Haiwei Li, Flora

Flora is a Ph.D. Researcher at the University of Cambridge, where she is a Cambridge Trust Scholar and a Research Assistant in the Sustainable Design Group. She is also a Visiting Scholar at ETH Zürich. Her Ph.D. viva is scheduled for April 2026, and she is expected to graduate in June 2026.

Flora holds an M.Sc. in Integrated Building Systems from ETH Zürich and a B.Eng. (Hons) in Building Services Engineering from The Hong Kong Polytechnic University.

Her research focuses on understanding the mechanisms of urban climate and heat mitigation, with a particular emphasis on translating engineering knowledge into future urban planning and design. She is especially interested in environmental design, retrofitting, and optimizing blue and green infrastructure across microclimate, neighbourhood, and regional scales to support more sustainable, climate-resilient cities.

Haiwei Li

Contact Me:

Email: hl618@cam.ac.uk

LinkedIn: www.linkedin.com/in/flora-haiwei-li-flora

Website: Link

ORCID: 0000-0003-2263-7224

ResearchGate: Link

Google Scholar: Link

Featured First-Author Work

Li, H., Zhao, Y., Wang, C., Ürge-Vorsatz, D., Carmeliet, J., & Bardhan, R. (2024). Cooling efficacy of trees across cities is determined by background climate, urban morphology, and tree trait. Communications Earth & Environment, 5(1), 1-14. https://doi.org/10.1038/s43247-024-01908-4

Cooling efficacy of trees across cities is determined by background climate, urban morphology, and tree trait

Urban trees are widely seen as a solution to urban heat, but their cooling performance depends on tree traits, urban form, and climate. This study reviews 182 studies covering 110 cities or regions across 17 climate types. It finds that tree planting can reduce peak monthly temperatures below 26 °C in 83% of cases, with pedestrian-level cooling reaching up to 12 °C through shading and transpiration. In tropical, temperate, and continental climates, combining deciduous and evergreen trees in open urban areas provides about 0.5 °C more cooling than single-species planting. In arid climates, evergreen trees perform best in compact urban settings, guiding tailored greening strategies.

Li, H., Bardhan, R., & Debnath, R. (2024). Heatwave interventions must reduce invisible gendered challenges in the Global South. PLOS Global Public Health, 4(10), e0003625. https://doi.org/10.1371/journal.pgph.0003625

(Working paper) Li, H., Sen Sharma, K., Tiwari, A., Kapoor, R., Prabhu, J., Debnath, R., Bardhan, R. Co-creating retrofit-led interventions provide gender-specific heatwave adaptation in vulnerable communities. (2026)

Heatwave interventions must reduce invisible gendered challenges in the Global South

Heat-health risks in the Global South are intensifying due to climate change, urbanization, and inequality, with women facing disproportionate impacts. Limited data and underreporting make gender-specific risks hard to measure, complicating intervention design. The article highlights how “invisible infrastructures” — social, cultural, economic, and built systems — shape exposure, access to cooling, and resilience. Poor housing, inadequate ventilation, unsafe water access, and limited healthcare worsen risks, especially for low-income women. It argues for context-specific, gender-sensitive interventions such as passive cooling, reliable water and electricity, clean cookstoves, and improved emergency care to reduce vulnerability and strengthen community resilience.

Li, H., Raman, H., Abong, B., Bange, T., Zhao, Y., Brayne, C., Bardhan, R. (2025). Mosquito prevention strategies can improve indoor heat stress in hot climates: A case of traditional Kenyan homes. Environmental Research: Health, 3, 035004. https://iopscience.iop.org/article/10.1088/2752-5309/adedab

Mosquito prevention strategies can improve indoor heat stress in hot climates: A case of traditional Kenyan homes

This study examines the combined health risks of indoor overheating and malaria in rural settlements in Kisumu, Kenya, using data from 138 residents. It assesses a low-cost housing intervention—open eaves with screens—through measurements, simulations, and surveys, following the IPCC risk framework. Results show that overcrowding, cooking, and poor ventilation greatly increase heat and health vulnerability, with over 10% of homes experiencing extreme heat stress. Open eaves and windows can reduce overheating, but windows may also increase mosquito entry. Screened open eaves offer a promising balance, though design details must also address indoor air pollution and resident behavior.

