# FAQs ## General FAQs
What file formats are available for OS NGD data? OS NGD data is available in four easy-to-use formats: GeoPackage, CSV (comma-separated values), GeoJSON and vector tiles. The download service of OS Select+Build supports GeoPackage and CSV. OS NGD API – Features supports GeoJSON. OS NGD API – Tiles supports vector tiles. For more information about the file formats, please see the [File formats and naming page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/file-formats-and-naming).
Why are there TOIDs in the OS NGD which are not in OS Premium products? Features are now available to customers more quickly in the OS NGD than they are in OS Premium or OS OpenData products. OS NGD features that will also ultimately be supplied in existing products with a TOID will now appear in the OS NGD in advance of their publication in OS Premium or OS OpenData products. At a future refresh of OS Premium or OS OpenData products, these features will appear with the same TOID*.*
Is COU data available? Monthly and daily COU (Change-Only Update) supplies are available for CSV files; however, they are not available for GeoPackage files. For more information about COUs, please see the [COU supplies page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/cou-supplies) and the [Data ordering and currency page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/os-ngd-data-ordering-and-currency).
Why won't my CSV file open in excel / why won't my CSV file load in my GIS? Please be aware that CSV (comma-separated values) files are designed to be opened in a database or GI system, and opening them in other software applications could corrupt the data. In particular, Excel has a row limit which might be exceeded by some of our CSV files containing OS NGD data, depending on the order you placed and its size. We recommend that you load CSV files containing OS NGD data directly into a database or GI system, rather than trying to open these files in Excel. For more information about CSV files, please see the [File formats and naming page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/file-formats-and-naming).
Does the launch of the OS NGD mean that some existing OS Premium and OS OpenData products will be withdrawn? No. The launch of the OS NGD does not instigate withdrawal of any of the existing OS Premium or OS OpenData Download products, including APIs. These products will continue to be managed through their product lifecycle, including withdrawal / becoming End of Life, separately to the OS NGD. The OS NGD gives new and enhanced ways for customers to access the most trusted and up-to-date geographic information from OS. The [OS Product Roadmap](https://www.ordnancesurvey.co.uk/products/roadmap) is available to view on our website – it has details of upcoming OS NGD releases, new Download product releases (for Premium and OpenData products), and dates for planned product withdrawals as they reach their End of Life stage.
Why do I have different start dates when applying temporal filters to different feature types? Following the launch of the OS NGD in September 2022, we have and will continue to add new feature types as part of our [data enhancements](https://docs.os.uk/osngd/os-ngd-news/future-os-ngd-data-enhancements). These could be brand new feature types or enhanced versions of previously existing feature types (with new attribution added for a new version). In both instances, these new / enhanced feature types will have later start dates. The earliest date on which you can request a one-off snapshot of a date in the past for data in a feature type in most collections is 29 September 2022, but that date can be later for certain feature types (for example, it's 28 March 2023 for the Waterbody Catchment Feature Type). Each feature type page states the earliest start date available for temporal filtering on that feature type.
How is OS NGD file naming managed? Please refer to the [File formats and naming page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/file-formats-and-naming) for information about the OS NGD file naming convention.
Why am I receiving blank files? You could have received blanks files for one of the following reasons: * You are collecting data via a COU (Change-Only Update) and there has been no change. * You have selected a feature type from one of the following OS NGD collections which has a different update frequency from your selected COU: OS Boundaries Collection (updated biannually), OS Functional Areas Collection (updated biannually), OS GB Postcodes and OS NI Postcodes Collections (both updated monthly), OS Transport Features Collection (updated daily, with the exception of the Street Light Feature Type which is updated monthly), OS Transport Network Collection (updated monthly), OS RAMI Collection (updated monthly),or the OS Water Network Collection (updated quarterly). * You have requested data via a temporal filter (i.e. a one-off snapshot of a date in the past) that predates the start date for data from one or more of the collections in your order. The earliest date on which you can request a one-off snapshot of a date in the past for data in a feature type in most collections is 29 September 2022, but that date can be later for certain feature types (for example, it's 28 March 2023 for the Waterbody Catchment Feature Type). Each feature type page states the earliest start date available for temporal filtering on that feature type. * You have selected the Annual Full Supply order frequency option for a data package containing one (or more) of the new feature types or new data schema versions of existing feature types released in March and / or September of the current year. If you select an Annual Full Supply frequency for your OS NGD data order in OS Select+Build, we will provide you with the data as it was on 01 January of the current year. This means if a new feature type or a new data schema version of an existing feature type was released after 01 January and you order either of these as part of your Annual Full Supply, you will receive an empty data package for the newly released feature type / new data schema version of an existing feature type. The data for the new feature type / new data schema version of an existing feature type will then be included in your supply on the next 01 January after the release, after which point, data packages containing one of these feature types will no longer contain blank files. * For example, the March 2026 OS NGD data enhancements release contained six new feature types (Retail Area Aggregated, Retail Area Major, Retail Area Minor, Postcode Unit Point, Postcode Unit Point, and Postcode Unit Area) and one new data schema version for an existing feature type (Compound Structure v3.0). The data for the new feature types and the new data schema version for the existing feature type will not be part of Annual Full Supply orders until 01 January 2027. * You have an existing OS Select+Build recipe that uses the Site Routing Point Feature Type or transact with that feature type via an OS NGD API. As of 26 April 2023, this feature type no longer contains any features or data. It is in an 'end of life' state where it has stopped receiving updates; therefore, any OS Select+Build recipes that use this feature type or any OS NGD APIs that transact with it will return blank files. You can access historic data for Site Routing Point using the temporal filter option in OS Select+Build by selecting the tick box for a 'one-off snapshot'; historic data is available for this feature type from 03 October 2022 to 25 April 2023. As part of a data enhancement programme in late spring 2023, all Site Routing Points were recreated as [Road Nodes](https://docs.os.uk/osngd/data-structure/transport/transport-network/road-node) and are therefore now accessible via the OS NGD Transport Theme (from the OS NGD Transport Network Collection).
What is data schema versioning and how does it work in the OS NGD? Please see the [Data schema versioning page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/data-schema-versioning) in the OS NGD Fundamentals section for a full explanation of data schema versioning and how it works in the OS NGD.
Can I download OS NGD documentation for offline use? Yes. The OS NGD documentation can be saved to a PDF using the Export as PDF button on the right-hand side of the page. This tool allows users to download individual pages or groups of pages that make up a complete document. Users should limit their documents to a maximum of 100 pages per PDF to prevent issues when saving or opening the file.
What is the difference between 'Unknown' and 'NULL' attribute values? **'Unknown'** is used in two main circumstances: * Where an attribute is expected to be populated but Ordnance Survey does not currently have a value. For example, a feature captured by a remote sensed survey where the land use has not yet been determined. * Where an attribute is expected and Ordnance Survey has attempted to obtain a value, but this has not been possible. For example, the ‘use’ of an inactive or derelict site is not clear. In summary, 'Unknown' means the attribution value is in scope, but Ordnance Survey does not currently know what that value is. **'NULL'** is used in two main circumstances: * The value is not applicable to the attribute in question in this instance – typically, where an attribute value is optional. For example, non-mandatory references to other OS NGD features, such as from Land features to an associated Structure feature, or from a Building feature to a land use Site feature where those referenced features may not exist in the real world. * The information is currently out of scope for provision on this attribute for all instances of this feature – usually where there is no viable method of capturing reliable information. For example, [Status ](https://docs.os.uk/osngd/data-structure/land-use/land-use-features/site#newstatus)is only provided on a range of land use Site features with a specific function. The [Status](https://docs.os.uk/osngd/data-structure/land-use/land-use-features/site#newstatus) on land use Site features where provision of a reliable value is not possible is recorded as 'NULL'. In summary, 'NULL' means the attribution value either does not exist or is not in the current scope of Ordnance Survey capture.
