Dowding system

This wartime illustration shows the Dowding system reporting chains for a single Sector, light grey. It shows ROC reports flowing back through the Sector controls to HQ, but does not show the radar which reported directly to HQ. The communications flows from FCHQ to Group, between groups, and down to Sectors, and then to the defenses.

The Dowding system was the world's first large-scale air defence command and control system,^[1] built by the Royal Air Force just before the start of World War II. It used reports from Chain Home (CH) radar stations and the Royal Observer Corps (ROC) to produce a unified map that was relayed to the defensive interceptor aircraft and anti-aircraft artillery.

The Dowding system was developed after tests demonstrated problems relaying information to the fighters before it was out of date. Air Chief Marshal Hugh Dowding, commander of RAF Fighter Command, solved the problem through the use of hierarchical reporting chains. Information was sent to Fighter Command Headquarters (FCHQ) central "filter room" at Bentley Priory and used to prepare a map of the battle. Details of action in their areas was relayed to the Group and Sector HQs, where operators re-created the map at a larger scale. Commanders could make decisions on how to employ their forces quickly and without clutter. Instructions were relayed to the pilots only from the lowest level, normally co-located at the fighter's operating bases.

The Dowding system is considered key to the success of the RAF against the German air force (Luftwaffe) during the Battle of Britain. The combination of early detection provided by Chain Home and the rapid dissemination of that information acted as a force multiplier, allowing the fighter force to be used at extremely high rates of effectiveness. In the pre-war period, interception rates of 30% to 50% were considered excellent, that meant that over half the sorties sent out would return without having encountered the enemy. During the Battle, average rates were over 75% and continued to improve, and several raids were met with 100% success rates.^[2] Lacking their own direction system, Luftwaffe fighters had little information on the location of their RAF counterparts, and often returned to base having never seen them.

Development

Previous work

Dowding inherited the London Air Defence Area, a defensive system first built by E.B. Ashmore in World War I to protect London.^[3] Ashmore had divided the defensive weapons into three rings around London; the outermost ring held searchlights and anti-aircraft artillery, the middle ring was the operating ground for fighters, and the innermost, in the city itself, contained additional guns.^[4]

In order to communicate the location of the targets to the weapons, Ashmore set up a large plotting table in the Horse Guards in London. Information from spotters was relayed to this central control room within thirty seconds, where wooden blocks were placed on a large map in order to indicate the location and other information. Observers around the map could relay this information to one of 25 regional control rooms, who re-created the portions of the map of interest to them, and relayed that information to the various weapons in their region.^[4]

During the inter-war era, the system had been maintained and expanded, eventually under the direction of the Air Defence of Great Britain (ADGB). Throughout this period various commenters noted the serious problem that the system lacked a form of early detection, which was deemed essential to the flow of information in the system.^[5] Numerous experiments with acoustic mirrors - such as those at Denge - and similar devices were carried out almost continually, but these always proved unsatisfactory, with detection rates often as low as 5 miles (8.0 km). In spite of this, lacking any other solution, in December 1934 the Air Ministry made plans to deploy sound detection devices around London as part of the Thames Estuary plan.^[5]

Fine tuning

Robert Watson-Watt's work on radar set off a revolution in defence. Radar appeared to offer the desired warning time in any weather and day or night. On 27 July 1935 Henry Tizard suggested running experiments to test the practicality of the interception system, based on an estimated 15 minute warning time that radar would provide. This led to a seven-month long series of experiments at Biggin Hill, under the direction of Wing Commander Eustace Grenfell aided by navigation expert Squadron Leader Robert Linton Ragg.^{[N 1] [6]}

In these experiments, Gloster Gauntlet fighters were tasked with intercepting virtual aircraft, and later, civilian airliners. These aircraft were assigned virtual targets, and it was the task of the operators to intercept the attacker before it reached its target. At first the interceptions were calculated using trigonometry and mechanical calculators, but eventually Grenfell had enough of the "confounded machines", and proceeded to arrange a perfect interception by eye.^{[7][N 2]}

