An Analysis of Accessible Natural Green Space Provision for the SO50 Postal Code Area

Using the Digimap data download service, CodePoint data in .csv format was acquired for the SO50 District. The .csv file was amended; the column ‘Non-domestic_DPs’ was removed due to it containing a non-readable hyphen. The .csv file was then added to ArcMap using ‘add data’. A coordinate system was added via the tools menu and the ‘add XY data’ function. The X field was selected as easting, and Y northing. The British National Grid was chosen as the coordinate system. The table data was then converted to a shape file (.shp) using ‘export data’ and added to ArcMap. The original, now redundant, SO50 .csv file was then removed.

The shape file provided for the Eastleigh area was then added. From figure 1 it can be seen that a significant number postcode units are present beyond the shape file boundary. Additional areas were therefore acquired.

Figure 1 – Initial MasterMap data coverage (green) and SO50 postcode data (purple ) from Digimap

Using the MasterMap facility additional shape files covering the SO50 district were acquired. Given the size of the area covered, the data was ‘chunked’ and thus needed to be assembled onto one layer. The ‘merge’ function under ‘General’ in the toolbox combined all downloaded data sets into one output. The result can be seen in figure 2.

Figure 2 – SO50 postode data (purple) with initial MasterMap area (green) and ‘merged’ area data (blue)

The shape files contained date relating to natural, anthropogenic and ‘other’ features. Green space is ‘natural’ and a layer with only natural features was needed. With the shape file (figure 2) selected the ‘select by attributes’ function was used from the ‘selection’ menu.

An SLQ query was undertaken with the syntax: “Make” = “Natural” and a new selection was created. An issue still remained in that some features existed that had been classed as natural but were anthropogenic in nature as noted under the ‘Them’ column in the attribute data. Themes such as rail and roads needed to be removed. In order to remove these the same process was repeated but the ‘select by attributes’ function was this time used to extract only natural features according to their ‘Theme’. From the attributes table the ‘natural’ themes were judged to be:

  • Land
  • Land, Water
  • Water
  • Water, Land

The SQL syntax used was:

 “Theme” = “Land” OR “Theme” = “Land, Water” OR “Theme” = “Water” OR “Theme” = Water, Land”

The data was exported to a new layer. Under the symbology tab in the layer properties the data was grouped and colour coordinated via their ‘Descript_2’ values and their theme. ‘Descript_2’ contained the type of plant that each feature was but did not include any information on water based features and ‘general’ land (i.e. agricultural land). ‘Unique values, many fields’ was chosen under the categories heading (figure 3).

Figure 3 – Symbology chosen for the data set

Coniferous and non-coniferous trees, orchards and coppice features were grouped under woodland. All instances of marsh were grouped, as were grasslands, scrubs, water features and general land and the layer saved (figure 4). An appropriate colour scheme was selected.

Natural England’s Standards for accessible natural green space (ANGSt) is used as the benchmark for this exercise (McKernan & Grose, 2007).

These standards recommend that people living in towns and cities should have:

  • accessible natural green space of at least 2ha in size, no more than 300 metres from home
  • one accessible 20 hectare site within 2km of home
  • one accessible 100 hectare site within 5km of home
  • one accessible 500 hectare site within 10km of home
  • one hectare of statutory Local Nature Reserve per one thousand of the population
Figure 4 – Full colour coded map containing only ‘natural’ features

To determine whether or not a postcode meets any of the criteria, the areas of each green space feature are calculated and suitable features extracted, with the distance from each postal unit.

Figure 5 –Add field parametres

Firstly the areas are calculated. In the sorted layer’s attribute table a new field is created and labelled ‘layer’. The type is set to double (figure 5).

The area column is highlighted and in the right click menu, ‘Calculate Geometry’ is selected. The property field is set to area, the coordinate system set to the same as the data source, which should be the British National Grid, and the units set to sq. m. The area column should then show the area in m2 of each feature (figure 6).

Features that match the requirements of a minimum of 2, 20, 100 and 500ha can now be extracted from the layer.

Using the select by attributes function and using the SQL syntax ”Area” > = 20000 the features with only areas greater than 20000m2 are selected. 20000 is used as this equates to 2ha in metres.

Using the export data function a new layer is created using only the features of  greater than 2ha (figure 7).

The symbology and grouping of feature types of this layer was matched to the previous one using the import function in the symbology tab.

Next the process was repeated for areas of greater than 20ha. This time the sytax used was ”Area” > = 200000. The same export and symbology process was reperated for this new layer (figure 8).

Figure 6 – area column added to the attrubute table

When it came to determining areas greater than 100ha it was found that none existed for this data set and such no new layer was created. This rendered a query for 500ha redundant.

Figure 7 – Postcode data (purple) with areas of land greater than 2ha in size
Figure 8 – Postcode data (purple) with areas of land greater than 20ha in size

With the suitable areas defined, appropriate distances must then be selected as defined by ANGSt.

For 2ha an accepted distance is no more than 300m. Using the ‘Near’ function under proximity in the Analysis Tools section of the ArcToolbox this is calculated.

