These Technical Notes accompany the Green Space Index and supplement the published Frequently Asked Questions with more detail on how the Index has been calculated and compiled.

We continue our work to develop and expand the Green Space Index and would welcome any comment by email to info@fieldsintrust.org.

The aim

Through the Green Space Index we analyse publicly accessible local park and green space provision within Great Britain. The Index follows on from our Revaluing Parks and Green Spaces research released in 2018 and supplements our wider policy work in the parks and green spaces sector. The May 2019 release of the Green Space Index was the first time such a measure had been published to take stock of the quantity of local park provision across Great Britain and the annual updates support our evidence-led approach calling for parks and green spaces to be revalued for the benefits they contribute to society. We believe that green spaces are good, do good and need to be protected for good.

Having established baseline data through the May 2019 and May 2020 releases of the Green Space Index, future iterations of Ordnance Survey's Greenspace map and updates to population projections will enable us to track trends in provision over time.

Scope

The Green Space Index covers Great Britain and analyses publicly accessible local park and green space provision as mapped by Ordnance Survey, with a more detailed discussion below of how their data is used. Not within the scope of the Green Space Index is regional park and green space land, such as national parks and common land, or other aspects of green infrastructure such as canal towpaths and grass verges.

Northern Ireland is not included in the Index because park and green space data is not made available by Ordnance Survey of Northern Ireland. If such data becomes available in the future we will look to expand the Green Space Index to cover the province.

Green Space Data

The Green Space Index is compiled using the OS Open Greenspace product, with the May 2020 release of the GSI using the April 2020 release of OS Open Greenspace.

Within the OS Open Greenspace product we use the site extent polygons in all cases. From the site extent layer we exclude four typologies which are not deemed to be within scope of the Green Space Index. The excluded typologies are: Allotments & Community Growing Spaces; Cemetery; Religious Grounds; and Golf Course.

Typologies which are within our scope and which have been retained for analysis are: Bowling Green; Other Sports Facility; Play Space; Playing Field; Public Park or Garden; and Tennis Court.

We have also removed six areas of land, detailed below, which appear within the dataset but we do not deem to be within scope. We are reliant for the Index on the dataset published by OS and have not performed a nationwide review of the accuracy of this data as it is not within our scope to do so, but do believe these examples to be indicative of errors which may currently exist within the data. These have been fed back to OS, who are accepting feedback on errors within the data as they continue their verification process. We will continue to work with them to support this process where we are able.

Land removed from in-scope typologies as not deemed to be within scope of Green Space Index:

  • 9E891041-3B3D-2C71-E053-A03BA40A7146 - Whinlatter Forest Park
  • 9E89103E-2B7F-2C71-E053-A03BA40A7146 - Rutland Water
  • 9E89103D-E018-2C71-E053-A03BA40A7146 - City of Manchester stadium and surrounding tarmacked areas
  • 9E8911BB-0CE6-2C71-E053-A03BA40A7146 - National Cycling Centre
  • 9E89105E-867E-2C71-E053-A03BA40A7146 - Principality Stadium
  • 9E89105E-85C5-2C71-E053-A03BA40A7146 - Principality Stadium (Cardiff Arms Park)

Within the data there are widespread instances of overlap and duplication of park and green space land. These fall into two categories:

  • Overlap where small features have been overlaid on larger features.
    • For example a polygon representing a play area sitting on top of a larger polygon representing a park. This leads to double-counting of the area of the play area. We have run analysis to identify areas of overlap and erase the area of the smaller feature from the larger feature, thus removing double-counting but retaining the mix of typologies that may exist on a space.
  • Duplication of entire polygons where two or more identical polygons sit on top of one another.
    • Common examples of this are a polygon representing a tennis court layered on top or below an identical polygon representing an other sports facility. We have controlled for this by analysing identical polygons and only retaining the first one listed within the dataset.

We have taken every effort to remove all instances of overlapping and duplication from the data as accurately as possible within our scope. As previously noted we are reliant on the dataset published by OS and will continue to work with them to support their verification process where we are able. We have noted accuracy-related revisions between the April 2019 and April 2020 releases of the OS Greenspace data and it is for this reason we do not feel able to draw comparisons between the findings of the first two releases of the Green Space Index. If you find any inaccuracies whilst browsing the OS Greenspace product you can report these using the contact form on the OS website, selecting "Map errors and omissions".

Other data used in the Green Space Index

Population projection data
Whilst the majority of the Green Space Index is compiled using the OS Open Greenspace product, some measures are calculated by combining analysis of this with population data.

For the May 2020 release of the Green Space Index we have in all instances used data for the actual 2020 population, to provide a more accurate product. This is a change from the May 2019 release of the Index in which we used data from the 2011 Census.

For the population figures for 2020 and beyond, we used data provided by Geolytix, who specialise in data and demographics. The projections for years beyond 2020 are modelled from a household first perspective. They use OS address data and combine data on all large-scale planning developments with Census data on household compositions. This is done at postcode and Output Area level, but the figures are then constrained to 2016-based ONS population projections at local authority level.

In analysing the impact of projected changes in population of green space per capita, we used the projections data calculated by Geolytix. They have used the above methodology to provide the projections up to and including 2025 and for projections beyond 2025 apply change at local authority level in 2016-based ONS projections to give precise projections at local authority and national/regional level. We have not analysed effects of changing population below local authority level beyond 2025 as this data is not considered reliable at such a granular scale for such timeframes.