Tizard introduced the equal angles method of rapidly estimating an interception point, by imagining the fighters and bombers at the opposite corners on the base of a triangle. The bombers were flying at a particular angle to the base, and the commander simply had to send the fighters on the opposite angle, so the two would meet at the apex of the triangle. With this "Tizzy Angle", interception rates shot up, and by the end of 1936 they were consistently above 90% as long as the altitude was known and did not change. If the bombers changed their altitude, or the fighters arrived below them, repositioning for the attack reduced the success rate to about 60%.^{[8][N 3]}

Radar

Dowding inherited this system in 1936, when ADGB duties were taken over by the RAF, mostly by Fighter Command. By this time Watson-Watt's work on radar was well advanced, and Dowding became a major proponent of rapidly adding it to the system.^[3] To test this, information from the CH research station at Bawdsey was sent into the system in a series of tests starting in April 1937. These tests showed great promise, allowing the controllers to estimate the altitude of the aircraft within 2,000 feet (610 m), which greatly reduced the problems seen earlier.^[8] However, these tests also demonstrated several problems.

CH was also a fairly short range system, so in order to increase warning time the CH systems were sited on the shoreline, looking out to sea. This meant that it provided no location information for friendly fighters on the way to the attack, nor location information for any aircraft over land. To address this later problem, the newly-formed Royal Observer Corps (ROC) was tasked with maintaining information on the aircraft once they were over the British Isles. This led to another serious problem; using different systems, the reports from CH and the ROC were sometimes contradictory or wrong. Also, the volume of information supplied by the now-national network was overwhelming.^[9]

CH also, initially, had no way of identifying friendly from enemy aircraft. This problem had been considered throughout the development of CH and several solutions had been introduced that produced unique shapes for the radar returns of friendly aircraft. However, as the radars looked out to sea, these systems were of no use when the fighters were over land, or even near the coast. The ROC could not discriminate between friendly and enemy fighters either as they were too small to see clearly when flying at high altitudes.

Dowding immediately set about attacking these problems. To address the sheer volume of reports, and the potential for overlap, he instituted the policy of sending all location reports to the new direction centre at Fighter Command headquarters. Here the data was collated and filtered to produce accurate plots and tracks. The data was then sent, on demand, to the group headquarters, and then to the sectors (see below). To fix the problem of locating friendly aircraft, Dowding pressed for a supply of high-frequency direction finding ("huff-duff") sets, which could locate the fighters using their existing radio sets. This led to the introduction of the "pip-squeak" system that was widely available from early 1940, along with the first identification friend or foe (IFF) transponders which were widespread by October 1940.^[10]

By early 1939 the basic reporting system was in place, and from 8 to 11 August 1939, RAF Bomber Command was asked to launch a series of mock attacks on the system starting over France. Reports from No. 11 Group RAF were extremely positive, stating that "RDF information and plotting throughout the Exercise was consistently first-rate, and enabled interceptions to be effected on the Coast", to which Dowding added "daylight raids were normally tracked and intercepted with ease and regularity".^[10]

Description

Filtering

The first Dowding improvement was implemented by adding a "filter room" at FCHQ where all CH reports were sent.^[11] Here, operators checked new reports against previous reports in an effort to match them up. This helped reduce the instance of the same targets being reported in several locations due to slight differences in measured location, or due to their movement over time. For instance, two radar stations might report on the same target a few minutes apart, which would produce different locations; it was the duty of the filter room to realize these were the same target, and update the single set of location information with the new position. The filter room assigned each of these targets a number, and used that ID number to communicate information to group operations rooms.^[9]

Reporting hierarchy

11 Group's Operations Room in the Battle of Britain Bunker. On the plotting table are numerous plots. The sector clock on the wall behind the map has coloured regions matching the colours on the plots. Above the clock is the main tote board showing the status of the various airfields and their squadrons. The map covers the area near Bristol, near the mannequin's rake, to the French coast and inland areas.

Close-up of a marker for friendly fighters. It indicates that 20 aircraft from 92 and 72 Squadrons of Biggin Hill are being sent to 25,000 ft (7,600 m).

