please empty your brain below
|
|
Nice post. Where Biggin Hill's concerned, the lack of a nearby railway station used to be quoted as a reason why the property there was noticeably cheaper. Pre-Beeching, there was a branch line out to Westerham from Dunton Green, running along the foot of the downs - a route now covered by the M25. The Westerham would have been closer to Biggin Hill than current stations in the area, albeit down a very long and steep hill.
Waterhouse | 08.07.15 - 8:02 a.m. | #
|
By the way, Knockholt is still in Kent.
dg writes: Fixed, ta.
Waterhouse | 08.07.15 - 8:05 a.m. | #
|
Anyone that has to go to a station over a mile or two away is going to catch a bus (within London) or drive (to the station car park) unless they are feeling particularly spritely.
I expect that Oxted has a car park (behind the Wetherspoons it appears). Which can't be said for most rail stations within London. I also suspect the roads are a lot quieter than those within London making such a journey less frustrating. So maybe that area appears to be badly connected, but is, in fact, well connected if you can drive a short distance.
Sykobee | 08.07.15 - 9:23 a.m. | #
|
Didn't they gerrymander out the North Ockendon salient a few years ago by making it flush with the M25, in the same manner as the short point extending almost as far as Potters Bar?
dg writes: Nope, still very much London.
Matt | 08.07.15 - 10:22 a.m. | #
|
Surely they should be concave-sided polygons, (Hypocycloids) each side having a radius of 2 miles (or three, or four) and centred on the nearest stations?
@Matt
The Great Warley part of the "salient" outside the M25 was transferred to Brentwood in the 1990s, but the area south of the railway (Ockenden) chose to stay put.
Potters Bar was a salient of Middlesex into Hertfordshire (or Barnet into Hertfordshire) but this was tidied up with the formation of the GLC in 1965. There have been some minor changes since, mainly to align the boundary with the M25 and to accomodate expansion of the urban areas of Elstree and Barnet.
timbo | 08.07.15 - 10:55 a.m. | #
|
|
Knockholt was not always in Kent - it was part of the London Borough of Bromley from 1965 to 1969.
timbo | 08.07.15 - 10:58 a.m. | #
|
|
That is pretty impressive actually - that London has that many good connections, and also DG for finding this out!
AliE17 | 08.07.15 - 11:20 a.m. | #
|
The actual furthest point from a station must be equidistant from three or more (since three points are necessary to define a circle, and if you move away from the centre of that circle you must move closer to one of them). Within central London the furthest point is somewhere near the Albert Memorial, with Mount Pleasant and Cheyne Walk also quite distant.
The point you highlighted is equidistant from Oxted and Dunton Green, but Chesfield and Knockholt are a little further away. If you go a few hundred metres further north along that bisector, you will be receding from both Oxted and Dunton Green, whilst Chesfield and Knockholt get closer. Eventually you will reach a point where three (or maybe four) of them are equidistant. It looks to be about 200m N of where you identified.
(and now I must stop playing with the radius app)
timbo | 08.07.15 - 11:40 a.m. | #
|
|
OK timbo, I've relocated the final location approximately 200m to the north of Grays Lane. And that's four and a quarter miles from a station, so even better (or worse, depending). Thanks!
diamond geezer | Homepage | 08.07.15 - 12:16 p.m. | #
|
Couldn't resist
Mount Pleasant 850m from Kings Cross, Russell Square, Farringdon and Angel
On the Flower Walk NW of the Albert memorial 1km from High St Kensington, Bayswater, Lancaster Gate, South Kensington and Gloucester Road
South End of Battersea Bridge 1350m from Imperial Wharf, Battersea Park, and Sloane Square
Roehampton Vale (A3 bridge over Beverley Brook) 3.4km from Southfileds, Raynes Park, Norbiton, barnes
Grays Road, Biggin Hill(ish) (near Woodlands cattery! 6.9km from Dunton Green, Chesfield and Oxted
timbo | 08.07.15 - 12:31 p.m. | #
|
|
What you said - I'm not as good with the radius tool as you and didn't go off-road!.
timbo | 08.07.15 - 12:36 p.m. | #
|
timbo says "The actual furthest point from a station must be equidistant from three or more"
Only true if London was of infinite extent! Otherwise there could be points on the London boundary equidistant from only two or one station. In a manner of speaking. (Actually I cannot find any such points, but they could exist in theory).
Malcolm | 08.07.15 - 12:55 p.m. | #
|
@Malcolm - indeed, there are boundary conditions, (the boundary intersecting the hypercycloid, and the centre of that hypercycloid being outside the boundary). Indeed it's surprising (to me) that the centres of all eight of DGs two-mile red zones are within the GLA boundary.
There will be many places on the GLA boundary equidistant from two stations: to find them, draw a line between any two stations, find its bisector (crossing the first line at its midpoint, at right angles). That bisector is infinitely long, and every point on it is equidistant between your two stations. If it crosses the GLA boundary, it will do so an even number of times (although it is unlikely that more than one of them will be at a point where no other station is closer than the two used to define the bisector).
Anywhere not on such a bisector is necessarily closer to one station than any other.
