When I first published this article I was keen to illustrate spatial variations in confirmed cases of Covid-19 across England when adjusted for local populations. I used data for April 1st to show these variations. Much discussion was generated but what was apparent at this relatively early stage in the spread of the virus was that local variations in testing capability was having an impact on the spatial variations and as such it was difficult to identify places as genuine case hot-spots.
I leave the 01-Apr map here for reference. Outside of London (already recognised at the time as the centre of the UK's outbreak) places such as Cumbria and Tyne & Wear appeared to be hot-spots - the former somewhat unexpectedly (given its rurality) though subsequent media reports have suggested that the Cumbria outbreak was caused by people returning from skiing holidays.
I have subsequently been analysing the Public Health England (PHE) data weekly (using every Monday's data) for 10 weeks since 23-Mar. I set the scale of the map using a log scale in anticipation of an exponential growth profile in cases. The length of the scale is necessarily wider than the current data fit as I wanted to future-proof the maps and allow for greater saturation of cases within the general population.
The maps scale converts to cases/100k as follows:
0 - 1 = 0.00 - 2.72
1 - 2 = 2.72 - 7.39
2 - 3 = 7.39 - 20.08
3 - 4 = 20.08 - 54.60
4 - 5 = 54.60 - 148.41
5 - 6 = 148.41 - 403.43
6 - 7 = 403.43 - 1096.63
7 - 8 = 1096.63 - 2980.96
8 - 9 = 2980.96 - 8103.08
It is apparent from the animation that the exponential growth in cases has abated. The past two weeks have seen little movement in local authorities case incidence with respect to the map scale. Much of England is in two shading categories 4-5 and 5-6 with just a handful of local authorities having crept into the 6-7 band. With the exception of two London Boroughs (Brent and Harrow) all of the spikes in cases are in the North of England (parts of metropolitan Greater Manchester, Merseyside, South Yorkshire and Tyne & Wear as well as Middlesbrough, Cumbria and Blackpool.
The more insulated parts of England are clearly the South West along with East Sussex, Brighton & Hove, Isle of Wight and the more rural Lincolnshire and Rutland. A real oddity is Calderdale in West Yorkshire which bucks the trend for the towns and cities of the M62 corridor.
In order to shed light on more subtle variations in case incidence I have produced a static map using data from Monday 25th May. This helps us to see what is likely to happen to the map in the coming weeks. It again uses 9 shading categories but this time the scales are from 4.00-6.25 in 0.25 increments. Now it is possible to identify those authorities that are closer to moving into the 6-7 bracket (over 403 cases/100k) currently occupied by Cumbria, Tyne & Wear etc. Large parts of Greater Manchester, Merseyside, South Yorkshire and North Cheshire, a good number of London Boroughs, parts of the West Midlands, Bedford, Luton & Reading are all relatively close to moving into the 6+ shading category.
