Thank you; these are interesting to look at.
One thing I notice; there seems to be quite a big variation in the traffic light data between different dates. So for example if I compare the same sites between your Railton map and the Ferndale one, which seem to be based on sampling weeks about a month apart. The Branksome Rd junction is +11% on one and -15% on the other.
Or, Villa Rd/Brixton Rd: +43% on the Railton map (September 2020), +51% on the Ferndale map (October) and then -22% on the Oval map (December).
I assume similar variations would be seen in traffic counts.
It's probably a mix of the variations in traffic, as you said, and the fact that Lambeth have used different weighting from LTN to LTN and week to week.
Railton week of the 21 September - 93.59%
Ferndale week of the 19th October - 78.02%
Tulse Hill week of the 29th October - 89.21%
Oval month of December - 76.53%
There are a few inconsistencies within the Lambeth data that make you wonder how closely it was checked
In the Oval one
Site 2: Meadow Road - North uses a weighting of 99.55% and Site 4: Meadow Road South uses 75.21%.
Site 11: Stockwell Park Road uses 96.47%
Ferndale is pretty consistent as is Tulse Hill.
Railton has a couple of strange ones, Site 8: Railton Road (Daily Flows) at 86.57% and Site 9: Regent Road (Daily Flows) at 86.5%
and the table of data for Site 17: Coldharbour Lane (Daily Flows) is scramble and repeats the PreCovid numbers as the Baseline.
it's difficult to know why neighbouring areas vary in weighting within a relatively short period of time, let alone 2 halves of one street (Meadow Road).
I've tried to use the averages for each LTN and week as above.