Can I try out OS NGD data for free using a Data Exploration Licence? Yes, you can try out most OS NGD collection data for free for three to six months by applying for [a Data Exploration Licence](https://www.ordnancesurvey.co.uk/licensing/data-exploration-licence) (DEL). This licence is primarily designed for organisations either considering becoming an OS Partner or who want to trial OS Premium data before working with an existing OS Partner. [More information about the licence is available from the OS website](https://www.ordnancesurvey.co.uk/licensing/data-exploration-licence#agreement). The licence is available for all OS NGD collections apart from the OS Islands Address and OS Functional Areas Collections. It will let you use data from the OS GB Address, OS GB Postcodes and OS NI Postcodes Collections data for three months and data from all other collections (apart from the OS Islands Address and OS Functional Areas Collections) for six months. | OS NGD Collection | Length of DEL | | --------------------- | ------------- | | OS GB Address | 3 months | | OS Boundaries | 6 months | | OS Building Features | 6 months | | OS Functional Areas | N/A | | OS GB Postcodes | 3 months | | OS Islands Address | N/A | | OS Named Features | 6 months | | OS Land Features | 6 months | | OS Land Use Features | 6 months | | OS NI Postcodes | 3 months | | OS RAMI | 6 months | | OS Structure Features | 6 months | | OS Transport Features | 6 months | | OS Transport Network | 6 months | | OS Water Features | 6 months | | OS Water Network | 6 months |
How can I verify my data download is correct? Feature types downloaded via OS Select+Build will be supplied with an Order Summary File, a JSON manifest that provides metadata for that specific supply. You can use the manifest to validate the data to ensure the currency, record count, schema version, feature name and coordinate reference system (CRS) match your expected order. You can find more detail on the contents of the order summary file and how to use it to validate your data order on the [Data ordering and currency page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/os-ngd-data-ordering-and-currency) in the OS NGD Fundamentals guide.
## API FAQs
How is OS NGD API – Features different from OS Features API? The [OS Features API](https://docs.os.uk/os-apis/accessing-os-apis/os-features-api) is based on the OGC standard Web Feature Service (WFS) and provides access to OS OpenData and OS Premium Product data. It does not give access to any OS NGD data. OS NGD API – Features gives you simple access to the OS NGD using the latest in API standards (based on the [OGC API – Features specification](https://ogcapi.ogc.org/features/)). As the new API uses the latest standards, it's easier than ever before to quickly discover what OS NGD data is available and create your own bespoke data selections with more understandable filtering options. Ultimately, this will help accelerate your time-to-value by making it easier to build awesome things with our trusted geospatial data. Please see the ['What data is available?' page](https://docs.os.uk/osngd/access-the-os-ngd-api/os-ngd-api-features/what-data-is-available#what-feature-types-are-available-in-os-ngd-api-features) for the full list of OS NGD feature types included in the OS NGD API – Features.
Why does OS NGD API – Tiles only include a subset of OS NGD feature types and their attributes? A key benefit of vector tiles is that they comprise layers of vector features optimised for caching and scaling that can render quickly. As a result, to ensure the vector tiles are lightweight and offer a smooth user experience, we only use the OS NGD feature types required to create a complete contextual basemap. The OS NGD API – Tiles basemap does not contain every OS NGD feature type, nor the complete list of attribution available within the feature types that are included; we have purposefully only selected feature types and a subset of attribution from them that are useful for visualisation as this keeps the tiles lightweight and quick to render. Even with minimal attribution, map developers can apply styles and interactivity to vector features that is just not possible with raster maps. The inclusion of unique identifiers (where available) allows you to cross-reference with the full product (for example, with [OS NGD API – Features](https://docs.os.uk/osngd/getting-started/access-the-os-ngd-api/os-ngd-api-features)). Please see the ['What data is available?' page](https://docs.os.uk/osngd/getting-started/access-the-os-ngd-api/os-ngd-api-tiles/what-data-is-available) for the full list of OS NGD feature types included in the OS NGD API – Tiles basemap. [Additional data overlays of OS NGD datasets](https://docs.os.uk/osngd/access-the-os-ngd-api/os-ngd-api-tiles/what-data-is-available#data-overlays) are available to add additional information to the basemap, depending on your needs.
When is OS NGD API – Tiles updated? The OS NGD API – Tiles basemap is updated weekly. This means that the OS NGD data is updated with the latest available data every week, usually by Monday. The data used to update the API is taken from the previous Wednesday, ensuring that we’ve updated the vector tile cache with the latest OS OpenZoomstack and OS NGD data available at the time.
{% hint style="info" %} Additional general FAQs and answers about OS APIs are available on the ['OS Data Hub FAQs: Plans' page](https://osdatahub.os.uk/support/plans) and ['OS Data Hub FAQs: Account and API' page](https://osdatahub.os.uk/support/faqs/account-and-apis), for example, 'What's a project?', 'How do I generate an API Key?', 'What throttling is applied to the APIs?'. {% endhint %} ## OS Select+Build FAQs
Which formats can be taken using OS Select+Build and at what frequencies? The download service of OS Select+Build supports GeoPackage and CSV file formats (more information on these file formats is available on the [File formats and naming page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/file-formats-and-naming) from the OS NGD Fundamentals section). For information about data ordering and currency, please see the [Data ordering and currency page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/os-ngd-data-ordering-and-currency).
What are the default coordinate reference systems (CRS) used by the OS NGD collections in OS Select+Build? Please see the [Coordinate reference systems page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/coordinate-reference-systems) in the OS NGD Fundamentals section for full details. It is possible to select what CRS you'd like to receive data in for feature types in your data package. There's more information about selecting a CRS on the [Getting started with data packages page](https://docs.os.uk/osngd/downloading-with-os-select+build/getting-started-with-data-packages#how-do-i-create-a-new-data-package). If you don't choose a particular CRS for your data package, OS Select+Build will automatically select the default CRSs for the feature types in your data package for you.
What is a recipe and how do I create one? A recipe is a bespoke selection of OS NGD data which is made by a user within OS Select+Build. Recipes allow you to choose the OS NGD data that best fit your requirements. Please see the [Getting started with recipes page](https://docs.os.uk/osngd/getting-started/downloading-with-os-select+build/getting-started-with-recipes) for more information about recipes and step-by-step instructions for creating them.
How do I edit a recipe? Please see the [Editing OS Select+Build recipes page](https://docs.os.uk/osngd/getting-started/downloading-with-os-select+build/getting-started-with-recipes/editing-os-select+build-recipes) for step-by-step instructions on how to edit a recipe.
