Colorado Meteorite Recovery
The amazing story of how we found three Colorado meteorites from a fireball and computed the original meteoroid orbit around the Sun
Meteorite Recovery from the Elbert, Colorado Fireball of 11 January 1998
Frank Sanders
10 December 2000, revised 23 May 2007
Executive Summary
A bright fireball was widely observed in central Colorado skies on 11 January 1998 at 12 minutes past local standard midnight. Accompanying sonic booms were heard in the Black Forest area, and were recorded in Colorado Springs and at Erie, CO by a consumer-grade video camera and a U.S. government infrasound array, respectively. Subsequent investigation and analysis by staff and volunteers of the Denver Museum of Nature and Science (DMNS) Meteorite Survey fused triangulation information from numerous eyewitnesses, the video surveillance camera and the infrasound acoustic array. Those data were used to establish the exact fireball flight-line parameters (including heading, descent angle, altitude, and the vertical projection of the flight path over the earth’s surface), the mean velocity of the bolide, its exact atmospheric entry time, and the most likely elements for the meteoroid’s original orbit around the sun. In the spring of 2000 and the fall of 2002, three meteorites (one in 2000 and two in 2002) were found from the January 1998 bolide near the towns of Elbert and Treadway, respectively.
Soon after the 1998 fireball event a public information campaign was conducted in the vicinity of the point of retardation (POR) to alert residents to the possibility of finding a meteorite locally. Ground searches were undertaken by the DMNS Meteorite Survey at that time, without result. In 2000, a meteorite specimen was discovered by local residents and brought to DMNS for identification and evaluation. The public education campaign had paid off, as the people who brought the specimen to the museum had been alerted by the campaign. The specimen’s fall location (which was beneath the calculated 1998 bolide flight path), its lightly weathered but mostly intact fusion crust, and the nearly unweathered condition of exposed ground mass below the fusion crust led to the conclusion that the specimen was a fragments from the 11 January 1998 bolide. It was originally hoped that the collected specimen would be transported to a NASA laboratory in Houston where it could be subjected to sodium-22 examination and mineralogical analysis to determine its precise age since its fall and its proper classification. But as events finally unfolded, the meteorite’s owners terminated all communication with researchers and withdrew the meteorite from research access before such studies could be performed.
In November 2002 two more pieces were collected from the Elbert fall of January 1998, this time from a location near Treadway. Under hand-lens examination the Treadway pieces are visually identical to the Elbert pieces, and like the Elbert pieces they were found in the predicted strewnfield area for the January 1998 bolide. Unfortunately, to date these pieces have not undergone laboratory examination, either.
The failure to be able to perform laboratory examination of the collected meteorites notwithstanding, the collected data have been analyzed to yield the flight trajectory, fall coordinates, and original orbital elements of a meteorite that shattered and fell in several pieces near Elbert, Colorado. This is only the sixth meteorite specimen in history to have been correlated with original elements. The computed orbital elements indicate that the meteoroid’s orbit carried it to the far edge of the asteroid belt at aphelion and just grazed the Earth’s orbit at perihelion. The meteoroid’s orbital inclination was somewhat larger than the Earth’s, however.
1.0 Introduction and chronology
1.1 The fireball event: At 12:12 AM (12 minutes past midnight) MST (0712 UT) on January 11, 1998, a bright fireball was widely observed along the Colorado central Front Range and the adjoining eastern high plains. As subsequently reconstructed by Denver Museum of Natural History staff in the Geology Department, the fireball travelled southwest to northeast and was visible for several seconds. It shattered into at least two major pieces prior to the point of retardation. At or just before the final point of retardation, one or both of the major fragments shattered into a number of smaller pieces. The dark flight of those pieces was observed by a number of eyewitnesses in the Elbert-Kiowa-Elizabeth area.
1.2 Collection of initial eyewitness reports: Following the event, local television and newspaper outlets gave extensive coverage to the fireball event, and announced a telephone number at DMNH for witnesses to call. Within a few days, over 650 phone calls were received at that number. From those calls, some 50 stationary witnesses were identified by DMNH staff. Of those, about 36 people believed to probably be the best witnesses were interviewed at field locations, each interview consisting of a verbal description of the event and the acquisition of azimuth and elevation angle bearings on the observed tracks across the sky. Members of the museum’s Meteorite Survey, a group of trained volunteers, collected these data in a timely fashion.
