The excerpts are from pp. 20-31 of Project Checo Report (18), Air War in the DMZ, September 1967 - June 1968.  These excerpts focus on the efforts to set up an anti-infiltration barrier near the DMZ between North and South Vietnam, and extending west across the main infiltration routes in Laos.

of fence. This maximum emphasis on DYE MARKER ended abruptly in late January with the siege of Khe Sanh and the Tet Offensive. Only scattered parts of the barrier were completed. _71/

While the Marines worked on DYE MARKER, the Air Force made preparations for initial operations in DUMP TRUCK. This operation, whose eastern area is shown in Figure 8, was the weakest link in the chain of three subsystems. Detecting troops moving along narrow jungle trails was far more difficult than finding trucks on Laotian roads, or monitoring enemy movements in the open lowlands of eastern Quang Tri. As conceived in the fall of 1967, DUMP TRUCK had three main components:         (1) seismic, acoustic, and magnetic sensors to locate troops and air emplaced button-bomblets designed to explode underfoot and activate acoustic sensors; (2) EC-121s to relay the sensor signals; and (3) the Infiltration Surveillance Center (ISC) at Nakhon Phanom to analyze the data and identify targets. _72/  Mine fields would be laid to block or channel infiltration. The planners had reservations about the antipersonnel subsystem having much effect on the enemy in view of the inherent difficulty of visually spotting and attacking men under tropical canopy. The operations orders recognized that:  _73/

"Slowing or stopping troops walking through the country by air alone is a more difficult and complex problem. The aim of the antipersonnel operations will be to reduce and slow infiltration, inflict casualties, force infiltrators into more difficult terrain, and demoralize porters and troops."

Lt. Gen. Alfred Starbird, in charge of the overall DOD planning for the anti-infiltration system, noted to the 7AF commander that tests in Florida and Panama revealed that FACs given the location of sensor-located infiltrating troops could not spot them to initiate strikes "Rather, strikes had to be made on the basis of sensor signal       alone if any success was to be achieved. A similar situation is expected to apply in DUMP TRUCK". _74/  Significantly, even in January 1968, 7AF was talking not about implementing DUMP TRUCK but about beginning the operational test, suggesting again the lack of optimism over the antipersonnel system. _75/^­

The establishment of DUMP TRUCK entailed complex technical and political coordinations. Ground surveillance teams were planned for DUMP TRUCK to observe trail activity and plant sensors in both Laos and South Vietnam . Hand emplacing sensors provided known sensor locations (to the extent of map accuracy). These "Spike Teams" of Americans and Vietnamese were infiltrated and extracted by helicopter with FACs present at both operations. Coordination among 7AF, MACV, and the U.S. embassies at Bangkok and Vientiane were lengthy and complex. _76/

Unlike the ROES for Spike Teams, the 7AF/III MAF overlap of authority in northern Quang Tri has not been formally resolved to date, though actually 7AF has not implemented DUMP TRUCK. A glance at Figure 8 shows a gap between the strong point barrier and the eastern boundary of DUMP TRUCK. Although strong point Alpha 5 was never built, the original barrier was to run to the foothills near Dong Ha Mountain. West of there, III MAF and MACV talked of "defiles" through the mountains and the need for a portion of DYE MARKER to be tailored to the area. A December 1967 plan on DYE MARKER published by MACV defined the III MAF Defile System as running from Route 9 north to the DMZ and west from Alpha 5 to the Laotian border. Battalion base camps along Route 9 would anchor ground operations to the north: sensors would greatly aid in locating infiltration routes. Additionally, at the end of 1967, the division of authority between III MAF and 7AF remained decidedly unclear in the overlapping area of the Defile System and DUMP TRUCK. _77/