How do I delete a recipe? Please see the [Managing OS Select+Build recipes page](https://docs.os.uk/osngd/downloading-with-os-select+build/getting-started-with-recipes/managing-os-select+build-recipes#how-do-i-delete-a-recipe) for step-by-step instructions on how to delete a recipe.
How do I know which recipes are associated with which data package? Please see the [Getting started with data packages page](https://docs.os.uk/osngd/downloading-with-os-select+build/getting-started-with-data-packages#how-do-i-find-and-download-an-existing-data-package) for step-by-step instructions on how to check which recipes are associated with each data package.
How do I create an OS Select+Build data package? Please see the [Getting started with data packages page](https://docs.os.uk/osngd/downloading-with-os-select+build/getting-started-with-data-packages#how-do-i-create-a-new-data-package) for step-by-step instructions on creating an OS Select+Build data package.
How do I delete an OS Select+Build data package? Please see the [Getting started with data packages page](https://docs.os.uk/osngd/downloading-with-os-select+build/getting-started-with-data-packages#how-do-i-delete-a-data-package) for step-by-step instructions on deleting an OS Select+Build data package.
How does AOI ordering work with OS Islands Address Collection data? The OS Islands Address Collection contains data for Northern Ireland, the Isle of Man, and the Channel Islands. When ordering data from this collection, you will be provided with a full supply of the data regardless of whether you select 'All of Britain or Predefined Area', or an 'area of interest' (AOI)​. For example, if you order the OS Islands Address Collection and the Building Line Feature Type selected with an AOI, then you will receive the Building Line Feature Type data for your selected AOI and all of the OS Islands Address Collection data​ (i.e. data for Northern Ireland, the Isle of Man, and the Channel Islands, and not just data for your AOI).
How do I use attribute filtering in OS Select+Build? Please see the [Getting started with attribute filtering page](https://docs.os.uk/osngd/getting-started/downloading-with-os-select+build/getting-started-with-recipes/getting-started-with-attribute-filtering) for step-by-step instructions on how to filter attributes using OS Select+Build.
How do I identify the correct OS NGD collection for calculating usage? The OS NGD collection can be identified within the file name of your .zip file: themeshortcode\_collectionshortcode\_featuretype.zip. For example, bld\_fts\_buildingpart.zip is the file name which would be created for an order of the Building Part Feature Type within the OS Building Features Collection of the OS NGD Buildings Theme. For more information about the file naming convention for OS NGD data, please see the [File formats and naming page](https://docs.os.uk/osngd/getting-started/os-ngd-fundamentals/file-formats-and-naming) in the OS NGD Fundamentals section.
## Theme-specific FAQs ### OS NGD Address Theme FAQs
Can I use OS NGD Address Theme data to update my AddressBase Premium holding? We wouldn't recommend trying to use OS NGD Address Theme data to update existing product supplies of AddressBase products. The different data format and attribution mean that there are no direct mappings available back to AddressBase products, and there is a risk of updating your existing holding incorrectly.
Why are there historic addresses in the Alternate Address data? The Alternate Address Related Component provides alternative address labels where they exist for a ‘current’ address, independent of whether these alternate labels are current / live or historic. Where a historic address label is present but there is also a live / approved address label, then the historic version will appear in the Alternate Address data.
How does lifecycle work in OS NGD Address Theme data? As addresses move through different stages of their lifecycle, you will see different update types dependent on your chosen data package. In OS NGD Address Theme data, pre-build, built, and historic addresses are kept in separate feature types. This means that as an address moves from one stage to another (for example, pre-build to built), it will be removed from one feature type and added to the other. This differs to AddressBase Premium, where, when an address changes from one stage to another, the record is updated, i.e. not deleted from one table and inserted into another. This will also mean that, if, for example, a user has a data package that includes pre-build addresses but not built addresses, as an address moves from pre-build to built in its lifecycle, the user will see a deletion to the record that they hold of this address.
Why are there addresses flagged as 'Non-Postal' that are not in the Non-Addressable Object Feature Type? Addresses that cannot receive post are classified as such for various reasons. Therefore, there are a number of different ways of extracting addresses that occupy different classes of 'Non-Postal'. The terms 'Non-Postal' and 'Non-Addressable Objects' used in OS NGD Address Theme data represent categories of addresses which are defined by differing logic. Non-Addressable Objects are features which OS deem to always be non-postal. This is our most certain view of addresses that are non-postal, for example, Bandstands, Telephone Boxes, and Nature Reserves. The vast majority of Non-Addressable Objects will also have an 'N' value in the Postcode Source attribute. 'Non-Postal' in the Description attribute is essentially an amalgamation of all addresses that are defined by Local Authorities to be either 'N: Non-Postal' or 'L: Maybe Postal', and therefore is most relevant for customers who are only interested in using it as a filter to get to the 'Postal' addresses. The Postcode Source attribute is the best attribute to use if you are interested in a more nuanced representation of Non-Postal addresses. This splits out 'Not-postal' and 'Maybe Postal' into two different code lists. The vast majority of addresses that have Non-postal in the description field but are not in the Non-Addressable Object Feature Type are likely to be 'Maybe Postal' addresses. This is a Local Authority representation of addresses that may be able to receive post, such as Car Parks.
What does parent-child relationship mean? The parent-child relationship is used to relate multiple addresses to each other. These are most commonly used to relate addresses within a building, for example, a block of flats. A parent Unique Property Reference Number (UPRN) is assigned to the building shell to give an indication that the flats within that building are related to the overarching address, for example, Example Court and the flats within. A good example of this is: * Parent: 1 High Street (UPRN: 1234) * Child: Flat A, 1 High Street (UPRN: 2345) * Child: Flat B, 1 High Street (UPRN: 3456) However, parent-child relationships also work beyond the level of a single building footprint and can be used to relate addresses in a larger addressable site to each other, for example, University buildings on the same campus site. An example of this is: * Parent: Exeter University (UPRN 1) * Child of UPRN 1: Block C, Exeter University (UPRN 2) * Child of UPRN 2: Flat 1, Block C, Exeter University (UPRN 3) Please note: Parent records are only inserted where child addresses share part of an address, for example, '1 High Street' in the above example. If two flats occupy the same building but are addressed 12A and 13A, then no parent record would be created.
What's the source of the improved level information for the Floor Level, Lowest Floor Level and Highest Floor Level attributes? These attributes are populated by data captured by local authority address custodians. The data previously had been available within our AddressBase products, but there was a large variation in the textual format. The level information has been standardised into an ordered numeric list (for example, ‘First Floor’ is now ‘1’, and ‘Ground Floor’ is now ‘0’ and remains a string field).
How is the Full Address attribute populated in OS NGD Address? In OS NGD Address Theme data, the Full Address attribute is populated entirely from address data sourced from local authorities. This differs from the logic behind the Full Address attribute used in the AddressBase Core product, which is populated with information from Royal Mail and data from local authorities.
### OS NGD Administrative and Statistical Units Theme FAQs #### Boundaries data FAQs
Can I easily identify wards linked to a specific unitary authority? Yes. This can be done non-spatially by interrogating the new Boundary Parent Reference attributes.
