Centre for Climate Communication and Data Science

At C3DS, we combine insights from climate communication research and computational social science. We use both qualitative and quantitative research methodologies.

Our research specialisms include misinformation, visual communication and AI, and understanding narratives and discourse. Using our strengths in computational social science techniques - such as natural language processing, network analysis and application of machine learning tools to novel contexts - we offer new perspectives and solutions for climate communication.

We are committed to working with practitioners: with media organisations, the tech sector and NGOs and charities, as well as with academic collaborators. We work with a growing range of partners, who bring their climate communication challenges to us to co-produce research projects together.

Climate visuals and AI

Images are an important but often overlooked aspect of communication. Our research combines qualitative visual analysis with computational approaches to understand the meaning and role of imagery. For example, our work has investigated media imagery accompanying news of European heatwaves; and emerging research into the use of colour for visualising climate change stock imagery. Methods include object detection models and image meta-analysis, as well as using qualitative insights to inform and develop innovative machine learning techniques.

Live monitoring of climate narratives

A major interest of the research centre lies in the real-time tracking of climate information and narratives. Our goal is to offer tools that empower stakeholders to monitor ongoing discussions and gain timely insights. These tools facilitate swift adaptation to emerging trends, enabling effective responses to the dynamic landscape of climate-related discourse. Through our work, we aim to enhance the understanding of current climate narratives and provide valuable resources for informed decision-making and strategic actions.


Detecting and combating climate misinformation presents a notable challenge. Our research focuses on developing computer-assisted systems to identify and categorise such misinformation, aiming to highlight the extent of the issue. We are dedicated to raising awareness about these strategies, to empower individuals and organisations to use them as analytical tools in understanding the ongoing debate. Moreover, our objective is to provide a user-friendly infrastructure that facilitates easy monitoring of discourses and, in the long run, implement automatic debunking to counter false narratives instantly.

Networks, actors, and impact

Climate change discourse in the media is often presented as a battleground, wherein competing actors use various tactics to sway public or political opinion towards their point of view. Our research interests cover the characterisation of these actors (such as their importance or position within the media ecosystem) as well as a deeper understanding of their strategic communications (such as the uptake of a particular message). To achieve these aims, we combine a range of tools across natural language processing and network science to develop new tools to answer questions emerging from social science perspectives.