In late December, 7AF requested of III MAF "that 7AF be cleared to conduct unrestricted air operations under control of the ABCCC" in the overlap area except around Khe Sanh. The ABCCC would "limit artillery fires through the fire support center for minimum time periods when orbit, strike, mission aircraft or ground teams would be jeopardized by artillery". Also, the ABCCC would request artillery fire on sensor-located targets when "artillery is considered most advantageous...." Marine control of any areas within the South Vietnamese portion of DUMP TRUCK would be reestablished by giving 7AF 48 hours routine notice or one hour emergency notice. _78/

To this request, III MAF said, "No," The Marines were directly responsible for "the anti-main force war" along the DMZ and Laos and could not surrender control of the area in question, Further, the present coordination procedures between the ABCCC and the FSCC at Dong Ha worked and would continue to work well._79/

Seventh Air Force thought these were "complex coordination procedures which would impair operations of both commanders". Consequently, 7AF outlined to MACV a proposal to implement the Laotian part of DUMP TRUCK on 20 January and to maintain a capability to extend it into South Vietnam if MACV desired that. _80/. This reflected 7AF's intention to concentrate DUMP TRUCK on infiltration from the west end of the DMZ down the Santa Fe Trail . COMUSHACV quickly responded by giving 7AF "primary control ... for the purposes of controlling and managing DUMP TRUCK activities" in the extreme northwest corner of Quang Tri where the Marines hardly, if ever, entered. (Fig. 8.) This in-country portion equalled perhaps one-sixth of the overlap area in question. _81/  

DUMP TRUCK and Khe Sanh

In assigning this small corner to 7AF, COMUSIIACV stressed the possible need for sensors around Khe Sanh to monitor an increasing enemy threat. On 19 January 1968 , he ordered the diversion of DUMP TRUCK assets to the trails and roads approaching the combat base. On 20 January, the siege began with the first mortar and rocket attack. On 21 and 22 January, a total of 104 sensors went in, of which 75 worked.  On 25 January, Khe Sanh received the first message or Spotlight Report on sensor-located targets. _82/

In these few days, the IGLOO WHITE sensors became operational as a battle­field surveillance system, rather than an infiltration monitoring system. It remained so for many months until the threat around Khe Sanh lessened. Figure 9 shows just how densely the sensors clustered around Khe Sanh rather than in Laos . The eventual success of the sensors at the combat base attracted wide interest in the use of sensors and in April and May 1968 brought the birth of several projects, such as seeding enemy Base Area 101 and monitoring truck traffic in the A Shau Valley. _83/  Seventh Air Force foresaw this possible dis­sipation of resources into many small projects and took precautions to conserve the sensor assets. _84/  In September, 7AF, suspended indefinitely the testing of the antipersonnel system and allocated all assets to the antivehicle role for the upcoming COMMANDO HUNT. _85/  Thus, the antipersonnel subsystem, like the strong point obstacle subsystem to the east, was postponed under the impact of Tet/Khe Sanh and never fully implemented when those crises passed.

Much knowledge came out of Khe  Sanh, however, about the use of sensors in an antipersonnel role. Most importantly, the air-delivered sensors proved their value beyond any doubt, and eventually, the Marines stopped attacking the single sensor-generated targets. There were, of course, several problems as individual sensors did not provide good specific target accuracy. Although the Marines received target locations to a grid position of plus or minus five meters, the targeting officer at Khe Sanh learned the sensor locations were known within only 200 meters on the average. Accordingly, the Marines could not be certain about the specific accuracy of Spotlight Reports. Other weaknesses were false alarms from artillery and aircraft, delays in getting the sensor reports to Khe Sanh, and the inaccurate positioning of sensor strings. In the latter case, because heavy foliage hid the exact trail locations, and because many areas had a web of interconnecting trails, many of the strings were placed perpendicular to the suspected trails to cut across the foot traffic at some point. However, by not being parallel to the trail, the sensor-strings could not provide good data on speed or direction of personnel movement. _86/

While of limited value for individual targets, the air-delivered sensors provided spectacular assistance against  target complexes and massed troops. To attack, the enemy had to concentrate forces and supplies and, in doing so, he set off heavy area sensor activity, these massed concentrations made inviting targets to air and artillery. The Khe Sanh target officer described an example: _87/ 

 ". . . We shifted   from firing at point targets to firing at area targets, and we used the sensors to tell us when a major troop movement was occurring or a supply mission was taking place. The sensors now did not tell us where to shoot, but when lucrative targets would most likely be in known locations.  

