model – Cabot Institute for the Environment blog

In March 2020, the third and final instalment of my PhD research made its way into Climate of the Past. In that article, I do my best to synthesise all I learnt over 5 years about an event that occurred 34 million years ago called the Eocene-Oligocene Transition. (Just so we are all on the same page: palaeoclimate scientists are interested in this period of the Earth’s history as it is when the first major Antarctic Ice Sheet appeared; before then Antarctica was warm and at least partially forested.)

An image of what Antarctica might have looked like at the onset of the Eocene-Oligocene Transition.

Four and a half years ago I wrote a piece for the Cabot Institute Blog about using a climate model to understand this point in the Earth’s history, and how many questions remained in our understanding. Why was the Earth so hot beforehand? What caused it to cool and eventually for Antarctica to glaciate? What other important changes would have occurred around the world at this time? At the time, I focussed particularly on the latter question.

The more time I spent trying to answer some of these questions, predictably (as is the way with science), the more complex some of them became. In the end, for my own peace of mind, I simply tried to bring together as much information as I could from lots of different sources to try to create a picture with some sort of clarity. I focussed on the high latitude Southern Hemisphere, because that is where a lot of the action was occurring at the Eocene-Oligocene Transition and it is also where models potentially have some difficulties in reproducing the climate.

To build up this picture, I used multiple climate model simulations of the period from two different modelling groups and compared these to the biggest dataset of proxy records of Southern Hemisphere climate 34 million years ago that I could compile by myself. Just reading and compiling all of the data from papers took me around a year. Not solidly (I had lots of other things to do too), but even still, reading papers solidly is very difficult in my opinion. Synthesising all of that different information into something coherent in my head is also something that I cannot force to happen quickly. It comes when it is ready.

Some of the complied proxy data for the high latitude Southern Hemisphere the Eocene-Oligocene Transition included in Kennedy-Asser et al. (2020).

In the end for this paper, I generated no primary data myself. It is all secondary data, either provided by other researchers I work with or taken from this very slow and lengthy review of scientific literature. Maybe, back at the start, that is not how I had pictured the finale of my thesis might look. Maybe the plan was to build up to some exceptional new result that I discovered, with data I produced with my own hands. But that wasn’t the case and, to be honest, I think it is better the way it is. Science is, and should be, a collaborative effort. In the spirit of this, I put all of the data I compiled and used, including all of my analysis scripts and detailed notes of where I obtained secondary data, up on the Open Science Framework. This way I hope the science can keep collaborating and continue growing.

Two thirds of my thesis were based on ‘my own’ data, messing around with a climate model, trying out new ideas, seeing if anything revolutionary popped out. This was really important too: for me to grow as a researcher, to learn about how the model works and to try to generate some outside-the-box ideas. Occasionally, of course, something truly revolutionary will be discovered. In the end, however, my conclusion is that model results often lack meaning by themselves: they need observations or proxy records to go with them to provide some sort of truth of what really happened, whether that is outside right now or 34 million years ago.

My new paper finds very similar things about why the Earth changed so much at the Eocene-Oligocene Transition to earlier research carried out nearly 20 years ago. It doesn’t challenge or rewrite everything we know, but that’s okay. The main scientific conclusion from my paper is that incorporating all of this data is actually essential to coming to the same conclusion as the research from many years ago. Without the inclusion of the boring, extensive data review, I might have quickly, excitedly jumped to a different conclusion that, on balance, seems less likely to be correct.

Much like this paper brings together different existing scientific data to compliment research built up over many years, it also brings together my own work and thoughts. It took many years, but it wouldn’t make sense to rush it: the conclusions take a bit of time, even if all of the data and answers are already out there.

————————————

This blog is written by Cabot Institute member Alan Kennedy-Asser, a Research Associate at the School of Geographical Sciences, University of Bristol. You can follow Alan on Twitter @EzekielBoom.

Alan Kennedy-Asser

Tag: model

Frozen in time: reflections on a PhD and the history of Antarctica