The Tajikistan superbolide of July 23, 2008. I. Trajectory, orbit, and preliminary fall data



The results of the atmospheric trajectory, radiant, heliocentric orbit, and preliminary strewn field calculations for an extremely bright slow-moving fireball are presented. In the evening hours of July 23, 2008, a bright object entered Earth's atmosphere over Tajikistan. The fireball had a −20.3 maximum absolute magnitude and a spectacularly long persistent dust trail remained visible over a widespread region of Tajikistan for about 28 minutes after sunset. The fireball was also recorded by a visible-light satellite system at 14 h 45 min 25 s UT, and the dust trail was imaged by video and photocameras. A unique aspect of this event is that it was detected by two infrasound and five seismic stations too. The bolide was first recorded at a height of 38.2 km, reached its maximum brightness at a height of 35.0 km, and finished at a height of 19.6 km. The first breakup occurred under an aerodynamic pressure of approximately 1.6 MPa, similar to the values derived for breakups of the scarcely reported meteorite-dropping bolides. The fireball's trajectory and dynamic results suggest that meteorite survival is likely. The meteoroid followed an Apollo-like asteroid orbit comparable to those derived for previously recovered meteorites with accurately known orbits.


Meter-sized and larger meteoroids of both asteroidal and cometary origin may penetrate deep into the Earth's atmosphere, and could be potential meteorite producers. The importance of compiling good fireball data is obvious, as from the atmospheric trajectory and preatmospheric velocity, the heliocentric orbit of the meteoroid can be computed, and then attempts can be made to link the meteorites directly to their parent bodies (Trigo-Rodríguez et al. 2007, 2009; Madiedo et al. 2013). By studying the behavior of the luminosity of meteoroids in the atmosphere, as well as the ablation mechanism and physical properties (mass, density, porosity, etc.), we can infer important information to establish under which circumstances meteoritic materials give rise to potential meteorite falls. However, observational data capable of providing the determination of precise heliocentric orbits of meteorite-dropping bolides are still scarce (Ceplecha 1961; McCrosky et al. 1971; Halliday et al. 1978; Brown et al. 1994, 2000, 2004; Grady 2000; Borovička et al. 2003; Trigo-Rodríguez et al. 2006; Bland et al. 2009; Jenniskens et al. 2009; Spurný et al. 2010, 2011; Haack et al. 2012).

Meteorite-dropping bolides with orbital information are such rare events (Table 1) that different detection systems, including photo, seismic, infrasound, and satellite, need to be considered to reconstruct, with enough accuracy, the atmospheric trajectory and heliocentric orbit, and also to assess carefully the ending mass depending on the particular initial velocity, deceleration, and entry geometry. Bright fireballs are produced by large meteoroids, which may penetrate deep into the atmosphere and generate a cylindrical blast wave during their hypersonic passage, which decays to low-frequency infrasonic waves when propagating over large distances (ReVelle 1976; Edwards et al. 2006; Ens et al. 2012). Here, we present the data of multi-instrumental observations of a superbolide registered on July 23, 2008 over Tajikistan, with a radiated energy of 0.05 kilotons of TNT (1 kT = 4.2 × 1012 J) (Brown 2008).

Table 1. Main orbital data of recovered meteorites.
Meteorite nameMeteorite typeYear of fallOrbital dataReferences
a (AU) e i(°)
  1. References list: [1] Ceplecha (1961); [2] Spurný et al. (2003); [3] Grady (2000); [4] McCrosky et al. (1971); [5] Halliday et al. (1978); [6] Brown et al. (1994); [7] Borovička et al. (2003); [8] Brown et al. (2000); [9] Brown et al. (2004); [10] Llorca et al. (2005); [11] Trigo-Rodríguez et al. (2006); [12] Bland et al. (2009); [13] Jenniskens (2009), [14] Haack et al. (2012), [15] Brown et al. (2011), [16] Spurný et al. (2010), [17] Spurný et al. (2011)

