Official Statistics

Access to blue space in England

Published 29 May 2025

Applies to England

Last updated: 2025

Latest data available: 2025

Contact

Enquires on this publication to: access.statistics.feedback@defra.gov.uk

Tel: 03459 335577 (Defra enquiries) Find out more about call charges at – GOV.UK

Responsible Statistician: Rhidian Thomas

Website: Access to Nature statistics – Gov.UK

Key messages

• 86% of households in England were found to have access to blue space within a 15 minute walk, when using a broad definition of blue space.
• However, only 28% of households were found to have access when the definition was limited to including substantial blue spaces only.
• When using a broad definition of blue space, a higher percentage of rural households (94%) had access to blue space compared to urban households (85%), though this was reversed (21% in rural and 30% in urban) when considering only substantial blue spaces.
• These statistics build upon previous work to determine the percentage of households that have access to green space in England within a 15 minute walk. We will continue to develop this work with the aim of producing combined access to green and blue space statistics in future.

Official statistic in development

As this is an official statistic in development, we welcome feedback on the novel methods used in the development of this indicator. For example, feedback on whether this new indicator measures something users feel should be measured, and how well it does so. To give feedback, email the production team at access.statistics.feedback@defra.gov.uk.

Introduction

These statistics describe the percentage of households that are within a 15 minute walking distance of various types of blue space for small neighbourhoods in England. For this publication, blue space is broadly defined as natural and artificial water environments which can provide health and wellbeing benefits, although there is no universally recognised definition and the specific benefits are not yet fully understood. This work represents the first time a comprehensive dataset of blue space access points has been created for the whole of England using a standardised methodology and is a continuation of our recent publication exploring Access to green space in England. Importantly, these access points identify areas where walking alongside or near blue space is possible, not where the water can be entered for recreation. In this document we refer to these walkable areas alongside blue spaces as ‘accessible waterside’. Accessible in this context means reachable by foot along a path, road or public right of way, it does not mean the watersides are necessarily accessible to people with health conditions or impairments. Previous mapping by Natural England identified water sources that were likely to be accessible based on proximity to public rights of way, urban paths or nearby green space, but did not define points of access or calculate the distance from households to blue space access points.

In producing these statistics, we calculated the distance to blue space for every household in England measured along a network of walkable streets and paths. The results are summarised by the Census for England and Wales Middle layer Super Output Area (MSOA), although results by smaller Output Area are available in the published datafile (both using Census 2021 geographies). For further information regarding census geographies see the Method for generating statistics section or the ONS Census 2021 geographies page.

The network model used to calculate walking distances from households to blue space access points was run several times with different types of blue space included. We refer to these different model runs as Scenarios (further detail on scenario parameters can be found following Figure 1):

• All blue space (Scenario 1): includes all blue spaces that can be accessed on foot which are at least 50 m² (0.005 ha) in area or at least 50 m in length. Access points were included irrespective of the length of walkable waterside route which they provide access to.
• Substantial blue space (Scenario 2): includes blue spaces that can be accessed on foot which are at least 0.5 ha in area and for which the length of walkable route alongside the waterbody is 250 m or longer. This is a subset of blue space access points used in Scenario 1.
• Substantial blue space along paths and smaller roads (Scenario 3): includes substantial blue spaces as defined in Scenario 2, but filtered to remove those which require walking on A or B roads to experience them.

Figure 1: Access points included in Scenarios 1, 2 and 3.

Notes for Figure 1:

• Each panel in the figure shows the same set of blue spaces and routes, along with a changing set of access points to show which were included in each scenario.
• In the Scenario 1 panel, all access points are included. This includes access points to waterside along A and B roads, smaller roads and paths as well as access points to small (under 0.5 ha) pieces of blue space and small sections (under 250 m) of accessible waterside.
• The Scenario 2 panel is the same as the Scenario 1 panel, except some of the access points included in Scenario 1 have been removed. Two access points are removed as they provide access to a piece of blue space under 0.5 ha, whilst a further two are removed as they provide access to a section of accessible waterside under 250 m. This leaves six access points which provide access to larger areas of blue space and longer sections of accessible waterside along A and B roads, smaller roads and paths.
• The Scenario 3 panel is the same as the Scenario 2 panel, except some of the access points included in Scenario 2 have been removed. Two access points are removed as they provide access to a section of accessible waterside which requires walking on an A or B road to experience the blue space. This leaves six access points which provide access to larger pieces of blue space and longer sections of accessible waterside along smaller roads and paths.

Households with an access point within a 1 km distance threshold were considered to have access to blue space. The 1 km threshold was chosen to represent an approximated 15 minute walk at a typical walking speed of 4.0 km/h; this aligns with the distance threshold used in the Access to green space in England publication. Blue spaces smaller than 50 m² in area or at least 50 m in length were not used to generate access points in any scenario, as these could include small and temporary waterbodies (see Caveats for more details).

Scenario 1 includes all blue space access points that were generated. These access points can be associated with relatively small waterbodies and very small sections of walkable route. For example, where a footpath includes a small bridge crossing a narrow stream there will be an access point at the bridge, even if the rest of the footpath is not alongside the stream. We believe the inclusion of these access points, regardless of the length of accessible waterside (i.e. the area alongside a waterbody that can be accessed on foot), is important to reflect the potential sedentary nature of blue space use. However, it is not appropriate to include these if the aim is to consider active blue space experiences.

Scenario 2 only considers larger blue spaces (at least 0.5 ha in area) where a waterbody can be walked alongside for a minimum distance of 250 m. There is no accepted standard for blue space access equivalent to the Accessible Greenspace Standards for green space, and evidence on the relationship between blue space size and its impact on health and wellbeing is limited and inconsistent. Therefore, when selecting the waterbody size and accessible waterside length thresholds for Scenario 2, a combination of specialist knowledge, local experience and understanding of the data was applied. Ultimately, a 0.5 ha filter for waterbody size (roughly equivalent to four Olympic swimming pools), and a 250 m filter for accessible waterside that can be accessed on foot (equivalent to a circular walk around a 0.5 ha waterbody, and providing roughly a four minute walk at a typical walking speed of 4 km/h) were applied. These filters were chosen to be roughly in line with those used in the access to green space statistic, but with a smaller size requirement to reflect that in this context, blue space is experienced from around and alongside a waterbody, rather than within it, as for green space.

