Integrating Data Mining, AI and Multi-Criteria Analysis for Water-Related Climate Impacts
Background
Climate change impacts are unevenly distributed, creating geographic and sectoral hotspots of social, environmental, and economic risk. Water-related risks, like droughts and floods, disproportionately affect vulnerable populations and drive socio-economic disruptions, including migration and reduced livelihood security. Addressing these challenges requires integrating diverse data sources (e.g., climate projections, hydrological models, and socio-economic indicators) into actionable insights for decision-makers. This research develops a decision-support framework combining data science techniques (text mining, web scraping, geospatial analytics) with multi-criteria decision analysis (MCDA) to assess and communicate water-related climate risks. The goal is to build interactive dashboards that help policymakers prioritise interventions in climate-sensitive sectors, especially water security and population resilience.
PhD opportunity
Research Questions:
• How can structured and unstructured data (e.g., climate projections, water availability, socio-economic vulnerability, policy texts) be collected, cleaned, and integrated to represent multiple dimensions of water-related climate risk?
• What multi-criteria decision-making approaches best capture trade-offs between water security, social vulnerability, and environmental sustainability under uncertainty?
• How can these insights be effectively communicated through interactive dashboards to support evidence-based decisions on climate adaptation and population resilience?
The candidate will:
• Identify and acquire data relevant to water-related climate risks and population impacts (e.g., flood risk, drought vulnerability, migration patterns).
• Apply data science techniques including web scraping, text mining, and geospatial analysis, alongside behavioural decision science methods such as expert elicitation and preference modelling.
• Develop risk aggregation models using MCDA and decision-making under uncertainty frameworks.
• Build a prototype decision-support tool (e.g., a dashboard) for visualising integrated risk scores and scenario comparisons.
The project offers hands-on experience with cutting-edge tools in AI, data analytics, and decision science, while addressing urgent climate resilience challenges. It provides opportunities for collaboration, interdisciplinary training, and contributions to open-source platforms. The research aims to support climate-resilient development by informing policies that protect vulnerable populations and ensure sustainable water resource management.
Applicant Profile
– A strong background in statistics, data science, or applied mathematics.
– Skills in data wrangling, integration of heterogeneous datasets, and handling missing data.
– Interest in decision analytics, sustainability, and climate risk.
Desirable but not essential (training will be provided):
– Experience with Python/R, dashboard development, or machine learning.
Other information
Met Office Climate Data Portal
https://www.metoffice.gov.uk/about-us/news-and-media/media-centre/weather-and-climate-news/2023/new-climate-data-portal-makes-data-more-accessible
https://doi.org/10.1109/ICDMW.2008.30
https://doi.org/10.1016/j.eiar.2024.107607
https://doi.org/10.3390/w13101358
https://doi.org/10.3390/w12092379
https://business.leeds.ac.uk/research-cdr/staff/355/dr-sajid-siraj
https://environment.leeds.ac.uk/sustainability-research-institute/staff/1874/dr-andrea-taylor