PříbramH519592.401 ± 0.0020.6711 ± 0.000310.482 ± 0.004[1], [2], [3]
Lost CityH519701.66 ± 0.010.417 ± 0.00112.0 ± 0.1[4]
InnisfreeL519771.872 ± 0.0010.4732 ± 0.000112.27 ± 0.01[5]
PeekskillH619921.49 ± 0.030.41 ± 0.014.9 ± 0.2[6]
Tagish LakeCM220002.1 ± 0.20.57 ± 0.051.4 ± 0.9[7]
MorávkaH5-620001.85 ± 0.070.47 ± 0.0232.2 ± 0.5[8]
NeuschwansteinEL 620002.40 ± 0.020.670 ± 0.00211.41 ± 0.03[2]
Park ForestL520032.53 ± 0.190.680 ± 0.0233.2 ± 0.3[9]
Villalbeto de la PeñaL620042.3 ± 0.20.63 ± 0.040.0 ± 0.2[10], [11]
BunburraAchondr.20070.8514 ± 0.00090.245 ± 0.00139.07 ± 0.07[12]
Almahata SittaUreilite20081.308201 ± 0.0000090.312065 ± 0.0000032.54220 ± 0.00004[13]
GrimsbyH4-0620092.04 ± 0.050.518 ± 0.01128.07 ± 0.28[15]
JeseniceL620091.75 ± 0.070.431 ± 0.0239.6 ± 0.5[16]
Mason GullyH520102.47 ± 0.0040.6023 ± 0.00070.832 ± 0.013[17]

Observational Data and Description

On July 23, 2008 at 14 h 45 min UT, 28 min after sunset, an extremely bright slow-moving fireball and its spectacular dust trail were witnessed by numerous casual witnesses in a widespread region of Tajikistan (Konovalova et al. 2010). The most distant known observations of this event in Tajikistan are from about 300 km from the epicenter. The fireball was also seen probably in parts of Uzbekistan, but we received no reports from this country. Evening twilight was still light and the sky in the area of the fireball was completely clear. The intensity of the flash was so high that, at about 100 km from the epicenter, it was noticed through the windows by people located inside buildings. The majority of the fortunate eyewitnesses agree that suddenly the fireball became extremely bright. Some observers reported an intense rumbling sound, similar to thunder, just after the fireball appearance and it was audible for several seconds. The sonic boom was heard as far as 100 km away from the burst location. It was very strong in the area of about 30 km around the ground projection of the burst location. It is now a well-established fact that sonic booms are generated by the hypersonic flight of large meteoroids in the atmosphere and, probably, also by meteoroid fragmentation at the height of the brightest flare (Ens et al. 2012). Many witnesses watched the resulting thick dust trail, which was perceptible for about 20 min and was later distorted by atmospheric winds. Initially, the fireball trail was very bright, exhibiting a blue-white color that turned into orange-red by the end of the event (see Fig. 1). Several distinguishable strong knots in this persistent train, probably produced by successive meteoroid fragmentations, are similar to dust trails of other fireballs observed in a similar manner, and suggested to us, from the very beginning of this study, a meteoritic explanation.

Figure 1.

Selected images of the persistent trail. a), g) Photographs obtained from 82 m-d. (Dushanbe) by A. Yusupov. b) Photograph obtained from Zarafshon (Dushanbe) by N. Dorosheva. c) Photograph obtained from Hissar by M.Z. Hudonazarov. d, f) Photographs obtained from Vose by M. Ahmetzjanov. e) Selected video frame obtained from 91 m-d. (Dushanbe) by A. Alimov.

The bolide was bright enough to be recorded by a visible-light satellite system, and video- and photocameras. The satellite observational system (U.S. Department of Defense satellites [Tagliaferri et al. 1994]) is able, optically, to detect bright light flashes in the atmosphere, which are caused by superbolide entries. The visible-light satellite sensors of this system are the only source of data on the fireball radiation and provided the irradiated energy for the brightest part of the fireball light curve (see Fig. 2). From these data, the total radiated flash energy was 1.87 × 1011 J (6000 K blackbody model) (Brown 2008). Based upon the visible-light intensity-time signature, the peak brightness of the strong flare was determined to be approximately 3.6 × 1010 watts steradian−1 (Brown 2008). Using the conversion factor of Ceplecha et al. (1998) for a 6000 K black body model, a value of 3.6 × 1010 watts steradian−1 corresponds to an equivalent peak absolute visual magnitude of −20.3 for the July 23, 2008 fireball.