The centre points of the circles, which define the most remote point, are where these bisectors intersect.
I am intrigued by the idea of being "equidistant from one station", but I known what you mean - Falmouth is the closest station to the most rail-remote station on mainland Cornwall - but the centre of the circle it falls in, defined by Falmouth, Penzance and Roscoff, is a long way out to sea.
timbo | 08.07.15 - 1:55 p.m. | #
|
I had a go at doing what timbo suggested (at 10:55) using the rail and metro info from the National Public Transport Data Repository. Think there may be another site just to the north of Chaldon Church near Caterham? The map also suggests that points 2 and 4 are within 2 miles of a station. Of course it depends on data position quality and I'm not exactly a GlobalMapper superuser!
https://www.google.com/maps/d/edit?mid=zpHPeLx_vkUk.kj0YQrm5sg0c&usp=sharing
Geo_rich | 08.07.15 - 3:50 p.m. | #
|
Blimey that's a good (and useful) map, thanks.
I think a lot of the detail depends on precisely where you think a station is...
(My ruler thinks the site to the north of Chaldon is with a mile of Caterham station)
...but your curves are far more accurate than my approximate blobs, cheers!
diamond geezer | Homepage | 08.07.15 - 4:30 p.m. | #
|
@Geo_rich: Using the tried and trusted method of Explorer map and dividers, there seems to be a point north of Chaldon church as near as dammit 2 miles from Coulsdon South, Caterham and Merstham stations.
Knockholt is in Kent, but Knockholt Station is (just) in Greater London.
Steve Bird | 08.07.15 - 4:37 p.m. | #
|
A few human beings, if not people, live in Richmond Park.
https://en.wikipedia.org/wiki/Thatched_House_Lodge
Mike D | 08.07.15 - 4:55 p.m. | #
|
|
There also must be a single "prime" spot (very likely in central London) that offers walking distance to the most travel options by rail? i.e somewhere at equal distance to access services out of London to say north of England/Scotland, the west and Wales, the East Coast and South Coast and of course access to Heathrow/Gatwick express. Where this "golden point" is I don't know. Somewhere near Hanover Square perhaps?
OnTheBus | 08.07.15 - 6:32 p.m. | #
|
|
The Royal Ballet School for students aged 11-16 is in Richmond Park too so the Park is populated after all
Sara | 08.07.15 - 6:36 p.m. | #
|
I too am impressed by geo_rich's map; forgot to say earlier that I was also impressed by DG's research. Just a comment for anyone who takes a look at geo_rich's map;the flower-like structure which you first see is an artefact of the limited subset of station used. The interesting bits come when you zoom in and see the hypocycloids.
In principle the same thing could be done using the same technique for Lonodon bus stops. I believe the data must exist, it must be part of how traveline produces maps with bus stops on it. But I've no idea whether the data is freely available.
Malcolm | 08.07.15 - 7:49 p.m. | #
|
Malcolm, yep the bus location data is in the NPTDR, along with river boat services, DLR stations etc. Its nationwide too, so one could run a similar map for the whole country. If you look at the viewer at the link below, it should be possible to find the point furthest in the UK from a source of public transport. Sounds like a challenge DG?
http://tables.googlelabs.com/DataSource?snapid=85105
Geo_rich | 08.07.15 - 8:30 p.m. | #
|
|
Until recently I used to live v near grays lane and commute into the city by train. It's a 12 minute drive to Orpington which at points during the day has a fast train into town. I could do door to desk in less than an hour, and that included 15 min or more of walking. Not bad as the day started out looking over fields.
Tom | 08.07.15 - 11:15 p.m. | #
|
|
Seeing as DG was kind enough to mention my map on the main post I've come over all proud and had to tidy it up a bit! Data within London is still the same, but now a consistent group of stations within five miles of the outer boundary are contoured. Also added is the 4 mile buffer around each station to help pin down area 8.
Geo_rich | 09.07.15 - 6:58 p.m. | #
|
Aww thanks.
And, as has been suggested, a 1 mile line would be fascinating too, in particular because it would start to show up gaps in accessibility in Central London.
(but probably a separate map, because I suspect that'd make a completely separate post)
diamond geezer | Homepage | 09.07.15 - 10:01 p.m. | #
|
What happens to "4-mile" section in the south east corner if you take into account the stations on the line between Redhill and Tonbridge?
As some stations on that line actually fall within the areas covered by the >2 mile radius, surely that'd have an effect?
D-Notice | Homepage | 09.07.15 - 10:09 p.m. | #
|
D-Notice - yep it would have an effect but only on the contours around the edge of the plan; those stations are sufficiently far outside London (> 5 miles) that the radius wouldn't affect the detail within the London area. I could have cropped to where the contours are valid, but kept it that way to show how the buffers are formed and for aesthetics.
It really comes down to where to set the edge of the plan; the current buffer considers several thousand data points and took about 90mins to calculate. When I first tried to do it for the entire UK dataset it crashed my rather old desktop PC!
Geo_rich | 09.07.15 - 10:53 p.m. | #
|
|
|
TridentScan v2.10.0 | Privacy Policy
|