"An example of how the sensors gave us our only warning was in the attack on hill 881 south. To the southwest of hill 881 south, there is a ridge line running from southeast to northwest with a trail along the top. The nearest approach of the trail to the hill top is slightly over 1 km. Sensor strings had been placed all along the trail, and to the northeast of the trail along the ridge from his base areas in Laos into Khe Sanh. A sensor string up the trail about 3 km from hill 881 was usually our first indication of traffic.  

One night no activity was detected anywhere along the trail even though for the previous two nights there had been heavy activity from all the sensors in the area. On this particular night around midnight, all the sensors on the northwest side of the trail began to activate, not just once, but many times until by adding all the troop estimates given to us by the ISC, there seemed to be at least 2,000 to 3,000 enemy in a small area. It appeared that a regiment had moved in during the last two days and was now assembling in a position to attack hill 881.  

"This was the only intelligence we had of the impending attack, and it took about 45 minutes for the targeting and intelligence shop to convince the S-3 that an attack was imminent. We then took all our artillery resources and for about 30 minutes directed them all into an area of about 500 by 1,000 meters between the road and hill 881. When we stopped the fires, we asked the ISC for a readout. We were told that they had heard our fires, and that now there seemed to be incredible confusion in the area. There are screams, yells, panics and orders. To us it sounds like a regiment, or what is left of it, in perfect confusion trying to pull out in a hurry.”

This officer also described the defensive tactics used against enemy assaults on the camp. Varied intelligence would usually indicate when an attack was imminent and the sensors would tell precisely where the enemy was located and identify the exact time the attack was starting. When the assault began, the 175-mm guns at Camp Carroll put down fire on the east flank of the enemy, while the 105-mm and 155-mm guns at Khe Sanh laid barrages on the west flank and rear. Machine guns, small arms, and 60-mm and 81-mm mortars stopped the assault force at the perimeter, while airstrikes hit the probable reserves. In five such attacks, the body count totaled 1,600 enemy dead on the wire. The Bru Montagnards in the area reported many enemy dead from air and artillery strikes, and intelligence reported several assault forces refused to attack when, commanded for fear of certain death., In describing these events, the target officer continually emphasized that it was the sensors that provided the time and place of attack. _88/

A memorandum written in September 1968 by the Analysis and Reports Branch, Task Force Alpha, came to many conclusions similar to those of the Marine target officer: _89/

"DUMP TRUCK/Khe Sanh sensor fields have given ISC the capability to detect both moving and stationary activity in the antipersonnel area. However, this capability is limited by the mobility of personnel and the alternatives of routes offered by a complex trail network.  In order to provide maximum coverage of such a network, sensor strings are often emplaced across the trails rather than parallel to them. While this extends detection capability over a number of alternative routes, it diminishes the capability to determine direction and speed of movement. Furthermore, when direction and speed can be determined, the multiple choice of routes makes it impractical to track the target ahead. On the other hand, experiences in Battlefield Surveillance have demonstrated the ISC's capability to provide indicators of enemy movement patterns and battlefield tactics."