Li, H., Zhao, Y., Bardhan, R., Chan, P. W., Derome, D., Luo, Z., Ürge-Vorsatz, D., & Carmeliet, J. (2023). Relating three-decade surge in space cooling demand to urban warming. Environmental Research Letters, 18(12), 124033. https://doi.org/10.1088/1748-9326/ad0a56

Relating three-decade surge in space cooling demand to urban warming

This study examines 30 years of climate-driven increases in urban cooling demand across Hong Kong, Sydney, Montreal, Zurich, and London. Using cooling degree hours, it shows a clear upward trend caused by climate warming, urban heat islands, and extreme heat events. The analysis also finds that raising indoor cooling setpoints can significantly cut energy use: increasing the setpoint by 1°C can save around 20% energy under many conditions, though savings decline as background temperatures rise. For example, shifting from 26°C to 27°C saves about 10%, while 22°C to 23°C saves over 20%. The study calls for behavioral adaptation and stronger policy measures to curb cooling demand.

Li, H., Zhao, Y., Sützl, B., Kubilay, A., & Carmeliet, J. (2022). Impact of green walls on ventilation and heat removal from street canyons: Coupling of thermal and aerodynamic resistance. Building and Environment, 108945. https://doi.org/10.1016/j.buildenv.2022.108945

Impact of green walls on ventilation and heat removal from street canyons

This study shows that heat removal from street canyons can significantly improve when the approaching airflow is buoyant due to upwind surface heating. Using water tunnel experiments with simultaneous velocity and temperature measurements, the researchers found that buoyant flow can double heat removal by reducing suppression at the canyon opening. The effect depends on street canyon geometry and the distance from the upwind heated surface. Conversely, cooler upwind surfaces can create a local cooling effect by sending cooler fluid into the canyon. The findings emphasize that approaching-flow buoyancy is crucial for understanding urban heat island processes and designing effective mitigation strategies.

Li, H., Zhao, Y., Bardhan, R., Kubilay, A., Derome, D., & Carmeliet, J. (2023). Time-evolving impact of trees on street canyon microclimate. Journal of Physics, Vol. 2654, No. 1, p. 012145. https://doi.org/10.1088/1742-6596/2654/1/012145

Time-evolving Impact of Trees on Street Canyon Microclimate

This study investigates how the cooling effects of linden street trees change with age during heatwaves. Using CFD simulations of an idealized street canyon in Zurich, it models fully coupled airflow, heat, moisture, radiation, shading, and transpiration for trees aged 10–100 years. The results show that young trees aged 10–20 years provide limited pedestrian-level cooling during extreme heat. The strongest heat mitigation occurs with trees aged 30–60 years, when leaf structure is most effective. Because tree growth also affects street ventilation, the study highlights that the cooling benefits of urban trees evolve over time and should be considered in long-term urban planning.

Li, H., Zhao, Y., Liu, J., & Carmeliet, J. (2021). Physics-based stitching of multi-FOV PIV measurements for urban wind fields. Building and Environment, 108306. https://doi.org/10.1016/j.buildenv.2021.108306

Physics-based stitching of multi-FOV PIV measurements

This study presents a new physics-based method for stitching together multiple particle image velocimetry (PIV) flow fields in large urban wind tunnel experiments. Called regional-flow stitching, it matches neighboring fields of view by comparing local flow patterns using a vorticity-based reference window and minimizing normalized root-mean-squared error. Verified with planar and stereo PIV data, the method is more efficient than traditional stitching, reducing processing time to about 46% while maintaining accuracy. It also supports 3D visualization with ParaView. The approach offers a faster and more robust tool for combining large urban flow datasets in experimental fluid dynamics research.