Spatial and policy urban indicators for Naarm / Melbourne, Australia,

Full description

Description

This collection contains the input data, analysis configuration and derived outputs for a spatial and policy urban indicator analysis of Naarm / Melbourne, Australia (2025), undertaken using the Global Healthy and Sustainable City Indicators (GHSCI) software with optional Google Earth Engine (GEE) indicators enabled.

The analysis measures neighbourhood-level liveability and sustainability for the urban portion of Greater Melbourne: walkability (walkable neighbourhood population density, street intersection density and a daily living destination score), access within 500 m along the pedestrian network to fresh food, convenience destinations, public open space and regularly serviced public transport, plus the optional Earth Engine indicators for large public urban green space (LPUGS) availability and accessibility, and the Global Urban Heat Vulnerability Index (GUHVI). Results are reported as a population-weighted city summary and as a 1000 m population grid, and are accompanied by a completed urban policy checklist assessing the alignment of Melbourne's urban planning and transport policies with healthy-cities evidence.

Indicators were estimated at sample points spaced 30 m along the pedestrian network within populated grid cells, using a 1000 m walkable neighbourhood and a 1600 m study region buffer, then aggregated to the 1000 m grid and to the city. The pedestrian network was derived with OSMnx from an OpenStreetMap extract; street intersections were taken from Victoria's official road infrastructure dataset in preference to OSM-derived intersections.

Key findings are summarised in the accompanying reports. In brief: the majority of Melbourne neighbourhoods have low walkability and poor access to fresh food and to frequently serviced public transport, with the distribution of walkability and transport access favouring inner and middle suburbs; while 90% of residents have some public open space within 500 m, this falls to 66% for public open space of 1.5 hectares or larger; and approximately 20% of the population live in areas most vulnerable to urban heat, concentrated in the north-western and south-eastern suburbs. Melbourne's policy framework had good coverage of relevant urban planning and transport policies, but lacked measurable targets for over one third of the policy indicators.

What is included

Inputs (AU.zip, ~1.25 GB uncompressed) — the complete set of source data files and the study region configuration required to reproduce the analysis, laid out as expected by the GHSCI software (see "How to replicate", below):

• AU_Melbourne_2025_1000m.yml— the study region configuration: data paths, coordinate reference system, network and analysis parameters, data provenance, licences and citations for every input, and the reporting configuration.

• australia-260112.osm.pbf— OpenStreetMap extract for Australia.
• GCCSA_2021_AUST_SHP_GDA2020.zip— study region boundary (Greater Melbourne).
• SOS_2021_AUST_GDA2020_SHP.zip — urban region definition (Sections of State).
• Australian_Population_Grid_2025_in_GEOTIFF_format/ — 1000 m population grid.
• DTP PTV GTFS Feeds - September 2025/ — public transport timetable feeds.

• DTP-Vicmap-Transport-Road-Infrastructure-2026-02-14/ (and .zip) — road intersections.
• Melbourne 2025 report - 2026-06-16/— the completed urban policy checklist and a supporting policy document.
• images/— the photographs of Melbourne used to illustrate the PDF reports, credited to Sonja Broersen and Carl Higgs (2026).

Outputs (AU_Melbourne_2025_1000m/) — the indicator results and documentation produced by the analysis:

• AU_Melbourne_2025_1000m_indicators_region.csv — population-weighted city-level indicator estimates (one row).
• AU_Melbourne_2025_1000m_indicators_1000m_2025.csv— indicator estimates for each populated 1000 m grid cell.
• AU_Melbourne_2025_1000m_1600m_buffer.gpkg— GeoPackage of all spatial layers used and produced, including the study region and urban boundaries, population grid with indicators, pedestrian network nodes and edges, destinations, public open space and sample point estimates.
• AU_Melbourne_2025_1000m_metadata.yml/ .xml— ISO 19115 metadata recording the full analysis configuration, data provenance and lineage.
• output_data_dictionary.csv / .xlsx — plain-language descriptions of every output variable, its scale (city or grid) and its category.
• AU_Melbourne_2025_1000m_scorecard_statistics.yml— contextual summary statistics.
• PDF reports: policy indicators, spatial indicators, and combined policy and spatial indicators for Naarm / Melbourne 2025; supporting figures and maps.

