Scott Hosking

From the tropics to the Antarctic, Senior Research Fellow Scott Hosking's work applies machine learning techniques to the environmental sciences

Fact file

  • Environmental Data Scientist at the British Antarctic Survey
  • Has over 15 years’ experience in atmospheric and climate research
  • Co-founder and board member of the Cambridge Centre for Climate Science

Tell us about your research?

I develop AI and machine learning approaches to environmental data, to shed new light on how our planet works. Over the years, my research has shifted from specific locations – first the tropics, then the polar regions – to working on a global scale. I’m now focused on assessing resilience in ecosystems, as well as food, energy and water security.

What do you hope is the impact of your research?

To use the incredible power of AI and machine learning to analyse the vast amounts of environmental data we have gathered over the decades through field campaigns and observations. Pushing the boundaries at this new intersection of environmental and data science can help inform policy and strategic decision-making. Ultimately this will help to improve our quality of life, through conserving biodiversity and ecosystems, and protecting resources.

What’s the most surprising thing to come out of your research?

A few years ago I published a paper showing that, if global temperatures reached 1.5˚C above pre-industrial levels, the change in wind patterns would mean that UK wind energy generation could increase by 10 per cent. While warming on this scale would be problematic for other reasons, this scenario could see wind providing a greater proportion of the UK’s energy mix than previously assumed.

What’s been the most challenging aspect of your career so far?

When I first became interested in applying machine learning to environmental science, it was seen as risky. Trying to win research funding at the intersection between two very different disciplines – “traditional” environmental sciences and the newer AI approaches to data sciences – has been tough. The tide is quickly turning, though, and I’m excited to play my part in bringing these two communities together by leading research activities for both the British Antarctic Survey and the Turing.

Your work in three words…

Diverse. Energising. Interdisciplinary.

When not working what can you be found doing?

Getting my nails painted by my 4-year-old.