Figure 2.

Satellite optical light curve of the July 23, 2008 fireball. One count on the vertical axis corresponds to 2.35 × 108 watts/steradian (Brown 2008). The following features are highlighted on the light curve and also on the dust trail of the fireball (see Fig. 4a): B = beginning measured point of the fireball dust trail; 1F = the main flare; 2F, 3F = the second and third flares; 1D, 2D = the first and second depressions of brightness.

Infrasound signals associated with the fireball were captured by two CTBT stations, IS31 (Kazachstan) and IS46 RU (Russia), of the global International Monitoring System (IMS) at distances up to 1530 km and 2130 km. According to the infrasound data, the total energy of the event yield based on the average period between the two stations is 0.19 kT of TNT (0.09 kT for IS46 RU and 0.34 kT for IS31 KZ) (Popova et al. 2011).

In addition, we undertook a careful investigation of available records of all seismic stations of the geophysical Survey of Tajikistan to search for possible signals from the meteoroid disruption blast. Good seismic signals of the fireball were found in the recordings obtained by four digital and one analogical seismic stations ranging from the immediate vicinity of the fireball trajectory up to a distance of 210 km (Table 2). The closest analog seismic station “Hissar” was less than 50 km away from the epicenter of the meteoroid explosion. The seismic records from two digital stations are shown in Fig. 3.

Table 2. List of the seismic stations that detected the July 23, 2008 fireball.
NameLongitude (°E)Latitude (°N)Altitude (m)Arrival times (UT)
Figure 3.

Seismic records of the July 23, 2008 fireball obtained from a) Garm and b) Gezan stations.

Seismic waves, which represent transformed sonic booms, were recorded 1 min and more after the fireball (Konovalova et al. 2011). Therefore, the time of the fireball passage is reliable, as it was derived from all these observations. In this paper, only the preliminary results of the seismic data obtained for the July 23, 2008 fireball are presented. A more detailed analysis of the seismic detection of this bolide will be made in a future paper.

One day after the event, we interviewed the inhabitants and a few witnesses furnished numerous photographs of the dust trail. Two of these, which exhibited clear references for being calibrated (see Figs. 4a and 5a), were taken from different locations immediately after the flight of fireball. We also obtained tens of eyewitness reports, but unfortunately the accounts were of insufficient quality to provide additional insight on the bolide trajectory. The fireball dust trail was, in fact, photographed only during the later stages of the fireball's trajectory path. The bolide followed a trajectory with a low zenith angle, and consequently fell almost vertically. The terminal point of the fireball was seen close to the western horizon from the town of Dushanbe and station HisAO.

Figure 4.

a) The trail of the July 23, 2008 fireball from HisAO (Tajikistan). The pointers the next features on the dust trail of the fireball are shown: B = beginning measured point; 1F = the main flare; 2F, 3F = the second and third flares; 1D, 2D = the first and second depressions of brightness. The photograph by U. Hamroev was taken at 14:45 UT. b) One of the calibration pictures containing stars from the constellation of Canes Venatici.

Figure 5.

a) The trail of the July 23, 2008 fireball from Dushanbe. The photograph by N. Dorosheva was taken at 14:47 UT. b) One of the calibration pictures containing stars from the constellations of Bootes and Canes Venatici.

