Greenspace#

Greenspace metrics have been developed by the University of Leicester for residential addresses/postcodes for cohorts partnering with UK LLC, where the appropriate permissions are in place.

Some variables cover the entire United Kingdom (UK), while others are limited to Great Britain (GB). Please refer to the table below for geographic coverage of variables:

Caveats:#

• Network polygons are always smaller than circular buffers of an equivalent size. i.e. a network polygon of 300m will always be smaller than a circular buffer of 300m.

• Network polygons are substantially more complex to calculate. Hence, running network polygons for large sizes is not recommended and for distances over 2km, travel is likely to be by car/public transport instead of on foot, and hence network polygons are less likely to be significant. In contrast, for small distances usually travelled on foot, network polygons are likely to give a better representation of accessible greenspace.

• Note that these metrics have been designed for running from address points rather than generalised postcode locations. If only postcodes are available then researchers should take note of the following. In urban areas, postcode polygons are small and hence, results will not vary significantly compared with using an address point. In rural areas, postcode polygons can be very large and hence there could be some differences between an address point and the result using the postcode alone but it depends on the character and variability of the landscape.

• The Ordnance Survey provide an Open Code Point product which uses the delivery point nearest to the mean of all address points within the postcode polygon. This works reasonably well as an address point substitute but will not give exactly the same as using the actual address point. The table below gives a summary of the key greenspace metrics. This represents a core set of metrics that researchers may find most useful.

Note Network polygons are proposed for smaller distances and circular buffers for larger ones. For larger distances of 5km and 10km, only the maximum greenspace size is calculated. This is because for large areas, most other metrics (e.g. NDVI, percentage greenspace) converge to a single value and hence variability decreases. These larger distances are only included to allow for application of Natural England’s ANGSt model on greenspace accessibility (which uses distances of 300m, 2km, 5km and 10km).

Table 2 – Best recommended options of greenspace metrics if exact residential locations are available (NP = network polygons, CB = circular buffer)

Limitations:#

• The greenspace metrics were developed for residential addresses in the Longitudinal Linkage Collaboration (LLC). These can be re-run for any cohort of participants where latitude / longitude point locations are available for participant addresses. They can also be run for postcode mean delivery points but this approach may not be suitable for rural households as, in these cases, postcode mean points can be significant distances from specific households.