The likely Roswell debris culprit remained hidden from the UFO investigators for over a decade. Despite thousands of man-hours dedicated to investigating the case, these UFOlogists seemed to have missed obvious clues. Interestingly, as early as the late 1980s, British researcher Andy Roberts was suggesting the potential source, "Secret balloon launches were taking place at that period and there seems to be good reason to suspect anything other than a secret test object of balloon type was retrieved" (Spencer and Evans 101). This comment and other objections were ignored by proponents who plunged headlong into the quagmire of confabulating witnesses and age-old memories. It took dogged work by investigators who did not have a preconceived notion of the source to discover the link between Roswell and the New York University (NYU) projects balloon flights for project MOGUL. By 1994 UFO researchers Robert Todd and Karl Pflock had come to the conclusion that the likely source was a balloon flown by the NYU team of scientists out of Alamagordo, New Mexico. The USAF looked further into the leads generated by these two researchers. As a result, they published their report, The Roswell Report: Fact vs. Fiction in the New Mexico Desert. The source of the Foster Ranch debris was determined to be from the NYU project's flight #4 that had been launched one month earlier. Mack Brazel's own description of the debris, published the day after the world heard about the crashed "disc" described in detail the materials that Professor Charles Moore, who was the project engineer, stated were in the flight. This seemed to be the answer for the source of the Roswell mystery.


Needless to say, many UFOlogists who were deeply involved in the case dismissed the possibility that flight #4 could have been the source of the debris at the Foster Ranch. Professor Moore's discussion about the NYU project and flight #4 resulted in him being declared a "debunker". Doctor Moore's experience with UFO researchers up to this point had been one of dismissal and quoting him out of context in regards to the Roswell case. Now that he was presenting the evidence for flight #4 being the source of the debris, UFOlogists took a different approach. Pro-crash UFOlogists had to positively show it was impossible that flight #4 could have made it to the Foster Ranch. Among these arguments was Kevin Randle's contention that the winds were blowing in the wrong direction to send the balloon flight to the Foster Ranch. However, it was as simple as Randle thought it might be. Professor Moore knew that many of his balloon flights experienced different wind directions and speeds as they rose into the upper atmosphere. It would take careful examination of all the weather data available to see if his balloon flight could have "landed" at the Foster Ranch.


To evaluate the flight path, Professor Charles Moore had to obtain as much information that was possible about the winds for the dates surrounding the flight. From the book UFO Crash at Roswell: The Genesis of a Modern Myth we find that he was able to obtain:


1. Northern Hemisphere surface and 500mb atmospheric pressure maps


2. Surface weather observations for West Texas and southern New Mexico


3. Pilot balloon measurements for winds aloft taken from White Sands (Orogrande), New Mexico (32 miles to the SW of the Alamogordo launch station).


4. Flight characteristics for flights #5 and #6 from the NYU records


While some critics may find this information insufficient to make an accurate determination, Professor Moore, who has decades of experience dealing with atmospheric physics, felt he could make a reasonably accurate reconstruction of flight #4's journey.


Initially, Moore reevaluated flights 5 and 6 and their data for several important and key reasons. The first had to do with flight #5, which had two interpretations in the NYU time-altitude plot. Moore found this to be inaccurate when compared to the ground track of the balloon. Therefore he created a third interpretation of the data and then attempted to see if it more accurately fit the actual ground track:


To verify its worth, I have examined possible ground tracks that would result from the use of the greater altitudes suggested in the midportions of interpretation 3 with the reconstructed theodolite elevation angles extracted from interpretations 1 and 2. Both of these sets of data would place the balloons, while still at their maximum altitude, almost directly over the spot where they later landed. With such a location over the recovery site at the start of the final descent, despite the strong tropospheric winds observed during the ascent. Such a cessation of winds aloft was unlikely, but the ground track shown in NYU Figure 32, and here in Figure 2, is consistent with the actual recovery site, about 25 miles (40 km) east of Roswell... The plot makes more sense than do the time-altitude curves given for interpretations 1 and 2 in NYU figure 33; their presentation of balloons floating at some intermediate altitude after some of the lifting balloons burst is unrealistic. From our later experience, we know that these balloons would either continue to rise or to descend without floating at some constant level. In fact, the slow stratospheric ascent depicted in interpretation 3 is exactly the performance that we were trying to achieve with the use of the dribbler ballast dropper. (Saler, Ziegler and Moore 87-8, 91,93)