We generated access points to blue space where both paths and roads (excluding motorways) are alongside waterbodies. However, the included roads may not provide equal pedestrian experiences depending on how much traffic a road services. Scenarios 1 and 2 both treat a walk along all types of roads (except motorways) as equal blue space experiences. Scenario 3 attempts to address this by discounting blue spaces along A or B roads. These roads are defined under the Government’s road classification guidance as being ‘of significance to traffic’ and are therefore likely to service a high volume of vehicles. Blue spaces along smaller roads of low significance to traffic were still included in this scenario, as these were thought likely to provide a reasonably positive walking experience.

As there were no pre-existing standards for blue space access that could be applied, we particularly welcome feedback on the different scenario parameters used.

Access to blue space in England

The percentage of households in England with access to blue space varied between 28% and 86%, depending on the definition of ‘blue space’ used (Figure 2). This wide range reflects the impact of varying the size and type of blue space included within the models. For instance, in Scenario 1, a very high percentage of households in England were found to have access to a broad definition of blue space (86%) due to the very broad definition of ‘blue space’. This percentage was lower when the data were filtered to focus on substantial blue spaces, as in Scenario 2 (28%). Removing substantial blue spaces which involve walking on A or B roads, as in Scenario 3 (28%) had a very limited impact on the percentage of households with access when compared to Scenario 2. Reported percentages were rounded to the nearest whole number and the difference between Scenarios 2 and 3 was 0.6%.

Figure 2: The percentage of approximately 25,800,000 households in England with access to blue space, calculated using three definitions of ‘blue space’.

Across MSOAs, there was a great deal of variation in access to blue space, even at the very local level. The median percentage of households in a Census of England and Wales MSOA with access to blue space was 97% for Scenario 1, 14% in Scenario 2 and 13% in Scenario 3 (Figure 2). The distribution of access by MSOA was highly skewed in all three Scenarios; but in different directions for Scenario 1 compared to Scenarios 2 and 3. In Scenario 1 where all blue space access points are considered, many MSOAs had high percentages of households with access. However, in Scenario 2 (substantial blue space) and Scenario 3 (substantial blue space with no walking along A or B roads roads) many MSOAs had low percentages of households with access.

Figure 3: The distribution of percentage of households within MSOAs with access to blue space, calculated using three definitions of ‘blue space’.

Notes for Figure 3:

• The area under the curve represents the distribution of MSOAs across the gradient of percentage of households with access to blue space. The area under the curve is coloured to highlight this distribution, with the darkest area representing the central 50% of the data, medium blue 75% of the data, and the lightest blue 95% of the data distribution.
• The median of each distribution is marked by a vertical black line.

In terms of differences between rural and urban areas, when a broad definition of blue space was used (Scenario 1), the percentage of households with access was higher in rural areas (94%) than urban areas (85%). When only substantial blue spaces were considered (Scenarios 2 and 3), access to blue space tended to be slightly higher in urban than rural areas (Table 1).

Table 1. The percentage of households with access to blue space, calculated using three different definitions of ‘blue space’ and split between rural and urban areas.

Scenario	Rural	Urban
1. All blue space	94	85
2. Substantial blue space	21	30
3. Substantial blue space along paths and smaller roads	21	29

Discussion

These results show that the way in which ‘blue space’ is defined has a significant impact on the percentage of households that are deemed to have access in England. Large differences in access were found between Scenario 1, which broadly includes smaller blue spaces, and Scenarios 2 and 3, which only include substantial blue spaces. This may be due to the uneven distribution of access points to substantial blue spaces, which are often associated with river networks or the coast. This pattern of distribution means many households are unlikely to have access to any substantial blue spaces within 1 km (approximately a 15 minute walk). By contrast, smaller waterbodies are common throughout England, and so many more households have access to one within 1 km, resulting in the high percentage of access reported in Scenario 1. However, the smaller waterbodies included may not provide the same blue space ‘experience’. For example, a small, narrow stream will provide a different sensory experience to a wide, open river or lake.

The exclusion of blue spaces which require walking along A and B roads made limited difference to the percentage of households with access to blue space. This suggests that most blue spaces which have access points that require walking along A or B roads also have access points along paths or smaller roads. Importantly, the A and B roads were still included as possible routes for households to get to access points.

In Scenario 1 (all blue space), a slightly higher percentage of households had access to blue space in rural areas compared to urban areas. However, this difference was reversed in Scenarios 2 and 3 when only substantial blue spaces were included. Historically, settlements were often formed around waterways due to the opportunities for transport, agriculture and industry. Therefore, urban areas are more likely to be near larger waterbodies (rivers, canals, estuaries or the coast). This means urban areas are less affected by the uneven distribution of larger waterbodies throughout England when it comes to access as defined in Scenarios 2 and 3.

Scenario 1 used a smaller size threshold for blue space than what was used for green spaces in the Access to green space publication statistics. The broadest scenario in the green space publication included a 2 ha minimum for green space, compared to a 50m² (0.005 ha) minimum for blue space used in the broadest blue space scenario. Although these size thresholds are not directly comparable, as blue space and green space may be used in different ways (the use of blue space in this context may be more sedentary, as the spaces are walked around, not within), the difference is important to account for when interpreting the results. Additionally, the 2 ha threshold used for greenspace is backed by evidence from the Accessible Greenspace Standards for green space. Whereas, for blue space very limited evidence was available and the 0.5 ha threshold was selected using only specialist and local knowledge. We welcome feedback on the threshold used for blue space, please contact us at access.statistics.feedback@defra.gov.uk.

When interpreting the results, there are important data and methodological constraints to consider (see Caveats for more details). The access points that were generated are based on likely points with which access to the blue space can be inferred from the data, though access is not guaranteed. Therefore, this may lead to an overestimation of access if some of these points do not provide access in reality.

Scenarios 2 and 3 may also underestimate access due to blue spaces appearing fragmented in the data even though in the real world, users would experience them as a larger, uninterrupted blue space. For example, someone walking along a river with bridges may have an uninterrupted blue space experience; whereas in the data, each bridge separates the blue space into distinct smaller blue spaces. These smaller blue spaces may be removed when filtering for only substantial blue spaces, whereas if the river was counted as a single combined blue space it would qualify for inclusion in the scenarios which select for substantial blue spaces (see the Development plan for proposed improvements).

In this analysis, blue space access points were considered equal in terms of both quality and the potential number of people they could serve. This approach enables us to baseline the national picture on access to blue space. However, we acknowledge that the quality of blue spaces and the experiences they provide vary greatly depending on factors such as their capacity, popularity and recreational use. Quantifying the quality of blue space in a universal way across the different types used in this work is challenging, but could be explored in the future. Further analyses could be conducted to understand how many households share a particular blue space as their nearest in the network model.

