Overview
The Sahel region faces drastic climate shifts, including rising temperatures, excessive droughts and large-scale floods. These shifts disrupt traditional pastoral livelihoods as pastoralists move seasonally in search of fresh forage and water for their herd. Environmental changes force them to alter routes, depart earlier, or settle permanently.
Existing mobility modelling remains divided: social science research uses qualitative methods to study highly localised strategies while quantitative models prioritise predictive performance but abstract away contextual mechanisms. The proposed conceptual framework addresses this gap by explicitly bridging social science and data-driven modelling. Local perceptions of environmental change and movement decisions are collected through interviews and stakeholder engagement, translated into hypotheses linking climate indicators to mediating variables (e.g. pasture availability) and then tested using historical datasets and ML models. Northern Ghana, a key dry-season destination for cross-border Sahelian pastoralists, serves as a pilot project to test this methodology.
Rather than correlating climate data with mobility outcomes in a black-box AI-manner, this approach models the underlying causal chain. This directly supports IPCC’s call for more context-specific, evidence-based assessments of climate-related mobility while offering tools to anticipate (im)mobility shifts, support early-warning efforts and inform adaptive policy responses across the Sahel.
Speaker
Aline Van Driessche is a second Year PhD student, based in both the Architecture and Computer Science department at the University of Cambridge.
This interdisciplinarity is also reflected in her background, both professionally and academically speaking. She graduated from the KULeuven in Belgium with an Engineering-Architecture degree (specialised in Urban design) as well as a masters in AI. She combined those two passions during an extra masters year in Barcelona on Sustainable Emergency Architecture, where she studied the use of AI methods in the context of post-earthquake damage assessments.
Similarly, during her professional career she continued to explore the intersection between urban space and data-driven methods, while working as ML engineer in a startup that focuses on assessing carbon storage in natural assets based on remote sensing data.
Details
This event is part of the Lunchtime Seminar Series organised by the Wolfson College Senior Members.
It is open to all, free to attend and there is no need to book.
Tea and coffee will be available for the audience.
Access
This event will take place in the Combination Room on the first floor of our main building. It has step-free access with a lift and there is an accessible toilet located on the first floor of the building.
Contact
If you have any questions, please contact our events team - events@wolfson.cam.ac.uk
Lunchtime Seminar Series
The Lunchtime Seminar Series provide an opportunity for students, Fellows and Senior Members to share their expertise in a friendly and supportive environment over lunch.
The seminars are held in the Combination Room from 13.00 for an hour on most Wednesdays during full term. The audience is encouraged to bring their lunch in on a tray and take part in the discussion.
Most people who attend are non-specialists, so the talk has to be aimed at a general audience and speakers are warned to avoid technical jargon. We very much encourage students to offer a paper and use this as an opportunity to try out their thesis ideas on listeners who come from all different academic disciplines.