The widening margin

Some more plots from the talk I gave the other way. I was trying to think of a way to illustrate in concrete terms the problem of speed for the air defence of Britain. I came up with the following:

Bomber time to London vs. fighter time to intercept height

Simply put, it shows the length of time it would have taken for an attacking bomber to fly from the coast to London (in blue) -- call it the crossing time -- and the time it would take taken for a defending fighter to climb high enough to intercept (in red) -- call it the intercept time. And how these changed over time, obviously. As can be seen, the fighters generally had enough time to climb high enough to intercept the bombers before they got to London, but the margin decreased over time, from 15 or so minutes during the First World War, to less than 5 in the Second.

But all this is not straightforward so I'll explain further. To begin with, the data is slightly dodgy. It's mostly drawn from the same source as this, which is fine as far as it goes. But that means that I'm showing how long it would have taken British bombers to penetrate from the coast to London, which was not really a great worry. Having said that, it's probably reasonable to assume that the performance of British bombers was roughly in line with those used by Continental air forces. (And the RAF's own air defence exercises had to make this assumption, too, because borrowing somebody else's air force for a day wasn't feasible.) One day I'll create a dataset for European aircraft ...

How are the numbers derived? First, the bombers (blue). This is just the distance from the coast to London divided by each bomber's maximum speed (which is not necessarily realistic). Why the coast? Because it was only when the incoming raiders crossed the coast that they could be detected by ground observers, and fighters dispatched to intercept them.1 What is the distance from the coast to London? Well, obviously it varies, depending on which direction the enemy came from (and some writers expressed fears that they would fly up the Thames Estuary and avoid detection). Looking at a map, 50 miles seems like a reasonable approximation.

Next, the fighters (red). The time it takes for a fighter to climb meet the bombers is the height of the raid divided by the climb rate of each fighter. This climb rate is a bit of a problem. I don't a good source for this number and had to plunder Wikipedia. That's bad enough in itself, but it's worse because the data is inconsistent. Sometimes -- when it's not missing -- it's expressed in feet per minute, and sometimes in the number of minutes to reach a given height. Obviously one can be turned into the other, but actually both are only approximations, and I've had to extrapolate and interpolate from these to get a usable number.2 What height would the bombers be at? Well, that varied -- it was higher on average during the Second World War than in the First because aircraft were more capable, and also because bombers tried to climb higher to escape the fighters. I've assumed that this height was 10000 ft in the 1910s, 15000 ft in the 1920s, 20000 ft in the 1930s, and 25000 ft in the 1940s.3 I just plucked these numbers out of the air, more or less, but they seem to work well in terms of keeping the red and blue trends in touch with each other. If anything they are probably underestimates.

Some other points. Firstly, the fighters would generally have to move horizontally to intercept the bombers, as well as vertically. This plot says nothing about that. But given the edge fighters had in speed and the location of their aerodromes, they should be able to cover that distance while climbing. Secondly, the data points are for the year each aircraft entered into RAF service. But since they remained in service for several year, at least, the data points should really be horizontal lines.4

Thirdly, I'm assuming a perfect command, control, communications and intelligence system. Fighter Command (and its predecessors) was good, but it still took a finite but non-zero amount of time for sightings to be reported, sifted, collated and reported, and then for squadrons to be allocated, given orders, and take off. Also there was a chance that raids might not be observed, that squadrons could be given the wrong vector, that the enemy could be missed in cloud -- so the greater the gap between the red data points and the blue ones the better. The more inefficient Fighter Command, the narrower the margin for error.5

Bomber time to London vs. fighter time to intercept height (radar)

Now we can show what difference radar made. The Chain Home system came into operation in 1939 and had an effective range of 120 miles. What this means here is that instead of only having to cross 50 miles from the coast to London after being detected by the observers on the coast, the bombers now had to cross 170 miles after being detected. As the above plot shows, this pushed up the crossing time dramatically: from 1939, the defenders could generally expect to have around 40 minutes' warning of any raids. The margin for error increased dramatically, from only 5 minutes or less, to more than half an hour, which is far better. In theory, the defending fighter squadrons would now have plenty of time to get in position before the enemy arrived. Of course, that's not the whole battle, but it's a good start!

