* In the summer of 1943, the Allies finally used Window against the Reich. The action sent the countermeasures war into high gear, with the Allies generally obtaining and keeping the edge.
* The US RRL organization learned about Window from the British, and put their own experts to work on how to best make use of it. A noted antenna expert, Dr. L.J. Chu, worked out the theory, and then Fred Whipple, an astronomer well-known in his field for his definitive work on comets, figured out how it should best be fabricated and used. Whipple, like many of the boffins pressed into war service, would retain fond memories of his military work and kept a prototype Window-cutting machine in his office for the rest of his career.
Window was being stockpiled in quantity, but there was still no commitment to its use. R.V. Jones kept lobbying for the go-ahead. On 23 June 1943, Churchill called a meeting of Chiefs of Staff to consider the question of Window. Jones made his case; Watson-Watt was still opposed, but Leigh-Mallory, in charge of Fighter Command, agreed that reducing the suffering of Bomber Command outweighed his concerns over the air defense of Britain. Churchill thought the matter over for a moment, and then called on his instinct for theatrical phrases: "Very well. Let us open the Window."
Window went into operation in July 1943, during Operation GOMORRAH, the devastating raids on Hamburg. The Window strips were 30 centimeters long and 1.5 centimeters wide (12 by 0.6 inches) and were manually tossed out by crew members in packets containing 2,000 strips each. 46,000 packets were dropped. Window completely disrupted German air defenses. Night fighter radar operators reported ghostly British bombers approaching them at high speed and then disappearing abruptly, over and over again. Out of 700 bombers in the raid, only twelve were shot down.
Through most of the rest of the war, Window was dispensed by hand, though late in the conflict bombers were fitted with automatic dispensers. Bomber forces not only used Window for concealment, they also used it for decoying the defenses, dumping Window screens at a right angle to the actual direction of the main bomber attack, or even flying in circles around a target and dropping Window bundles to baffle the defenders.
In another one of the many ironies of the Wizard War, the Japanese had actually opened the Window in May, the month before Churchill's decision, when they dropped strips of "giman-shi (deceiving paper)" to jam SCR-268 radars during a raid on Guadalcanal. The Japanese had discovered the same obvious trick, but the news failed to reach the Allied high command before they decided to let the cat out of the bag. To add to the irony, the Japanese never managed to obtain a real advantage from their giman-shi. One problem appears to have been the scarcity of aluminum, which meant they couldn't produce enough giman-shi to make a real difference.BACK_TO_TOP
* For a time after the introduction of Window, RAF Bomber Command conducted night raids over Germany with relative impunity. Field Marshal Erhard Milch, in charge of German aircraft procurement, commented: "I am beginning to think that we are sitting on a limb, and the British are sawing that limb off."
Hermann Goering was disgusted. "In the field of radar they must have the world's greatest genius. They have the geniuses and we have the nincompoops ... The British would have never dared use the metal foil here if they had not worked out 100% what the antidote it. I hate the rogues like the plague, but in one respect I'm obliged to take off my cap to them. After the war's over, I'm going to buy myself a British radio set, as a token of my regard for their high-frequency work."
Goering's comments about Germany's radar "nincompoops" were revealing, since it was something of a self-fulfilling judgement. Allied leadership clearly gave greater priority and respect to their technical resources than their German counterparts, who seemed to regard their own boffins as a resource that would provide miracles spontaneously, to be browbeaten when they failed. In fact, many German radars were ingenious, well-designed, and highly capable given the limitations of longwave systems. In one of the many illuminating ironies of the story, most German microwave work had been canceled by the authorities just before the capture of the Rotterdam Geraet earlier in the year.
Goering was also dead wrong about the British having countermeasures against Window. The only actual countermeasure they had was the fact that they and the Americans had large numbers of heavy bombers and the Germans did not, making the use of Window a good gamble.
The Germans would in fact use Dueppel themselves, beginning in October 1943 and particularly during the "Baby Blitz" on England in early 1944, but Luftwaffe bomber formations were too small to achieve the densities of Dueppel needed to be effective. The British decision to open the Window proved completely justified by later events.