Moore's modification of flight #5's time-altitude plot seems justified in this context. Because of the start and stop locations being known, he never altered the ground track. Evaluation of flight #6's data also revealed some errors but this time it was with the ground track and the maximum altitude reached by the balloon. According to Moore:


If the ground track were correct and errors were made in reporting the recovery site, the actual site would have been deeper in the mountains, in an area almost accessible from High Rolls, where Sid West, who recovered the equipment, resided. However, there was a road directly south from High Rolls that passed near the location where West reported having recovered the debris from this flight...On examining the track, I noted that all of the azimuths appeared to be rotated clockwise about 12° or 13° , suggesting that perhaps someone made an unnecessary correction to the original azimuths measured during the flight. The magnetic declination for Alamogordo at that time was about 12.5° . It appears to me that someone in the NYU data reduction section may have added the magnetic declination to the azimuths reported for the theodolite measurements when the analyst learned that the theodolite orientation was made using g a compass but he or she did not know that this correction had already been made. In any event, replotting the ground track data after subtracting the 12.5° from the azimuth for each location rotates the track counterclockwise so that the last point of the track is almost on the azimuth of the reported recovery site. However, the last point is about 20 percent more distant than the launch site than is the recovery site. But the maximum altitude of 72,500 feet (22.1km) given in NYU Figure 35 is about 20 percent greater than the peak altitude of about 60,000 feet (18 km) given in Crary's diary for this flight. Figure 3 in this chapter was prepared by use of the corrected azimuths and by scaling the horizontal distances by the factor (60,000-4069)/(72500 - 4069), where 4,069 is the altitude in feet of the theodolite at Almamogordo. With these corrections to the Flight #6 ground track, the last point on the new ground track lies just upwind (i.e., southwest) of the probable actual recovery site...I think the revised plot is probably more nearly correct than the one given in NYU Figure 34. The origin of the altitude error in plotting is not clear; Crary obviously obtained his value for the maximum altitude from us, because we received the radiosonde signals during the flight, but for some unknown reason, the NYU data analysts chose a different interpretations, one that now seems unlikely... (Saler, Ziegler and Moore 94-5)


In both cases, the data had to be modified to be consistent with the known values of the launch and landing sites. Such modifications, in light of this information, can be considered justified so that Moore could accurately determine what was happening in the upper atmosphere during this time period.


Flight #5 and #6's performance at levels over 50,000 feet was nothing new to Professor Moore as he states in the text of his book:


In our early ballooning, we learned about an interesting phenomenon: During the summer months, the winds in the stratosphere over midlatitudes and northern latitudes invariably flow from the east and are not affected by tropospheric weather changes...Most of the high altitude balloons launched during the summer months were initially carried to the east by the winds in the troposphere, but after ascending into the stratosphere, they were blown to the west, often for hundreds of miles. And as they descended at the end of the flights, they were carried back toward the east, again by the winds at lower levels. (Saler, Ziegler and Moore 104)


Once flights 5 and 6 reached an altitude in the upper 50,000-foot levels, the balloons took a turn towards the WSW after passing through a transition area. There can be little doubt that Professor Moore's analysis of these conditions is accurate.


With the upper stratosphere conditions established, Moore had to tackle a much more difficult analysis, the tropospheric wind conditions. He had no data for the conditions at Alamogordo on the morning of the 4th of June but did have wind data for the lower troposphere from a station roughly 30 miles away. An extract of the data was made available in his published work but he seems to have more information than what was published. He interpolated the wind data from the balloon launchings at White Sands on 3 and 4 June to produce an estimate of the conditions for the launching time for flight #4. In addition, he could determine how the atmosphere was affecting the winds based on what happened to flights 5 and 6. After examining the wind conditions and plotting them against each other, it is easy to see that between June 3 and 5 the source of mid and upper tropospheric winds were shifting from the southwest to a more westerly direction. Lower tropospheric winds seem to be fairly consistent from the SW or WSW. This is shown from data in flights 5 and 6 as well as the balloon data. Moore's evaluation of the atmospheric pressure maps produced other critical information. The surface maps showed a low-pressure cell over northern New Mexico, while the 500mb maps showed a low-pressure system over Oregon and high-pressure cell to the southeast in Texas. What this all meant was that any winds in southern New Mexico up to the 18,000 foot level were going to be predominantly from the southwest on 4 June. This verifies his interpolation of the balloon data. With all this information, Professor Moore was able to create a fairly accurate model of what was happening in the troposphere.


