsunami of similar size along the eastern coast, building standards were matched to something smaller – the largest recorded events. But is planning for the largest historic event always feasible? Another disconcerting possibility – what if the largest possible event is a ‘black swan’ event – something so big that it’s beyond our comprehension?
Following on from the idea of using previous historical events to plan the future, how does this change with the human landscape – for example, can a historical record centuries old be reliably applied to the present day? I’m inclined to say yes, given that matching of observational data and accounts of historical eruptions have been used to estimate the magnitudes of historical events – the impacts of which can be extrapolated to the modern day.
I touched on this earlier when I mentioned ‘personally acceptable’ levels of risk, versus what is acceptable when that risk is integrated over a whole group. How can we quantify risk to human life, and determine what is an acceptable level? Perhaps this is not the role of scientists, but it is worth consideration all the same. At the research day, I was intrigued by one speaker’s cost-benefit evaluation for introducing vertical evacuation structures in Japan to mitigate tsunami risk; essentially, what was the cost of the structures compared to the impact of the loss of life in a tsunami, based on the productivity of the individuals concerned? Perhaps it’s ultimately the economics that determine what risk we can prepare for.