University of Edinburgh logo British Antarctic Survey logo National Oceanographic Centre logo University of Leeds logo

Projects

Urban forests through space and time

 

Project Summary: 

Urban trees and forests play an important role in both the physical and cultural environment of the UK’s towns and cities [Davies et al., 2017]. However, our understanding of the way this role has evolved over time is limited by a lack of direct observations of canopy coverage [Doick et al., 2020]. This project will shed new light on changes to the extent of tree canopy cover, and access to the benefits provided by trees, in UK cities through a fusion of data sources spanning the past two centuries. 

The project will focus on two case study cities, Leeds and Edinburgh, with contrasting development histories. The local authorities governing both cities have declared climate emergencies; as part of their climate change mitigation and adaptation strategies, Leeds has ambitions to double its current tree canopy cover by 2050 and Edinburgh has a commitment to become a ‘million tree city’ by 2030. Understanding the historical evolution of canopy coverage, and access to it, in the two cities will help inform the purpose, scale and location of future tree planting.  

Aims / Objectives: 

1) Compilation of a detailed chronology to identify data availability for the case study cities over time.  

2) Development of machine learning processes for automatic extraction of tree and woodland feature data from historical sources identified in (1). 

3) Development of machine learning processes for automatic extraction of tree and woodland feature data from earth observation data identified in (1). 

4) Development of digital tree atlas combining historical and modern information for Leeds and Edinburgh. Identification of survivor trees present in both historical and modern records. 

5) Ground truthing of modern digital tree atlas using a citizen science approach, and in-depth surveying of survivor trees identified in (4), in collaboration with project partners. 

6) Development of canopy cover projections for the case study cities that meet the aspirations of local government and maximise equitable access to ecosystem services delivered by trees. 

Central Edinburgh on the OS Town Plan of Edinburgh, Sheet 35, Surveyed: 1852, Published: 1854. Courtesy of the National Library of Scotland.

Methodology: 

Historic maps provide a key source of information on the evolution of both the natural and human-modified environment. The National Library of Scotland (NLS) have a unique collection of digitised historic maps from across the UK, dating back to the 1550s.  Of particular interest for this project are town plans dating to the last half of the 1800s (e.g., here and shown above), on which individual trees have been carefully represented with details that point to species, composition, arrangement and stature. Cataloguing and understanding the data content of these maps by hand is time-consuming and the volume of data makes this a technically challenging process. Modern deep learning tools are now sufficiently highly developed for image segmentation that they may be used to build a data processing chain to automatically locate and geo-reference individual symbols and markings within the map images. The project will also explore the use of satellite data (e.g., from Sentinel-2) to enhance the geo-referencing of the historic maps at a finer scale than is currently performed in order to better geolocate the historic record of individual, or groups of trees, and allow their comparison across different data sources. 

Machine learning techniques will also be applied to remotely sensed information collected in recent years (e.g., aerial imagery, LiDAR) to quantify changes in tree canopy cover over the past decades. By combining present-day remote sensing with the historical records, local authority information on planting dates, and field data collected in the two cities, the student will develop a digital tree atlas for each case study city. The modern digital atlas will be calibrated using a combination of field data collection and hand mapping of high-resolution aerial imagery.  

The digital tree atlas will be used to identify survivor trees that are present in the historic records and still exist in the modern landscape. These survivor trees will be surveyed by skilled citizen scientists who will record their location, biophysical characteristics and accessibility. This information about the survivor trees will be used to determine the benefits provided by these trees through time. The digital atlas will then be used in conjunction with tree planting opportunity mapping to inform future tree canopy expansion in the two cities. 

The student will be based and registered at the University of Leeds but will spend substantial time at the University of Edinburgh and make visits to the project partners. The student will join the Biosphere-Atmosphere Group (BAG) and the Leeds Ecosystem, Atmosphere and Forest (LEAF) centre at the University of Leeds, as well as the Urban Forest Research Group (UFoRG) at Forest Research.