What's the difference between a boundary and a functional area? Features in the [OS Boundaries Collection](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/boundaries) have specified geometries, often detailed in law; OS documents these geometries, but is not responsible for saying *where* the boundary goes. In contrast, [functional areas](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas) are algorithmically derived notional geometries – which means OS will place them based on the information we know about each area, such as the address, site and topographic data we hold in the rest of the OS NGD.
#### Retail areas data FAQs
Why is a Retail Area Minor feature not automatically part of a Retail Area Aggregated feature? [Retail Area Aggregated](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-aggregated) features are derived from areas that contain at least one [Retail Area Major](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-major) feature and any neighbouring addresses with a classification that is indicative of retail use. As a result, Retail Area Aggregated features might be a single parade of shops or a large section of a city. Not all [Retail Area Minor](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-minor) features are located close enough to a Retail Area Major feature to become part of a Retail Area Aggregated feature. For example, in the image below of Aylesbury, the northeasternmost Retail Area Minor feature doesn't appear in a Retail Area Aggregated feature as it's not located close enough to a Retail Area Major feature. The other five Retail Area Minor features over the centre of the town do form part of Retail Area Aggregated features as they are all located close enough to Retail Area Major features to automatically become part of Retail Area Aggregated features.
Example showing Retail Area Aggregated, Retail Area Major and Retail Area Minor features in the centre of Aylesbury. A background map has been included for context.

Example showing Retail Area Aggregated, Retail Area Major and Retail Area Minor features in the centre of Aylesbury. A background map has been included for context.

Why is this address not in a retail area? * Our retail area features are created based on a combination of count and proximity thresholds. If the retail addresses are not close enough together, they will not meet our capture requirements. We are keen to optimise the thresholds used by the algorithm, so please [contact us](https://www.ordnancesurvey.co.uk/contact-us/business-government/use-os-data) or your OS Account Manager to report any instances of this problem as an area for us to learn from. * Currently, we require an address to have an [Address Classification Code Value](https://docs.os.uk/osngd/code-lists/code-lists-overview/addressclassificationcodevalue) which starts ‘CR’ (i.e. Retail) to be classified as a retail address. Particularly large warehouse-type stores are classified as ‘Industrial’ (i.e. as 'CI' in the [Address Classification Code Value Code List](https://docs.os.uk/osngd/code-lists/code-lists-overview/addressclassificationcodevalue)), so will not be included as a retail address in a potential retail area feature. If an address has a different classification code value and you believe it should have a Retail (‘CR’) classification code value, please report this via our Errors & Omissions Tool on the [OS Data Hub](https://osdatahub.os.uk/). If a change is made, this will be incorporated into the next release of retail areas data.
Why is there not a retail area where I expect to see one? Our retail areas are based on several factors, including address classification, density and proximity thresholds. If an area doesn't meet these criteria, then a retail area feature won't be produced for it. We know this means that certain areas customers expect to see marked as retail area features are not always present in the data, but we are keen to optimise the thresholds used by the algorithm. If you would like to be part of our user engagement about what makes a retail area, please [get in touch](https://www.ordnancesurvey.co.uk/contact-us/business-government/use-os-data).
Will more feature types be added to the OS Functional Areas Collection in future? Why is the collection not called the OS Retail Areas Collection? We plan to release more feature types in future for other kinds of functional areas, including office areas, leisure areas and more. If you would like to be part of our user engagement about which kinds of functional areas would be of most interest and what should be included in each, please [get in touch](https://www.ordnancesurvey.co.uk/contact-us/business-government/use-os-data).
Why do retail area features overlap? Overlapping features are expected in [functional areas](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas) (both overlaps of different feature types and of features in a single feature type). This is because a single building might have characteristics that relate to more than one area. For example, a building on a street corner could have retail addresses related to two streets both containing numerous retail address; therefore, that building on a corner would be included in two [Retail Area Major](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-major) features. As the name suggests, [Retail Area Aggregated](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-aggregated) features are an aggregation of Major Retail Area features and sometimes [Retail Area Minor](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-minor) features, so will always overlap other features.
In Retail Area Major features, why does an address with a Unique Street Reference Number (USRN) not always have the USRN as the dividing feature? If the address falls within a Site (for example, a shopping centre), the Site is used as the dividing feature instead of the USRN (in such cases, the [Dividing Feature USRN attribute](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-major#dividingfeature_usrn) value would be NULL and the [Dividing Feature Site ID attribute](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-major#dividingfeature_siteid) value would be populated).
Example image showing a dividing feature for West Quay Shopping Centre in Southampton. A background map has been included for context.

Example image showing a dividing feature for West Quay Shopping Centre in Southampton. A background map has been included for context.

Why have you provided counts of addresses with classifications other than 'Retail'? The proportion of addresses with different types of classifications can help you to understand the characteristics of an area. It enables use cases where you can look at the changes in an area over time or between different areas which started with similar characteristics. For example, if one area received funding for an initiative and another did not, you can use these characteristics as a counterfactual to see the impact the initiative made.
Why can non-retail buildings be included in retail areas? Our [functional areas](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas) are deliberately generalised polygons to allow for how different people perceive the same area. This is achieved by adding a buffer to the specific retail locations. Sometimes buildings which don't have a classification of 'Retail' are encapsulated by that buffer (for example, a detached residential house between two shops being included).
Why have you included point coordinates on polygon features for retail areas? We have chosen to include weighted centroids for the retail area features to enable a number of use cases where a point is easier to handle. Using a position weighted on the retail addresses makes this a better representation for examples like routing where it increases the accuracy of estimations.
Example image showing the location of an address weighted centroid generated from the Easting/Northing or Latitude/Longitude attribution for a Retail Area Major in Bristol. Attribution for the Retail Area Major feature is displayed. A background map has been included for context.

Example image showing the location of an address weighted centroid generated from the Easting/Northing or Latitude/Longitude attribution for a Retail Area Major in Bristol. Attribution for the Retail Area Major feature is displayed. A background map has been included for context.

Why is the weighted centroid for my retail area feature outside the polygon for the feature? We have chosen to include weighted centroids for the retail area features to enable a number of use cases where a point is easier to handle. Using a position weighted on the retail addresses makes this a better representation for examples like routing where it increases the accuracy of estimations.
Example showing the location of an address weighted centroid generated from the Easting/Northing or Latitude/Longitude attribution for a Retail Area Major in Bristol. Attribution for the Retail Area Major feature is displayed in a pop up box. A background map has been included for context.

Example image showing the location of an address weighted centroid generated from the Easting/Northing or Latitude/Longitude attribution for a Retail Area Major in Bristol. Attribution for the Retail Area Major feature is displayed. A background map has been included for context.

However, weighted centroids don't always fall inside the polygon (for example, in donut-shaped or crescent-shaped polygons). Where this happens, the information that the polygon inherits from the weighted centroid point is not always valid for the polygon. For example, in Chippenham, the weighted centroid falls on a river and so is classed as non-urban, despite the polygon being fully urban. This impacts subsequent calculations, such as which [Retail Area Aggregated](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-aggregated) feature is defined as the town centre. The root cause of a weighted centroid not being in a polygon affects all three retail area feature types; currently, under 2.1% of all features in the three retail area feature types are impacted by this issue.
