About our Datasets

The Geographic Data Service (GeoDS) obtains, creates, maintains and supplies datasets and research platforms to a wide range of users who wish to conduct research using smart data. Our key audience is UK academic researchers and students, however we aim to deliver data for bona-fide research purposes where our agreements allow.

Please see our main about page to learn about the Geographic Data Service, its context and focus.

Dataset Categories

We provide data with three different tiers of access:

Open

Data which are freely available to all for any purpose following a simple registration. Products include those created and acquired by the GeoDS and products where the GeoDS have added value. We also host a number of online tutorials, these typically use our Open data. For convenience, these are shown as "Tutorial" rather than "Open" in the data catalogue.

Safeguarded

Data to which access is restricted because of license conditions, but where data are not considered ‘personally-identifiable’ or otherwise sensitive. Data is delivered to successful applicants on a time-limited basis following registration, project approval requirements, training and signing of a "user agreement" data licence by the user, their institution and UCL (acting for GeoDS).

All academic applications need to have ethics approval (or statement of exemption) from their institution in place before the data is delivered. In addition, if the dataset is anonymised data about individuals (aka microdata) then all applicants must supply proof of safe researcher training completed - we generally recommend the MRC course "Confidentiality and Data Protection in Health Research" at https://bygsystems.net/mrcrsc-lms/course/view.php?id=95.

To apply please look at each individual dataset and following the "Request Data" link there. The complete process, from initial application to final data delivery, typically takes 1-3 months.

For some Safeguarded datasets, which relate to aggregated datasets which we have produced ourselves, or where review provisions by our data partners no longer apply, and we are comfortable based on the full application description that the project does not present ethical or other concerns, we may be able to fast-track our review and delivery process.

Secure

Our Secure datasets are retained in our secure lab for data protection and/or commercial confidentiality requirements, and access is available to researchers with ac.uk and nhs.net email addresses, virtually through a web browser, with registration and project approval requirements, training and signing of a "user agreement" data licence by the user, their institution and UCL (acting for GeoDS).

All applications need to have ethics approval (or statement of exemption) from their institution in place before the data is placed within the user's share in our lab. In addition proof of safe researcher training (SRT) completed is required. We accept ONS SRS, UKDS or similar SRT, and can book you on a UKDS SRT course if needed. You will need to spend some time completing the course and assessment. Finally, our lab is run by UCL's Advanced Research Computing service and they also have their own separate online data security training requirements, which we will advise you of in due course.

To apply please look at each individual dataset and following the "Request Data" link there. The complete process, from initial application to final data delivery, typically takes 2-4 months.

Internal datasets

A small number of datasets cannot now be accessed by external researchers through the service, however these are retained for internal GeoDS research and data product creation.

Removed datasets

Many of the datasets that we resupply from data suppliers (as opposed to those we have created or derived ourselves) come with a time-limited upstream data licence agreement which means we are no longer able to supply them on the GeoDS Data Service once the agreements expire, and normally are required to delete all copies we hold ourselves. To facilitate academic understanding and service historic references to the data, we retain some information about these datasets.