Providing data at a fine granular level, such as Output Areas, allows for the investigation of spatial patterns in the data, and ultimately understanding where interventions may be targeted. For example, identifying clusters of Output Areas that have very little blue space provision may provide the best opportunity to either improve access to existing blue spaces or introduce a new blue space. However, it is worth noting that reported low levels of access could be due to limitations in the data available at the time this analysis was done (see Caveats for more details). Our Development plan aims to build on this work and improve the quality and coverage of data used.

Official statistics in development designation

Our statistical practice is regulated by the Office for Statistics Regulation (OSR). OSR sets the standards of trustworthiness, quality and value in the Code of Practice for Statistics that all producers of official statistics should adhere to. You can read about how Official Statistics in Defra comply with these standards on the Defra Statistics website.

This publication is an official statistic in development. Official statistics in development are official statistics that are undergoing a development; they may be new or existing statistics, and will be tested with users, in line with the standards of trustworthiness, quality, and value in the Code of Practice for Statistics.

Details of how we plan to develop these statistics are laid out in the Development Plan. We particularly welcome feedback from users on the methodology and presentation of the statistics set out in this release, and our future plans for development. For example, feedback on whether this new indicator measures something users feel should be measured, and how well it does so.

Background and methodology

A shortest path algorithm was used to find the closest blue space access point to each household, given the walkable routes available. This type of model has three major components:

Source locations – the places where the shortest path calculation begins. For this statistic, source locations are households.

Destination locations – the places where the model aims to find the shortest path to. For this statistic, destination locations are blue space access points.

Network – consisting of links and nodes. Links represent transport infrastructure in line form, e.g. roads and paths, and nodes show the locations at which they interconnect. Together they form a traversable network, where each link has a length and two associated nodes. The path with the smallest sum of lengths between source and destination locations forms the shortest path.

More detail for each of these components is provided below.

Source locations - households

Unique Property Reference Number (UPRN) data available in the Ordnance Survey’s AddressBase Plus (Table 3) product were used to identify the location of residential properties. UPRNs were selected with the class codes given in Table 3 (found in the Technical Annex) and where the ‘state’ was given as ‘In Use’. Each UPRN has an associated geometric point used when integrating them as source points in the network model. See Households and node relationships for more information on integrating households into the network.

Destination locations - access points

Access points are point locations that show where a person is likely to access a particular type of blue space. Importantly, access in this context is defined as being able to see and experience blue space; not to enter it for recreational activities.

There was no pre-existing England-wide blue space access point dataset to use for this analysis and associated statistics. Therefore, one was created.

The first step to generating a blue space access point dataset was identifying blue space of interest and preparing the blue space datasets. Three blue space datasets were used:

• Ordnance Survey water features – water
• Ordnance Survey water network – water link
• Ordnance Survey water features – tidal boundary

The ‘water’ and ‘water link’ datasets were used to represent non-coastal inland blue space, and the ‘tidal boundary’ to represent coastal blue space. All three datasets were filtered to remove non-relevant features and features which did not meet the size threshold. Where available, the ‘water’ dataset was used in preference to the ‘water link’ dataset. For further details on the above, see the Technical annex.

To identify where access is available route data is also required. Four route datasets were used:

• Ordnance Survey (OS), Paths
• Ordnance Survey (OS), Roads
• Natural England, Public Rights of Way (PROW)
• Natural England, England Coast Path

The paths, roads and PROW datasets were combined to form a single general routes dataset and the England Coast Path was kept separate as a coast only route dataset.

To locate accessible waterside, sections of routes which passed within a defined proximity of blue space were identified (Figure 4). The threshold proximity used was changed depending on the type of blue space and route. These different thresholds were applied by buffering the blue space by different amounts:

• Inland blue spaces (all blue space excluding the tidal boundary) were buffered by 10 m. General routes (OS paths, OS roads and public rights of way) which intersected this 10 m buffer were identified as inland accessible waterside.
• Coastal blue spaces (tidal boundaries) were buffered by 50 m. Sections of the England Coast Path which intersected this 50 m buffer were identified as coastal accessible waterside. The rationale for this is that as the purpose of England Coast Path is accessing the coast, we can assume that in general along this path you are accessing coastal blue space. This justifies the 50 m buffer zone.
• Coastal blue spaces (tidal boundaries) were also buffered by 20 m. General routes (OS paths, OS roads and public rights of way) which intersected this 20 m buffer were identified as coastal accessible waterside. The rationale for using a 20 m buffer rather than the 10 m buffer as used for inland blue spaces, is because the coast is an expansive blue space which provides visual access from a greater distance. It is also likely that coastal routes, such as those along clifftops or promenades are further than 10 m from the tidal boundary, although they arguably provide access to the coast. Desk-based validation found using a 20 m buffer worked well to limit issues with barriers to visual access such as walls and buildings caused by larger buffer distances but still capture the majority of route sections providing an actual blue space experience.

The created inland and coastal accessible watersides were combined into a single accessible waterside dataset, which was used to calculate accessible waterside length (see accessible waterside section in Technical annex) and to produce blue space access points. Access points were produced by generating a point at each end of a section of accessible waterside. The produced access points were refined (see Technical annex for details) to produce the final blue space access point dataset.

Figure 4: Approach used to generate blue space access points

Notes for Figure 4:

• Figure shows the process used to generate access points, with each panel showing a different stage of the process.
• In panel A, a diverging route and two blue spaces are shown. Some sections of the route pass nearby to the blue spaces. In panel B a buffer zone around each of the blue spaces is identified by a hatched area. In panel C only sections of routes which pass within the blue space buffer zone are shown; these sections are identified as accessible waterside. In panel D, an orange square is shown at each end of the accessible waterside sections. These orange squares are the blue space access points.

Network

A network consisting of nodes and links was constructed using the pandana package in Python. Figure 5 shows an example network. Each link in the network has an associated length (in meters), as well as a start and end node. Nodes are used to connect links, and also represent start and end points for the shortest path analysis. Two datasets were used to construct the network; Ordnance Survey MasterMap Highways Network Paths and Roads. These data are provided in a network ready format. Roads, paths and connecting links (which indicate when roads can be accessed from paths and vice versa) are provided with corresponding start and end node identifiers. Lengths of each link are provided for roads and paths, while for connecting links a length of 1 m was manually added. Connecting links are often not physical features, so 1 m was deemed an appropriate corresponding value, as a value >0 was required for the model to run. Motorways were excluded from the roads data before building the network.