Are temporary retail locations included in retail area features? As a rule, no, as we don't have information about temporary retail locations. However, sometimes temporary retail locations happen to be included in retail area features because they are located in an area which also has permanent retail locations. For example, a market held on a high street would automatically be in a [Retail Area Major](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/functional-areas/retail-area-major) feature by virtue of the neighbouring shops. However, an ice cream van in a cliff top car park is unlikely to be in a retail area feature.
How do I use the Related Entity Related Component table with OS NGD Address Theme data? Cross-reference identifiers in the Related Entity Related Component table for retail area features are stored as a string. This means you need to change the character type and potential filter by feature type when trying to join the Unique Property Reference Number (UPRN) to other datasets (for example, the Built Address Feature Type) where they are stored as integers.
Example image showing the relationship between a Retail Area Minor feature and the Related Entities table. A background map has been included for context.

Example image showing the relationship between a Retail Area Minor feature and the Related Entities table. A background map has been included for context.

#### Postcodes data FAQs
Why isn’t the polygon for my Postcode Unit Area feature the shape I expect? [Postcode Unit Area features](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area) are created with input from a variety of data sources, the most important of which is the coordinates of addresses with that postcode. After the coordinates have been determined, an algorithm fills in the gaps between them until all of the landmass of Great Britain is part of at least one Postcode Unit Area feature. Sometimes, particularly in areas of low address density, this creates very irregular shapes.
For Postcode Unit Area features, why do the counts on the postcode parts attributes not equal the figure in the Total attribute? Sometimes a delivery point has a postcode but we don’t have enough other information about it to place it in a particular part. However, we include these delivery points in the total to ensure all records in Royal Mail's Postcode Address File (PAF) are accounted for. For example, if a postcode has three parts which have five delivery points each plus a delivery point which cannot be placed in any of those parts, the attribution on each feature will be: | Attribute Name | Count | | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----- | | [`postcodepartcount_total`](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area#postcodepartcount_total) | 3 | | [`postcodepartdeliverypointcount_total`](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area#postcodepartdeliverypointcount_total) | 5 | | [`postcodedeliverypointcount_total`](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area#postcodedeliverypointcount_total) | 16 | | [`postcodedeliverypointcount_matched`](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area#postcodedeliverypointcount_matched) | 15 | | [`postcodedeliverypointcount_unmatched`](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area#postcodedeliverypointcount_unmatched) | 1 |
Why are there postcode parts for Postcode Unit Area features which do not have any delivery points in them? [Postcode Unit Area features](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area) are created to cover the whole landmass of Great Britain. In places where the landmass fragments (particularly along the coast), a postcode part is created for each piece of land, and that polygon is attributed to the postcode of the nearest delivery point.
How do I find multiple postcodes in the same building? Why can’t I see vertical streets? The easiest way to find a postcode in the same place is to look for [Postcode Part Count Coincidental Geometry attribute](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area#postcodepartcount_coincidentalgeometry) values of over 0 for your [Postcode Unit Area features](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area).
The Postcode Unit Area Feature Type was designed to include all postcodes which can have a polygon, for those polygons to align to real-world features, and for full complete survey coverage of Great Britain. Where multiple postcodes are in the same building, they will share a geometry which covers the whole of that building; we call these ‘coincidental geometries’. This is different to the [Code-Point with Polygons product](https://docs.os.uk/os-downloads/products/areas-and-zones-portfolio/code-point-with-polygons), which has visible squares where the vertical streets are to be linked to in a separate table to view the postcodes, which was a pain point for many customers. More information on this is available on the [More than Maps site](https://docs.os.uk/more-than-maps/data-demonstrators/addressing-and-location-demonstrators/os-ngd-administrative-and-statistical-units/os-ngd-postcodes#difference-between-os-ngd-postcodes-and-code-point-with-polygons).
Why isn’t this Building feature within a Postcode Unit Area feature? Postcode polygons ([Postcode Unit Area features](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-area)) have been designed to align to the [Building features](https://docs.os.uk/osngd/data-structure/buildings/building-features/building) visually and are accurate to 1:1 scale; however, they don't use the same vertices as the Building features, so tiny slivers of a building could be in the neighbouring postcode polygon. As a result, analysis using ‘wholly within’ tests will get inaccurate results. Note, for many use cases, Address points are more suitable than Building features.
Why are some Postcode Unit Point features not in buildings? Most [Postcode Unit Point features](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-point) will be at the same coordinate as a suitable address with that postcode. That address does not have to be located in a building, either because the address is in place ahead of a [Building feature](https://docs.os.uk/osngd/data-structure/buildings/building-features/building) being created or because it is for a feature other than a building (such as a caravan). Additionally, we give a point for all postcodes in Royal Mail’s Postcode Address File (PAF). When there are no suitable addresses with a postcode, that Postcode Unit Point feature is given an approximate location based on other addresses in the postcode sector. Sometimes that approximate coordinate is outside of a building and could even be in water, especially in coastal areas.
Why is the Postcode Unit Point feature at a different coordinate than the Code-Point point for the same postcode? OS NGD postcodes ([Postcode Unit Point features](https://docs.os.uk/osngd/data-structure/administrative-and-statistical-units/gb-postcodes/postcode-unit-point)) and [Code-Point postcodes](https://docs.os.uk/os-downloads/products/areas-and-zones-portfolio/code-point) are created in different ways. Most of the time, the different methodologies produce the same result. Both results are equally valid as postcode points are notional geographies, meaning unlike our other data, they are not defined in law or surveyed to be in an exact position. In both datasets, the point is intended to be used as an easy-to-process approximation of the postcode's theoretical position. More information on this is available on the [More than Maps site](https://docs.os.uk/more-than-maps/data-demonstrators/addressing-and-location-demonstrators/os-ngd-administrative-and-statistical-units/os-ngd-postcodes#difference-between-os-ngd-postcodes-and-code-point-with-polygons).
### OS NGD Buildings Theme FAQs
How are Building features derived? Building features are derived by using the Building Part Feature Type and merging connected Building Parts that are within the same Land Use Site and have a contiguous border greater than 0.5m. If a Building Part Feature Type does not belong to a Site, then it will retain the same geometry in the Building Feature Type. If there are multiple Sites, the smaller Site has hierarchy over the larger site (i.e. Building Parts only merge within the smaller Site, and if they share a border with a Building Part outside of the smallest Site, then the Building Parts features will not merge).
Is height data available on building features ? Yes. Height data is available on [Building](https://docs.os.uk/osngd/data-structure/buildings/building-features/building#height_absolutemin_m) features and [Building Part](https://docs.os.uk/osngd/data-structure/buildings/building-features/building-part#height_absoluteroofbase_m-formerly-absoluteheightroofbase) features.
How have Building address counts been derived ? The various address count attributes that have been derived from the Built Address Feature Type in the OS GB Address Collection where the Build Status attribute value is classified as ‘Built In Use’. Residential, commercial and other address counts have been derived where the primary code in the address classification is equal to ‘R’, ‘C’, or anything not ‘R’ or ‘C’ respectively.
What is a ‘Main Building’ and how has the data been derived ? The main building aims to easily identify the building a casual visitor to a site would identify as the most useful building to represent the entire site. For example, the house on a residential site would be identified as the main building rather than the shed in the garden, or the ticket office of a train station rather than the waiting room on the other side of the platform.