 Toward the end of the siege, Microtale direct readout equipment was installed at Khe Sanh. By putting these with forward observers who could monitor several sensor strings within line of sight, the Marines had an organic and workable system for using sensors to generate targets. The readout was real-time and the observer could exclude probable false alarms from artillery and aircraft. Being familiar with the terrain, the forward observer could guess the probable target location in relation to the sensors and call for artillery at Khe Sanh without the usual time-consuming coordination steps that data processing via Nakhon Phanom (NKP) required, According to the Khe Sanh target officer, "Activations from the sensors read by the microtales were almost always taken under fire, even though the other sensor reports, from NKP, rarely were toward the end," The microtale was too limited in capability to identify mass activity. _90/

Sensor data relayed by the EC-121 went into Air Force intelligence channels as well as to the Marines, The CHECO report on the siege of Khe Sanh detailed ':he unprecedented in-country intelligence effort run by 7AF at Tan Son Nhut to provide targets for NIAGARA --the SLAM in defense of the combat base, Sensor data joined photo and human intelligence in providing an all-source target data base, Not only did 7AF target its own aircraft based on this data, but the Marines used it in the form of Hot Item Reports and computer runs to target their artillery and close air support. _91/

Marine Mobility and DUEL BLADE II

With the end of NIAGARA on 31 March, the Air Force participation in northern Quang Tri operations lessened, NIAGARA had had several in-country zones where OF could frag airstrikes without coordinating with the ground commanders, These zones lapsed with the end of the operation, At the same time, the massive B-52 strikes shifted to Laos and II and III Corps. Even the debate over the Forward Bomb Line slipped into the background for a while, because the siege of Khe Sanh had caused COMUSMACV to place the 1st Marine Air Wing (less helicopters) under operational direction of the 7AF Commander, thus reducing the command and control questions regarding the FBL to part of a larger debate Then the November bombing halt closed TALLY HO to offensive air opera­tions.

By mid-summer 1968, many NVA units had withdrawn from northern Quang Tri, allowing the 3d MarDiv and Marine aircraft to cope more easily with the remain­ing threat. Though the Marines went ahead with their sensor program, the 7AF DUMP TRUCK operation was not implemented           Seventh Air Force chose to delay the antipersonnel system and redouble efforts in the antivehicle system. North of the DMZ, a major choke point interdiction campaign received heavy emphasis from July through October, but its center of operations lay north of TALLY HO. Even the previously described interdiction campaign against Route 1036 was against a road angling west around the DMZ, rather than south toward the forward Marine positions. Then in November the bombing halt in North Viet­nam and the DMZ brought the "understanding" that the NVA would not mount attacks from the zone. Partial observance of this tacit understanding by the enemy led to a further waning of military operations around the DMZ.

From the perspective of mid-1969, the sieges of Con Thien and Khe Sanh marked the peak of Air Force participation in operations around the DMZ. A year later, the area received few Air Force sorties. A tabulation of tactical air­ strikes in the 3d MarDiv area, including the DMZ, for l March through 15 April 1969 showed that the Air Force furnished one sortie in ten. _92/           In recognition of the predominant Marine effort, this report will sketch the military trends of the 15 months prior to July 1969, and describe the Marine sensor program in particular. It will then examine two subjects of special interest to the Air Force: the enemy helicopters in the DMZ and the renewed Forward Bomb Line debate revitalized by Operation THOR.

Just prior to the Tet Offensive and the siege of Khe Sanh, the 1st Cavalry Division entered Thua Thien, the province below Quang Tri. Then in February, the 101st Airborne Division arrived to help liberate Hue, Not only did these deployments permit the 3d MarDiv to concentrate forces in northern Quang Tri, they made the U.S. Army the numerically superior force in the two northern provinces. By April 1968, there were 31 Army maneuver battalions versus 24 Marine battalions._93/  In March, COMUSMACV created the Provisional Corps Vietnam (PCV), later renamed XXIV Corps, and put an Army general in command. Also at this time, heavy Army artillery moved near the DMZ to enhance counterbattery fire for the Marines, who were light on artillery compared to the NVA. _94/

This U.S. concentration of forces countered the enemy buildup. Intelligence officers at PCV in late March 1968 identified one NVA division in the DMZ and three around Khe Sanh. 95/  Later, the enemy threat at Khe Sanh abated, but it continued high around Dong Ha and Quang Tri City . For instance, in late April 1968, an estimated 17 enemy battalions operated within the so-called Leatherneck Square._96/