Figure 5: The components used in the shortest path analysis.

Notes for Figure 5:

• A diagram showing the relationships between households, network links, network nodes, blue space and blue space access points.
• Links and nodes are joined together to make a continuous set of roads and paths, households are located along one edge, and a blue space access point is shown where a link meets a blue space.

Method for generating statistics

To quantify the distance between households and access points, using infrastructure such as roads and paths, a network-based approach to analysis was undertaken.

Given the large amount of computing resource required for the shortest path calculations (due to the density of roads and paths in some parts of England), a tiling approach was used (Figure 6). A regular 5 x 5 km polygon grid was created across the extent of England. This grid was then clipped to the boundary of England (buffered by 10 km) to identify and retain only the terrestrial grid squares. The 10 km buffer was used to remove any ambiguity in coastal areas, retaining grid cells with any likely household locations.

For each grid cell, the following was undertaken:

1. Extract the households that intersect the grid cell.
2. Extract the blue space access points (relevant to the given scenario) that intersect the grid cell and any of the 8 neighbouring grid cells. For coastal grid cells, there will be fewer than 8 neighbouring grid cells.
3. Extract the network data that intersect the grid cell and any of the 8 neighbouring cells.
4. Create a network with the extracted network data using the python package pandana.
5. Identify the nodes in the network that relate to the households. These are now referred to as source nodes. See section on Households and node relationships.
6. Identify the nodes in the network that relate to the blue space access points using the closest straight line distance. These are now referred to as destination nodes.
7. Calculate every unique pair of source nodes and destination nodes.
8. Run the shortest path algorithm for every unique pair of source and destination nodes. This algorithm sums the length attribute of each link in the network that it traverses when calculating a path, resulting in a total length (in meters) for each shortest path.
9. For each source node, select the shortest distance recorded to any of the destination nodes.
10. Join the shortest distances back to the household data, so that every household has an associated shortest distance.

For a given scenario, once all the above steps were undertaken for every grid cell, the outputs were combined, producing a table of all households in England and their shortest distance to a blue space.

Households and associated model output data were then spatially joined to Output Area geographies and summary statistics per Output Area were calculated.

The ONS produces a rural/urban classification scheme for Output Areas, the most recent of which uses the 2021 Output Areas. We have used the 2021 Output Areas to aggregate the results of the shortest path analysis and mapped this on to the corresponding ONS rural/urban classification.

Figure 6: Tiling approach taken in the shortest path modelling.

Notes for Figure 6:

• A 5 km by 5 km box surrounded by eight boxes of the same size in a grid.
• In the central box are data points showing households, network links, network nodes and blue space access points.
• In the surrounding boxes, only network links, network nodes and blue space access points are present.

Scenarios

Varying combinations of different types of access points and the size of corresponding blue spaces were used to provide destinations in each of the scenarios. Below are the details of which combinations were used, which are summarised in Table 2.

All scenarios use a 1 km distance threshold between household and blue space access point. This distance was chosen to represent an approximated 15 minute walk at 4.0 km/h.

Table 2. A summary of the thresholds and types of blue space included in Scenarios 1, 2 and 3.

	1. All blue space	2. Substantial blue space	3. Substantial blue space along paths and smaller roads
Blue space size threshold	50 m2 or 50 m in length	0.5 ha or 250 m in length	0.5 ha or 250 m in length
Distance threshold	1 km	1 km	1 km
OS water link	Included
OS water feature	Included	Included	Included
Access points near A and B roads	Included	Included

Caveats, limitations, and uncertainties

Access points

Access points represent likely access and we cannot be certain there is public access at every point. The access point dataset set was created from national datasets, and only limited ground truthing has been carried out using local knowledge. There are several reasons access points may not represent actual access:

• There may be barriers between the route and blue space, for example a wall, fence, trees or small building.
• Roads labelled as restricted access have been removed, but no such tagging is available for paths. This means access points associated with paths which are not Public Rights of Way (PRoW), may not be accessible to the public. The existence of depicted blue space access points does not create any right of access.
• The access point may relate to a small or narrow water body or course which on the ground may not be physically or visually accessible and may be temporary in nature.

Conversely, some genuine blue space access points may be missing from the generated access point dataset for the following reasons:

• Routes alongside blue space may provide visual or auditory access to the blue space from further than the distance threshold used (10 m for inland and 20 or 50 m for coastal waterbodies). In these cases, no access points would be generated where in the real-world access is possible. This is more likely to be an issue for substantial blue spaces, which are more likely to provide visual or auditory access from greater distances.
• Blue spaces which are in Accessible Green Infrastructure but not in close proximity to a route do not show up as accessible by foot. In this case the blue space access point dataset could be improved using specific methods to generate access points for blue space inside Accessible Green Infrastructure. See the Development plan for proposed improvements.
• In Scenario 3, some access points may be ‘missing’ where long routes (in particular PRoW and coast path routes, which tend to have long uninterrupted line representations in the data) have an access point generated on each end, but can also be accessed via an A or B road between those two end points. In this case, the access point on the A or B road would provide secondary access to an included stretch of waterside along a path or smaller road. However, in Scenario 3 this access point would not be included.

For all scenarios, filtering has been based on the size of the blue space itself and the available length of blue space experience. Desk-based validation showed this worked well, however there are some caveats:

• Blue space size filters were applied indiscriminately to the shape and type of blue space. This may not be appropriate as, for example, the blue space experience walking alongside a long, narrow river and around a circular lake which have the same area may be quite different.
• The blue space size filter applied to the data may not be truly representative of a real-world size filter as waterbodies can be fragmented in the data where they are not in the real world. For example, by a bridge, lock or other change in typology or type of blue space.
• Waterside length that is accessible on foot was used as a filter for certain scenarios, and this was based on the summed length of collections of accessible by foot watersides within 20 m of each other. For a limited number of collections, the calculated accessible by foot waterside may be overestimated, and some blue space access point included erroneously. This overestimation may occur where parallel accessible by foot watersides both contribute to the accessible by foot waterside collection length. For example, a path and a road both running parallel to a river for 100 m would generate two 100 m sections of accessible by foot waterside. This would translate to a 200 m accessible by foot waterside length although in reality the blue space experience is only 100 m in length.
• For Scenario 1 in particular, some access points and waterside access may be identified where a bridge crosses a watercourse or where a route terminates near a waterbody. In such circumstances the implied access may be restricted to being visual only, with no opportunity to walk alongside the waterbody.