1.3 Public education campaign near the point of retardation: The investigation subsequently took two paths: A public education program was conducted by the DMNH Meteorite Survey in the area where the fall was believed to lie, and a set of ground searches were performed in areas believed to lie under the flight path. The public education program consisted of three elements: An evening lecture at a local church; a display of meteorite specimens and a description of the 11 January event at a local grocery store; and an article about the event in a local newspaper. The public education materials described to local residents the likely appearance of the meteorite falls, and encouraged contacts with the DMNH Meteorite Survey in the event that any fall should be discovered.
1.4 The education campaign bears fruit: Neither the public education program nor the ground searches yielded any specimens in 1998-1999. But in August 2000, following a March 2000 telephone call to the DMNH Geology Curator, a local family brought a suspected specimen to DMNH. It had been found by the family’s five-year-old boy in the spring of 2000 whilst he was playing on his family’s property. The boy had given it to his mother as a curiosity. She recognized it as having an unusual appearance (black fusion crust with light gray ground mass showing through in a few places), and recalled the newspaper article and grocery store display of 1998. She and her husband subsequently brought the specimen to DMNH for identification.
1.5 The Elbert and Treadway meteorites are found: The Elbert specimen (Figures 1-4) was identified by the DMNS Geology Curator as a meteorite. Its fresh but slightly weathered fusion crust was consistent with having lain on the ground for two years. Its fall location was below the calculated flight path of the 11 January 1998 bolide. Based upon these facts, the specimen was considered to be a fragment from that bolide. Its mass was 550 gm. It was recovered several kilometers uprange from the likeliest location of the main fall.
In November 2002 a new pair of meteorite specimens came to light from the nearby location of Treadway. Like the Elbert specimen, the Treadway pieces were collected within the area that the author had predicted for the Elbert bolide strewnfield. The ground mass characteristics and fusion crust preservation of the Treadway pieces were identical to the Elbert meteorite under hand-lens examination. The Treadway specimens showed little if any weathering beyond that of the Elbert pieces that had been found 18 months before. It was concluded that the Treadway specimens (Figure 5) also originated from the Elbert bolide of 11 January 1998. Unfortunately those pieces have not been made available for laboratory examination. The slow rate of meteorite fusion crust weathering in the Treadway and Elbert meteorite specimens may be due to the relatively dry conditions that prevail in eastern Colorado.
2.0 Raw Data
Raw data were collected from eyewitnesses, a video surveillance camera, two sound recording systems, and a Department of Defense (DOD satellite). The raw data are described in this Section.
2.1 Eyewitness data: From three dozen eyewitness observations, the following details were reported with some consistency: The visual event occurred between ten and fifteen minutes after local standard midnight. It lasted between three to five seconds. The color was initially white, and the object flickered rapidly. Illumination cast on the ground at a distances in excess of 23 kilometers was comparable to a full moon. As the event progressed, the color shifted to red-orange. Disintegration into two major pieces was observed late in the flight, with an accompanying shower of sparks that rapidly decelerated relative to the major pieces. Figure 6 shows an eyewitness drawing of this break-up. The disintegration was described by some witnesses as appearing similar to an exploding firework. The two main late-stage pieces faded gradually as they travelled toward the point of retardation (POR). A second major break-up may have occurred just prior to the POR; some accounts indicated that a third major piece was observed at the POR.
A number of witnesses in the fall area observed pieces falling at a steep angle from the POR, during dark flight. Their glow faded to invisibility while still above the observers’ the local horizon. As many as six pieces were observed during dark flight.
2.2 Camera data: No cameras are known to have recorded the bolide directly. But a VHS-technology consumer-type video surveillance camera pointed toward a parking area in Colorado Springs did record ground illumination and some reflections from the bolide. The recorded data were inadequate for an analysis of the trajectory across the sky, but do provide a record of the duration and flicker characteristics of the event. Figure 7 shows two frames from the camera. The first frame was recorded just prior to the event and the second frame shows the brightest ground illumination during the event. Comparison of the frames shows the magnitude of illumination due to the bolide. The bolide’s closest approach to the camera was approximately 23 kilometers slant range (as described in Section 3). The bolide elevation angle relative to the camera site was between 10° to 20° above the horizon, to the left (roughly north).
