All times in this article are in Coordinated Universal Time (UTC).

It passed close to Witney, Kidlington and Bicester in Oxfordshire. Detailed site investigations of damage are still being undertaken so the details given below are provisional. However, initial reports suggest there may have been three separate tornadoes near Eynsham, Kidlington and Bicester. There was extensive damage to trees, though property damage appears slight. The strongest damage suggests that winds reached a strength of F1 or T2-3 on the two scales of tornado intensity (Fujita and TORRO scale). Numerous photos and/or video footage of the rotating updraft and funnel clouds are available on the internet. Hail of marble size, or slightly larger, was reported at Witney and many other locations along the storm's path. Eyewitness accounts of the storm can be found in this BBC News article

The Chenies and Dean Hill Doppler radars detected rotation on a scale of about 5 km - larger than an individual tornado but nevertheless an indication of a rotating storm cell or mesocyclone (see first figure below). During the afternoon lightning strokes were detected moving from SW to NE along the track from south of Gloucester, near the county boundary, towards Bedford between 15:00 and 18:00 (see the second figure below).

Doppler radial winds measured by the weather radars at Dean Hill (left) and Chenies (right). Positive values are towards the radar site and are emphasised by the arrows.

Observations of sferics (lightning flashes) during the preceding hour on 7 May 2012, showing 15:00 - 133 sferics (top left), 16:00 - 166 sferics (top right), 17:00 - 101 sferics (bottom left), 18:00 -114 sferics (bottom right).

As shown in the first figure below, the UKV T+6 forecast for 15:00 and the 15:00 analysis both had some indication of the location of the storm which produced the tornado, the 15:00 NDP analysis had a more accurate depiction of the location of the precipitation although it didn't clearly identify the two separate storm centres with too much precipitation W-E rather than SW-NE in the forward cell, the precipitation further west was better represented in the NDP analysis.

The second figure below shows that the NDP forecasts from 12:00 onwards all had a good forecast of the location of the storms at 15:00 with the forecast skill improving with time. However, these forecasts could not resolve the actual tornado.

Further information

More Information on the NDP system and example forecasts can be found by following links below:

• The Nowcasting Demonstration Project for the London 2012 Olympic and Paralympic Games
• The NDP Data Assimilation system
• NDP forecasts of the 28th May 2012 thunderstorm line from Kent to Northamptonshire/Leicestershire border

A comparison of the analyses valid at 15:00 on 7 May 2012 for the UKV (left) and NDP (centre) models with the radar-observed rain rate.  

Forecasts or analyses valid at 15:00 on 7 May 2012 of the rain rate from the NDP forecast cycles between 12:00 and 15:00.