Daily data on the maximum simultaneous usage numbers for the larger UK bikeshare systems. Used in this CDRC Data Story on how bikeshare has been impacted, both negatively and positively, from changes in traveller behaviour due to COVID-19 and the associated lockdown in the UK, between 1 March 2020 and 21 April 2022.

Column definitions: Aggregated activity scores

systemgroup: allbig is all the systems in the dataset that had over 100 bicycles as of early 2020. londonbig are the ones that are in London (this was only the central London Santander Cycles at the time), while restofukbig are the others (Bournemouth, Brighton, Cardiff, Edinburgh [now closed], Glasgow, Hereford, Liverpool [now too small] and Watford.

Note: The population of the systemgroup categories was intended to be stable to allow a long-term comparison, however some (e.g. Belfast and Cardiff) closed for significant periods during the last two years, while Edinburgh permanently in 2021 and Liverpool reduced below 100 bikes in later 2021, so is no longer included. Depending on the desired time period of study, users may wish to construct their own stable grouping using selected cities from the individual data.

daytype, theinterval, therange: See below.

therange_wrtbaseline: This is therange_pc with respect to the average seen for that systemgroup/saytype/theinterval combination during the first two weeks of March 2020.

Column definitions: Individual city activity data

System is the name of the bikeshare fleet. In most cases this corresponds to the city it is in, although this can be a simplification – the bournemouth entries include the full extent of the Beryl system there which also is in neighbouring Poole and Christchurch, for example. london is the Santander Cycles in central London. Only systems with over 100 bicycles are included.

Note: Several systems opened (e.g. Leicester Ride-On, Norwich Beryl, Dundee Ride-On, Forth Valley, London Human Forest, London Jump, London Beryl, London Freebike, Stirling Nextbike, West Midlands Cycle Hire, Bristol YoBike), or became larger than this threshold, after 1 March, therefore they do not feature in the aggregated file above, but do have individual rows in this file.

daytype shows whether it’s a weekend/holiday day or not. Note that the English bank holiday calendar is used, including for the non-English systems in the list.

theinterval indicates whether the numbers on this row relate to the 6am to 10am period only (ampeak) or are fullday which runs from 6am to midnight. The thinking behind this was to separate commuter peaks from an overall peak. Some bikeshares are commuter dominated (e.g. london) while many are not.

first, low, high and last are the number of bikes available to ride in each system, at the start of the interval period, the lowest number seen during the period, the highest number, and the number at the end of the period. The idea is if the difference between low and high was 100, then a maximum of 100 bikes were out at the same time, during that reporting period.

Firstts, lowts, hights and lastts are the time of day that the first, low, high and last values occurred, respectively. Firstts should always be around 6am, and lastts should therefore always be just before midnight or just before 10am.

Finally, therange and therange_pc list the difference between high and low, both in the number of the bikes, and as a percentage of the largest number. This final statistic gives you an idea of how busy the system has been that day.