The camera data provide a quantitative record of the event’s length and of the early flickering described by witnesses. The camera recorded the event on a total of 85 frames at a rate of 30 complete frames per second. Flickering occurred during the first 61 frames. The event faded gradually during the next 24 frames. Thirteen flicker peaks were recorded, and three were notably bright. The flicker intervals were (counted to the nearest millisecond): 300; 233; 167; 200*; 67; 67; 200; 67*; 200*; 167; 200; 100; where asterisks denote intervals preceding especially bright flickers. Flickering occurs for two secs with an average flicker interval of 0.15 sec. The event fades gradually for another 0.8 sec. The total recorded ground illumination interval lasts 2.8 sec, consistent with the event intervals of three to five seconds described by witnesses.
Because the illumination fades gradually toward the end of the video record, and because structures and trees in the camera vicinity were low-lying, only extending a few degrees above the ground from the camera’s vantage point, it is unlikely that the bolide was eclipsed by nearby objects at this location. The gradual recorded fade is therefore most likely a record of the bolide’s diminishing brightness as it approached the point of retardation. The implication is that the bolide faded from significant brightness to dark flight within an interval of 0.8 sec.
2.3 Acoustic data from a video camera and an infrasound array: The bolide generated a short, loud boom succeeded by several seconds of low-pitched rumbling sounds that were heard by many witnesses in the Parker and Black Forest areas between Denver and Colorado Springs. Residents in the fall area were startled by the boom even if they were indoors. Those who did not see the fireball in some cases believed that an object had struck either their residence or an outlying building, and went outdoors to investigate. Two microphone-equipped systems collected records of the bolide sounds. One was the video camera described in Section 2.2. The other was a National Oceanic and Atmospheric Administration (NOAA) infrasound system located at Erie, CO.
Figure 8 shows audio data from the video camera microphone. As heard by the human ear, the recording contains a distinct double boom that lasts a fraction of a second. The peaks in Figure 5 are interpreted as corresponding to this multiple boom. The interval between the end of the visual event and the arrival of the acoustic impulse is 68.7 sec. A sonic velocity of 333 m/s indicates a distance of 22.9 km traveled in that interval. Only a single point along the bolide’s computed flight path (see Section 3) passes within that distance of the video surveillance camera. That point lies at an azimuth of 337 degrees from the camera, on a line connecting the camera to the flight path at precisely 90° to the calculated 67° heading of the bolide (see Section 3). This point is 32 km uprange from the computed point of retardation. It is therefore concluded that the video camera most likely recorded a sonic bow shock wave (a sonic boom) rather than the sound of an in-flight break-up or explosion.
The NOAA infrasound acoustic array at Erie, CO is used to detect and record the sounds of such phenomena as tornadoes, avalanches, reentering space vehicles, and bolides. The Erie array recorded three distinct sounds from the 11 January 1998 event. They were detected at 17, 20, and 26 minutes past local midnight (0717, 0720, and 0726 UT). Two of the sounds came from the direction of the witnessed bolide, but the third sound, which was by far the most energetic, came from nearly the opposite direction. All three sounds had the acoustic signatures of explosion events. The source of the third sound, in particular, has not been explained. Telephone calls from DMNH personnel to law enforcement agencies in areas as far north as southern Wyoming yielded no witness reports of additional bolides in those skies that night. The source of the first two sounds is assumed to be the witnessed bolide, but the cause of a three minute interval between two similar sounds from approximately the same direction is likewise not understood. The Erie array is located approximately 96 km NNW (bearing 340°) from the calculated bolide flight track, and the propagation of sound from the bolide flight path to the Erie station would have been at 87° (counterclockwise) relative to the computed 67° bolide heading of the flight path (see Section 3).
At an acoustic speed of 333 m/s, the nominal distance of 96 km between Erie and the bolide flight path would be traversed in 290 sec. Thus the moment of bolide entry is inferred to have been about 290 sec prior to 0717 UT, that is 0712 UT or twelve minutes past local standard midnight.
The acoustic data imply that a sonic bow shock wave (sonic boom) was heard in Colorado Springs, as interpreted from geometry (above). Although the Erie station lay at nearly a right angle to the flight path, a bolide break-up prior to the point of retardation may have been the source of the first or second boom recorded at Erie (based upon the acoustic signature recorded by the NOAA facility). The reason for a three minute interval between the first two booms is not known. The Erie data indicate that a second bolide not noted by eyewitnesses may have entered the atmosphere some distance north and approximately nine minutes after the entry of the witnessed object. Its high energy implies that it may have been generated by a more massive object than the one that was observed.