The blue space access points dataset does not include any access points which are directly associated with motorways. Similarly, for Scenario 3 access points directly associated with A and B roads were removed. However, in both instances, access points in the vicinity of these roads have not been removed.

Walkability

Where possible, walkability has been factored into the path and road network data used. Motorways have been removed from the data so that roads of these types are not deemed as possible routes when the shortest path is calculated. This therefore leaves the assumption that all other road types are suitable, which may not be the case, for instance dual carriageways. In the future we would like to include data on pavements to make our assumptions of ‘walkability’ more accurate.

Definition of a household

Unique Property Reference Numbers (UPRN) were filtered such that only residential properties were included (see the Technical Annex). As described by the Office for National Statistics in their census geographies description. Each Output Area in England is made up of between 40 and 250 households and usually has a resident population of between 100 and 625 people. MSOAs are made up of groups of Output Areas. They comprise between 2,000 and 6,000 households and usually have a resident population between 5,000 and 15,000 people. There are over 6,800 MSOAs in England. In this analysis, in a very small number of Output Areas, there is a greater number of households due to the presence of certain types of accommodation. For example, some Output Areas encompass halls of residence at a university campus and therefore contain a greater number of households than many other Output Areas in the dataset. Equally, a small number of Output Areas contain very few households.

Distance from households to nodes

A methodology has been implemented in these analyses to account for the distance more accurately from the ‘front door’ of a household to blue space access points, even if nodes in the network are some distance away (see Technical Annex for more details). While this correction has been applied to a majority of households, there are some which do not have the corresponding network relationship information to make the calculations and therefore the adjustment is not applied. In this analysis, 0.09% of households do not use this method to adjust distance to blue space calculated, and the straight line distance from household to the nearest associated road node was used instead. For these households this likely leads to an underestimation of the distance to the nearest blue space.

Distance from access points to nodes

The blue space access point dataset created includes route attribution. However, due to complexities in this relationship which are not present in the address to node relationship it was more difficult to apply a method to calculate access point to node distance using this information. Instead, the nearest node (using straight line distance) to each access point is used as the destination node. In some cases, this may add excessive distance to the results while in other cases it may be an underestimation as a straight line distance does not account for the layout of infrastructure such as paths and roads. See the Development plan for details on plans to improve this using the available route attribution.

Border areas

These statistics are produced for England. However, as the borders with Scotland and Wales are fully accessible, in reality, occupants of households may cross national borders to access blue space. The generated blue space access point dataset covers England, and a 5 km cross border coverage into Wales and Scotland. These therefore provide potential destinations in the model for houses near a national border.

Rights of way integration

Ordnance Survey Mastermap Highways (Roads and Paths) data does partially represent the public rights of way network in England, especially in urban and suburban areas where rights of way are likely to be physical paths and roads. However, it does not provide full coverage, and therefore some areas will have a lack of coverage in the network part of the model.

Rights of way data collated by Natural England are more complete, but are not ‘network ready’, i.e. they are not provided with topological relationships to Ordnance Survey link and node data. These data have been used to generate access points where relationships with rights of way are present, but the rights of way have not been integrated into the network. See Development plan section for plans to pursue this work. Consequently, the model likely overestimates distances between households and blue spaces, especially in rural areas.

Disconnected parts of the network

Small parts of the roads/path network are disconnected from the wider network as a whole. This could be a natural feature of the data, or due to the tiling approach taken or due to the removal of links such as motorways. If a household or access point is associated with a node in a disconnected part of the network, then shortest path calculations in these areas may be incorrect.

Development Plan

As an official statistic in development, we intend to continue to improve the datasets and methodology used in the analyses as well as the outputs produced. The development of the statistics will be guided by feedback from users, while we also plan to address the areas listed below.

Developments planned for the next statistical release provisionally to be published in 2026:

• This work was built upon the existing Access to green space in England publication by investigating access to blue space. In the next update, we plan to consolidate and finalise this work into a combined Access to green and blue space statistic.
• Currently, the roads network used in all scenarios includes restricted access roads. We plan to add weightings to the network to strongly discourage the use of such roads in shortest path calculations. In this release, the decision to include them has been based on the fact that, for some households, the nodes that form part of restricted access roads are the source nodes for the shortest path analysis. Therefore, removing these features would reduce the accuracy of shortest path calculations for those households.
• Ordnance Survey road network data contains an attribute which describes the presence of pavements alongside roads. We plan to assess the quality and then integrate this into the network model, weighting roads of certain types that have pavements more favourably than those without. An assessment of the coverage of these data is required.
• Ordnance Survey road and path network data contain an attribute which describes the elevation gained when traversing each segment of path or road. We plan to assess the quality and then integrate these data to weight routes which involve elevation gain, accounting for the extra time required to walk up hill.
• Currently, distances between blue space access points and network nodes are not incorporated into the shortest path calculations. We plan to implement a similar method used for address-to-node relationships to better account for the distance between these features.
• Currently, blue space size filtering is done based on the size of a single piece of blue space in the data. However, in reality a blue space experience may involve walking amongst many nearby but disconnected smaller blue spaces. In order to capture these interconnected blue space experiences we plan to add size filtering based on a blue space ‘collection’ size, where a blue space collection is a grouping of blue spaces in close proximity. This will improve alignment with the method used to calculate accessible waterside length (see Technical annex for details).
• Currently, where blue space exists within green space, access points are only generated to the blue space if documented routes pass near the blue space. However, in many green spaces a pedestrian is able to walk around freely and would be able to access the blue space without using a specific route. We plan to add access points to these blue spaces which exist inside freely accessible green spaces.

Longer term development plans:

• In this access to blue space statistic, we used access points that represent publicly accessible locations alongside or nearby waterbodies as destinations in the network model. We would like to explore the possibility of developing an additional access point dataset which represents public access onto or into a waterbody for recreational activity across England.
• Currently, results from the analyses are presented in terms of household units. We are keen to explore whether we can incorporate the latest 2021 census data to produce statistics at the population level. In future development of this work, we plan to explore household to population relationships and will assess whether a robust relationship between the two can be established and applied to the data.
• We would like to further refine our household classification by exploring whether the Ordnance Survey AddressBase Plus data can be linked with other data such as Valuation Office Agency data to further refine the assumption that an address is an occupied residential property.
• Rights of way data are only partially represented in the Ordnance Survey Mastermap Highways (Paths and Roads) data. We plan to convert rights of way data compiled by Natural England into a network ready dataset, remove duplications in the Ordnance Survey data and run scenarios that utilise a fuller complement of rights of way.
• We plan to explore if it is possible to integrate activity data from third-party sources to potentially generate new access points, but also validate the assumptions made in our access point generation.