Data Processing and Analysis

Fortunately, two casual witnesses separated by a distance of 11.3 km were alert enough to capture the dust trail immediately after the flight of the fireball with photocameras. These two records provided an exceptional opportunity to carry out a detailed study of the event. So, the situation was much better than if only visual sightings were available as the primary sources for determining the fireball trajectory. Fortunately, the apparent position of the fireball during the last part of the recording was low enough to include some terrestrial objects in the images (TV aerials of some close buildings, roofs, and also electric columns and electric wires). These could be used as the fiducial objects to derive the horizontal coordinates of the bolide. In a first step, we interviewed the photographers at the places where the pictures were taken, determining the exact position of the cameramen. Then, to calibrate original photorecording, the nighttime stellar calibrations of the same field were made from the location of the original photorecording. These images included the above-mentioned terrestrial objects (see Figs. 4b and 5b). These stellar calibrations were then digitally mapped to the original photofield. To achieve a reliable calibration, the position of the photocameras at the time of the fireball had to be known with good precision. Fortunately, this could be done, as one of the casual photographers was taking the picture through his apartment window, and the other located for us his position very precisely.

The photographs provided enough comparison stars in the field of the fireball dust trail to perform a good astrometry. Stellar calibration was necessary and possible because the landmarks were not far from the observer. Each photograph was calibrated separately using the reference points seen on the image. After a careful calibration, the rectangular coordinates x, y of several points including any feature points (beginning, terminal, and all flares and depressions) of the fireball dust trail were measured on both photographs by using a Zeiss-Ascorecord device. Although the dust trail within the first 1–1.5 min was already distorted, we assumed that the axis line of fireball trail did not have a large transverse displacement during this time. The fireball trajectory was assumed to be straight. This assumption is justified because any curvature is indistinguishable within the precision of the data. Approximately 20–30 stars on each photograph (their apparent positions) were used for the definition of the conversion of x, y into a, z, the azimuth and zenith distance.

For each measured point along the trajectory, including the beginning, maximal brightness, middle point, ending, as well as three flares and two depressions of brightness of the fireball, we could compute the height above the sea level, the distance from the places where the pictures were taken, and the length along the fireball trajectory. The first step to achieve all this was the computation of right ascension and declination (as well as azimuth and zenith distance) of these individual points of the fireball trail on both photographs. We used stars as fiducial points in this respect. The astrometric calibration of the fireball apparent trajectory in reference to the stars was made by following the standard procedure and method described in Babadzhanov and Kramer (1963) by using the METEOR software package developed by the Meteor Department (Institute of Astrophysics of AS, Tajikistan). This software allows us to calculate the equatorial (right ascension and declination) and horizontal (azimuth and zenith distance) coordinates of each measured point of the fireball trail and the standard deviation for each measured point from the computed great meteor circle (Table 3a and 3b).

Table 3a. Data for each measured point along the fireball dust trail at the station A (HisAO): n is the number of measured points; RA (right ascension), DEC (declination), A (+astronomical azimuth), and Z (zenith distance) are the coordinates of any measured point of fireball trail; σ is the standard deviation for each measured point from the computed great meteor circle.
n RA (°)DEC (°)σ (arcsec)A (°)Z (°)
Table 3b. Data for each measured point along the fireball trajectory at the station B (Dushanbe): n is the number of measured points; RA (right ascension), DEC (declination), A (astronomical azimuth), and Z (zenith distance) are the coordinates of any measured point of the fireball trail; σ is the standard deviation for the each measured point on the computed great meteor circle.
n RA (°)DEC (°)σ (arcsec)A (°)Z (°)

The maximum accuracy, which expresses the internal precision among the set of reference stars on both photographs that can be reached, is of approximately 3.4 and 2.9 arcmin, respectively. The standard deviation for all the measured points on both computed great meteor circles was estimated as σ = ±9.6 arcsec and σ = ±14.4 arcsec accordingly. The results are consistent as demonstrated by maximal deviations of the computed great circle of the apparent trail of the bolide and confirmed by the absence of large errors in the measurements of both photographs. We have also estimated the average systematic difference, which can be connected with a probable error in the overlapping of calibration and original images containing the bolide trail. For this purpose, the compass measurements of foreground fiducial objects, which were at about 200–300 m away from the location of the original photo recording, were made. Comparing the measured and calculated azimuths and zenith distances of these objects, we obtained the average systematic difference between the original and calibrated photograph a less than of 0.4° in azimuth and less than 0.3° in zenith distance and taken to be representative of the random internal error between the absolute values for individual measurements. Finally, by using the standard procedure given in Babadzhanov and Kramer 1963), the length and height of each measured point and the fireball's radiant were calculated.