Bomber time to London vs. fighter time to intercept height (acoustic)

Lastly, here's a counterfactual which I've long wondered about. Between 1933 and 1935, the Air Ministry put a fair amount of effort into researching the feasibility of using acoustic mirrors as a comprehensive early warning system. The acoustic mirrors were, mostly, concrete hemispheric dishes for focusing sound, which had been used as early as 1916. The biggest ones, at Dungeness in Kent and Maghtab in Malta, were 200 feet long curved walls. Land was actually purchased along the Thames Estuary for the beginnings of a national acoustic mirror system, but work never started because radar came along. But if it hadn't, then in 1940 Fighter Command might have relied upon a network of these acoustic mirrors all along the coast.6 How useful would they have been?

The experimental mirrors had a maximum detection range of 22 miles (on very windy days it was a lot less). I'll be generous and call it 25 miles, which is then added to the 50 miles from the coast to London for a total distance of 75 miles. The Thames Estuary acoustic mirrors probably would have come online in 1936, and so again I'll be generous, and assume that London at least would have a working early warning system from that year.

Taking all this into account, the results can be seen above. And sadly the acoustic mirrors wouldn't have made much difference -- a margin of only about 10 minutes, not much improved on the 5 minutes with no warning system. Of course, even a few minutes' extra warning was worth having, but the Air Ministry was right to terminate development of the acoustic mirror network in order to concentrate on the far more promising radar.

John Ferris has argued against the idea that 'Air defence in Britain began during 1934 and only because radar was developed', and that the importance of the C3I system -- ultimately a legacy of the First World War -- has been underestimated by historians: it was 'ideally preadapted to radar'.7 And he's right. Even without effective early warning, as long as the enemy bombers could be intercepted and shot down on their way back home, air defence could still work by inflicting prohibitive casualties. Except, that is, when the casualties from bombing were predicted to be massive, and then a failure to stop the bomber getting through would have devastating consequences. Radar was part of the antidote to the fear of the knock-out blow. Or rather it could have been, if it hadn't remained secret until 1941 ...

(Just to repeat: the data and assumptions underlying these plots are on the dubious side, and are not fit for any purpose, probably including this one!)


  1. I'm neglecting radar, obviously, but see below. I'm also neglecting the fact that sound detectors, of the type that had been developed during the First World War, had a range of about 5 miles. But see even further below. Distant patrol aircraft were also used as a kind of picket line. 

  2. What I really need are curves showing climbing time vs. height because the higher an aeroplane flies, the harder it is to climb in the thin air. I assume these are available somewhere, but digging them up is too much work for a quick and dirty plot like this! 

  3. Fighters got a lot better at climbing very rapidly by the late 1940s, but as that happens I'm shifting the goalposts ever higher, as it were, and so the above graph is understating the rate of climb of fighters. 

  4. E.g., the two red triangles in the late 1930s are the Hurricane and Spitfire, which between them were the RAF's primary interceptors throughout the war. This plot makes it look like there wasn't anything able to catch raiders in 1940, which was not the case! 

  5. I could model this inefficiency by adding a fixed number of minutes to the climb time of the fighters -- call it the response time -- but I don't know what a reasonable number is and it might vary a fair bit. For instance, in 1918 LADA (London Air Defence Area) had a response time of 2.5 to 5 minutes, according to John Ferris, “Fighter defence before Fighter Command: the rise of strategic air defence in Great Britain, 1917-1934”, Journal of Military History 63 (1999), 853 (JSTOR). But it presumably rose after LADA was dismantled after the war. David Zimmerman, Britain's Shield: Radar and the Defeat of the Luftwaffe (Stroud: Sutton, 2001), 25, seems to suggest that 5 minutes was the time it took in 1933 just to transmit observations to ADGB (Air Defence of Great Britain) HQ, but that's for the big acoustic mirrors which probably required more computation than normal acoustic detectors. So, pending more comprehensive figures, I'll just leave the response time out of it. 

  6. See ibid., chapter 2, for more on the acoustic mirror research of the 1930s. 

  7. Ferris, ibid., 845, 884. 

CC BY-NC-ND 4.0 This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Permissions beyond the scope of this license may be available at https://airminded.org/copyright/.