The second improvement introduced by Dowding was aimed at reducing the amount of information being sent to the pilots, and speeding its delivery. For this task, Dowding introduced a hierarchy of control and information flow which ensured that only the information the pilots needed was forwarded to them. At the topmost level was FCHQ, located at Bentley Priory. FCHQ maintained an overall view of the entire battlespace, by plotting the tracks from the filter room on a large map overlaid with the Ordnance Survey National Grid.^[3]

During the Battle of Britain the defensive fighters were split among four active groups, 10 through 13. 11 Group, which handled most of the fighting, was located in the Battle of Britain Bunker (RAF Hillingdon) at Uxbridge, not far from FCHQ. 13 Group, covering the Midlands, was located at Hucknall (and later Watnall), 13 Group in a quarry in Blakelaw outside Newcastle, and the late-formed 10 Group in the aptly named Box outside Bath.^[3]

At the Group HQs another plotting board, covering only the area of interest to that Group, re-created the plots from FCHQ via position reports sent to them over telephone links. This allowed the Groups to re-create the map, but on a scale and location more suitable to their area of operations.^[12] Small wooden blocks with tags were placed on the map to represent the location of various formations, indicated by the ID number created in the filter room. The tags indicated the ID number, friend or foe status if known ("H" meant "hostile, F for "friendly"), the estimated number of aircraft, and their altitude. The color of the altitude tags indicated when the report was updated, the colours matching a large sector clock on the wall. This allowed observers to quickly determine whether or not a plot was up to date. As each plot was moved, coloured arrows were left behind, thereby showing the track.^[13] If a target was assigned squadrons for interception, these were indicated with cocktail stick-like tags with the squadron number on them.^[14]

Each Group was split into several Sectors, which handled operations for one or more airfields. This formed the lowest level of the hierarchy, based at the Sector Controls, normally co-located at one of the sector's airfields. For instance, 11 Group was split into sectors A through F running anticlockwise from the south-west to north-east, along with Sector Z to the northwest. A typical sector, Sector C which saw considerable action, was controlled out of Biggin Hill, but also contained two other major airbases at Lympne and West Malling (Maidstone).^[3] Telephone links from Sector to Group allowed the Group's version of the map to be re-created at the Sector, again filtering out information that was not of interest in that area.^[15]

It was the Sector Controls that were responsible for communication with the pilots, as well as balloon barrage and anti-aircraft artillery batteries, providing early warning of the approach of the enemy or warning them not to fire on friendlies.^[3] In order to combat the problem that the fighter pilots tended to ignore orders, Sector Commanders were normally former pilots themselves, either retired or on medical leave. Dowding, Patrick Blackett and Henry Tizard personally drove home the fact that the pilots could not simply hunt for their targets, and had to follow the instructions of the operations centers.^[9]

By creating this system, information flow was primarily in one direction, and continued to be split out on the way down. For instance, the filter room might be receiving 15 reports per minute from various CH sites, but these would be about formations that might cover the entire coast of Great Britain. Portions of these reports, say those over Kent, would be sent to 11 Group, while others, the attack on RAF Driffield for instance, would only be relayed to 13 Group. In turn, 11 Group's operations room would relay only those tracks of interest down to the sectors, filtering out tracks that were out of their range, or being handled by other sectors. Finally, the sector operators only had to relay information to the pilots that actually had an effect on their flight - data about other formations was not relayed, freeing the radio time.

Information sources

The tall towers of the Chain Home system allowed them to detect targets as far as 100 miles away, over France.

A ROC Post Instrument plotter, with Mickelthwait height adjuster attachment. ROC reports began to flow in as the enemy approached the shoreline.

Chain Home offered an enormous improvement in early detection times compared to older systems of visual or acoustic location. It was not uncommon for CH stations on the southeast coast of Kent to detect enemy raids as they were still forming up over their airfields in France. Additionally, this information was available day or night, in any weather. However, in order to provide maximum warning time, the CH stations had to be placed as close to the shore as possible. As they could only locate objects in front of them, this meant that CH provided no service over land. This required two additional locating services.^[3]

The task of tracking enemy aircraft over land fell to the Royal Observer Corps. Because of the large number of ROC stations and the relatively small areas they covered, information duplication and overload was a concern. For this reason, information from the ROC was sent to a parallel system of Observer Centres, who acted as both filter and communications stations. This pre-filtered information was then sent into the Dowding reporting chain.^{[3][N 4]}