On reading a report about the Rotterdam Geraet, Goering later commented: "We must admit that in this sphere, the British and Americans are far ahead of us. I expected them to be advanced, but frankly I never expected them to get so far ahead. I did hope that even if we were behind, we could at least be in the same race!" Kammhuber, who had antagonized his superiors with his single-minded focus on his own agenda as the ultimate priority of all activities of the Reich, would be fired in mid-September after several failures to adequately deal with RAF raids, with General Josef "Beppo" Schmid taking his place as commander of night defense operations.
* By late spring 1943, the RAF and USAAF were dumping hundreds of tonnes of Window a month, but the Germans were beginning to recover. They would never defeat it completely, but they were gradually able to take back some ground they had lost. In fact, RAF night bomber losses continued at a high rate, which leads to the ugly question of just how bad they would have been if the Allies hadn't developed countermeasures.
The German defense was particularly assisted by a stroke of luck. By coincidence, the improved Lichtenstein SN-2 radar began to go into service in July 1943, at about the same time that the Allies started using Window. SN-2 turned out to be the right thing at the right time, since at the time the Allies were using Window cut to lengths appropriate to shorter wavelengths, and the SN-2 could cut through the interference. Although the British were jamming Freyas and other early-warning radars, ground controllers could still use the unjammable Y-Dienst direction-finding system to direct night fighters into bomber streams, where they could use SN-2 to hunt down targets. The weak link in the Zahme Sau scheme was the ground controller communications channel, and so the Allies stepped up their efforts to jam or spoof the communications.
In the late stages of the war, the Germans would also deploy an improved longwave AI named "Neptun", derived from a simple tail warning radar of the same name, and which apparently was originally designed as an ASV. The Neptun AI operated over a relatively wide range of frequencies and had a maximum range of up to five kilometers (three miles) and a minimum range of a hundred meters (330 feet), but it was too little and too late.
The Luftwaffe also used another tactic to complement Zahme Sau, known as "Wilde Sau (Wild Sow)", in which Luftwaffe day fighters conducted night attacks, with the cities below left lit up so the fighters could spot Allied bombers. It seems the first RAF crews to find their targets were not blacked out immediately suspected something was dangerously not right. Wilde Sau proved effective, but day fighters, not having been designed for night operations, suffered from landing accidents and a simple, dangerous tendency to get lost at night.
Some of the German countermeasures were clever. The Germans built a system named "Flammen (Flames)", that could home in on IFF Mark III signals, a trick that was helped along by the fact that some RAF crews superstitiously thought that their IFF could jam Wuerzburg and so left it on all the time. RAF losses climbed until December, until the Allies got wise to the trick and ordered pilots to turn off their IFF over hostile territory. In turn, the British would hunt Luftwaffe bombers using German IFF during the Baby Blitz. Similarly, when the RAF deployed a tail warning radar named "Monica" on their bombers in June 1943 as a means of warning the pilot that an attacker was on his tail, the Germans quickly invented a device named "Flensburg" to home in on Monica emissions. That was particularly ironic, since Monica had been so prone to false alarms due to other bombers in the stream that it was of little use in the first place.
* While the Germans tried to deal with Allied countermeasures, they were also working on centimetric radar, using the Rotterdam-Geraet they had captured early in 1943. They had quickly determined that the device operated at centimetric wavelengths, and Professor Leo Brandt of Telefunken was assigned to reverse-engineer the device. The Telefunken factory in Berlin was bombed on 1 March 1943 and the device was destroyed, but that same night a Halifax bomber was shot down over the Netherlands, providing a replacement. This time, the device was taken to a flak tower, one of the huge reinforced concrete "castles" used to house anti-aircraft batteries.
As it turned out, the Germans did not have time to deploy centimetric radar themselves. They were so desperate for components that the wreckage of Allied bombers was scavenged for magnetron parts. 25 "Berlin" 10 cm (3 GHz) AI sets were built late in the war, but only a few were ever fitted to night fighters and they saw little or no action. The improved "Bremen" variant never got beyond a single prototype. A microwave ground-based search radar named "Marbach" was developed and saw some use near the end of the war. Marbach had a peak power of 20 kW, a pulse period of 0.6 microseconds, a PRF of 500 hertz, and a maximum range of about 50 kilometers (31 miles). A targeting radar named "Kulmbach", with similar specifications but a tighter beam and half the range, was also built. The two radars were linked to form the "Egerland" fire-control system, but only two Egerlands were built before Germany's surrender.