Moore also had to establish flight #4's configuration and tracking methods. All Moore had to go with was his memory and Crary's diary, which read:


Jun 4 Wed. Out to Tularosa Range and fired charges between 00 and 06 this am. No balloon flights again on account of clouds. Flew regular sonobuoy up in cluster of balloons and had good luck on receiver on ground but poor on plane. Out with Thompson pm. Shot charges from 1800 to 2400. (HQ USAF Attachment 32/Appendix 17)


Moore remembers that they had lost radar contact with the balloon while the balloon was being tracked with a theodolite. Flight #4 did not make it into the NYU records because, "Only those flights in which an attempt was made to control the altitude of the balloon are included in the summary. Excluded are flights made to test special gear and launchings which were not successful" (HQ USAF Attachment 32/Appendix 13 p. 27). The lack of any record (other than Crary's journal entry) indirectly verifies some of Moore's memory of the flight having problems being tracked and that no useful data was obtained from the flight. The theodolite measurements would only be used for mapping the ground track of the flight as stated in the NYU records, "The flights made in the early part of this program were tracked optically with theodolites. Coupled with the height data, theodolite readings provided a fairly reliable horizontal locus of the balloon" (HQ USAF Attachment 32/Appendix 13 p. 13). The theodolite measurements would not provide enough information to make an analysis of the flight. Vertical measurement would have been made via radiosonde or RADAR. Professor Moore recalls that the initial flight from Alamogordo did not use a radiosonde:


As I remember, in place of the bulky and heavy receiving station needed to record the pressure signals from balloon-borne radiosondes for our preferred height-determination method, we took to New Mexico several lightweight ML307 B, corner-reflecting radar targets to be carried on our balloon trains so that they could be tracked by the SCR584 radar. (Saler, Ziegler and Moore 81)


The end result would be that the SCR-584 radar would be used to provide vertical tracking of the balloon. The NYU project had already been using the reflectors as shown in the diagram for flight #2 and there would be no reason to discontinue their use. According to Moore, he recalls that the operators had difficulty tracking the reflectors. Even if they were able to track the reflectors, the balloon train's flight path could have exceeded the RADAR's limited range of about 40 miles. There was no other RADAR downrange towards Roswell to pick up the continued track and there was no apparent coordination with other RADAR outside of White Sands. Examining the track of flight #5, it was obvious that some of the balloons were exceeding the SCR-584's range. It appears that the NYU team did not expect their balloons to exceed this limit. The use of a radiosonde for flight #5 reveals that the team had learned from flight #4 and no longer used the reflectors in subsequent test flights that summer. The radiosonde would be the method subsequently used to provide altitude data for the balloon flights.


The last piece of the puzzle for Professor Moore was to determine when flight #4 was launched. He had to look closely at Crary's journal entry and the weather conditions in New Mexico at the time. Early on, the NYU project was required to have optimum clear conditions for their launch. However, due to problems encountered meeting these conditions the team requested less stringent standards. Examining Crary's diary, we see that they had planned to conduct a launch around midnight because they were setting off charges starting at midnight. On June 2, Crary remarked "Changed shooting plans to coordinate with balloon flights"(HQ USAF Attachment 32/Appendix 17). The timing of the charges with the launching indicates a launch had been planned at 0000 but then clouds interfered. Dr. Moore's examination of weather conditions at the time show that skies were cloudy (most of these being mid and high altitude clouds) through most of southwest New Mexico at midnight but conditions changed quickly. Both El Paso and Columbus to the southwest of Alamogordo were indicating clearing conditions around 0200. Additionally, Engle to the northwest of the launch site indicated conditions clearing a short time later with the exception of high altitude cirrus clouds, which were dissipating by 0300. Since Alamogordo did not make any observations until after sunrise, it is hard to tell when the conditions actually improved to the state where a balloon flight could be attempted. Comparing the Engle with the El Paso/Columbus observations, it appears there was a general overall clearing of the skies in the region around 0300. Consider the fact they did not launch initially because of clouds and that it was very difficult to manipulate the balloon train once it was arrayed for launch. Once conditions were clear enough, it is likely that they would have launched their balloon cluster right away. Using this line of reasoning, Professor Moore assumes flight #4 was launched at 0300 and there is enough evidence to support this conclusion.


