Data Science of the Natural Environment Conference A DIGITAL FUTURES EVENT - 23 November 2023

Q: "Given the job/role you have now have, what advice would you give to your younger self?"
Gavin Shaddick: "To attend more events like these and go talk to more people... Meeting new people and building a network is where new ideas spark."

On the 23rd November 2023, The University of Manchester’s Digital Futures platform along with the Alan Turing Environment and Sustainability special interest group, and the Sustainable Futures platform, hosted a one-day conference centred on the interface between data science and the natural environment.

Global climate change, pollution, the move to net-zero and widening economic-social divides are some of the large-scale global pressures that demand data drive approaches to mitigate impacts on the environment. Whilst wider narratives tend to focus on application of powerful algorithms in this space, we must consider the entire breadth of data science from underlying digital infrastructure to stakeholder engagement.

The conference aimed to bring together researchers and various stakeholders partners together to share developments at the data science – natural environment interface.

The day featured numerous talks from both internal to the University and external participants who all focused on themes consisting of but not limited to:

  • Computer vision approaches
  • Development of hybrid physics-ML models
  • Low power/edge applications
  • Stakeholder engagement
  • Policy adaptation
  • Trust and standards

After a series of talks provided by academics and industry experts alike, an insightful panel discussion occurred at the end of the day along with a networking session intended to bridge key interdisciplinary gaps recurrent in this field.

Introduction and First Keynote Speaker

Professor David Topping kicked proceedings off in earnest, welcoming attendees and providing a breakdown of the conference's aims, speakers and themes. Professor Topping is Professor of Earth and Environmental Sciences at The University of Manchester as well as the co-lead for Digital Futures' Cities and Environment theme.

After providing the conference with a fitting introduction, Professor Topping welcomed the first of the day's keynote speakers, Blair Edwards. Edwards is a Senior Technical Staff Member at IBM who presented a talk on Geospatial Foundation Models. The presentation centred around Geo AI, data and modelling at scale and their applications to addressing the climate crisis.

Watch Blair Edward's presentation here:

1st Round of Talks

After a short break, attendees returned to the room to listen to Michael Tso's presentation on 'State tagging for improved contextual understanding of environmental data'.

Elena Fillola Mayoral followed with her talk 'Accelerating GHG Emissions Inference: A Lagrangian Particle Dispersion Model emulator using Graph Neural Networks'.

Zahra Mohammadnazari was up next with a presentation on 'Hybrid machine learning-metaheuristic model for sustainable agri-food production and supply chain planning under water scarcity'.

Songyan Zhu rounded off the first round of talks with their talk named 'Machine learning boosts the quantification of interactions between air pollution and land ecosystem carbon sequestration'.

Second Keynote Speaker

After a lunch break, the conference welcomed the second keynote speaker of the day, a special presentation by Mat Basford and Sophie Walker from Open Data Manchester. Basford works as a User Researcher whilst Sophie is a Project Manager.

Their talk was titled 'The Users, and Barriers to Use, of Environmental Data: a User Research Perspective' which focused on vast amounts of unused environmental data held by the Natural Environmental Research Council (NERC) and how the Digital Solutions Hub (DSH) plan to use this data with the support of Open Data Manchester and The University of Manchester. The talk went through how Open Data Manchester were tasked with designing and delivering user research to understand who the likely users of environmental data are, what their needs are, and what barriers – institutional, technical and human – they would face.

2nd Round of Talks

Orlando Timmerman commenced the 2nd round of talks with a presentation on 'Understanding and predicting past, present, and future coral reef distribution via multimodal machine learning'.

Riza Theresa Batista-Navarro followed by presenting their talk 'Harnessing Natural Language Processing to Raise Awareness of the Carbon Footprint of Food'.

Maria Sharmina was next with a talk named 'Digital Security and Sustainability: Synergies and Trade-offs'.

After that, Meghna Asthana presented 'Automated monitoring of ocean floor health – a step towards creating deep seabed digital twin'.

A presentation by Weiying Zhao on Soil Organic Carbon Estimation from Climate-related Features with Graph Neutral Network succeeded.

Lastly, the Alan Turing Institute provided the conference with an update on their Environment and Sustainability Grand Challenge.

Panel Discussion

After a series of intriguing and pioneering talks, the conference concluded with a panel discussion. Chaired by Professor Topping, the panel members included:

The panel discussion consisted of four questions all to do with data science and its contributions and effects on environmental sciences. Below are the questions along with some highlighted responses given by panel members:

What are the upcoming trends in data science that could be pivotal in tackling environmental issues?

How can data scientists work effectively with environmental scientists, policymakers, and other stakeholders?

The panel members in discussion.

What are the upcoming trends in data science that could be pivotal in tackling environmental issues?

Given the job/role you have now have, what advice would you give to your younger self?

Watch the panel discussion here:

After a deeply insightful and intriguing day, the conference was brought to a end with some closing remarks provided by Professor David Topping.

The University of Manchester's wider Digital Futures network is highly interdisciplinary and operates across the whole range of the University’s digital research - connect with us and keep in touch: