‘It’s actually quite scary’: data analysis of sea level changes

‘We don’t want it to get much higher than 4 or 5mm a year as this would cause incredible damage’

Science Lives: Prof Andrew Parnell of Maynooth University on his life’s work

As a statistician you reckon statistical models offer “scientific superpowers”. Why is that?

We live in a complex world where we are trying to understand big messy issues like climate change or how cancer develops, and when we analyse the data we always want to create the best predictions and reduce the levels of uncertainty as much as possible.

Statistical models help us to do that by allowing us to quantify the parts of the problem we know well and separate them from the parts we know less well. We want to make the best possible predictions, but also admit that we don’t know everything, and provide a window of uncertainty that will allow us to make the best future decisions.

Tell us about your work in statistics helping to analyse data in climate change …

Right now I’m working a lot on sea-level change. We have tide gauges from around the world and cores taken from marshes, which have ancient fossilised organisms stored in them. These organisms live at specific parts of the tidal range and are made of carbon, so we can radiocarbon-date them. This gives us a history of sea-level change at a site. We use the statistical methods [that] my team has developed to estimate rates of sea level change around the world.

What do the statistical analyses reveal?

It’s actually quite scary. We know that during the 20th century we had a rise of about 1.7mm per year. In the early part of the 20th century this [reached] about 3mm per year. We don’t want it to get much higher than 4 or 5mm per year as this would cause incredible damage to infrastructure around the world. For example, we know that New York city has over $25 billion of infrastructure at under one metre above current sea level.


Very little is known about sea level rise in Ireland, so I'm delighted to be part of the new Marine Institute project A4 [The name stands for aigéin (oceans), aeráid (climate), agus athrú Atlantaigh (Atlantic change)], led by Dr Gerard McCarthy at Maynooth University, which will look at this important problem. Hopefully this is the start of much more funding for climate change research in Ireland, as we consistently rank as one of the worst countries in Europe for climate action.

How else have you been applying the statistical superpowers?

We have worked with colleagues and used statistical analyses to help figure out the local history of tsunamis from sand sediments in a cave in Indonesia, to find patterns of proteins that turn up in the blood when a person has an aggressive prostate cancer (in the hope of developing a rapid and easy blood test in the future), and to measure the performance of 3D-printed medical implants. I have also just become a partner in a new EU project that aims to use the power of machine learning and statistics in improving the genetic breeding of wheat.

What are you working on now?

I’ve just moved to the Hamilton Institute at Maynooth University to take up a Hamilton professorship, so I’m in the process of building up my research team and getting my teeth into all of these important problems.

If there's anyone out there who wants to do a PhD project in any of these areas, I would encourage them to get in touch. While I can't match the pay scale of Google or Facebook, I can promise a rewarding career that could really make a difference to the lives of millions of people.

Do you ever take a break? If so, what do you like to do?

I have two small children, so they tend to keep me busy most of the time. Otherwise, all my favourite sports involve me chasing a ball: football, tennis, squash, etc. I’m very competitive, but not very good at any of them!”

Claire O'Connell

Claire O'Connell

Claire O'Connell is a contributor to The Irish Times who writes about health, science and innovation