Professor Moore determined that the balloon flight rose at a rate of 600 ft/minute to an altitude of 46,000 feet. Flight #5 rose at a rate of 600-800 feet/minute until it reached 35,000 feet but it lost two balloons during the ascent (one shortly after reaching 35,000 feet). Based on these values, it seems Professor Moore has assumed that the entire train was intact on its ascent. However, he assumes they did not rise too fast because of the lack of the sun heating up the air inside of the balloons (which occurred on flight #5 launched around sunrise the next day). Dr. Moore's estimation of a rise rate for flight #4 appears acceptable when compared to flight #5's performance. As the balloons rose higher and the sun rose, it seems logical that the ascent rate would increase. Moore does not state this but when the lifters were supposed to cut off at 35,000 feet, he did not change the ascent rate. One might assume the ascent rate of the balloons due to heating/expansion would compensate for the loss of the lifter balloons.


As the balloon rose Professor Moore began his ground track. From his wind interpolation of the White Sands data and the data from flight #5, he assumes the balloon took a northeasterly course. Flight #5 initially took this track until it reached levels over 13,000 feet, where it shifted to a more easterly course. However, Dr. Moore's track for flight #4 continues in a northeasterly direction. Moore's wind data and surface map information confirms that the winds in the mid to upper troposphere for the fourth of June were more towards the northeast and not easterly. This was due to the "trough aloft" moving across the New Mexico region. Comparing the projected wind direction Moore computed with the flight paths of flights 5 and 6, one can see the shift in wind direction towards the east (for flight #5) and then southeast (for flight #6) in this area of the atmosphere. Additionally, the wind speeds Moore uses to compute his flight path also are pretty consistent with those experienced by flight #5. Taking into consideration these factors, there is little doubt that the flight would take a northeast course until it began to reach the upper levels of the troposphere somewhere near Arabella, New Mexico


Once the balloon train reached the upper troposphere, new wind systems began to take control of the balloon flight. The balloon train now took a shift of direction to the northwest as it passed through the Tropopause (the boundary between troposphere and stratosphere). These are the same winds experienced by flight #5 the next day. However, unlike flight #5, Moore figures that flight #4 continued its ascent to levels higher than that reached by flight #5. Again, the early launch time plays a key role such that the balloon flight would not experience too much loss of lift by balloons bursting from solar heating. In flight #5, the flight moved northwesterly until it began to lose lift. However, just before descending the flight did manage to reach the lower levels of the stratosphere and shift towards the WSW for a short period. This is the same wind direction experienced by flight #6 a few days later. Therefore, when Dr. Moore shows the flight making a turn towards the WSW, he is not arriving at this value arbitrarily. The reason that flight #5 did take this course was because the flight stayed at levels lower than that flight #4 apparently achieved. Again, all of these values are consistent with information from flights 5 and 6. Flight #4 now traveled WSW for almost three hours in Moore's estimate. Compared to flight #5, flight #4 spent an extra two hours in the upper atmosphere. Once again, the time of launch explains the difference and it does not seem that Professor Moore is taking too many unnecessary assumptions in this model.


The final descent of flight #4 was rapid. Considering the amount of time the train was in the upper atmosphere, it seems that there would be a significant amount of balloons bursting during the end of the flight. When one examines the description of balloon debris by weight reported by Mack Brazel to the Roswell Daily Record, it seems that this was the case. According to Professor Moore, there was only enough material for three or four balloons. Flight #5 had 12 balloons intact when it landed and it fell at an average rate of 800-1000 feet per minute until it reached 25,000 feet. Assuming that some other balloons were intact to pull some of the upper train away from the Foster Ranch, the number of balloons left after the descent would have been less than a dozen. This is consistent with the rapid descent calculated by Moore. During the descent the train again resumed a northeast flight path as it passed back into the troposphere. This path took it into the vicinity of the Foster Ranch. Upon touching down, the train would have been dragged in a northeast direction scattering balloon and reflector material (as well as other items in the train such as parachutes, launching rings, etc) over a good size area. Once this ballast had been shed, the remaining balloons would have taken the remains of the train airborne and further to the northeast.