Likewise, a separate reporting chain was tasked with tracking friendly aircraft through the use of radio direction finding (RDF) on their radio transmissions, using a system known as "Huff-Duff" (receivers) and "pip-squeak" (transmitter).^[13] Developing a fix using RDF requires two or more observation locations spread apart by some distance and then using triangulation on their reports. In this case three stations were typically used, located about 30 miles (48 km) apart. This information was reported to a selected Sector Control, who used this to plot their locations and pass on that information to the main operations plot. Controllers could then give directions the squadrons. Locations of the fighters were sent up the chain only as required.^[16]

Finally, information on the status of the fighter squadrons was known to the Sectors but needed by the Groups in order to choose which squadrons to commit. For this task, operations rooms also contained a series of blackboards and electrical lamp systems indicating the force strengths of the fighter squadrons and their current status. Known as the "tote board", this allowed commanders to tell at a glance which units were available to receive commands. The statuses were Released (not available); Available (able to be airborne in 20 minutes); Readiness (airborne in 5 minutes); Standby (pilots in cockpit, airborne in 2 minutes); Airborne and moving into position; Enemy sighted; Ordered to land; Landed and refueling/rearming. Next to the tote board was a weather board. It was the responsibility of the Women's Auxiliary Air Force (WAAF) plotters^{[N 5]} to continually update the tote and weather boards, and relay that information up the chain.^{[17][18][N 6]}

Command and control

The Sector Control room at Duxford controlled the area from London to The Wash. Radio and telephone operators sat on the upper level, with plot operators in the white chairs.

The most visible part of the Dowding System were the "operations rooms". The most advanced of these were located at FCHQ and Group HQ's. These rooms typically consisted of three layers; a large plotting table on the lowest level, communications operators located at desks around and above the plot but in the same room, and finally a second story located above the plot and sometimes isolated behind glass.^[13]

Actual command control over the battle was devolved to the Group level.^[11] Observing the map from above, they could track the movements of enemy aircraft through their airspace. When the time was right, they examined the tote board, selected an available squadron, and called their Sector to have them scramble. The orders could be as simple as "Squadron nine-two, intercept hostile two-one".^[15] The Sector would then phone the office for that squadron (often nothing more than a small shack) and ask for the planes to be scrambled.

After forming up the Sector Controller would ask the squadron to set up their pip-squeak system so they could be tracked accurately. Once their location was measured and transferred to the sector plotting table, the controller would begin to relay instructions to the squadron leader. The vectoring was accomplished using the methods developed in the Biggin Hill exercises in 1935, with the Sector Commander comparing the location of the friendly aircraft with the hostile plots being transferred down from Group, and arranging an interception.^[7] As the fighters were faster and often flew to the interception point first, they were instructed to orbit the location, which allowed them time to climb and seek an advantageous position.

Squadrons, bases and enemy formations were all assigned code words aimed at easing communications speed and accuracy. Updates from sector control at Biggin Hill to 72 Sqn would take the general form "Tennis leader this is Tophat control, your customers are now over Maidstone, vector zero-nine-zero, angels two-zero". This indicated that Biggin Hill's controller (Tophat) was asking No. 72 (Tennis) to fly due east (vector zero-nine-zero) at 20,000 feet (angels two-zero) to intercept their target. The sector was also responsible for local gun and balloon defences, as well as ensuring the fighters could be safely returned to an airfield after their sortie.

Physical construction

The entrance to Group 11's operations room, which was completely underground.

FCHQ and Group Control centres, in keeping with their importance in the battle, were located in bomb-proof bunkers away from airfields. Most of these were built just before the war. Communications were ensured by hundreds of miles of dedicated phone lines laid by the GPO and buried deep underground to prevent them from being cut by bombs.

Sector Control centres tended to be relatively small, and were mostly housed in brick, single-storey, tile-roofed structures above ground, where they were vulnerable to attack. By 1940, most were semi-protected by an earth bank or "blast wall" surrounding them which reached as high as the eaves. Fortunately, Luftwaffe Intelligence was unaware of the importance of these rooms and most were left alone.

