EGU 2014 Day 4 – Global catastrophes, uncertainty and can you ever know your audience?

Day four in the big geology house…. Today started well, I took a little time in the morning to work on my presentation for tomorrow, but because I forgot to reset the clock on my computer I missed the workshop on applying for funding that I was aiming to attend. Nevertheless I made it to my first session of the day ‘Volcanism, Impacts, Mass Extinctions and Global Environmental Change‘ (SSP1.2/GMPV41) which is a session that has to win the prize for BEST NAME OF A SESSION EVER. I bumped into a lecturer from my University there, who seemed a little surprised to see me – he asked why I was there and I said ‘Global catastrophes? Of course I’m coming to this one!’ and he replied that he was there for the isotopes. Strangely the organisers seemed to underestimate the interest factor of such an epic session title, and had put the session in a really small room. People were crammed in all over the place, sat on the floor, standing by the walls, and every seat was full. You just can’t deny the pulling power of massive disasters.


The talks themselves were great, a series of presentations on Tunguska and the Siberian trap flood basalts and their associated infrastructure. There were questions of whether magamatism could trigger a mass extinction, and if the dates of massive flood basalt eruptions did actually precede the extinctions? Seth Burgess did actually present data that suggested the main body of the eruptions did actually commence AFTER the extinction, but also recognised the problem of sampling bias – a common problem in the geological sciences – that you can’t always get the data you want because, oh, a mountain is on top of it. So you have to predict as best you can based on incomplete data. From the data that Seth Burgess had, he suggested that there was more than one phase of the eruption and that the lavas that couldn’t be sampled may actually contain sills that predate the extinction.


Later on the session, the extinction moved from the end-Permian to the Cretaceous. Now if you are unfamiliar with that name – think dinosaurs. It’s the one where the famous film should really have been called Cretaceous Park instead (and why it wasn’t I really don’t know!). The thing with the end-Cretaceous extinction was that it happened not only at the same time as a massive flood basalt, but at the same time as a massive (and famous) meteorite impact called Chixulub. Mark Richards from the University of California, Berkeley, told us that there was a 1 in 5 chance of all of these events being a coincidence, which isn’t really that bad. But when you include the size of the Deccan eruptions, the chance becomes 1 in 50, so it is hard to dismiss this as a coincidence. He suggested that another reason for the correlation, was that the impact could have triggered massive worldwide magmatic activity – in the same way that seismic triggers have been shown to induce magmatic activity on a much smaller scale. A question was asked however, if there was any evidenced for this systemic increased activity and although at the moment there is not, Professor Richards thinks that geochemical data could be available to support this hypothesis.

So from volcanoes and massive meteorite impacts I thought I would move on to uncertainty and the ‘Communication of Uncertainty about Information in Earth Sciences‘ (SSS11.1/ESSI3.6), convened by R Murray Lark from the BGS. This session was all about how we, as geoscientists, represent unceratinty. This is a really big deal, especially when you relate it to what I was saying above about sampling bias – a large amount of geological information is interpreted on, what the researchers would see as, less than perfect data sets. Now a lot of uncertainty work is based on how to represent a statistical analysis of the uncertainty to other geologists, but there is a growing interest in how we represent uncertainty to the public. Robert Kirby talked about how using ensemble data (think hurricane tracks) can help people to understand different types of data simultaneously, but by using means and standard deviation statistical analysis you can extract more meaningful data (which may be harder to understand). What was interesting to me, was the suggestion that non-experts would have a better understanding of the value of ensemble data that statistically analysed data, but this hadn’t been tested yet. In fact a lot of the work on public understanding of uncertainty seemed to be based upon assumptions – so perhaps these were initial results of studies that were ongoing.


The day ended with the Townhall meeting ‘You don’t know your audience! A ClimateSnack debate‘ (TM8). This meeting drew together a number of science communications experts from a variety of backgrounds to discuss ‘knowing your audience’ a central concept in sci comm and one that is often under debate. The panel consisted of:

Dr Sam Illingworth, a lecturer in Science Communication at Manchester Metropolitan University (read more about him here)
Christina Reed, an independant science journalist
Liz Kalaugher, from environmentalresearchweb
Prof David Shultz, a lecturer in Synoptic Meteorology at Manchester University and the author of Eloquent Science
Mathew Reeve, co-founder of ClimateSnack (the moderator)

The discussion covered a wide range of subjects relating to audience – can you ever know your audience, how do you know what your audience wants, where is ‘the room’ in a digital age? We even discussed the seemingly opposing views of should we even be attempting to communicate all forms of science (as some parts are genuinely too difficult to understand without four years of university education) and do we seek to maintain the ‘aura of mystery’ to preserve our academic importance? What was interesting here was the idea that as science communicators we all WANT to communicate every aspect of our science, but that is just unrealistic – and most people genuinely wouldnt care. What we have to do is make our science AVAILABLE instead.


Finally I asked about training our undergraduates in science communication and Prof Schultz raised an interesting point – that our undergraduates often have a hard enough time writing scientifically first and that writing for a general audience from a scientific perspective – especially as a scientists – often means you need to understand scientific writing before you can communicate it back to the public. Also, he said, in his experience students already think they can communicate with non-scientists without training!

This session finished at 8, so I trundled myself back to the hostel to prepare for my oral presentation tomorrow at 9.15am (eek!).