‘Input features’ is defined as the postcode layer and ‘near features’ as the 2ha layer, output units are left as metres. When the function has finished running a new column will have been added to the attribute table for the postcode layer entitled ‘Near_Dist’. The ‘select by attributes’ function is then used to select only features that are no more than 300 metres away from a unit postcode.

The syntax used is “Near_Dist” <= 300. This data is then exported to a separate layer showing only features that are greater than 2ha and postcodes no more than 300m away from a unit post code (figure 9). The postcode selection data is cleared so as not to interfere with the other queries.

Figure 9 – Map layer of areas 2ha in size or more. Postcodes within 300m (blue) and those more than 300m (purple) of an area

The select by attributes function is applied with the syntax “Near_Dist” <= 2000. A new layer is produced showing postcodes within 2km and features over 20ha in size (figure 10).

Figure 10 – Map layer of areas 20ha in size or more. Postcodes within 2km (blue) and those more than 2km (purple) of an area

A summary of accessibility to green space can be produced using the attribute tables and the summary statistics function.

There were 1238 postal code units in total. 857 have access to areas of green space within 300m and 1019 within 2000m. This means that there are at most 381 postcodes without access to any green space of more than 2ha in the SO50 district. This represents 30.8% (3sf) of the total which is a statistically significant amount.

There were 841 unique areas of land of 2ha or more compared to 11 of at least 20ha.  The greatest unique feature had an area of 580,010m2. No postcodes therefore are able to satisfy the third criteria for ANGSt based on this data set. The average size of green space was 50,186m2. This puts it above the first ANGSt requirement of 20,000m2.

From figure 9 it can be seen that general natural land is the dominant land type comprising 692 areas out of the total 841 representing 82.3% (3sf). Water bodies made up just 0.9 % and ‘natural environment’ land made up16.8% (3sf). Whilst a number of postcodes do have ‘access’ under the criteria used here there is no way of determining whether they would in reality. The general land category will include agricultural land, the ownership of which is indeterminate. Access therefore is questionable. A better gauge would use the ‘natural environment’ theme but at just 16.8% it would mean fewer postcodes would meet the access criteria. ‘Natural environment’ will also encompass areas such as peat bog, cliffs, and sand dunes for example, the accessibility of which may be non-existent.

Figure 10 indicates that postcodes have access only to ‘general land’ of 20ha or more under the methodology. As detailed above using general land type can be flawed and in reality it could be that no postcode has access to green space of a minimum of 20ha.

To improve upon the methodology further attribute data is required.  Actual land usage such as whether the land is an SSSI or agricultural in nature would be desirable. Data on whether the land is private or public and access rights should be sought along with details on times and dates that land can be used. For example a 20ha piece of green space would be virtually useless to people if it can only be accessed once a month and whilst would satisfy the ANGSt criteria, it wouldn’t strictly be seen to be accessible to the local population.

With regards to land usage the themes and types selected for use in the accessibility measure was highly subjective. The decision to include coppices and scrubland for example was based on the idea that they are green space. Different outcomes will be produced based on the initial classifications. Determining what constitutes green space is not simple and different studies suggest different criteria. The Urban Green Spaces Task Force (2002) defines green space as one of the following:

  • parks/gardens
  • country parks
  • semi natural urban green space
  • outdoor sports facilities
  • amenity green space
  • allotments
  • cemeteries

Kit Campbell Associates (2001) defines a green space as any land ‘any vegetated land or structure, water or geological feature within urban areas’. In this they include children’s play areas, bowling greens and tennis courts. Under the approach taken in the SO50 study these would have been removed as their ‘make’ would have classed as manmade. This highlights the importance of choosing appropriate criteria for study.

This study has shown that areas of land of greater than 100ha do not exist within distance for the SO50 district. The ANGSt criterion requires such areas of land be within 5km of households. MasterMap data should therefore, in future studies, be acquired to cover such distances up to 10km to determine whether SO50 postcodes satisfy the third and fourth criteria.  The fifth criteria states that there should be ‘one hectare of statutory Local Nature Reserve per thousand of the population’. The data that was used in this study does not contain any information relating to population for each postcode and as such this criterion cannot be tested for.

Further improvements could be made with the inclusion of an integrated transport network (ITN). Appropriate sized areas of green space may exist with 300-2000m of postcodes based on Euclidian distance but in reality it is impossible to travel in a straight line to a given destination. Using an ITN it would be possible to analyse the distances that people would actually have to travel to reach a green space. The use of roads, rail and footpaths could be more accurately taken into account when determining if postcodes do really have ‘access’ to green space.


If you have found any of my essays helpful or interesting please consider making a donation here. Thank you and hope you have enjoyed my writing.

References

DTLR, 2002. Green Spaces, Better Places, Final Report of the Urban Green Spaces Task Force. London.

Kit Campbell Associates, 2001. Rethinking Open Space, Open Space Provision and Management: A Way Forward. Edinburg.

McKernan, P. & Grose, M., 2007. An analysis of accessible natural greenspace provision in the South East. Lewes: Natural England.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s