However, the Germans were able to develop effective countermeasures against centimetric radar, and in fact the Germans were so focused on countermeasures that radar development was necessarily a lower priority. Telefunken built a simple detector named "Naxos" that could pick up 10 cm (3 GHz) H2S transmissions, and a more sophisticated detector named "Korfu" with greater range and accuracy. Korfu saw little use, but Naxos saw widespread service.
There were two different types of Naxos. "Naxos Z" was developed for night fighters and mounted in a blister on top of the fighter's canopy. It could detect an RAF bomber from much longer range than Flensburg Monica-homing system. Another version of Naxos, "Naxos U", was provided to U-boats to allow them to detect 10 cm (3 GHz) ASV, though by that time the U-boats were entirely on the defensive and it did them only a little good. Naxos was further hobbled by the fact that it proved very fragile in field conditions, and working out the bugs ended up being troublesome.
Although the kind of technical improvisation by the bottom ranks that in particular characterized the US military was not encouraged by the strictly hierarchical German military, it did happen. A captain and a sergeant in the air-defense system came up with the bright idea of hooking up a Naxos to a Wuerzburg dish system, resulting in "Naxburg". Naxburg had a directional accuracy of about 1 degree and range limited only by line of sight. It became an important addition to the Y-Dienst signals intelligence network.
* The British continued to develop countermeasures of their own. RAF bombers were equipped with primitive "passive" radar-warning receivers to warn them they were being hunted by night fighters. The first was "Boozer", introduced in 1943, which used a tail-mounted antenna to pick up Lichtenstein transmissions and turn on a warning light to alert a bomber pilot. Boozer did not generate emissions to give away the bomber, but it was also not very discriminating and gave continuous false alarms. Bomber crews soon learned to turn it off and ignore it.
The British also fought back directly. The RAF equipped Mosquito night intruders with a device called "Serrate" to allow them to track down German night fighters from their Lichtenstein B/C and SN-2 radar emissions, and also fitted Mosquitoes with a device named "Perfectos" that tracked German IFF. The secrets of Lichtenstein SN-2 and German IFF had been dropped into the hands of the Allies in July 1944, when the pilot of a Ju-88 night fighter flew the wrong way against a landing beacon and landed in the UK by accident.
Countermeasures led to more countermeasures. The Germans built a tail-warning version of Naxos, known as "Naxos ZR", to warn their night fighters that they were being tracked by centimetric radar. The Luftwaffe went to great lengths to hunt down the hated Mosquitoes, though with limited success. With the Reich crumbling, the Germans were increasingly unable to even find fuel to keep their fighters flying.
* As the Luftwaffe began to run out of steam, the Germans relied more and more heavily on anti-aircraft artillery to defend the Reich. This led to an Allied emphasis on jamming Wuerzburg and Mannheim gun-laying radars, and a German emphasis on developing counter-countermeasures. The contest went on to the end of the war.
The Germans managed to overcome Carpet jamming by providing Wuerzburg and Mannheim with a second band in the fall of 1943. The new band was centered around 58 cm (520 MHz), in contrast to the original band of 54 cm (560 MHz). A year later, they added a third band, around 66 cm (455 MHz). An adapter named "Wismar" was introduced in the summer of 1944 to allow rapid switching between bands. Of course, as mentioned, the Allies improved the jammers to cover the new bands. Eventually, SIGINT receivers were carried on some bombers to determine enemy radar operating frequencies so the jamming could be focused on those bands for maximum effectiveness. Electronic warfare specialists, or "Ravens", became an essential member of the bomber force crew.
The Germans tried more sophisticated counter-countermeasures as well. They devised an enhancement to Wuerzburg called "Wuerzlaus" that was introduced in the fall of 1943. Wuerzlaus could perform a limited amount of discrimination of targets on the basis of their motion. Moving objects caused a frequency shift, or "Doppler shift", in the radar waves reflected off them, and this frequency shift could be measured to sort out drifting Window clouds from the bombers that dumped them. The "laus" suffix was derived from the German word for "louse", and so the name basically meant "Wuerzburg delouser". Wuerzlaus was an early attempt at what would become "Doppler radar" after the war.
