Codeword Overlord

by Nigel West

  On 12 April the FHW issued a circular based on a recent analysis of Allied air operations, declaring that ‘points of main effort’ had been registered, listing the Pas-de-Calais, Paris, Tours and the Loire estuary. The same document announced the detection of the US Ninth Army in England, and again referred to the likelihood that General Patton would command an invasion army.

  Much of the FHW’s traffic analysis was conducted by the OKW/Chi’s intercept sites staffed by a handful of FunkAufklärung regiments, of which KONA 5, commanded by Baron Maximilian von Oer, compiled an Allied order of battle in April 1944 that purported to identify the call signs of thirty-four divisions, even though eleven of them had been fabricated, with the US Army’s 23rd Signal Company transmitting bogus signals. From March 1944 the great majority of phantom broadcasts were made by the 709 technicians and thirty-two officers of the 3132nd Signal Service Battalion. The epicentre of this effort was in Kent, where the First United States Army Group’s wireless traffic, with one Canadian and one American army, making a strength of up to eighteen divisions, was emulated. Initially the units were genuine, but after the attachment of the notional US XXXVII Corps and the 59th Infantry Division, all the subsequent components were imaginary.

  The extent to which the Germans depended on traffic analysis and accurate direction-finding was demonstrated in Edinburgh on 26 March 1944 when a lone Messerschmitt Bf 109 reconnaissance flight machine-gunned Edinburgh Castle, a transmitting site associated with the British Fourth Army, which had obviously been found by radio triangulation. This entirely notional formation, composed of the VII Corps in Dundee and II Corps at Stirling, supported by the real US XV Corps in Northern Ireland, had begun transmitting on its sham radio network, organised by SHAEF’s signals expert Brigadier Richard Barker, just four days earlier, on 22 March, and was not fully operational until 4 April. The sudden appearance of the Luftwaffe was considered proof positive of the enemy’s reliance on SIGINT. Nevertheless, there was also evidence that opinions on the subject of an Allied objective were varied. For example, on 6 April the Japanese naval attaché transmitted a signal to Tokyo in the recently broken SJA cipher containing the interesting prediction that the Second Front would not occur until June or July, when the Russian mud would have dried out to make a summer offensive practical.3 Yet, on 17 April, the Kriegsmarine’s commander-in-chief Admiral Karl Dönitz issued a general alert:

  A large scale landing in western Europe may be expected at any time. The success or failure of this invasion will be decisive for the issue of the war and for the existence of the German people.

  Then, on 21 April, the FHW published an assessment, although this did not become available to the Allies until after the war. Significantly, it provided evidence that the Germans had fallen for one of FORTITUDE’s major planks, which was the existence of the British Fourth Army, reported to be in Scotland.

  … news of the concentration and readiness of the strong group of forces in Scotland has continued … The English Army Command believed to be in Scottish Command is, according to a credible Abwehr source, the Fourth.

  The ‘credible Abwehr source’ could have been several people, because the creation of this fictitious unit had been the handiwork of several double agents. The narrative had begun with FREAK on 22 March when he signalled the transfer of the British 2nd Corps from Yorkshire to the Gartur House estate, west of Stirling, and a couple of days later reported that he had met a US Army officer, currently on the staff of the US XV Corps, who was attached to General Sir Andrew Thorne’s headquarters. The garrulous American allegedly had also let slip to the Yugoslav that the British 55th Infantry Division was now in Northern Ireland, a genuine deployment that was already known to Berlin because its sighting had been reported by the German legation in Dublin on 20 January in the compromised PANDORA channel.4

  Thus, in two messages, FREAK had created the Fourth Army and located the US XV Corps’ 2nd, 5th and 8th Infantry Divisions in Ulster, all under the command of the general who would have been known to the enemy for the defence of Dunkirk in 1940. FREAK’s Abwehr controller responded on 12 April:

  Your latest wires very satisfactory. Please continue. Congratulations. Wire as often as possible and watch out carefully in the course of the next few weeks. Please state exact number of divisions, etc, belonging to Fourth Army. Under General Thorne, is anything pointing towards intended landings in German Bay, Denmark, or Sweden?