Linking outputs to spatial ONS data

The data published as part of this statistics release are non-spatial. However, we have provided it in a format that can be joined to various spatial datasets, specifically census geography data provided by the Office for National Statistics. Once joined, the data can then be used for further spatial analyses and viewed in Geographic Information System (GIS) software. Below is a short, worked example on how to acquire and join the non-spatial statistics with geographic data:

1) Download the data file provided as part of this stats release. 2) Extract the relevant dataset and export as a CSV file. 3) Acquire the relevant spatial dataset from the ONS Open Geography Portal. For example, Medium Layer Super Output Area (MSOA) 2021 spatial data can be downloaded in various spatial formats such as Geopackage, Shapefile or GeoJSON here. 4) Given the statistics are provided at Output Area (OA) level, aggregate the data accordingly so that the statistics are presented at MSOA level. For example, group by MSOA code and sum the total number of households and total number of households with access. 5) Join the grouped statistics data to the spatial data using the MSOA21CD as the joining variable. 6) Export the joined data as a new spatial file, and use for visualisation or further analyses.

Acknowledgements

Thank you to the many people and organisations who have contributed by providing data and useful insights which contributed to developing the method behind these statistics. Specific thanks to The Rivers Trust for their proof-of-concept work on blue space access point generation.

Technical Annex

Datasets

This section describes the datasets that were used in the analyses and where appropriate, how they were transformed.

Blue space access points

The blue space access point dataset was produced and used to produce these statistics. Provisionally, we anticipate this dataset to be published on the Defra Data Services Platform later this year. Once published, the link will be added here. In the meantime, please contact access.statistics.feedback@defra.gov.uk to request a copy of the data.

Ordnance Survey AddressBase Plus

Ordnance Survey published version 12/09/2023 (Epoch 103).

These data consist of point geometries with a variety of attributes allowing users to classify and subset addresses, and link the data to other Ordnance Survey products such as MasterMap Highways. More information can be found at https://www.ordnancesurvey.co.uk/products/addressbase-plus.

This dataset was used under the Public Sector Geospatial Agreement.

© Crown copyright and database rights [2023] OS [AC0000805307]

© Improvement and Development Agency for Local Government copyright and database rights [2023]

Table 3 shows the AddressBase Plus class codes used to create a subset of the Unique Property Reference Number data.

Table 3. AddressBase Plus class codes used to create a subset of UPRN data.

Class Code	Class Description
RD	Residential Dwelling
RD01	Residential Dwelling - Caravan
RD02	Residential Dwelling - Detached
RD03	Residential Dwelling - Semi-detached
RD04	Residential Dwelling - Terraced
RD06	Residential Dwelling - Caravan
RD07	Residential Dwelling - Caravan
RD08	Residential Dwelling - Self Contained Flat (Includes Maisonette/Apartment)
RD10	Residential Dwelling - Privately Owned Holiday Caravan/Chalet
RH	Residential House in Multiple Occupation
RH01	Residential House in Multiple Occupation - HMO Parent
RH02	Residential House in Multiple Occupation - HMO Bedsit Other Non SelfContained Accommodation
RH03	Residential House in Multiple Occupation - HMO Not Further Divided
RI01	Residential Institution - Care/Nursing Home
RI02	Residential Institution - Communal Residence
RI03	Residential Institution - Residential Education

Ordnance Survey MasterMap Highways Network - roads

Ordnance Survey published version 03/10/2023.

These data are a unified network consisting of links (linestring geometries) and nodes (point geometries). The ID naming convention also allows for convenient integration with Ordnance Survey MasterMap Highways Network – Paths and Ordnance Survey AddressBase Plus data. More information can be found at https://www.ordnancesurvey.co.uk/products/os-mastermap-highways-network-roads.

This dataset was used under the Public Sector Geospatial Agreement.

© Crown copyright and database rights [2023] OS [AC0000805307]

Ordnance Survey MasterMap Highways Network - paths

Ordnance Survey published version 03/10/2023

These data are a unified network consisting of links (linestring geometries) and nodes (point geometries) for both paths and connecting links. Connecting links are features that link paths and roads. More information can be found at https://www.ordnancesurvey.co.uk/products/os-mastermap-highways-network-paths.

This dataset was used under the Public Sector Geospatial Agreement.

© Crown copyright and database rights [2023] OS [AC0000805307]

Ordnance Survey National Geographic Database – transport network

Ordnance survey published version 18/06/2024

These data represent a topologically structured link and node representation of Great Britain’s road, ferry, rail and path networks. Both the road link and path link component datasets were used for this analysis.

This dataset was used under the Public Sector Geospatial Agreement.

© Crown copyright and database rights [2023] OS [AC0000805307]

Ordnance Survey National Geographic Database - water features

Ordnance survey published version 18/06/2024

These data represent topographic water area features such as watercourses, lakes, drains, springs and intertidal watercourses across Great Britain. Both the water and tidal boundary component datasets were used for this analysis.

This dataset was used under the Public Sector Geospatial Agreement.

© Crown copyright and database rights [2023] OS [AC0000805307]

Ordnance Survey National Geographic Database - water network

Ordnance survey published version 18/06/2024

These data provide a network representation of the general alignment and connectivity of permanent water, including rivers, lakes, and canals. Only the water link component dataset was used for this analysis. This dataset was used under the Public Sector Geospatial Agreement.

© Crown copyright and database rights [2023] OS [AC0000805307]

Natural England Green Infrastructure - public rights of way

Natural England published version 1.2.

This dataset is a compilation of rights of way (linestring geometries) from highway authorities across England. Note: data for 48 authorities are not included. More information can be found at https://designatedsites.naturalengland.org.uk/GreenInfrastructure/UserGuide/Section03.aspx#prow.

This dataset is provided under the Open Government Licence.

Natural England - King Charles III England Coast Path Route

Natural England published version November 2023. This a line dataset showing approved stretches of the England Coast Path was used to represent coastal routes which could provide access to blue space. More information can be found at https://www.data.gov.uk/dataset/2cc04258-a5d4-4eea-823d-bf493aa31eef/king-charles-iii-england-coast-path-route.

This dataset is provided under the Open Government Licence.

