https://www.findaphd.com/phds/project/leverhulme-doctoral-programme-for-regenerative-innovation-regnr8-i-the-city-as-a-living-lab-novel-monitoring-and-machine-learning-based-modelling-of-trees-for-the-bioengineering-of-the-urban-environment/?p170121
The city as Living Lab – Novel monitoring and ML based modelling of trees for the bioengineering of the urban environment.
Supervisory team: Dr. Matteo Ciantia (SSEN) Dr Alexandra Morel (School of Humanities, Social Sciences and Law) Kevin Frediani (UoD Botanic Gardens) Dr. David Boldrin (James Hutton Institute)
Urban trees are an important nature-based solution to improve the social and environmental benefits of urban ecosystems, which are often not equitably distributed across urban landscapes. Applied research to date has largely focused on the biodiversity, carbon sequestration and wellbeing benefits of urban trees; however, their structural and physiological resilience to an increasingly stressful environment needs to be better understood to ensure safer infrastructure for public spaces. These stresses are manifold, including: elevated temperatures, air pollution, poor soil aeration, soil pollution, high pH and elevated soil salinity due to application of deicing salt in northern latitudes(1). Many of these stresses are expected to worsen with changing climate conditions, therefore, it is essential to improve practitioner understanding and monitoring of the physiological and structural responses of urban trees. Currently, hydraulic monitoring is being used to quantify tree water use in urban parklands to determine their contribution to urban hydrology as well as capture incidences of tree stress.
The tree populations of our urban environments have been largely chosen based upon visual properties and their ease of cultivation. Apart from their ability to grow in the local soil and environmental conditions, there has been a lack of awareness or consideration of their functional traits (2). This project builds on existing research activities in the University of Dundee’s Botanic Gardens, which currently spans testing the residual properties of urban trees as resilient functional structures for the built environment to documenting the ecosystem services and functional traits of the University’s extant tree population. Already these research activities are bridging civil engineering and ecological understanding; however, the new learning from this project will help to develop metrics at the localised level that can inform decision makers and developers to select trees able to withstand these overlapping pressures, particularly in areas of high social deprivation. The project also pioneers an automated monitoring system employing dendrometer growth, soil properties, tensile strength sensors and rootplate deflection data (3) to provide early warnings of structural risks from environmental stressors, contributing to safer public spaces.
By leveraging the interdisciplinary collaboration facilitated by the Living Laboratory at the University of Dundee Botanic Garden, as well as our close working relationships with Dundee City Council, the Eden Project, and the James Hutton Institute, this project will consider:
- 1. Establish whether observed stress responses can provide an early signal of changes in their biomechanical stability.
- 2. Develop an early warning system for urban tree stress that would utilise real time measurements relevant to structural risk and physiological condition. Machine learning (ML) based modelling is envisaged.
- 3. Assess the distribution of urban trees with tolerant traits across measures of social deprivation.
(1) Dmuchowski et al (2020) Strategies of urban trees for mitigating salt stress: a case study of eight plant species. Trees 36: 899-914
(2) Watkins et al (2021) Can Trait-Based Schemes Be Used to Select Species in Urban Forestry? Frontiers in Sustainable Cities. 3: 654618
(3) Marsiglia, et al (2023). Uprooting Safety Factor of Trees from Static Pulling Tests and Dynamic Monitoring. Geotechnical Engineering in the Digital and Technological Innovation Era (1 ed., pp. 218-225). https://doi.org/10.1007/978-3-031-34761-0_27
