Big data / machine learning approaches to de-risk urban heat islands to enhance urban sustainability

Please use reference number: SCEBE/22S/003/RE

This project is available as a 3 years full-time PhD study programme with expected start date of 1 October 2022 

Landcover modifications by cities lead to the urban heat island (UHI) effect, leading to increased economic expenditure, energy consumption, and adverse health impacts. Current planning approaches rarely account for the phenomenon and, lack of detailed data at fine scales on heat risk is a major stumbling block. Additionally, the estimation of heat vulnerability and exposure at local levels as well as local adaptive capacity, are not well developed. The overall effect of these research and practice gaps is that UHI measures are missing in global climate models and metropolitan planning methodologies, leading to weak sustainable development governance.

At the same time, data on proxy measures to estimate heat risk, vulnerability, exposure and adaptive capacity are proliferating. Wearable sensors, crowd-sourced data on weather, fine scale socio-economic data, publicly available land cover / land use data, and protocols for local climate mapping are on the increase. This ‘big data’ along with detailed understanding of urban planning contexts could help de-risk cities of heat vulnerability and contribute to local action to adapt to climate change and/or mitigate the negative consequences of urban growth.

Building on work already completed by the BEAM Centre (https://www.gcu.ac.uk/assetmanagement/beamresearch/sustainablecitiescommunities/) and in collaboration with the SMART Technology Centre, this project will bridge the data science – built environment divide for the common good and sustainable urban environments to solve the problems around urban heat identification and explore their mitigation using nature based solutions (NBS), water and sustainable materials.

The successful candidate will have a urban planning/environmental sciences/sustainability/civil engineering degree and/or data science background (First Class or 2:1 Honours) and a Master’s degree (ideally at least Merit) in a related area or at the interface between the two (e.g. AI, machine learning or sustainability, urban sciences). They will have experience of neural network techniques or willing to learn it as well as experience and knowledge of some quantitative research methods. Prior work in urban sustainability mapping is highly desirable.

Candidates must include an outline of their ideas for exploring big data/machine learning approaches to de-risk urban areas of heat/climate change vulnerability, drawing on relevant literature (via the ‘research proposal’ section of the application form; maximum of 750 words excluding references.