The Germans introduced two other enhancements at the same time, named "Nuernburg" and "Taunus". Nuernburg tried to sort out radar reflections that had an audio-frequency component, due to the engine vibrations of the target. Taunus was similar to one of the counter-countermeasures schemes developed for Chain Home, a filtering scheme that emphasized persistent targets (targets) and deemphasized transient ones (jamming). Yet another counter-countermeasures scheme, "Stendal", tried to zero in on the jammer itself to target the carrier aircraft. Stendal turned out to be too inaccurate to be useful.
* The Germans tried to improve on and combine these techniques, developing a scheme named "K-laus" near the end of the war that combined Doppler measurements with filtering, taking a bigger step towards Doppler radar. However, in general, the German counter-countermeasures were not very effective. In those days, radar operation was an art form, since radars effectively returned unprocessed data that a skilled operator had to interpret, picking out signals from the noise. Jamming only made matters worse. The German counter-countermeasures required very skilled operators to make good use of them, but the skill level of their operators was never very high on the average, and was degrading as more and more manpower was drained away to the front to hold back defeat.
Even when their radar was blinded, the anti-aircraft batteries generally kept up a heavy rate of fire, focused on the best guess for where the intruding formation was, and actually scored enough kills to make Allied countermeasures officers wonder if the countermeasures were actually working. Such "predicted barrages" also helped reassure the local population by at least giving the sound of putting up a strong defense. However, predicted barrages wasted enormous amounts of ammunition. During a raid through overcast by 720 USAAF bombers on Hamburg on 25 October 1944, the Germans fired over 24,400 rounds of heavy anti-aircraft ammunition, and shot down one bomber. Shortly after that, faced with using up ammunition faster than it could be manufactured, the practice was restricted to defense of high-priority installations. Bomber losses to flak then fell by 75%.BACK_TO_TOP
* The countermeasures war reached its highest form during the Allied invasion of Normandy on 6 June 1944. R.V. Jones, working with a colleague in photo-reconnaissance named Claude Wavell, compiled a map of German radar assets to help pave the way for the assault. Some of the stations were located by photo or electronic reconnaissance, as well as reports from resistance groups, but subtler measures were used as well. The TRE set up a network of direction finding stations in England, codenamed "Ping Pong", that could each pin down the location of a radar transmitter to within a quarter of a degree, with triangulation giving the exact location. In a particularly devious trick, the RAF flew solitary reconnaissance missions on precisely-defined tracks, with the German reports on the missions intercepted and decrypted, revealing locations of radar stations that had tracked the aircraft.
Three weeks before the invasion, Allied bombers and strike aircraft began to attack critical radar stations. Rocket-firing RAF Hawker Typhoon strike fighters proved most effective, but encountered heavy anti-aircraft defenses. The TRE fitted a few Typhoons with a device named "Abdullah" that could home in on radar sites; it was the ancestor of the modern "radar homing and warning (RHAW)" receiver used on "Wild Weasel" type defense suppression aircraft. Abdullah worked fine, but it proved to have a serious drawback. German radar sites that observed Allied aircraft flying straight down the boresight at them immediately put their flak defenses on full alert, and the effect of Abdullah was to simply make the attacks harder. It was set aside, and attack plans were modified so that the Typhoons flew an oblique course toward a radar site and only turned directly on it at the last moment.
Squadrons of countermeasures aircraft screened the airborne assault force and the naval force. The countermeasures aircraft carried Mandrel radar jammers and Airborne Grocer radio jammers, and dropped "Rope", essentially the same thing as Window but cut to 1.7 meter (5 feet 6 inch) lengths to jam Freya. Window and Rope were collectively known as "chaff". Incidentally, Rope led the Germans to introduce a Doppler detection system for Freya, of course known as "Freya-laus", which was also used on Mammut. A similar device, "Wasserfloh", was used on Wassermann.
The invasion fleet also carried 800 jammers, with some landing ships fitted out as dedicated jamming platforms. Of course, the landing force also heavily relied on radar, with radar beacons set up by pioneer teams on landing beaches and passive radar "corner reflectors" or "Angels" set up to mark obstacles. Since there were fears that the Germans might try to disrupt the longwave Eureka beacons with airborne Kettenhund jammers, a handful of British Mosquito night fighters were fitted with a TRE gadget named "Lucero" that could home in on Kettenhund transmissions.