Office for National Statistics Output Areas

Office for National Statistics published version December 2021 (V8).

The “Output Areas (December 2021) Boundaries EW BFC V8” dataset was used to aggregate model outputs. This is the full resolution dataset, clipped to the coastline of England.

More information can be found at https://geoportal.statistics.gov.uk/datasets/ons::output-areas-december-2021-boundaries-ew-bfe-v9/about

The “Output Area (2021) to LSOAs to MSOAs to LEP to LAD (May 2022) Best Fit Lookup in EN” dataset was used to map the 2021 Output Areas to 2021 Middle layer Super Output Areas.

More information can be found at https://geoportal.statistics.gov.uk/datasets/ons::output-area-2021-to-lsoas-to-msoas-to-lep-to-lad-may-2022-best-fit-lookup-in-en/about

The “Output Area (2011) to Output Area (2021) to LAD (December 2022) Best Fit Lookup in EW” dataset was used to map 2021 Output Areas onto 2011 Output Areas.

More information can be found at https://geoportal.statistics.gov.uk/datasets/ons::output-area-2011-to-output-area-2021-to-lad-december-2022-best-fit-lookup-in-ew/about

Source: Office for National Statistics licensed under the Open Government Licence v.3.0

Contains OS data © Crown copyright and database right 2024.

Office for National Statistics Rural/Urban classification

Office for National Statistics published version January 2016.

The “Rural Urban Classification (2011) of Output Areas in EW” dataset was used to map urban/rural classifications on to the 2021 Census Output Areas.

More information can be found at https://geoportal.statistics.gov.uk/datasets/53360acabd1e4567bc4b8d35081b36ff/about

Source: Office for National Statistics licensed under the Open Government Licence v.3.0

Contains OS data © Crown copyright and database right 2024.

Office for National Statistics Countries

Office for National Statistics published version December 2023.

The “Countries (December 2023) Boundaries UK BFE” dataset was used to define the areas in which the shortest path calculations were run, in this case, England.

More information can be found at https://geoportal.statistics.gov.uk/datasets/ons::countries-december-2023-boundaries-uk-bfe/about

Source: Office for National Statistics licensed under the Open Government Licence v.3.0

Contains OS data © Crown copyright and database right 2024.

Office for National Statistics Built Up Areas

Office for National Statistics published version December 2022.

The built up areas dataset was used in the production of access points. This is a generalised dataset, created using 25m grid squares (GB BGG).

More information can be found at https://geoportal.statistics.gov.uk/datasets/ad30b234308f4b02b4bb9b0f4766f7bb/

Source: Office for National Statistics licensed under the Open Government Licence v.3.0

Contains OS data © Crown copyright and database right 2024.

Access points

When creating the blue space access points for this publication, three Ordnance Survey datasets were used to represent waterbodies. These were filtered to exclude waterbodies not considered to be blue space. The filters used can be seen in Table 4 below.

Table 4: Blue space datasets used to create blue space access points, and the filters which were applied to exclude data which did not meet the blue space requirements

Dataset	Filter applied	Reason
Ordnance survey water features - water	Underground features and features in containment removed	Underground and contained features are unlikely to be accessible
	Removal of certain blue space types (based on blue space description field), Types removed were: - Buried Open Reservoir - Buried Open Water Tank - Cascade - Collects - Cooling Pond - Dew Pond - Fish Ladder - Fish Lock - Fish Trap - Open Tank Reservoir - Open Reservoir With Solar Panels - Open Water Tank - Overflow - Oyster Pit - Paddling Pool - Reed Bed For Waste Water - Sea - Settling Pond - Sinks - Spreads - Spring - Spring As Source Of Watercourse - Spring And Trough - Static Water As Source Of Watercourse - Still Water With Solar Panels - Swimming Pool - Watercress Bed	Removed blue space types were not considered likely to provide an adequate blue space experience. This was assessed based on knowledge of the blue space type or assessment via satellite imagery. Types were removed due to: - commercial or industrial nature - likely lack of access - insufficiently natural - insubstantial size - better data available elsewhere (‘Sea’ type only)
OS water network - water link	Underground features and features in containment removed	Underground and contained features are unlikely to be accessible
	Removal of certain blue space types (based on blue space description field), Types removed were: - Still Water - Reservoir - Foreshore - Leat - Canal Feeder - Overflow	Removed blue space types were not considered likely to provide an adequate blue space experience. This was assessed based on knowledge of the blue space type or assessment via satellite imagery. Types were removed due to: - commercial or industrial nature - likely lack of access - insufficiently natural - insubstantial size - better data available elsewhere (‘Still water’ and ‘reservoir’ are better represented by the water feature - water dataset)
	Remove non-surveyed data	Data listed as inferred was found to be unreliable. This was based on validation using local knowledge, satellite imagery and street view.
OS water features - tidal boundary	All non-high tide line boundaries removed

Two of the datasets (‘water link’ and ‘water’) have significant overlap, and a single waterbody in the real-world may be represented in both datasets. To avoid duplication of access point creation, the ‘water link’ dataset was ‘cookie cut’ to remove any spatial overlaps with the ‘water’ dataset. Where available the ‘water’ dataset representation was considered preferable as it provides a polygon representation (the ‘water links’ dataset provides line representation) and was found to more reliably represent real-world features when compared to local knowledge of an area.

Access point refinement

Due to the nature of the input data the outlined access point generation method results in an excess number of access points. Three methods were used to reduce the number of access points.

Firstly, for sections of accessible waterside shorter than 20 m, access points were generated only at a single randomly selected end. 20 m can be walked in less than 20 seconds at a typical walking speed, and therefore not generating multiple access points for these smaller sections of accessible watersides will not have a significant impact on walking distances from households.

The second method used was weaving correction. Where a route weaves into and out of the buffered blue space zone, the weaving generates access points at each entry and exit to the buffer zone. Along a single route with no branching, these are surplus access points. Instead, it is better to have just the ‘first’ and ‘last’ access points along the route. For each access point a location on the route line was calculated. The access points with the maximum and minimum line locations for each blue space ID – route ID combination were then selected and kept; all other access points along the route were removed.