Geographic and statistical boundaries
When visualising data we have produced a separate WebApp for each NUTS1 statistical region in Great Britain. Within each App we use 2011 Lower Layer Super Output Area (LSOA) boundaries for England and Wales and 2011 Data Zone (DZ) boundaries for Scotland. All analysis and calculations to compile the Index and data for display on WebApps have been carried out using full resolution clipped boundaries. To aid load times and provide as smooth a user experience as possible, however, we have presented the data on the WebApps using generalised clipped and super generalised clipped boundaries in England and Wales, which toggle based on the zoom extent of the map. Generalised and super-generalised Data Zone boundaries are not available so we have simplified the full resolution boundaries using the Douglas-Peucker algorithm with the same tolerances as used in the generalised and super-generalised LSOA boundaries.

To aid users' geographic understanding we have overlaid December 2019 Ward boundaries at a generalised clipped level and December 2019 Counties and Unitary Authorities boundaries at generalised clipped and ultra generalised clipped levels (toggling at same zoom extents as LSOAs as detailed above).

To help visualise the data and ensure consistency of comparison across nations and regions we have shown the data using graduated colours classified by geometric interval for the entirety of Great Britain.

Population not living within a ten-minute walk of a park or green space

A ten-minute walking distance is a well-established measure of an acceptable distance for a resident to be from their nearest park or green space. Calculating where and how many people are not served by a park or green space within a ten-minute walk can offer a useful addition to judging provision levels with overall quantity alone not capturing distribution of green space across communities. The ten-minute walking distance is referenced in our Green Spaces for Good strategy document, whilst in the United States the "10 Minute Walk" campaign led by The Trust for Public Land in partnership with the National Recreation and Park Association and Urban Land Institute aims to ensure that every US citizen has a park within a ten-minute walk of where they live.

In order to calculate a ten-minute walking distance we use 800m as a proxy. This indicative walking distance is defined within our Guidance for Outdoor Sport and Play document as well as being approximately equivalent to the half-mile distance used by the Trust for Public Land in their ParkScore product.

We calculate the number of people not living within a ten-minute walk of a park or green space by buffering from the edge of the site extent polygons used in the analysis and working out the population living in postcodes which do not sit within the buffers created.

We have chosen to buffer from the edge of the site extent polygons rather than use the access point data also provided within the OS Open Greenspace product as we feel this more accurately captures both formal and informal access points in publicly accessibly local green spaces, which often will be fully accessible from the majority of the perimeter.

We have buffered as the crow flies from the edge of the site extent polygons because performing route analysis on the entire dataset would have been inhibitively complicated. Consequently we expect the figures calculated for number of people not within a ten-minute walk of a park or green space may in fact represent an under-estimate and we will look at ways the accuracy of this figure can be improved for future releases of the Green Space Index.

GSI Score

The GSI Score is our unique measure of green space provision and analysis of whether a geographic area meets minimum quantity guidelines. Our 2015 Guidance for Outdoor Sport and Play document sets out recommended benchmark guidelines for formal and informal outdoor space.

Guidance for Outdoor Sport and Play gives minimum guidelines of 1.90 hectares of outdoor sports provision per 1,000 population, 0.25 hectares of equipped/designated play area provision per 1,000 population and 1.40 hectares of park, public garden and amenity green space provision per 1,000 population. The natural and semi-natural guideline of 1.80 hectares of provision per 1,000 population is not a typology that is deemed to be within scope of the OS Open Greenspace product and thus not included in calculations.

To calculate the GSI Score we analyse the quantity of outdoor sport provision, being the Bowling Green, Other Sports Facility, Playing Field and Tennis Court typologies within OS Open Greenspace, per 1,000 people in a geographic area and divide this figure by 1.90 to give a measure where 1 is a minimum level of provision. We do the same for parks and play using the Public Park or Garden and Play Space typologies, dividing the provision per 1,000 people by 1.65, in line with benchmark guidelines for these two typologies in Guidance for Outdoor Sport and Play. We combine parks and play into a single measure of provision to take account of both formal (e.g. equipped play) and informal (e.g. incidental play) opportunities which may feature in parks but which may be captured within either typology.

The measure of outdoor sport provision and the measure of parks and play provision are combined with equal weighting to give an overall GSI Score where 1 represents a minimum level of provision. A geographic area having a GSI Score greater than one indicates sufficiency in quantity of provision (based on minimum benchmark guidelines) across different typologies of park and green space, but as already discussed variances will still remain within the geographic area when considering population within a ten-minute walking distance of a park or green space.

Developing the Green Space Index

This release represents the second iteration of the Green Space Index and continues work to establish a baseline measure across the main indicators of park and green space provision. Due to accuracy-related improvements in Greenspace data and a change in our base for population data, both detailed above, we do not yet feel able to draw comparisons between releases. We will continue to refresh the Index annually following each April release of the OS Open Greenspace product. Whilst, as already discussed, it is appreciated that there is an ongoing verification process of the data within the OS Open Greenspace product, we hope that the Green Space Index will in due course allow us to begin to track trends and changes over time in provision of parks and green spaces across Great Britain.

We also aim to make available greater geographic information on the portfolio of parks and green spaces protected in perpetuity with Fields in Trust through an open data portal, so that this data can be used by others within the sector and wider when analysing the impact of parks and green spaces and the importance of protecting these spaces in perpetuity. Over the last year we have improved the accuracy and precision of the data we hold on our protected estate, although more work is needed on fully digitising this data and we continue to seek funding to enable us to fully realise our ambitions in this regard.

Attributions and credits

The Green Space Index is compiled using National Statistics data © Crown copyright and database rights (2020), NRS data © Crown copyright and database rights (2020), OS data © Crown copyright and database rights (2020) and Royal Mail data © Royal Mail copyright and database rights (2020).

We would like to place on record our thanks to Esri UK, Geolytix and Ordnance Survey for their continued support of our policy and advocacy work.