The following depiction is a representation from the ACES (Adaptation Controlled Environment System) printout for Los Angeles Center, from September 27, 1983. The ACES is a thick computer printout that is a listing of parameters specific to the computer that processes the radar data. In this document one finds, among other things, all the radar sort box assignments.
As you recall from slide 37 of my "Slide Show," the radar sort box that the Aero Commander was in shortly after it made its turn from the Paso Robles 196 radial to track the localizer into San Luis Obispo, was RSB #1114.
In the ACES listing below, RSB #1114 is adapted so that PRB (the Paso Robles Air Route Surveillance Radar) is first on the list, QLA (the long range radar near San Pedro Hill...which is the Los Angeles area) is the number two radar, and QSR (Boron) is number 3. Considering the distance the San Luis Obispo area is from Los Angeles (145 miles), and Boron (150 miles), it is understandable (and expected) that PRB would be the number one radar site for sort boxes in this area. As you recall from the "slide show" discussion, if for some reason PRB would fail, QLA would automatically take over. It would provide radar coverage for the area, but low-altitude coverage would suffer because of its distance. If both PRB and QLA were off-line, QSR would take over.
RSB ADAPTATION OPERATING SYSTEM JOB=OPS284 09/27/83....
23717 I 1109 N PRB A PRB C 00
23818 F 2802
23819 I 1110 N PRB QLA A PRB C 00
23820 F 2802
23821 I 1111 N PRB QLA A PRB C 00
23822 F 1500 2800 2500
23823 I 1112 N PRB QLA QSR A PRB C 00
23824 F 1500 2800 2500
23825 I 1113 N PRB QLA QSR A PRB C 00
23826 F 1500 2800 2500
23827 I 1114 N PRB QLA QSR A PRB C 00
23828 F 1500 2800 2500
23829 I 1115 N PRB QLA QSR A PRB C 00
23830 F 1500 2500 2600 2700 2800
23831 I 1116 N PRB QLA QSR A PRB C 00
23832 F 1500 2500 2600 2700 2800
23833 I 1117 N PRB QLA QSR A PRB C 00
23834 F 2600 1300 1500 0300 2500 2700
As you learned from my slide show and the papers I have written, the mosaic radar system, which is another name for "selective rejection" of data, is the way radar data has been processed in ARTCCs for quite some time. In this selective rejection processing method, data from one-and-only-one radar site at any given time is utilized, and other radar data is discarded (selectively rejected), especially in the case of untracked (typically VFR code 1200) targets.
So now, if somehow the first two radar sites in line 23827 were swapped, through whatever strange set of circumstances, to where the line read as the follows...
23827 I 1114 N QLA PRB QSR A PRB C 00
...it would be obvious that low-altitude, untracked targets would not be presented for display within the confines of this sort box, as the nearby Paso Robles radar data would be discarded. Yet, low-altitude aircraft that were already tracked (i.e. with a Full Data Block) prior to entering RSB 1114 would be retained for display as aircraft entered this sort box, as Paso Robles radar data would then be used as "supplemental." Such a swap in these radar sort box assignments would be extremely difficult to notice, as it would appear that some low-altitude aircraft (those already tracked and having then entered RSB 1114) would appear on the display just fine (supplemental radar data used), yet untracked low-altitude VFR targets within 1114 would not be displayed whatsoever (supplemental radar data not utilized).
While it may seem highly unlikely that RSB 1114 would be programmed in such a fashion, understand that if it was that way it would be easy for it to go unnoticed due to this difference in how tracked versus untracked aircraft are processed for display. It would give the appearance of intermittent problems with some aircraft transponders. And the chances of noticing such a glitch would be minimal, as approximately 75% of sort box 1114 lies over the Pacific Ocean where very little low-altitude flying takes place.
How could such mis-programming occur? Well, I suppose there is the human factors possibility of the data being accidentally switched. There is also the possibility of a computer memory element or a computer register which had a glitch that could contribute to such a critical mis-assignment. To say that it is impossible for the computer to have failed, or that it is impossible for a human error to have occurred in adapting these parameters, is, in my opinion, not realistic.
The above discussion does not take into account the fact that when radar data is discarded, it is not supposed to be stored on the Systems Analysis Recording tape (such a practice doesn't contribute to fulfilling the Search & Rescue mission!). However, when one considers a computer glitch, who knows what can happen. Nevertheless, the manner in which the radar data processing works is to reject data for display. If the 4 controllers didn't see the VFR Rockwell Commander's target, possibly sort box 1114 was inadvertently programmed incorrectly.
Note: I obtained over 1,000 frames of microfiche copy on this accident from General Microfilm. I also obtained nearly 1,000 frames of microfiche data on this accident from ALPA. I could not find any exhibit, nor any reference whatsoever, which showed or alluded to the Adaptation Operating System (ACES) for Los Angeles Center. Neither did I find any mention of the process of "selective rejection," "radar mosaicing," nor were "radar sort boxes" ever brought up.
[Return to "Reasonable Doubt" Home]
This page created Feb 25, 2000
© 2000 by Lusch's Midair Collision Investigations. All rights reserved.