  The proposition was enhanced by GARBO’s deputy, BENEDICT, on 28 March, who told GARBO that he had seen the 52nd (Lowland) Division in Dundee, where he had also spotted soldiers wearing another insignia, described as a shell on a dark background. GARBO remarked, ‘This insignia is completely unknown to me,’ but the message served its purpose and the division was duly assigned to the imaginary Fourth Army by FHW’s order of battle analysts.

  Further confirmation of the Fourth Army’s location and status was provided by GELATINE, who wrote a letter in secret ink to Lisbon, her fifth, on 12 March that duly reappeared in an ISOS message to Berlin on 6 April. In her letter, actually dictated by MI5’s Hugh Astor, Friedle Gärtner described how she had intended to stay over Easter with an American girlfriend in North Berwick, but had been obliged to cancel the visit because of imminent security restrictions to be imposed on the district. Her friend had explained that her English husband worked for General Thorne, and she hoped that her brother, an officer presently posted to Northern Ireland, might be appointed Thorne’s deputy. This rather feminine embroidery was intended to add a verisimilitude to GELATINE’s message, which conformed to her personality.

  Another piece of the intelligence jigsaw was contributed by BRUTUS, who told his controller on 31 March that he was about to travel to Scotland to confer with the Polish Army. He returned to London on 12 April and three days later submitted a typically detailed report on the various units he had encountered. The Fourth Army, he claimed, was headquartered in Edinburgh and was composed of VII Corps in Dundee and II Corps in Stirling, a component of which was the 58th Division, some of whose soldiers were to be seen wearing the Africa Star decoration. This artifice was intended to convey the impression that the 58th Division had been raised in Scotland from battle-hardened regiments that had seen action in the Western Desert campaign in Libya.

  The finishing touches to the Fourth Army were added by COBWEB in Iceland, who reported on 20 April that the local US Army garrison, the (notional) US 55th Infantry Division, was under the command of VII Corps in Scotland.

  ‡

  On 22 April Rommel’s Army Group B circulated a morale-boosting bulletin to the 15th Army in which he emphasised the need to defeat the Allies on the beaches:

  In the short time left before the great offensive starts, we must succeed in bringing all our defences to such a standard that they will hold up against the strongest attacks. Never in history was there a defence of such extent with such an obstacle as the sea. The enemy must be annihilated before he reaches our main battlefield …We must stop him at the water, not only delaying him, but destroying all of his equipment while it is still afloat … The high water line must be the main fighting line.

  A few days later, on 27 April, the Japanese naval attaché reported on a briefing given by Admiral Wilhelm Meisel, the Kriegsmarine’s chief of the operations staff, on 4 April:

  The Allies have completed their preparations and had concentrated their main force in southern England; taking the weather into account, they were more likely to move in May–July than in April, but they might choose April from the wish to coordinate the invasion with the Russian offensive in the south. Although it was so heavily defended, northern France remained the most likely place as it gave the Allies the shortest distance; but there would probably be a pincer movement from the Mediterranean and Biscay and fairly strong commando operations to cut sea communications in Norway, and possibly a landing in Greece to secure air bases. There was, however, no information on which to base a reliable forecast. Asked about the possibility that the Al
lies might desist from invasion and rely on intensified air attack while the [Soviet] invasion tied down large German forces, Meisel replied that they were bound to invade; they would wish to forestall Soviet penetration of central Europe and the British could not withstand a long war of attrition.

  This message was followed by another on 27 April that described a further interview with Meisel held on 13 April in which the German had revised his views and now felt that the invasion would happen soon, perhaps within the next four weeks. Meisel asserted that his change of opinion had been made as a result of the latest information, which had been drawn from four indicators: the concentration of troops and shipping in southern England; the marked increase in Allied daylight bombing raids on French targets; the escalation of espionage in France; and the temporary nature of the tough restrictions recently imposed on the civilians. Allegedly, Meisel also had been impressed by (undefined) items seized from captured enemy agents that had pointed to ‘a direct second front’ to be established in France without any diversionary attacks in Norway or the Mediterranean.

  The suggestion that recent arrests had yielded relevant information indicates that the evidence was either sabotage materiél or, more likely, written notes of SOE’s coded messages personnels (see Chapter 9).