* Along with direct countermeasures, the Allies conducted a massive deception campaign with many facets. Signals deception efforts were used to create simulated armies that seemed to be poised for landings in the Calais area and in Norway. Another part was the construction of two electronically simulated landing fleets by a team of TRE boffins under Robert Cockburn. Developing the scheme involved clearing Cockburn for briefing on the invasion plan, and after he was briefed in detail in February 1944, he could hardly sleep for days after being entrusted with such a dangerous secret.
The goal of the effort was to spoof almost 100 German radar stations along the French and Belgian coasts. Seetakt was the primary target, but the plan was designed to fool other radars as well. Rope was to be used to simulate vessels, with bundles dropped at low altitude to form a ship-sized cloud that disappeared into the water before it spread, to be replaced by another cloud of Rope.
Two simulated invasion fleets set out on the evening of 5 June 1944. One, codenamed TAXABLE, consisted of eight Lancaster bombers and moved in the direction of Le Havre. The second, codenamed GLIMMER, consisted of six bombers and moved towards Bologne, 300 kilometers (185 miles) farther east.
The bombers simulated a surface fleet by flying in a racetrack pattern about 22 kilometers (14 miles) long at a speed of 290 KPH (180 MPH). The crew members dispensed Rope on a precisely timed schedule to ensure that the Rope clouds advanced at a rate consistent with the motion of a surface fleet. Aircraft with Mandrel jammers accompanied the two "fleets", but operated at low power to allow German radars to penetrate their "jamming".
The whole scheme required detailed planning and coordination, and had been rehearsed against British radar operators in Scotland in May. They reported it effective, but just to make sure, another test was performed against a radar station on the Yorkshire coast where the operators hadn't been briefed beforehand. They reported the biggest convoy they had ever spotted.
Following the rehearsals, Cockburn got hold of 18 launches and worked them into the deception plan. The launches were equipped with an enhanced version of the Moonshine pulse repeater tuned to the German Hohentwiel ASV band, and also towed floats known as "Filberts" that were in turn tethered to 9 meter (30 foot) long barrage balloons. The Filberts carried 3 meter (10 foot) diameter Angels to simulate a large vessel.
British signal operators in the mock fleet began to pick up contacts with radars of German air patrols about midnight. When the deception fleets got to about 16 kilometers (10 miles) offshore from their targets, they generated smokescreens and played the sounds of a large fleet in operation over big loudspeakers.
Elsewhere, RAF Stirling and Halifax bombers performed a fake airborne assault, releasing Rope, as well as little dolls on parachutes that looked like real paratroopers from a distance, fitted with firecrackers to suggest small-arms fire. A few British Special Air Service commandos went in with the dolls to create further confusion.
In the meantime, RAF Lancaster and Flying Fortress bombers blocked the real airdrops by dropping a wall of Rope and blasting out communications jamming. German night-fighters were dispatched to attack the fake airdrop, but couldn't even find it when the communications jamming cut off their connection to their ground controllers. One RAF bomber involved in the deception exercise was shot down, the crew successfully bailing out. None of the transports involved in the true airdrop were attacked by night fighters.
Cockburn's deception effort was successful as well, though it proved to be overkill. TAXABLE was not observed, largely because most of the German radar stations it was intended to fool had been knocked out by air strikes. GLIMMER, on the other hand, seemed to have been very successful, sowing confusion among the Germans. R.V. Jones does not seem to have been directly involved in these two exercises. Given his inclination towards pranks, he likely wished he had been.
* Following the German surrender, the Allies interrogated German radar systems operators to determine the effectiveness of countermeasures. To get hands-on data, in late June and early July 1945, the British performed Operation POST MORTEM, in which the air defense network dealt with flights of bombers simulating attacks. The British learned that Allied countermeasures had been highly effective, though not entirely perfect.
After the operation, some of the German gear was packed off to Britain and the US for analysis. Most of the rest was demolished, though some was quietly spirited away by organizations in the countries of what had been Occupied Europe for their own analysis, and items such as Wuerzburg-Riese antennas were used for purposes such as radio astronomy.BACK_TO_TOP
* While the Americans closed in on Japan, the Japanese tried to catch up with radar technology, but it was even more of a case of too little, too late.