How is ‘Building Use’ derived ? Building use is derived from an in-house algorithm that uses either OS NGD Address data or OS Land Use Sites data to determine what the building is being used for. In most cases, the attribute value has been derived from address data classification.
What’s the difference between the ‘Building Use’ and ‘OS Land Use Tier A’ value ? There are two main differences; a difference in what each attribute is trying to convey and a difference in their derivation. * Building Use attempts to convey what the Building is specifically being used for, whilst OS Land Use Tier A attempts to convey what activity the site that the Building is situated within is being used for. For example, Nurses' accommodation on a hospital site will have the Building Use value of ‘Residential Accommodation’ but the OS Land Use Tier A value of ‘Medical Or Health Care’. * Building Use is mainly derived from address data, whilst OS Land Use Tier A data is derived from Ordnance Survey sites data.
What is the difference in the derivation process between Roof Shape and Roof Material? The derivation of Roof Shape involves using automated methods and height models to delineate roof faces, while Roof Material is derived from imagery processed by a supervised ML model and an Automated Feature Extraction algorithm.
What is the difference between Building Access Locations and Land Use Site Access Locations? Building Access Locations identify specific entry and exit points for individual buildings, while Land Use Site Access Locations represent access points for broader areas (Sites), identified for a specific purpose, connecting them to the Road Network.
### OS NGD Geographical Names Theme FAQs
Why are settlement extents depicted differently by the Named Area Feature Type within the OS Named Features Collection and feature types in the OS Boundaries Collection? The OS Boundaries Collection provides a definitive dataset of administrative and electoral boundaries; whereas, the depiction of settlement boundaries in the Named Area Feature Type of the OS Named Features Collection is an OS geographic depiction of the extent of settlements and should not be mistaken for administrative or electoral uses.
Why is the OS NGD Geographical Names Theme not OpenData like OS Open Names? More features and attributes are provided in the OS NGD Geographical Names Theme than in OS Open Names. The theme also includes a new feature type (Named Area) which shows geometry never before published by OS. The rich attribution provided by the theme's data lends itself to helping users perform gazetteer searches as well as high-level visualisation and spatial analysis of named features and places.
#### Crowd Sourced Name Point features' FAQs
Why are there no Crowd Sourced Name Point features in my area? Coverage of Crowd Source Name Point features is dependent on submissions from expert third parties via the Vernacular Names Tool. The Vernacular Names Tool is a resource developed by Ordnance Survey to assist Emergency Services by incorporating user-submitted colloquial names for geographic features across Great Britain.
Why are the names of the Crowd Sourced Name Point features not in a standardised format? The names are published as they are submitted by expert third parties. OS doesn't make changes to the names and doesn't enforce standardisation at this time because casing and punctuation can be important elements in some names. Spelling mistakes are also expected as these can be common variants tried in searches.
Can I submit a new name (i.e. a new Crowd Sourced Name Point feature)? Crowd sourced names can be submitted via the Vernacular Names Tool on the [OS Data Hub](https://osdatahub.os.uk/). You'll need to be member of an approved organisation to submit a name. The Vernacular Names Tool is a resource developed by Ordnance Survey to assist Emergency Services by incorporating user-submitted colloquial names for geographic features across Great Britain.
Can I suggest a change to a Crowd Sourced Name Point feature's name, coordinate or classification, or suggest a match to it? Please report these via the Errors & Omissions Tool on the [OS Data Hub](https://osdatahub.os.uk/).
Why has my name not been published? Names usually appear in the Crowd Sourced Name Point features dataset within two months of being submitted, but they can be viewed in the Vernacular Names Tool on the [OS Data Hub](https://osdatahub.os.uk/) before then. The Vernacular Names Tool is a resource developed by Ordnance Survey to assist Emergency Services by incorporating user-submitted colloquial names for geographic features across Great Britain. Names will not be published if they have been flagged as containing personally identifiable information or potentially offensive language. They will also not be published if the name is identical to an existing OS feature or is of an unsuitable classification (for example, a defibrillator as their locations are not currently captured).
Why is the same name in the Crowd Sourced Names Point features dataset multiple times? OS is not looking for duplicates within Crowd Sourced Names Point features, so if the same name is submitted multiple times, all entries will be published.
### OS NGD Land and Land Use Themes' FAQs
How does OS NGD Land Theme data differ from OS NGD Land Use Theme data? OS NGD Land Theme data contains land cover attribution that describes the physical and biological surface of the land. OS NGD Land Use Theme data contains land use attribution that describes the human or economic activity taking place on areas of land (i.e. how areas of land are being used / their purpose).
How does the OS NGD Land Theme differ from the OS MasterMap Topography Layer, Sites Layer, and Greenspace Layer products? The OS NGD Land Theme contains some land information data which can also be seen in the OS MasterMap Topography Layer, Sites Layer, and Greenspace Layer products. The data in the theme is similar as that provided by the aforementioned OS products, but it has been structured to be more analytical and consistent across the six OS NGD features collections, allowing you to easily select the specific features you require and delivering greater value to users. (The six OS NGD features collections, which also include data from the three aforementioned OS products, are the OS Building Features, OS Land Features, OS Land Use Features, OS Structure Features, OS Transport Features, and OS Water Features Collections.)
What are the NGD land cover enhancements? Please refer to the [OS NGD Land cover enhancements](https://docs.os.uk/osngd/using-os-ngd-data/os-ngd-land-cover-enhancements) section for further information.
Why are Wind Farm site polygons the shape they are? Wind farms are a complex site to model as they are often constructed across large areas of open land with much of the ground around individual turbines being used for other activities, such as animal grazing or a solar farm. Some Wind Farm sites, in contrast, may be wholly contained within another existing site, such as a factory. Additionally, access roads may extend a significant distance from the turbines, or each turbine may have individual access. Consequently, what is recognised as a Wind Farm extent is heavily dependent on the surrounding geography, the nature of the wind farm, and the user’s perspective. OS has chosen not to include the entire extent of the land parcels below turbines as the definition of a Wind Farm extent because this can sometimes translate to many thousands of hectares in mountain and moorland areas and therefore misrepresent the real-world site. Other sources of land extent data such as land registration and planning area data were considered, but did not provide consistent high-quality coverage across Great Britain. Instead, a nominal 300m buffer was applied to the individual wind turbines associated with wind farm schemes referenced in the Renewable Energy Planning Database to indicate the general area of land used as a Wind Farm. Land Use Sites extents are not meant to represent specific concepts like land ownership, which is made clearer by this nominal geometry design choice. Example map showing the nominal 300m buffer around individual Wind Turbines. This particular Wind Farm site (Wear Point Wind Farm) has four turbines.
### OS NGD Structures Theme FAQs
How does OS Structure Features Collection data differ from structure features data available in the OS MasterMap Topography Layer product? In comparison to OS MasterMap Topography Layer data, OS Structure Features Collection data has been enriched with additional attribution, including the following: * Geometry change metadata * A Description attribute to provide a one-word or simple phrase description of a feature * Detailed land use classification * Detailed land cover classification * Capture specification to give details about the geography of the area a feature resides in (that is, urban, rural, or moorland) * Embedded height attribution for chimneys, roofed tanks, and glasshouses, enabling users to create simple height models of these structures. Field Boundary features also have a height value. In addition, other feature types are available in the OS Structure Features Collection: the Compound Structure Feature Type and the Field Boundary Feature Type.
