The analysis and interpretation of spatial data has become an important aspect of urban forest practice and management. Spatial boundaries divide land into manageable pieces. At the broader scale these may be local government areas, suburbs or wards, and at the smaller scale they may distinguish roads from parks and private lots. Urban foresters have typically used these boundaries and areas to summarise and present data related to the trees they manage. However, trees exist throughout the urban landscape, and management issues such as diversity, resilience, and equitable access extend across and override these artificial boundaries. New methods for spatial analysis used by the City of Sydney are explained and demonstrated, with the results providing a view beyond these boundaries, allowing greater insight, and better management of their urban forest for the entire community.
Spatial data and boundaries
Data is becoming increasingly important in every aspect of our lives, with a growing expectation for strategic decisions to be based on meaningful analysis and interpretation of data. For professionals managing urban forests it is no different. The ability to compile and interpret data is a core skill, essential for identifying and reacting to current needs, and planning future action to address long term management issues. The success of urban forestry programs also relies on our ability to present data and complex matters to the community, in ways they can easily comprehend and relate to.
If you are managing a large population of trees you may have a variety of spatial datasets available to you. Two common sources of data are aerial assessments of tree canopy cover and inventories of tree assets, with each able to be used in different ways gain insight into the urban forest. Both of these data sources usually have a spatial component, in that they are typically represented on maps as either areas or dot point coordinates.
Figure 1: Aerial assessment of canopy cover and tree inventories as examples of spatial data.
Source: City of Sydney Council.
Boundaries are also represented on maps, dividing land and data into manageable pieces. Boundaries are used to define land for a variety of purposes or contexts, including governance and administrative, legal, or environmental. They can provide a useful common basis for summarising and presenting all types of spatial data. Urban foresters have typically used these boundaries, and the land areas they define, to summarise and present data related to canopy cover or trees they manage. Boundaries also exist at a variety of different scales. At the broader scale these may be, local government areas, catchments, suburbs or wards. At a smaller scale, boundaries may distinguish roads from parks and private lots.
Boundaries and scale
Within urban forest plans or strategies produced by local governments over the last 10-15 years it is typical to find aerial assessments of canopy cover summarised to the suburb scale. For example, tables, charts or maps are used to present the percentage canopy cover within each suburb or neighbourhood, as a subset of the overall canopy cover for the entire local government area. While this is useful to portray large scale variations in canopy cover over a large area, it does not accurately reflect the amount of tree canopy an individual has in their local area. The percentage canopy cover is quoted for the suburb as a whole, with variations within suburbs not presented. It’s possible for most canopy within a suburb to be concentrated within parkland, with very little in the immediate area where people live or work. Similarly, if an individual lives close to a suburb boundary, they may question which canopy cover data is most relevant to them? These issues and questions prompt the analysis and summary of canopy cover data at smaller scales.
Mesh blocks are the smallest geographical area defined by the Australian Bureau of Statistics (ABS), with the aim in urban areas for each to contain about 30 dwellings (Australian Bureau of Statistics, 2016). It is becoming increasingly common for canopy cover to be analysed and presented at this scale. An example is the dataset NSW Urban Vegetation Cover to Modified Mesh Block 2016 (NSW Department of Planning, Industry and Environment, Figure 2). Data summarised at this smaller scale provides insight into canopy cover variability that exists within and across suburb boundaries, and allows for canopy cover data to be easily compared and contrasted with other socio-economic data available from the ABS. Mesh blocks also have disadvantages. Boundaries of mesh blocks typically follow the centreline of roads, causing problems if you also wish to assess the proportion of canopy provided by the street network or private properties as isolated land uses. There can also be great variability in the size of mesh blocks in urban areas due to variabilities in land use and population densities. This can make it difficult to compare all areas being assessed on a uniform like-for-like basis.
Figure 2: Sydney canopy cover per ABS mesh block. Accessed via online data platform: https://nsw.digitaltwin.terria.io/
There is also the option of presenting canopy cover at an even finer scale, such as per land parcel or lot. A USA based consultancy (PlanIT Geo, 2020) produces maps for clients that indicate the canopy cover for individual private lots (Figure 3). While this may help property owners to understand their relative contribution to overall canopy cover, as for mesh blocks the variability in lot sizes can make it difficult for a uniform comparison between areas.