3.0 Flight Trajectory, Data Analysis and Orbit Determination
Adequate data for determination of the three-dimensional orientation and location of the 11 January bolide trajectory were collected by the DMNS Meteorite Survey from nine witnesses spread across an area of about 4,800 square km. The positions of these witnesses ranged from north of Denver to as far south as the Black Forest (80 km range north-south); and from as far west as Golden to as far east as the longitude of Kiowa (60 km range east-west). Each witness provided two vectors to the bolide flight path prior to the POR, as shown in Table I. Each pair of vectors was used to determine a corresponding plane within which the bolide trajectory might lie. Intersections of each pair of these planes generated possible solutions for the bolide flight path. In principle, nine witnesses would generate ((9*8)/2) = 36 pairings and 36 corresponding flight path solutions. Unfortunately some of the witnesses were located close to one another (e.g., several were in the town of Castle Rock, separated by only a few kilometers each other), and others were colinear on the same side of the bolide path (resulting in solutions that were highly sensitive to the reported elevation angles). These cases resulted in some degenerate solutions that could not be used for flightline determination.
Still, a total of 24 non-degenerate paired solutions were finally computed for the bolide trajectory, as shown in Table II. The mean heading solution was 67° ± 18°. The mean computed descent angle was -24° ± 26°. Uncertainties in eyewitness data, and a lack of quantitative data from any cameras, result in the relatively large standard deviations for these values.
The coordinates of the sub-point where the bolide first became visible have been computed to be 39.111° N, 104.939° W, with a circular error probable (CEP) of ± 5 km. The mean computed altitude of the first visible point has been computed at 30 km ± 4 km above mean sea level (MSL). This places the terrestrial sub-point of initial visibility 6 km NW of the town of Monument, in NW 1/4 Sec. 7, R 67 W, T 11 S.
The coordinates of the sub-point of retardation (sub-POR) have been computed to be 39.251° N, 104.558° W, with a CEP of ± 1.5 km. The mean computed altitude of the POR has been computed to be 14 km ± 4 km MSL. This places the terrestrial sub-point of the POR 5 km NNW of the town of Elbert, in NW 1/4 Sec. 21, R 64 W, T 9 S. The altitude of the POR is near the lower limit of the probable altitude range that might be expected for bolides in general.
The distance over ground between the sub-points of first visibility and POR is 36.0 km. The direction of travel along this line was 67° relative to true north. (That is, the direction of travel was from WSW to ENE.) The computed descent angle between the first point of visibility and the POR was 24°; the distance between these two points is 39.4 km. The videotape data indicate that the elapsed interval to cover that distance was 2.8 sec (see section 2.2, above). The average speed computed for the bolide is thus 14.1 km/s. Using modeling described by Ceplecha [1], the bolide’s entry velocity is estimated to have been 14.4 km/s.
The first recovered meteorite specimen from the bolide was found 2 km SW of the sub-point of the POR, directly below the computed bolide flight path, at NW 1/4 of NW 1/4 Sec. 29, R 64 W, T 9 S. Computations and local witness reports of the dark flight of several fragments lead us to believe that the main fall lies near Sec. 8, R 63 W, T 9 S.
Provisional heliocentric orbital elements have been computed for the bolide, based upon the derivation for determination of meteor orbits provided by Dubyago [2]. These elements are presented in Table III. The computed orbit carries the original meteoroid to the far edge of the asteroid belt at aphelion, to a distance roughly equal to that of the Pribram meteorite. Perihelion lay just inside Earth’s orbit, at 0.89 AU, but the meteoroid orbit was inclined at 29° to the ecliptic.
4.0 Summary of the Elbert Fireball Event and Elbert-Treadway Meteorite Recovery
As reconstructed from analysis of the raw data, a bright bolide commenced in the skies of central Colorado at about twelve minutes past local standard midnight (0712 UT) on 11 January 1998. It first became visible to eyewitnesses as it passed above the Pike National Forest, east of the Rampart Range and 6 km northwest of the town of Monument at an altitude of about 30 km (19 miles) MSL.