  Finally, there was a third decrypt in the series, dated 27 April and based on a last interview with Meisel conducted ten days earlier. In this text the attaché explained more of his host’s reasoning: the Allied air forces had focused on French communications and airfields. Furthermore, open sources had detailed the imposition of the recently imposed coastal security zone and the ban on international mail. Meisel had also been impressed by a supposed change in public statements made by (unnamed) Allied leaders, the certainty that amphibious landing exercises had been conducted, and the news that war correspondents had been attached to specific units.

  Because the FHW’s weekly assessments of the enemy’s strength were largely based on SIGINT, enhanced by agent sightings but without the benefit of aerial imagery, they varied greatly, influenced by the nature of the Allied forces accumulating in Britain, which were in a constant state of flux. New units were arriving from across the Atlantic; seasoned troops were being redeployed from the Mediterranean; existing formations were amalgamated with others, and the deception operators were inventing imaginary divisions and exaggerating the status and readiness of training battalions. Of course, the FHW had started from a poor foundation, accepting a total of fifty-five divisions in January 1944, when the real figure was thirty-seven. By April the estimate had escalated to between eighty-five and ninety divisions, plus seven airborne divisions, as opposed to the reality of thirty-two divisions and three airborne divisions. In May FHW issued an estimate of seventy-five divisions, plus seven airborne divisions, and by June the true strength had been overestimated by forty-two divisions. In retrospect, Colonel Staubwasser conceded that ‘the whole intelligence service against the western Allies was exceedingly faulty.’ General Max-Josef Premsel, Chief of Staff of the 7th Army, remarked that ‘this overestimation may have played an unfortunate part in the Supreme and High Commander’s expectation of a second landing’.

  From the German perspective, the challenge was fourfold: the time and place of the attack, the strength of the Allied forces and the likelihood of a second wave. On 10 May von Rundstedt’s headquarters and Army Group B received a situation report from OKW:

  OKW foresees the start of the enemy attack in the middle of May. The 18th seems a probable date. But there is of course no certainty. Principal effort in Normandy, then in Brittany. It is thought the enemy will bomb the ground position with large-caliber bombs, concentrating on small areas, and will try to eliminate the coastal armament in the same manner as well by violent naval bombardment … important parachute landing will perhaps take place at nightfall … in Contentin, a particular alertness is demanded against airborne troops.

  On the same day, the OKM summarised the position as described by Hitler’s daily military conference:

  1. The Fuhrer is positively convinced that the enemy will commence the invasion by the end of May. Delay of the invasion is not anticipated unless a strong internal crisis develops in England. The establishment of a bridgehead appears possible without too great a risk, and is expected to be the first objective of the invasion. The especially threatened areas, such as the Contentin Peninsular, will be reinforced.

  2. In England 76 major formations are now ready. Landing transport in the Channel which until now were located in the vicinity of the Isle of Wight–Portsmouth for six or seven divisions; vicinity Yarmouth–Plymouth for eight and one half divisions; vicinity Falmouth for three-quarters of a division. Total transport capacity presently in England is estimated to be sufficient for twenty-seven divisions.

  3. Based on his latest inspection, Generalfeldmarschall Rommel is very confident of the strength of the Army in the Western Sector.

  Five days later, the OKM’s War Diary included a prescient entry: ‘The RSHA/SD in France reports that the Army will be on highest alert starting 20 May. The invasion should occur between 20 May and 10 June,’ and the next day the War Diary recorded:

  An army report discloses that the English fleet, the RAF and the U.S. Air Forces, including airborne troops, have completed invasion preparation. The army units stand ready at deployment points. The source maintains that in addition to the impending operations to be launched in Northern Europe, additional operations from North Africa and Corsica are to be expected.

  The OKW’s prediction that the Allied offensive would begin on 18 May, based on advice from the Kriegsmarine, turned out to be wrong, and so did the OKM’s revised assessment that the next dawn high tide and full moon would not coincide until August. In reality, of course, General Montgomery, alerted to the formidable beach obstacles constructed by the enemy, had decided in favour of a dawn attack on a rising tide, and this combination had dictated 5–8 June.

  Another influence on the OKW may have been GARBO, who at this stage, D-23, reported on a conversation with his (notional) mistress, code-named AMY, who was described as a secretary working in the Cabinet Office. GARBO claimed that she had told him that the invasion would only take place after sufficient American troops had arrived, and after a major Allied air offensive. The implication, of course, was that the Second Front was quite some time away, and would be protected by one or more diversionary landings.