Figure 3: Example of canopy cover per land parcel. Source: https://storymaps.arcgis.com/stories/7af8fdf671634f75bfcc17cb6c84c296
The above examples of spatial analysis, with land defined by administrative boundaries at a variety of scales, represent the current conventional practices within urban forestry. The remainder of this paper will present new options and examples of spatial analysis developed by the City of Sydney to look beyond these conventional boundaries, to gain greater insight, and to manage for more equitable and resilient urban forest outcomes.
An alternative approach – the “Urban Tapestry” method
A heat map is a data visualisation technique used to display spatial data over a uniform matrix of coloured cells. The cells are coloured to visualise aspects of the underlying data. The City of Sydney has adapted this general technique into something we have called the Urban Tapestry (or perhaps “Tapestree”) method. The method involves the establishment of a grid of uniformly spaced reference points spread across the entire local government area. A 100 metre grid spacing was chosen to provide a total of 2659 reference points (Figure 4). At each of the reference points, data is gathered from a buffer radius area surrounding it. The size of the buffer radius is variable and may be customised to suit the type of data and purpose of the analysis being undertaken. Data within each buffer area is analysed to produce summary statistics for each point location, with the reference points coloured to visualise the results. Where reference points are located close to the boundary of the local government area, the buffer areas may extend beyond our borders to consider the influence of the neighbouring council area.
The technique provides a consistent but also flexible method to summarise and compare data uniformly across an area, without any of the disadvantages associated with administrative boundaries and variations in the size of land being assessed. It allows us to look beyond these artificial borders, to explore issues such as equitable access to canopy cover or other greening, and the resilience of our urban forest.
Access to canopy cover for health benefits
Research published by Astell-Burt and Feng (2019) suggests that community health outcomes benefit from higher levels of tree canopy cover within 1.6km of people’s homes. Applying this research to calculate the percentage canopy cover within a 1.6km buffer radius of each of the sample points, we are able to visualise the community’s access to canopy cover at this broad scale, and compare the benefits available to the community at different points within our city area. The variable access to canopy can clearly be seen within the City of Sydney (Figure 5), but analysis at this scale may be better visualised across broader regions or entire metropolitan areas.
The results are presented for points within our local government area only, but the analysis includes an assessment of canopy cover beyond our borders where it falls within the 1.6km buffer areas, using publicly available canopy cover data published by the NSW Department of Planning, Industry and Environment. The urban forest does not end at our boundary, so it was important to assess the canopy cover that exists beyond it, that our residents also have access to and may rely on.
Access to canopy cover for urban heat mitigation
Numerous research studies detail the benefits of tree canopy in mitigating urban heat. Work by Ziter et al (2019) suggests that groups of trees that combine to provide greater than 40 per cent canopy cover, at the scale of a city block, can reduce local ambient air temperature by more than 1.3oC. In order to apply this research, and to visualise the community’s access to canopy at the smaller scale of a city block, a 100 metre buffer radius was used in the canopy cover analysis. The results (Figure 7) provide a much finer-grain assessment of canopy cover within local neighbourhoods and the immediate surrounds of where people may live or work. The effect of different sized mesh blocks or land parcels is eliminated, allowing for a more equitable representation of canopy cover across an entire area.
The results are relatively easy for the community to interpret. They can be directed to their closest reference data point based on their address to learn about the levels of canopy cover in their surrounding local area, and identify areas that are likely to be hotter or cooler during extreme heat weather events.
Applying the same analysis to historical canopy cover data provides insight into how access to canopy cover at this scale has changed over time (Figure 8). The effects of specific planting projects can be observed and compared to others. Increases and decreases in canopy cover can be visualised within suburbs and across suburb boundaries. The analysis also demonstrates how the composition of the urban forest can change within relatively short periods of time.
Figure 8: Access to tree canopy within a 100m buffer radius at 2008, 2013, and 2019. Specific examples of canopy cover change are highlighted, and can be related to management issues or project outcomes.
Tree inventory data can also be analysed and presented using the Urban Tapestry method. Data typically captured within tree inventories, such as species, tree size, and tree age, provides insight into the composition of the trees under management. It also allows for detailed analysis of diversity in regards to each of these parameters.