What is a compound structure? The Compound Structure Feature Type is a polygon feature which encompasses one or more components and represents a manmade construction that has been built for a specific purpose. Examples include a bridge, a dam, and an aqueduct. A compound structure essentially combines multiple polygons into a single polygon, making it easier for you to analyse and query data when you want to analyse a structure as a single part rather than its constituent parts.
Why don't all structures have names? Name attribution, where known, is provided for structure features to help you identify what a structure is called. This attribute is optional as not all structures have a name and, in some instances, OS may not yet have captured a structure's name in the data.
What structures will have height attributes allocated? Embedded height attribution is available for chimneys, roofed tanks, and glasshouses, enabling users to create simple height models of these structures. Field Boundary features also have a height value.
Where can I learn more about Field Boundary features? Please refer to the [Field Boundary](https://docs.os.uk/osngd/using-os-ngd-data/os-ngd-structures/field-boundary) section for further information.
Why are some London Underground stations missing from Tube or Metro Tunnel features in Compound Structures? OS's capture focus is on tunnels that reach the surface, meaning that the Metropolitan and District Lines of the London Underground are captured, whereas other London Underground lines are generally not captured.
Why are there temporal differences in data currency for bridge interactions? Bridges in the OS Structure Features Collection are updated daily, whereas networks in the OS Transport Network Collection and the OS Water Network Collection are updated monthly and quarterly, respectively. This can lead to temporal differences in data currency for bridge interactions.
Which third-party bridge information is available? OS only adds third-party bridge information as attribution to bridge features if it has been provided by an authoritative third party. Initially, this information has been provided by organisations for canal bridges, but OS looks to add information for other types of bridge, like rail, in the future.
### OS NGD Transport Theme FAQs
Can I use OS NGD Transport Theme data for routing? Yes. Data in the OS Routing and Asset Management Information (RAMI) Collection can be fed into routing software. The routing information provides details about potential routing hazards (for example, dangerous bends and low bridges) and vehicle restrictions (for example, one-way streets and road widths).
How has the Rail Network been generalised? The generalisation of Rail Networks should ensure full connectivity between relevant Rail Nodes. Rail track generalisation can be typically three or four Rail Links represented as one link, for example, multiple siding tracks. However, this may not always be the case, for example, tracks that pass either side of a station platform are normally included to ensure connectivity at a station.
How will tracks be distinguished from paths? Paths and tracks will both be represented in the OS Transport Network Collection. The Description attribute provides a one-word or simple phrase description of a feature, and you'll be able to use this attribute to easily distinguish between paths and tracks.
What routing and asset management information is included in the OS RAMI Collection? The OS Routing and Asset Management Information (RAMI) Collection provides in-depth routing and asset management information for roads across Great Britain. Rights and restrictions define properties that restrict, regulate or prohibit the use of the network by traffic which may affect the allowable route for users. Advisory information provides additional information that may either affect a driver’s preferred choice of route or highlight potential hazards, such as the presence of a toll or a level crossing. Asset management information provides additional information describing the network to support its long-term operation and maintenance.
#### Bus lane data FAQs
What is the coverage of bus lanes? There is full data coverage across Great Britain for [bus lanes](https://docs.os.uk/osngd/data-structure/transport/transport-network/bus-lane) data.
What is the update frequency of bus lane data? The [Bus Lane Feature Type](https://docs.os.uk/osngd/data-structure/transport/transport-network/bus-lane) is updated into the OS NGD monthly. Bus lane attribution on the [Road Link Feature Type](https://docs.os.uk/osngd/data-structure/transport/transport-network/road-link#presenceofbuslane_overall_m) is also updated monthly.
How have bus lanes been captured? Bus lanes are sourced from OS aerial imagery, for built-up areas only. Where features are not visible from aerial imagery (for example, where they're obscured by tree cover), these will not be included. You can record any omissions on the [OS Data Hub Error Reporting Tool](https://osdatahub.os.uk/errorsAndOmissions) for OS to investigate. Please see the [known limitations page on Using OS NGD Data](https://docs.os.uk/osngd/using-os-ngd-data/os-ngd-transport/bus-lanes/known-limitations) for more information.
Are Guided Busways, Bus Gates and High Occupancy Vehicle Lanes included in the Bus Lane dataset? Bus lanes are defined as a section of road that is segregated by a physical feature or painted section of road that indicates primary usage for buses. Guided Busways and Bus Gates have been captured in the dataset as they meet these criteria. However High Occupancy Vehicle (HOV) Lanes have not been captured because although buses can sometimes use these lanes, this is not their primary use. OS views HOV Lanes as being distinct from Bus Lanes and Bus Gates which have the primary use for buses, sometimes with exemptions for cyclists, motorcyclists and taxis.
#### Cycle lane data FAQs
What is the coverage of cycle lanes? There is full data coverage across Great Britain for [cycle lanes](https://docs.os.uk/osngd/data-structure/transport/transport-network/cycle-lane) data.
What is the update frequency of cycle lane data? The [Cycle Lane Feature Type](https://docs.os.uk/osngd/data-structure/transport/transport-network/cycle-lane) is updated into the OS NGD monthly. Cycle lane attribution on the [Road Link](https://docs.os.uk/osngd/data-structure/transport/transport-network/road-link#presenceofcyclelane_overall_m) and [Path Link ](https://docs.os.uk/osngd/data-structure/transport/transport-network/path-link#presenceofcyclelane_overall_m)Feature Types is also updated monthly.
How have cycle lanes been captured? Cycle lanes are sourced from OS aerial imagery, where they exist alongside publicly accessible Road Links and Path Links. Where features are not visible from aerial imagery (for example, where there is no visible evidence of a cycle lane or where it's obscured by tree cover), these will not be included unless they have been identified through other intelligence sources. In particular, the completeness of Unsegregated Shared Use Cycle Lanes will be lower as these features are typically indicated by blue signage at street level and don’t have a clear designation that can be seen from aerial imagery. Please see the [Known limitations page on Using OS NGD Data](https://docs.os.uk/osngd/using-os-ngd-data/os-ngd-transport/cycle-lanes/known-limitations) for more information. You can record any omissions on the [OS Data Hub Error Reporting Tool ](https://osdatahub.os.uk/errorsAndOmissions)for OS to investigate.
Are cycle routes included in the Cycle Lane Feature Type? The Cycle Lane Feature Type represents designated cycling infrastructure, not cycle routes. Features are only captured where a section of road or path is designated for cyclists, either segregated by a physical feature, identified by painted or other signage, or identified by a different colour of tarmac or surface type. Please note that routes that are part of the National Cycle Network are not captured as Cycle Lane features unless designated infrastructure is present that meets these criteria.
#### Pavement data FAQs
Why are the Pavement Links depicted in the centre of the road? The pavement data in the OS NGD uses the Road Link geometry in order to give the features a geographic location; therefore, pavements will be depicted using the road centre line.