The control rooms at Biggin Hill were destroyed by a raid on 31 August, but this was due to a chance bomb hit. As a further precaution, emergency control rooms were set up in different locations away from the airfields, with small loss in efficiency; Kenley, for example, could use an alternative room housed in a butcher's shop in nearby Caterham.^[18]

The vulnerability of the earlier rooms was appreciated, and new airfields built during the expansion programme of the 1930s used bombproof Mk. II, L-shaped structures.

Effects on the battle

Freya early warning radar. Early detection is not enough, the information needs to reach the pilots.

The effect of this system was profound, and remains a widely used example of the concept of force multiplication.^[19] Before the introduction of radar the task of interception appeared to be increasingly difficult, if not impossible. As bomber speeds and altitudes grew, even the increased performance of the fighters meant that the available interception time was dropping - certain aspects like getting the pilot into the aircraft and taking off introduced fixed overheads that did not improve along with the aircraft. With the only detection means being the observers and acoustic location with ranges on the order of 20 mi (32 km) under the very best conditions, the bombers would be over their targets before the spotters could call in the raid and the fighters could climb to their altitude.^[2]

The only way to counter this would be to maintain standing patrols - squadrons in the air, at altitude, ready to attack any incoming forces. Given the limited endurance of the fighters, about an hour, in order to keep a reasonable force in the air as many as ten times that number of aircraft would be needed. At any given time only a few would be in the air, the rest would be on the ground being repaired and refueled after their patrols. Likewise, thousands of pilots would be needed to crew the aircraft, as each pilot could only be expected to fly so long per day. Such a system was essentially impossible, and led to the concept that "the bomber will always get through".^[20]

Even with radar, the system wouldn't necessarily offer better efficiency. If, for instance, the radar was used to order fighters into the air with a suitable warning time, they would have to fly around and attempt to find the enemy. This might take so long that the fighters would run out of fuel, or simply spend the entire raid in the wrong area. It was expected that the majority of aircraft would not engage; during WWI interception missions, the vast majority of patrols returned home without ever having seen the enemy. The same was true for all forces during the Battle of France, where 30% interception rates were considered typical, and 50% excellent.^[2]

The Dowding system turned this on its head. By rapidly sending accurate information to the fighters, their ability to find and attack the enemy was dramatically increased, and the time required dramatically decreased. The fighters took off, flew directly to their targets, engaged and returned directly to their bases. In early operations interception rates of 75% were routine, and this number continued to improve as the operators became more familiar with their tasks. By the end of the Battle of Britain, rates of interception over 90% were becoming common, and several raids were met with 100% interception rates.^[2]

Peter Townsend later noted that:

The Germans knew about British radar but never dreamed that what the radar 'saw' was being passed on to the fighter pilot in the air through such a highly elaborate communications system.^[11]

In the Luftwaffe system of the same era, radar was seen primarily as an early warning device. The complexity of relaying the information from the radars to the fighters had not been addressed, and was apparently never considered seriously. Their counterparts in the army and navy had more advanced systems, but there was little inter-service sharing of information. During the Battle of the Heligoland Bight in 1939, over 100 German aircraft were scrambled to meet a small force of RAF bombers, but less than half of these found them. This was in spite of there being a Luftwaffe Freya radar that detected the raid while still one hour out; this unit was unable to communicate with the fighters. The only radar reports to reach them came from a naval Freya near the target, too late to improve matters.^[21]

Perhaps the highest praise for the system comes from one of the men that had to fight against it. After the war, Adolf Galland noted:

From the first the British had an extraordinary advantage, never to be balanced out at any time during the whole war, which was their radar and fighter control network and organization. It was for us a very bitter surprise. We had nothing like it. We could do no other than knock frontally against the outstandingly well-organized and resolute direct defence of the British Isles.^[22]

Problems

Although the Dowding system proved itself in combat, the system and the overall command of the battle presented several problems.