Traveling at a mean velocity of about 14.1 km/sec on a heading of 67 degrees true (that is, from WSW towards ENE), the bolide brightened rapidly and was observed for a total interval of about 2.8 seconds. It flickered rapidly for the first two seconds. About 2 sec into the visible flight it broke into two pieces, with an accompanying shower of sparkling debris. The two pieces were visible to distant observers for approximately another 0.8 seconds, with little or no flickering. Those pieces dimmed rapidly as they approached the point of retardation. At or near the point of retardation, the two main pieces disintegrated into three to six fragments that fell to the ground from an altitude of about 14 km MSL (about 46,000 feet MSL), placing this event near the lower limit expected for bolides in general. The subpoint of retardation has been determined to be 5 km NNW of Elbert, and the likeliest location of the main fall is estimated to be about 8.5 km NE of Elbert.
The bolide produced a sonic boom that was heard in Colorado Springs. Witnesses there were located at a right angle to the flightline at distances on the order of 20-50 km and thus were in a prime position to hear this phenomenon. In addition to the sonic boom the bolide generated an acoustic pulse from an explosive break-up event that was recorded by a Federal Government (NOAA) infrasound monitoring station at Erie, CO. Two additional, unexplained booms were recorded at Erie. One of these (the third in the overall sequence) was by far the most energetic of the three and originated to the north of Erie. This is remarkable because Erie is 96 km north of the fireball flightline. This indicates the possibility that an unwitnessed (due to local cloud cover north of Erie that night) and possibly larger companion to the Elbert meteoroid may have entered the atmosphere a few minutes after the witnessed Elbert bolide.
A video camera in Colorado Springs recorded the Elbert bolide’s illumination of the ground at a (minimum) slant-range distance of 23 km, resulting in a reasonably precise estimate of the interval of the bolide’s flight. This datum, combined with the known length of the bolide flight path between Monument and Elbert, provided a calculated velocity that was combined with other data to yield heliocentric orbital elements of the original meteoroid. The meteoroid is calculated to have occupied an orbit that took it to the far edge of the asteroid belt at its most distant point from the sun, and that brought it to just within the earth’s orbit at its nearest approach to the sun.
Three stony meteorites of undetermined classification have been collected from a location immediately below the computed the flight path, and about 2.2 km uprange from the point of retardation. The first specimen was collected two years after the 1998 event, in the spring of 2000. It has weathering characteristics in its fusion crust and ground mass that are consistent with having been exposed to the elements for that period of time. Its find location and its slightly weathered condition have led us to infer that it is a fragment from the 11 January 1998 bolide. Two more meteorites were found side-by-side at the nearby location of Treadway in November 2002. Their fusion crust, ground mass and weathering characteristics are identical to the Elbert meteorite under hand-lens examination, and like the Elbert meteorite they are obviously fresh falls. We conclude that the Treadway meteorite pieces did originate from the 11 January 1998 Elbert bolide.
Further ground searches in the probable fall area have not yielded additional meteorites. It was originally hoped that the collected specimens would be transported to the NASA laboratory at Houston, TX for Na-22 and mineralogical analysis and thin sectioning that would have determined its precise classification. It was believed that such analysis would have been valuable for basic research purposes while enhancing the financial value of the meteorites for their owners. Unfortunately none of the meteorites have been made available for laboratory examination or analysis.
5.0 References
1. Z. Ceplecha, et. al., Advanced Space Research, 84, 327 (1998).
2. A.D. Dubyago, Determination of Orbits, Chap. 12, Macmillan, 1961.
Fig 1. Scaled view of the recovered Elbert meteorite, found 19 months after its fall. The corner at right was never found.
Fig 2. Recovered Elbert meteorite with mostly intact fusion crust. This was what remained after 19 months on the ground.
Fig 3. Hand-lens view of the Elbert meteorite.
Fig 4. Microscope view of the Elbert meteorite.
Fig 5. Treadway meteorite specimens from the Elbert bolide.
Fig 6. Elbert bolide eyewitness sketch.
Fig 7. Elbert bolide video frame 216, taken at 49 km slant range.
Fig 8. Elbert bolide acoustic recording. Acoustic data were also collected from a NOAA array at Erie, CO.
Elbert bolide Table I. January 11 1998 fireball witness raw data
Elbert bolide Table II. Fireball witness correlation matrix
Elbert bolide Table III. Computed meteoroid orbital elements