When assessing the diversity of an urban forest, it is recommended for the assessment to be conducted at the neighbourhood level as well as across an entire municipality (Leff, 2016). If aiming for optimum tree diversity it is recommended that at the neighbourhood level no single species to represent more than 10% of total tree population; no genus more than 20%, and no family more than 30%. Applying this recommendation in a typical way would usually result in data being summarised within the confines of neighbourhood or suburb boundaries. But our suburbs are not all the same. They are of different sizes, and their boundaries may get in the way of gaining insight into our tree population that would help us plan and manage for the future. The problems that may arise from inadequate diversity (e.g. pest or disease outbreaks etc.) are not interrupted by suburb boundaries, so the analysis of diversity should not be interrupted by these boundaries either. The Urban Tapestry method allows for analysis of diversity at the recommended scale, but applied uniformly and consistently across boundaries and throughout an area.
An 800m buffer radius was determined to be the scale equivalent to an average suburb or neighbourhood in the City of Sydney. The City has a complete inventory of all trees within parks, streets, and City owned properties. The analysis involved identifying all inventoried trees within the buffer area surrounding each of the 2659 sample reference points across the entire LGA. A sum of the total number of trees was calculated, along with the sum of the most common family, genus and species. Percentages were then calculated to represent diversity, with each of the sample points coloured to reflect the result relative to the recommended limits. The results are presented in Figure 9.
It can be observed that achieving the optimum recommended diversity is a challenge but not impossible. The analysis highlights where the diversity problem areas are, those that are optimal, and those close to being optimal.
The analysis also informs us of the exact nature of the problem. The results of the complete diversity analysis at each reference point are available at the click of a button. The top ten most common family, genus and species within the buffer area surrounding each point are displayed, along with their respective totals and percentages. This data not only points out any problem, but can also assist with identifying species found in the local area that are not adversely affecting diversity and may be suitable alternatives for planting.
The analysis has a live connection to the tree inventory, so as the tree population changes over time as a result of routine inspections and planting, the diversity analysis and visual results are automatically updated. Tree age or size diversity may be assessed and presented in similar ways.
We suggest a critical look at the influence of boundaries in the analysis of urban forest data, and encourage practitioners to look beyond them when possible. The urban forest extends beyond these artificial borders, so our analysis must also. Alternative approaches (such as those presented here) can assist in identifying challenges and opportunities, and guide future action towards more equitable and resilient urban forest outcomes. They may also assist with communicating these issues to the local community and help them to engage in urban forest programs. These new approaches may also lead us to rethink how we measure success, with the potential for future canopy cover targets to reflect a community’s access to canopy cover within a specific scale from where they live.
The City of Sydney is proud to have developed these ideas and examples of data analysis. The authors thank Carl d’Entremont and Matthew Sund of the City of Sydney Spatial Information Team for their assistance in bringing them to life.
Astell-Burt T, Feng X. (2019) Association of Urban Green Space With Mental Health and General Health Among Adults in Australia. JAMA Netw Open. 2019;2(7):e198209. doi:10.1001/jamanetworkopen.2019.8209 https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2739050
Australian Bureau of Statistics (2016) Australian Statistical Geography Standard (ASGS): Volume 1 – Main Structure and Greater Capital City Statistical Areas, July 2016. Accessed online July 2020. https://www.abs.gov.au/ausstats/[email protected]/Lookup/by%20Subject/1270.0.55.001~July%202016~Main%20Features~Mesh%20Blocks%20(MB)~10012
Leff M. (2016) The Sustainable Urban Forest. A Step-by-Step Approach. Davey Institute / USDA Forest Service. Accessed online at http://www.itreetools.org/resources/content/Sustainable_Urban_Forest_Guide_14Nov2016.pdf
NSW Department of Planning, Industry and Environment. NSW Urban Vegetation Cover to Modified Mesh Block 2016. https://www.planningportal.nsw.gov.au/opendata/dataset/2b0dd699-9c23-40eb-b70f-1bcfdbc3f34a
PlanIT Geo (2020) Urban Tree Canopy Assessments https://storymaps.arcgis.com/stories/7af8fdf671634f75bfcc17cb6c84c296 Accessed 21 August 2020.
Ziter, C.D, Pedersen, E.J, Kucharik, C.J and Turner, M,G. 2019, Scale-dependent interactions between tree canopy cover and impervious surfaces reduce daytime urban heat during Summer. Proceedings of the National Academy of Sciences USA, 116(15) 7575-7580