A major issue was the volume of information that flowed through the system, especially into the FCHQ filter room. On 11 January 1940 an operations research report on the Dowding system pointed out that the filter room had originally intended to correlate radar reports, but had developed into something much more complicated. As a result too much control was placed in the filter room, which was producing results with "appallingly low standards". Additionally, it was well recognized that the filter room could be overwhelmed in battle due to the volume of reports. This actually occurred on several occasions and the filter room had to ask certain CH stations to stop reporting, but in practice this problem had little effect on their capabilities.^[23]

Another problem was due to the limitations of the TR.9D HF radio set equipping fighters at the time of the Battle of Britain. This system had two channels, and the operating frequencies of the two could only be selected before the mission. One of the channels was used for voice communications with their Sector Control, while the other was used for the "pip-squeak" system. With only a single usable voice channel, set to a squadron-specific frequency, the TR.9D limited the ability for friendly aircraft to coordinate. The set also had limited power, offering effective range of about 40 miles (64 kilometres) air-to-ground and 5 miles (8.0 kilometres) air-to-air (a function of the antennas used in air and on the ground), which presented numerous problems with reception quality. Making matters worse, the TR.9 had originally operated in a band that was relatively unused, but by the time of the war it had become widely used and interference was a constant problem.^[24]

Chain Home, while providing invaluable information, could only do so for aircraft in "front" of the antennas, typically over water off the shores of the UK. This meant that the system relied on the ROC once the raid was over land. The ROC, using visual means, had limited to no ability to provide information at night, in bad weather, or even through local overcast. This was not a major concern during the Battle of Britain, as the weather was unusually good,^{[25][N 7]} the so-called "fighter's summer", and the raids being attacked took place only when the weather allowed the bombers to see their targets.^[26] If the attacks took place in poor weather or at night, the system would be entirely ineffective. This came to be during The Blitz.

Finally, the system was dramatically affected by inter-group rivalries within the RAF command structure. This was especially true for the critical areas covered by 11 and 12 Groups, whose commanding officers were involved in a constant fight. Trafford Leigh-Mallory, commander of 12 Group, had originally been slated to take over the key 11 Group command before the war, but was instead replaced by Keith Park. Leigh-Mallory felt further slighted by the commander of Fighter Command, Dowding, because his 12 Group was tasked with defending 11 Group's airfields while 11 Group attacked the enemy.^[27]

Dowding had failed to set a command structure starting at FCHQ, which meant that every Group commander was able to control their own battle. Group commanders could request, but not demand, support by other groups. During the battle, Leigh-Mallory repeatedly failed to send his fighters to support 11 Group's airfields, preferring instead to build his "Big Wing" formations and directly attack the enemy. These formations took considerable time to build, and were often still forming as the enemy flew away. In the end they contributed little to the early stages of the battle, and 12 Group's actions represent a lost opportunity.^[27]

After the Battle

Throughout the use of the system, Dowding constantly had to fight off suggestions that the filter room be moved to the Group level. This would reduce demand on FCHQ's filter room, as well as provide redundancy. This issue came up several times, the last time on 27 September 1940, when the Air Council's request was once again dismissed by Dowding. This led to his summons to the Air Staff on 1 October, when he was forced to devolve command, although he delayed this until all fighters were equipped with new IFF systems.^[23]

Only a few weeks later, Dowding was removed from command, perhaps due in part to his failure to come up with solutions to The Blitz. Within months, the Groups had received their own filter rooms. The introduction of ground-controlled interception radars (GCI) in the winter of 1940/41 further devolved command by allowing sector controls to directly detect and control the aircraft in their area.^[23] Working in concert with IFF, now widely available, the GCI display eliminated the need for separate detection and plotting, as well as the communications links and manpower needed to operate the reporting system.

In the post-war era the entire fighter control system was quickly demobilized. However, the explosion of the Soviet atomic bomb in 1949, and the presence of Tu-4 Bull aircraft that could deliver it to the UK, led to the rapid construction of the ROTOR system. ROTOR re-used many of the existing GCI and CH systems, but built dramatically upgraded control rooms in hardened underground bunkers. ROTOR was in turn replaced by the Type 80 radar Master Control rooms, and then the Linesman/Mediator system in the 1960s.

References

Notes

Citations

Bibliography