Mapping Longitudinal Development Pressures in Northern Ireland Based on Machine Learning

The pattern of postwar rural development in NI, particularly the provision of housing in open countryside, has led to significant, but, as yet, undocumented environmental impacts, such as declining water quality and habitat fragmentation or destruction.

This research consolidates ongoing collaboration between QUB's School of Natural and Built Environment and the Office of Environmental Protection (OEP) to develop cutting-edge machine learning (ML) methodologies to create and test a longitudinal data set of one-off housing developments in the NI countryside from 1952 to 2023.

The study utilizes a wealth of data sources, including contemporary orthophotography, historical maps, and aerial photographs, to comprehensively assess the evolving rural landscape.

By employing ML algorithms, the project will identify and analyse the historical trends, the impacts of changing policy regimes and help identify the future challenges of rural development in NI, with a particular focus on environmental protection.

The ultimate goal is to support the OEP with their duties of scrutiny and assist future decision-making by other policymakers and planners, thus contributing to more sustainable development practices.

The project also aims to create transferable ML methodologies that could also be used to develop further important environmental datasets.

The research objectives include evaluating the efficacy of ML algorithms in portraying the temporal dynamics of rural development, developing context-specific ML algorithms tailored to NI's unique housing landscape, and utilizing the resulting dataset to shed light on critical environmental challenges.

Anticipated outcomes extend beyond academic contributions to encompass tangible benefits for government agencies, planners, and the wider society