  Apparently reassured that there would be a period of respite, Rommel reported on 27 May that, ‘the formation of the Anglo-American Schwerpunkt [focal point] in southern and south-eastern England is again confirmed by the location of Montgomery’s headquarters south of London’, perhaps an indication that the Abwehr had learned the location of SHEAF’s advance headquarters, at Southwick House, just 5 miles north of Portsmouth. This news, Rommel speculated, made a landing in Normandy and the Contentin Peninsula less likely, and seemed to suggest the Pas-de-Calais.

  On 30 May the OB West War Diary, compiled daily by the distinguished historian Professor Percy Schramm, reflected what von Rundstedt saw as the current situation, considering the thirty-five bridges over the Seine between Paris and Rouen that had been destroyed in what was obviously a calculated campaign of air strikes over the past six weeks.

  Systematic attacks, especially on all communications installations in the area carried out by the enemy air forces, reveal the enemy’s intention to throw into confusion the networks of communications and thereby paralyse troop movements and supply systems into the rear of the zone. The recent successful attacks on the Seine bridges resulted in considerable paralyzing of the traffic across the river between Paris and Rouen. Direct connection between the Channel front north of the Seine and the Seine Bay, and Normandy, was thus interrupted. This may indicate the enemy’s intentions against Normandy (establishing of bridgeheads). Enemy activities against our main defence lines are still comparatively small in spite of an increased number of attacks. The danger of invasion has come nearer but
it is not yet immediately imminent judging from the intensity of the air attacks.

  Another comment on the Allied air raids was offered, on 31 May, by Admiral Theodor Krancke, whose staff had undertaken an analysis of attacks on coastal artillery positioned between Dunkirk and Cherbourg. Damage assessments indicated that none of the Atlantic Wall’s hardened artillery bunkers, manned by the Kriegsmarine, had been seriously affected, which led the admiral to assess that the destruction of the gun batteries would be an essential precondition for any beach assault, and on that basis opined that no invasion was imminent.

  Early on 5 June, while preparing for a scheduled inspection tour of the beach defences with his son Leutnant Hans von Rundstedt, a university archivist in peacetime, the commander-in-chief signed off on his weekly report to the OKW:

  The systematic and noticeable intensification of enemy air attacks indicates a more advanced state of readiness. The probable invasion front remains the sector from the Scheldt to Normandy … and it is not impossible that Brittany might be included … and it is still not clear where the enemy might land within this total area. Concentrated air attacks on the coast defences between Dunkirk and Dieppe, and on the Seine–Oise bridges, in conjunction with the paralysing of supply services, and the southern flank between Rouen and Paris (inclusive) might be indicatory of a main front of a major landing intended by the enemy. However, the cessation of traffic across the Seine would equally affect troop movements required in case of an enemy attack on the western part of the Bay of the Seine, Normandy and the northern cast of Brittany … imminence of invasion is not recognizable.

  As D-Day approached, FHW assessed the US Army strength at ninety-eight divisions, instead of eighty-nine, and counted twenty-two US Army divisions in Great Britain, instead of twenty. As for the British and Canadians, fifty-seven divisions were located, whereas the true figure should have been less than half, a total of twenty-three plus one French and one Polish. An intercepted telegram BJ/508 sent by the Japanese ambassador in Berlin, Baron Oshima Hiroshi, on 28 May, addressed to Tokyo and encrypted in the compromised PURPLE cipher, contained a claim allegedly made to him by Hitler during an interview the previous day that the Allies had ‘about 80 divisions’ in Britain, an estimate provided by FHW, which listed nineteen armoured divisions and eight parachute battalions. This Japanese diplomatic reporting, of which there were 600 messages read during 1944, dated back to November 1943 when the ambassador completed a four-day tour of inspection of the Atlantic defences and submitted a very detailed account of his observations to Tokyo, listing every unit he saw and describing their weapons (see Appendix II). Another, dated 14 December, described a recent briefing given by Ribbentrop six days earlier in which he had mentioned the opening of a Second Front in the west that would be timed to coincide with a big Soviet offensive. Although the exact place of the attack was unknown, most likely it would be in Belgium or the Channel narrows. The foreign minister had then gone on to say that in view of the difficulty in that area, most likely the Allies would begin with a landing in Normandy or Brittany, and wait to assess the progress before launching the main attack.