The third method used was clustering of access points. Excess access points were generated where multiple routes meet or occur in proximity to blue space, as each route will generate separate access points. Similarly, excess access points were generated where multiple blue spaces occur in proximity to a route, as each will generate separate access points. In combination these factors result in an overly busy access point dataset, particularly in more built-up areas. To remedy this clustering of access points was done using density-based spatial clustering of applications with noise (DBSCAN) to partition the access points into clusters based their distance to other access points. Points within 20 m of each other were considered for clustering, and clusters were formed where 2 or more points were involved. For points identified to be part of a cluster, the middle geometry of all access points in a cluster was used to form a ‘clustered access point’. For each cluster, this clustered access point was used to represent all points in the cluster, and the original ‘raw’ access points were removed. Attributes associated with the original ‘raw’ access points were preserved as array values in the new clustered access point. Not all access points were identified as part of a cluster and for these the original ‘raw’ access point geometries were kept. This resulted in a refined blue space access point dataset which was used in the network analysis.

Blue space size

Blue space size was used to filter both the blue space features considered for access point creation and to filter the blue space access point dataset for the different scenarios.

Size filtering was done based on area for the ‘water’ dataset and associated access points and length for the ‘water link’ dataset and associated access points. For both of these datasets, filtering was applied to individual blue spaces within the data. For example, a collection of 5 nearby ponds each 0.2 ha in area would all be removed under under a 0.5 ha threshold, but a single pond with an area of 0.5 ha would be included. This may not be the best representation of real-world experience as the collection of smaller ponds with a total area of 1 ha may provide a better blue space experience than a single pond 0.5 ha in area. See the development plan for aims to improve this by using a blue space collection size in future.

For the ‘tidal boundary’ dataset no size filtering was applied as the coast, sea and estuaries associated with this dataset were considered a single large blue space, despite fragmentation of the data.

Accessible waterside length

Accessible waterside length was one method used to filter the blue space access point dataset. This was calculated using the accessible waterside dataset which was created during the blue space access point generation. The accessible waterside dataset consists of linestring representations of sections of routes in proximity to blue space. To simply calculate accessible waterside length, the length of each linestring can be calculated. However, the aim of filtering on accessible waterside length was to select for accessible waterside which provide a reasonable length of blue space experience, which could be provided by multiple connected or nearby sections of accessible waterside. To account for this, an accessible waterside collection length was used instead of simply using accessible waterside length.

To calculate collection length, accessible waterside collections were first identified:

• Accessible waterside was buffered by 10 m and then flattened into a single layer for each type of blue space (for example, still water or water course). This creates a flat accessible waterside collection layer for each type of blue space, where pieces of blue space within 10 m of each other are represented by a single large polygon (an accessible waterside collection).
• Each accessible waterside collection was given a unique collection ID.
• A spatial intersection was done between the original accessible waterside lines and the created accessible waterside collection polygons. Based on this intersection an accessible waterside collection ID is assigned to each original accessible waterside line.
• The original accessible waterside lines are then grouped by their assigned collection IDs and the sum of the accessible waterside lines is calculated for each collection ID group. This is the accessible waterside collection length
• Using the collection ID this accessible waterside collection length was linked back to each individual accessible waterside that forms the collection.

Figure 7: Method used to calculate accessible waterside length from sections of accessible waterside (waterside which can be walked along on foot).

Notes for Figure 7:

• The figure shows the process used to produce a total accessible waterside length from multiple sections of nearby accessible waterside.
• Panel A shows three lines of accessible waterside near to each other. In panel B, each of these sections is buffered by 10 m to form a hatched area around the accessible waterside. These hatched areas overlap, and are combined to form a single hatched area in panel C; this hatched area is identified as an accessible waterside collection zone. Panel D shows the individual lengths of the three accessible waterside sections within this accessible waterside collection zone are added together to form a combined length for the collection of nearby accessible waterside sections.

Following this we have an accessible waterside dataset with accessible waterside collection lengths included. When the blue space access points are created from the accessible waterside data (see the ‘Access points’ section in the Technical annex), this accessible waterside collection length is included in (and can be used to filter) the final blue space access point dataset.

Network

Households and node relationships

OS AddressBase Plus point geometries are typically not directly represented by nodes in the network. This means that there is a discrepancy in the distances calculated in the shortest path analysis, and the actual distance from the ‘front door’ of the household to a blue space. A two stage process was undertaken to account for these differences:

Stage 1, when available, the relationship between OS AddressBase Plus and OS Mastermap Highways data was used to identify the network node associated with the household (Figure 8a).

1. The links associated with each address were joined to the household data using the ‘osRoadLinkTOID’ attribute.
2. The two nodes associated with the link were joined to the household data.
3. The straight-line distances between the household and each of the two nodes were calculated.
4. The node with the shortest distance was identified as the starting node for a given household.
5. For households without an ‘osRoadLinkTOID’ attribute, the start node was identified by finding the nearest node (using straight-line distance) in the network. This differs from the previous steps, as the closest node may not be a node associated with a road/path used to access an address.

Stage 2, the distance between the household and its associated road/path, and the distance from the closest point on the road/path to the household to nodes at either end of the associated road/path was calculated. The three distance values were then used to adjust the shortest path distance once it was calculated (Figure 8b).

1. Using the same associated links and nodes as described in stage 1, the straight line distance from a household to its associated road was calculated.
2. From the point on the road/path identified in step 1, the distance along the road/path to each of the two associated nodes was calculated.
3. Using the shortest path model outputs to indicate which roads were traversed, the relevant node distance from step 2 was selected and added to/subtracted from the distance calculated in step 1.
4. Using the three values calculated in steps 1 and 2, along with outputs from the shortest path model that indicated which nodes were traversed, an adjustment was made to the shortest path distance, to account for the route taken from the ‘front door’ of the household.

For the small number of households with no associated road/path data, the straight-line distance to the node was used instead. Using the above calculations and adjustment, a more accurate distance to blue space access points was calculated for each household.

Figure 8: Household to node relationships. Firstly, the nearest associated node is identified (A), and then the distance from the household to that node calculated (B).

Notes for Figure 8:

• A diagram with two parts, A and B.
• In part A the diagram shows one household and a network made of nodes and links, represented by lines and circles. One of the links is annotated as the associated link for the household. The nodes (circles) at either end of that link are labelled as associated nodes. There are dashed lines between the household each of the associated nodes. An annotation text box explains that the distance to one of the nodes is shorter than the other and therefore that node is selected as the ‘starting node’ for the shortest path calculation.
• In part B, the same household and network links and nodes are shown. Additionally, there is a dashed line from the household to the associated link, and another dashed line to the associated node. These two lines are labelled as distance A and distance B respectively. From the start node, a yellow highlighted follows the network links to show the shortest path taken. An annotation text box described how distances A and B are added to the shortest path distance for a more realistic distance calculation.