The IJN did develop a longwave airborne search radar, the 2 meter (150 MHz) "IJN Mark VI Model 2", with some similarities to the British ASV.II. It appeared in service in early 1944 and was mounted on Mitsubishi G4M twin-engine bombers, with a large Yagi antenna in the nose and horizontal dipoles on the fuselage. About 2,000 sets were built. Of course, the IJA had to develop a comparable set in parallel, designated the "Taki 1", that operated on the same band. About a thousand of them were built. These sets did prove useful in action, but the Allies were striding far ahead in radar technology.
In the spring of 1944, Boeing B-29 Superfortress heavy bombers were beginning raids on the Japanese home islands from bases in China. Instead of using dedicated ferret machines, some of the B-29s were fitted with SIGINT gear and jammers to perform countermeasures tasks as part of their normal bombing mission.
The Ravens on board the B-29s picked up Japanese surface radars. While there was no evidence that Japanese night fighters were carrying AI radars, there were some cases where they seemed to lose the scent on a B-29 when the bomber turned off its IFF. That hinted strongly that the Japanese had a device that could activate Mark III IFF, possibly something they had obtained from the Germans, but in the end the incidents turned out to be coincidences. They had no such capability.
The China raids were given up in early 1945. They were logistically difficult, since all the supplies, fuel, and bombs for the operations had to be airlifted over the Himalayas. The capture of the Marianas that spring provided a much more convenient and secure base for Superfortress attacks on Japan. Superfortresses pounding the Japanese home islands all carried their own jammers. For additional protection, Superfortresses dedicated to the jamming mission, known as "Porcupines" for their collection of antennas, or "Guardian Angels" for obvious reasons, accompanied the formations. Window and Rope were also used intensively.
Interrogation of Japanese radar operators after the war showed that the effects of Window and Rope were mixed, but the active jamming systems were brutally effective, completely blinding Japanese radars. Ironically, since the Allies knew the IJA Tachi 3 radar was a variant of the British GL Mark II, the Americans tried to refine their jamming techniques by using GL.IIs obtained from the British as test systems. However, the information provided from this exercise wasn't available until the war was all but over.
* The thorough defeat of Japanese electronics technology was an embarrassment for the Japanese, who had shown flashes of brilliance that had been squandered.
Although the Japanese did eventually develop IFF, they never got it into mass production, and the IJN and the IJA naturally didn't cooperate on the matter, with both services pursuing separate IFF schemes. As their development of the cavity magnetron and early use of chaff showed, the Japanese had people with good ideas, but radar work remained organizationally muddled and crippled by inadequate resources. The shortage of good-quality electronic components was so severe that the IJA and IJN, despite all their antagonisms, managed to form a joint committee to try to work out their supply problems. Many radar engineers were drafted into the military to fight in combat, throwing valuable technical talent into the meat grinder.
Such improved technologies as were built amounted to nothing. For a striking example, with Germany and Japan and both at war with the US and Britain, the Germans sent a submarine to Japan with the design of the Metox longwave radar warning receiver, almost precisely at the time the Allies were moving up to centimetric sets. The Japanese built about 2,000 copies and found them more or less useless.
The IJN was working on AI radars for night fighters, but never got them into service, and attempts by the IJN to adapt the centimetric 10 cm / 3 GHz Mark 2 Model 2 as an air-defense radar met the same fate. Since the IJA Tachi 6 series warning radars had little height-finding capability, the IJA did try to develop height-finding radars, including the "IJA Tachi 20" and "IJA Tachi 35", but only a few ever saw action.
The Japanese developed jammer systems that never got into combat, and work was squandered on a futile attempt to build a radio-energy death ray. The Germans provided the Japanese with Wuerzburg radar technology by submarine. The Japanese were still trying to get Wuerzburg into production at the end of the war. Of course, the IJA and IJN had separate Wuerzburg development programs.
The Japanese last-stand defense was brave, determined, and doomed. In the electronic field, they were so far behind that even if they had managed to field their improved systems, the Americans would have quickly neutralized them.
American radar helped deal out the final blows. Both atomic bombs dropped on Japan at the end of the war were triggered to airburst at 580 meters (1,900 feet) by a proximity fuze system, though it was derived from the AN/APS-13 tail-warning radar. The fuze system included four redundant units, known as "Archies", to reduce the possibility of a premature detonation, with the bomb detonating only when two of the Archies agreed that they had dropped through the critical altitude. There was a backup mechanical fuze system in case the electronic system failed, but the Archies worked as advertised.BACK_TO_TOP