Input data sources

Contextual statistics reported in the outputs (Gini index, Human Development Index, GDP per capita) are sourced from the World Bank and UNDP, as recorded in the configuration file.

How to replicate the analysis

The analysis runs entirely inside Docker containers, so no local Python environment is required. Analysis of a city the size of Melbourne is computationally intensive and may take a number of hours.

See the following resources for directions on getting and using the GHSCI tools with optional Earth Engine indicators workflow:

• https://github.com/healthysustainablecities/global-indicators/wiki
• https://github.com/healthysustainablecities/global-indicators/tree/earth-engine#optional-indicators-using-google-earth-engine

1. Install Docker Desktop and ensure it is running.

2. Obtain the software. Download the latest version of the Global Healthy and Sustainable City Indicators (GHSCI) software (ee variant; here). We recommend using the most recent release rather than the exact version used here, as it will incorporate subsequent fixes and improvements; results may differ slightly between versions. The version used to produce these outputs is recorded in the metadata.yml/.xml lineage statement.

Because this analysis includes the optional Earth Engine indicators, use the Earth Engine variant of the software (the earth-engine branch/release, launched with global-indicators-ee.bat on Windows or bash ./global-indicators-ee.sh on macOS/Linux).

3. Register a Google Earth Engine project and authenticate, following the instructions at

https://github.com/healthysustainablecities/global-indicators/tree/earth-engine#how-to-generate-the-optional-spatial-urban-indicators.

This is required for the large public urban green space and urban heat vulnerability indicators. To skip these indicators instead, set `gee: false` in the configuration file and run the standard (non-Earth Engine) software.

4. Extract the input data. Download AU.zip from this repository and extract its contents into a new folder named AU within the software's process/data directory, so that the configuration file is located at process/data/AU/AU_Melbourne_2025_1000m.yml and, for example, the OpenStreetMap extract is at process/data/AU/australia-260112.osm.pbf.

5. Launch the software following the directions in the software guide (https://healthysustainablecities.github.io/software/): run .\\global-indicators-ee.bat (Windows) or bash ./global-indicators-ee.sh (macOS/Linux) from the repository root. This starts the analysis container along with a PostGIS/pgRouting database container.

6. Run the analysis. At the prompt inside the container, start Python (or Jupyter Lab via `lab`) and run:

python
import ghsci
r = ghsci.Region('data/AU/AU_Melbourne_2025_1000m.yml')
r.analysis()

7. Generate the outputs (indicator CSVs, GeoPackage, metadata, data dictionary, maps and PDF reports):

python
r.generate()

The photographs used to illustrate the reports are included in `AU.zip` at the paths listed under `reporting: images:` in the configuration, so the reports can be regenerated as published; substitute your own images there if preferred.

Progress and any warnings are written to a processing log in the region's output directory, process/data/_study_region_outputs/AU_Melbourne_2025_1000m, which is also where all outputs are written.

Licensing

Figshare permits a single licence per item. Open Data Commons (ODC) has been selected as the closest available match to the Open Data Commons Open Database License (ODbL 1.0) under which OpenStreetMap data — on which much of this analysis depends — is made available.

The individual input datasets are made available by their respective custodians under their own licences, which are recorded for each dataset in the configuration file (AU_Melbourne_2025_1000m.yml) and in the ISO 19115 metadata accompanying the outputs.

Australian Bureau of Statistics and Victorian Government inputs are provided under CC BY 4.0; the OpenStreetMap extract is provided under ODbL 1.0. Users must comply with the attribution and share-alike requirements of the respective source licences when reusing these data or any derived product. Derived outputs incorporating OpenStreetMap data are subject to ODbL.

The photographs of Melbourne included in AU.zip and reproduced in the PDF reports are the work of Sonja Broersen and Carl Higgs (2026) and are included to allow the reports to be regenerated as published; please retain the credits shown alongside each image.

Related resources

• GHSCI software: https://github.com/healthysustainablecities/global-indicators
• Software guide: https://healthysustainablecities.github.io/software/
• Global Observatory of Healthy and Sustainable Cities: https://healthysustainablecities.org