Scottish Government Urban Rural Classification 2007-2008

Annex A: Scottish Government Urban Rural Classification Methodology

Scottish Government Urban Rural Classification 2007 - 2008

1. Description

1. The Scottish Government ( SG) 6-fold and 8-fold urban rural classifications are intended to provide a consistent way of defining urban and rural areas across Scotland. Two main criteria have been used to develop the SG 6-fold and 8-fold urban rural classifications: population as defined by the General Register Office for Scotland ( GROS) and accessibility based on drive time analysis to differentiate between accessible and remote areas in Scotland.

2. Datasets Used

• Settlements 2006

The General Register Office for Scotland ( GROS) is responsible for the definition of settlements; a simplified methodology is as follows: GROSSmall Area Population Estimates ( SAPE) together with information from the Royal Mail Postcode Address File ( PAF) were used to classify 2006 postcode units as high or low density. This information was then used to identify areas of contiguous high density postcodes with a population of 500 or more that make up a Settlement. Details of the methodology used for the Mid-2006 Population Estimates for Settlements can be found on the GROSSettlements and Localities website.

• Road and ferry network

To calculate the drive times ESRI ArcGIS software extension Proterritory was used. This software requires the creation of a road grid; this section describes the methodology used to create this road grid.

Road grid input files:

• OSCAR Route Manager dataset 2005 (1:10,000) - road network for Scotland from Ordnance Survey
• Major roads and motorway in North England from Ordnance Survey Strategy 2005 (1:50,000)
• Ferry routes derived from Ordnance Survey Strategy 2005 and Scottish Government 2005
• Settlements 2006 from General Register Office for Scotland

To create a 20 metre square pixel road grid for SIMD 2006 the roads from OSCAR covering Scotland, the roads covering North England and the Ferry routes were merged and the Settlement 2006 dataset was used to define urban areas.

While the roads datasets only required reclassification (see table below). The building of the ferry routes required more work: reclassification, editing and speed calculation. A total of 101 ferry routes were created, ranging from 225 metres (Noss Sound - Gungstie) to 355 km. (Aberdeen - Lerwick).

Initially each ferry route was assigned a minimum estimated travel time. This information came from either Ordnance Survey Strategy 2005 or the individual ferry service internet sites. The next stage was to add ferry waiting times. A 30 minutes waiting time was selected after comparing the values for estimated waiting time for several different ferry routes. The ferry routes compared ranged from 10 minutes to 3 hours in estimated travel time but all had a 30 minutes waiting time.

From the travel time (including waiting time) and calculated length for each ferry route, the speed was calculated. Speeds ranged from between 2 and 22 miles/hour, where the slowest speeds in general correspond to the shorter ferry routes due to the relative long waiting time. The ferry routes were then classified into 10 speed classes.

The final class given to each ferry route and associated speed can be found in table 1. The roads were reclassified into similar road types resulting in the following classes and speeds for urban and rural areas, also found in table 1.

3. Methodology for creating SGUR Classification

The first stage in creating the SG classifications was to look at settlement populations. Settlements were categorized into the following 4 groups:

• Large urban areas: Settlements of 125,000 or more, which have a wide range of services.
• Other urban areas: Settlements of 10,000 or over and below 125,000, which have a good range of services.
• Small towns: Settlements of 3,000 or over and below 10,000. These are likely to be smaller towns which are unlikely to contain a full range of services.
• Rural: Settlements below 3,000 and those not living in settlements. These are unlikely to contain many services.

The next stage was to distinguish between accessible and remote areas. This was done by creating a 30 minute drive time for the 6-fold classification and a 30 and additional 60 minute drive time for the 8-fold classification from settlements of size 10,000 or more. Thus creating the following definitions of remoteness:

• Accessible: 30 minutes or less drive time from a settlement of 10,000 or over
• Remote: more than a 30-minute drive time from a settlement of 10,000 or over (or between a 30 and 60 minute drive time for the 8-fold classification)
• Very Remote: more than a 60 minute drive time from a settlement of 10,00 or over (8-fold classification only)

The 30 and 60 minute drive time boundaries were created using drive time software from the population weighted centroids of all settlements with a population of 10,000 or over; centroid co-ordinates were supplied by GROS. Settlements with a population of 10,000 or more located in England but within a 30 minute drive time of Scotland were also included in the in the analysis (i.e. Berwick-upon-Tweed and Carlisle).

Drive time calculations are based on average travel speeds. Each of the road classes (e.g. A roads or B roads) in the road network has an assigned speed. This speed is the average for that class (figures come from DETR) but the process does not take into account peak and non-peak travel times.

Table 1. Road class and average speed used in the analysis.

The 30 and 60 minute drive time boundaries were combined to form a single dataset.

Each settlement was initially classed depending on its estimated population. With the creation of a drive time dataset each settlement was also classed depending on its distance from a settlement with a population of 10,000 or more. In some cases drive time boundaries split a settlement. The method used here treats settlements as whole entities and classify them based on the drive time their centroid fell in.

Finally, settlements and drive time datasets were combined to form one national dataset containing both the 6 and 8-fold definitions.