Why within the Pavement Link Feature Type do I only see a centre line but the road has a pavement on both sides? When there is a pavement on both sides of the road, we create two features both with the same geometry; therefore, if viewed in a GIS, these will appear on top of each other, but each individual feature will have specific attribution relative to its own side of the road.
Why are there gaps in the depictions of pavements within the Pavement Link Feature Type? Often, the answer is there is no pavement adjacent to the Road Network. Some gaps appear, typically at junctions, where the end of Road Links have no adjacent pavement.
Why do pavements ‘stop’ at bridge intersections within the Pavement Link Feature Type? Currently OS does not capture large-scale depictions of features such as pavements when they are obscured by structures from above.
How do I interpret which side of the road is left and which is right? The left and right side of the road is determined by the direction of digitisation of the Road Link. In Direction of digitisation is from start node to end node, and Against Direction is from end node to start node.
Why do I have two Pavement Links for the same section of road? When there is a pavement on both sides of the road, we create two features both with the same geometry; therefore, if viewed in a GIS, these will appear on top of each other, but each individual feature will have specific attribution relative to its own side of the road.
#### Speed data FAQs
What is the source of indicative speed limit and average speed data? Indicative speed limit data is sourced from road signage and surveying. Average speed data is calculated based on historic speed information, which is based upon in-vehicle telematics devices; in this context, 'historic' means that the average speed data was collected over a six-month period for the selected road link.
What is the update frequency for average speed data and indicative speed data? Speed data is published monthly into the OS NGD. Indicative Speed data is updated by our third-party supplier four times a year, and Average Speed data is updated twice a year. In-between these updates, OS 'infills' speeds with inferred values to account for any changes in the network monthly.
Why is average speed provided in kilometres per hour (kph) rather than miles per hour (mph)? Average speed is provided in kph to support travel time analysis software in the market which predominantly uses kph as the unit of speed.
#### Street light data FAQs
How has Street Light data been sourced? Street Lights have been captured from OS aerial imagery.
What is the coverage and completeness of Street Light data? Street Light data is GB coverage. Features have been captured where they exist alongside publicly accessible Road Links and Path Links. Street Lights will not be included where they exist along restricted access roads. Where features are not visible from aerial imagery (for example, where obscured by tree cover or not visible due to urban shading), these will not be included.
Does attribute data for Street Lights include timings for when they are on or off? No, the Street Light dataset does not contain information on the times when they are on / off.
How has the Street Light coverage attribution been created? An algorithm is used to give an indication of how well-lit a link is; this algorithm is based on an inferred light radius and does not account for information such as the height of a lamp post, the type of bulb, or the presence of any light-blocking obstructions or buildings.
What is the update frequency for Street Lights? Street Light data is updated into the OS NGD monthly. This includes both the Street Light Feature Type, and Road Link and Path Link attribution.
Are there any known limitations for Street Light data? Yes. There are two known limitations: * **Street Light Completeness:** Street Lights have been captured from aerial imagery and therefore there are some limitations to completeness. Where features are not visible from aerial imagery (for example, where obscured by tree cover or not visible due to urban shading), these will not be included. * **Overlighting and Underlighting:** Street Light coverage attribution on Road Links and Path Links is indicative only, and there may be some instances of overlighting and underlighting. An algorithm is used to give an indication of how well-lit a link is; this algorithm is based on an inferred light radius and does not account for information such as the height of a lamp post, the type of bulb, or the presence of any light-blocking obstructions or buildings.
### OS NGD Water Theme FAQs
Why are more water courses depicted in the Water Links Feature Type than in the Water Link Sets Feature Type? Water link sets are only created where there are a number of water links that share the same name. Within the Water Link Feature Type, there are numerous water links which are unnamed; these unnamed water links therefore will not appear in the Water Link Set Feature Type as they cannot be joined accurately.
How is the In Country attribute determined for Waterbody Catchment features that straddle the border between two countries? Country attribution for catchments that straddle the Wales-England or Scotland-England border is as provided in the third-party data from the authoritative bodies and is aligned with the Geometry Source attribute (i.e. the source organisation that provided the data: Environment Agency for England, Natural Resources Wales for Wales, and Scottish Environment Protection Agency for Scotland).
Why are coastal catchments missing from Waterbody Catchment features? Coastal catchments for this feature type are not supplied in the third-party data from the authoritative bodies.
Why do River Basin District Catchment features include geometries for islands located far away from the mainland, including outside the extent of the British National Grid? The data for this feature type is as provided by the third-party data from the authoritative bodies.
What are the recommended scenarios for using Tidal Boundary features versus continuous tidelines, and how should I choose between them? [Tidal Boundary features](https://docs.os.uk/osngd/data-structure/water/water-features/tidal-boundary) should be used when segmented representations of mean high water and mean low water tidelines with richer metadata are required for detailed analysis or site-specific mapping purposes. Continuous tidelines (i.e. [Tidal Boundary Continuous High](https://docs.os.uk/osngd/data-structure/water/water-features/tidal-boundary-continuous-high) and [Tidal Boundary Continuous Low](https://docs.os.uk/osngd/data-structure/water/water-features/tidal-boundary-continuous-low) features) are recommended for applications that benefit from seamless, aggregated water boundaries with better intuitive visual experience, such as macro analysis or when updating boundary features to align with coastlines.
How are the watercourse widths calculated? In [Water Link Feature Type](https://docs.os.uk/osngd/data-structure/water/water-network/water-link) data schema version 2.0 onward, there are three possible methods for determining width values: * **Calculated From Polygon** – For Water Links which represent a body of water captured as a topographic area feature, calculations of average, minimum and maximum widths are based on the topographic area feature geometry. For some short links at junctions, the average width will be Modelled From Network instead. A small number of Water Links may fall outside of the capability of the Calculated From Polygon algorithm; in these cases, a Default Value for Geographic Region algorithm will be applied to determine width values instead. * **Modelled From Network** – For Water Links which represent a water feature captured as a topographic line feature, average width is a modelled width based on a known calculated downstream width and the number of upstream sources in the network. Where a calculated downstream width isn’t available, then a Default Value For Geographic Region algorithm will be applied instead. A small number of short Water Links at junctions which represent a body of water captured as a topographic area feature will also be Modelled from Network. A small number of Water Links may fall outside of the capability of the Modelled From Network algorithm; in these cases, a Default Value for Geographic Region algorithm will be applied to determine width values instead. * **Default Value For Geographic Region** – For Water Links captured as a topographic line feature (where no width has been processed), average width is a default width based on the capture specification. For Water Links which represent a body of water captured as a topographic area feature, there may be small number of features that fall outside the capability of the Calculated From Polygon algorithm and Modelled From Network algorithm to create; in these instances, a Default Value For Geographic Region algorithm will be applied to determine width values instead. The [Width Derivation Method attribute](https://docs.os.uk/osngd/data-structure/water/water-network/water-link#width_derivationmethod) indicates which method has been used to populate the [Width Average](https://docs.os.uk/osngd/data-structure/water/water-network/water-link#width_average), [Width Minimum](https://docs.os.uk/osngd/data-structure/water/water-network/water-link#width_minimum) and [Width Maximum](https://docs.os.uk/osngd/data-structure/water/water-network/water-link#width_maximum) attributes.