Ground tracks for flights 4 (estimated), 5, and 6 (Moore interpretation) from the book UFO Crash at Roswell: The Genesis of a Modern Myth




Time-Altitude plots for flights 4 (estimated), 5 (Moore interpretation), and 6 (Moore interpretation) based on information located in UFO Crash at Roswell: The Genesis of a Modern Myth and The Roswell Report: Fact vs. Fiction in the New Mexico Desert


All these interpolations, modifications, and analyses have come under criticism by UFOlogists. One in particular, David Rudiak, who is apparently an optometrist, documented his arguments on the UFO Updates mailing list. Among many things, Rudiak remarked that Moore uses too many variables in his calculations,


That's the problem with having lots of variables to play with and _no tracking data_ to pin anything down. You can make this balloon go and end up just about anywhere you want... Any critical analysis of Moore's model creates holes big enough to fly a fleet of flying saucers through. In brief, Moore makes lots of assumptions, has no tracking data, and only scanty weather data.... (Rudiak)


Moore's major assumption is that the Foster Ranch was the location that his balloon flight landed (which is verified by Mack Brazel's description of the debris) and then proceeded to show how his balloon could have gotten there. If the winds were blowing in the wrong direction, he could not have shown how his flight made it to the Foster Ranch. All of his assumptions are reasonable but Rudiak seems to think he knows better. It is obvious that Rudiak is either incapable of comprehending what Moore wrote or is deliberately writing misleading and inaccurate information in order to reinforce the belief that flight #4 could never be the source of the Foster Ranch debris. When it comes down to it, one has to wonder which individual has a better grasp of atmospheric physics and is being objective. A Professor with decades of experience in this field or an optometrist who seems more interested in vindicating a belief in extraterrestrial visitation than "critical analysis".


Ever since Dr. Moore began to argue that his balloon flight was the source of the debris, I have yet to read one professional critical analysis of Moore's writings. With all the money and resources, that these UFO groups possess, you would think they could find one individual, with the appropriate background and experience in atmospheric physics, to refute his findings. Their lack of ability to find such an individual speaks volumes. Now if only the Roswellites would apply a little "critical analysis" to the crashed alien spaceship theory. Nahhh.....never happen....


UPDATED November 2008 and July 2014


Since the fall of 2002, UFOlogists David Rudiak and Brad Sparks have pointed out that there are many errors with Professor Charles Moore's published work on flight #4. When one checks the math in the tables, they are correct that there are some problems in Moore's work that Moore should address. Many of the errors are minor and if one makes a few corrections, it still possible for the balloon to land in the vicinity of the Foster Ranch. One must remember that Moore was working with limited data and speculation about how the balloon flight performed. If one changes any variable such as ascent/descent rate or wind speed, then the balloon will overshoot or undershoot the ranch. As a result, there is a "footprint" for the landing area established somewhere to the northeast of Alamogordo AAF in the vicinity of the Foster ranch. This "footprint" could include the Foster Ranch since it is within the realm of possibility that the wind speeds/rate of ascent/descent could shift in favor of this area.


The most important factor to consider is that Brazel reports finding materials that match Moore's description of what probably existed on flight #4. The quantities reported are not what one would find from a simple weather balloon. Instead they point towards a multiple balloon array, which means Brazel saw or found remains from such a flight. The latest effort to "deflate" this testimony is to suggest that Brazel was shown the flight #4 debris which landed tens of miles to the northeast of his home. This explanation completely disregards the odd problem that the debris would have sat in somebody’s field for a month without anybody really noticing it. Of course, one might suggest that the location was extremely far from anyone’s ranch. This argument now assumes the military was scouring every acre of land of desert over the weekend with a limited pool of manpower. The probability of the military locating such a small debris field miles away from the Foster ranch and even further away from the second site location at the Hub Corn ranch stretches the limits of probability. Additionally, this location would have not been between the Foster Ranch and the Corn site, which would be the area first searched by the military for any remaining debris. Add to this, the likelihood that the military could decide that they needed to get Brazel to this location in order for him to create the description to the paper that evening. All of these "coincidences" are stretching the limit of credulity. A more likely scenario is that Brazel did find this debris at the Foster Ranch and it was the probable remains of flight #4.


As I stated in my webpage regarding much of what Sparks/Rudiak have written regarding this subject at, much of it is minor and can be explained if one looks at Moore's work and some of the records. Professor Moore is the one individual who might be able to answer some of these questions but based on the accusations being leveled by Rudiak/Sparks, he chose not to participate in an endless debate, which served no useful purpose. Based on what I was able to determine from my brief exchange with him, and reports from others, he had turned his attention away from Roswell and towards lightning research. He was perfectly willing to let his work stand, math errors and all, with no comment.

I have written a great deal more of some of the arguments about this balloon flight in my SUNlite newsletter. The following issues specifically address additional arguments made regarding flight #4 that are not addressed here or elsewhere on this web site:

SUNlite 4-4: The cluster of balloons P. 18-19

SUNlite 5-5: Crashology's last stand. P. 5-15

Future updates and information about Roswell and flight #4 will appear in SUNlite.