16 thoughts on “The widening margin

  1. I'm struck first by the fact that jet fighters had a longer climbing time than prop monoplane (or the fastest prop biplanes, but that was a lower field of combat): for quick responses, the propeller seems to have had an advantage for a while.

    Nice charts, for sure.

  2. Interesting as always Brett - you're doing a great job exploring your topic ("the bomber") from all sorts of angles. I have laboured under the impression that a thesis is a very narrow thing!

    Anyway, I'm reading Don Caldwell's JG26 War Diaries at present (plug "fascinating") and it's interesting that, even with a good detection and fighter control system, engaging the enemy fruitfully is always a fraught thing. Key for the fighter defence is getting up early and into the right relative position (eg. up-sun). Climbing into the middle of a raid was a recipe for disaster. So you need three things: good information, good capacity to derive intent from that information, and enough time to do something about it.

    Meanwhile of course the other side is doing everything it can to fox you on the first two points.

  3. Post author

    Jonathan:

    Yes, I noticed that too! The jet in question is the Meteor Mk I -- the fasting climbing aircraft on the graphs above is the Tempest, with a climb rate of 4700 ft/min, more than twice the Meteor's. I tried to double-check the Wikipedia figures ... all I can find in my books is that the wartime Meteors had a 'poor' rate of climb. Later marks of the Meteor did a lot better -- e.g. these numbers are from 1949 (Mk IV/F.4 I think), and in 1951 a modified Meteor took the world record for fastest climb, so there was nothing fundamentally wrong with the design. Climb rates are about excess thrust so I guess it was the powerplant? Maybe they streamlined the later marks better too.

    But whatever the reason, you're right, piston-engined combat aircraft still had some life in them.

    Don:

    Well, most of the things I write about here won't be in the thesis! The plots in this post, for example, though the WWII casualties one probably will be. That's part of the attraction of blogging -- I can write and think about things that don't fit anywhere else. The thesis is indeed fairly narrow, or at least threads a fairly narrow path through a fairly broad topic, and the blog is perhaps a bit misleading in that respect. But probably more interesting :)

    Yes, that's a good point, which emphasises the value of time but also that there are lots of other factors involved. Having enough squadrons available to actually be able to respond to the detection is also handy, for example!

  4. Post author

    Thanks, that's the sort of thing I mean! But to be churlish for a moment it's unfortunate that it's only for WWII-period aircraft.

  5. Chris Williams

    Wind. On Sept 15 1940 sez Derek Robinson, a serious headwind (90mph at altitude) slowed the Luftwaffe right down. It's reasonable to surmise that this could have worked both ways, making the potential margin that much narrower.

    ISTR that there's an article in _Technology and Culture_ from about 20 years ago, which looks at how the 'time to climb' problem spurred the development of the sleeve-valve engine.

  6. Post author

    That reminds me of the 'silent raid' of 19 October 1917, when high winds led to the butchering of a big Zeppelin raid ...

    Is this article by Andrew Nahum the one you're thinking of?

  7. George Shaner

    On the acoustic mirror angle, this work on the French air arm ("The Forgotten Airforce") had several comments on how that service created an operational system based on that technology; the difference as I recall was that the French built mobile detection units. Considering the French dispersion of effort when the war actually came it's hard to reach any conclusions as to whether they got value for the money; one suspects not.

  8. More wind ... rainy (Queen's Birth-) day yesterday and I came across a tape I'd made off the History Channel (missing it's 1st few minutes and title unfortunately). It skimmed lightly over US strategic bombing in WWII and included the comment that the initial high-level B29 raids on Japan were brought down to low level inter-alia o/a strong jet-streams. Allegedly the B29's were making net 50mph or so westerly transits across/to Japan, providing the defense with ample time to ready themselves.

  9. Post author

    George:

    Interesting! Were they acoustic mirrors, as such, or the horn type of detectors? The latter were quite common from WWI on.

    Don:

    Oh yes, the B-29! Quite ironic, as they were technological marvels, with fully-pressurised cabins and the capability of flying at the edge of the stratosphere ... just where the jet stream begins. Oops.

  10. George Shaner

    I'm pretty sure that they were horn-type detectors, but it's been about three years since I read the book and a copy is not close to hand.

  11. Pingback:

  12. Pingback:

Leave a Reply

Your email address will not be published. Required fields are marked *