The exact duration of the fireball was known from the data of a visible-light satellite system, which provided the brightest part of the bolide. Figure 2 (intensity versus time) in (Brown 2008) was used to determine the time marks for each feature point on the image of the fireball trail. Three distinct intensity features as strong flares and two as depression of brightness of fireball are really apparent in the satellite light curve. There is an excellent correlation of these main temporal features between the space sensor and ground-based photo data of dust trail, which were captured immediately after the flight of the fireball. The photo and satellite data were combined using the 1F, 2F, 3F bursts and 1D, 2D depressions of brightness (Figs. 2 and 4a) clearly observed in the photoimage of fireball dust trail and the satellite light curve as a fiducial point. This correlation allows velocity to be estimated over intervals along the fireball trail. Here, the assumption is made that the flares and depressions observed in the satellite optical light curve correspond to the bright features and depressions along the photoimage of fireball dust trail. With such a correlation, the average velocity of the meteoroid on the intervals between intensity features was derived. The accuracy of this method depends on the accuracy of measurement of coordinates of features on the fireball dust trail and the subsequent definition of the length of intervals between neighboring features. By applying this approach, finally the average velocity across the consecutive sets of several time intervals was derived. The average velocity from first detection (38.2 km altitude) to the second measured point (peak of brightness at 35 km altitude) produces a mean velocity of 14.3 ± 0.5 km s−1. This value, as well as average velocity over other time intervals, is found by simply taking the along-trail distance derived in the photo record between these heights and dividing by the elapsed time. The random along-trail distance error is close to approximately 5% and it is a largest amount of error for each measured point.

The minimum time lapse between two neighboring features is 0.08 s, and the maximum, 0.50 s. The time interval between the first and the last measurements is of about 1.51 s. Thus, the timing enables us to determine the average velocity Vm and the deceleration in the time intervals between the neighboring features. Note that the velocities derived directly from the terminal brightest part of the recorded trajectory using distances of only about 20 km are not so accurate. Moreover, we have excluded the ending interval to provide more robust velocity estimations in the measured portion of the path. The calculated velocity Vc was determined by fitting the smoothed average velocity Vm values by means of the least squares method. Using the exponential dependence of velocity versus time, the extrapolation yields the ending velocity of approximately 8.7 km s−1 and the deceleration of approximately −6.5 km s−2 at a height of 19.6 km. By taking all uncertainties into account, the standard deviation of the velocity value is not smaller than ±0.5 km s−1. Figure 6 shows both the measured average velocity Vm and the calculated velocity Vc versus time. Table 4 shows the altitude range, average velocity across the consecutive sets of four time intervals, and the error derived from the deviation of the average measured and computed velocity. We take a mean velocity of 14.3 ± 0.5 km s−1 to be the best estimate for the initial entry velocity for the 230708 meteoroid, which is necessary for computing the heliocentric orbit.

Table 4. Average calculated velocity between the features in several time intervals.
Altitude range (km)Velocity (km s−1)
38.2–35.013.8 ± 0.5
38.2–28.813.2 ± 0.4
35.0–28.812.8 ± 0.6
35.0–27.711.8 ± 0.4
Figure 6.

Velocity of the bolide as a function of time. Black squares correspond to the measured average velocity Vm; line corresponds to the calculated velocity Vc.

Results and Discussion

After a careful calibration, each photograph provided the celestial coordinates (azimuth and zenith distance) of the fireball as seen from the different locations as a function of time. From the data obtained from the double-station photos of the dust trail, combined with satellite fireball light curve, the astrometric measurements were able to determine the fireball atmospheric trajectory, radiant, preatmospheric velocity, and orbit. The fireball's dust trail was first produced at a height Hb of 38.2 ± 0.5 km, when the measured velocity Vb was 14.3 ± 0.5 km s−1. The fireball continued a 19 km luminous trajectory and finished its luminous phase at a low altitude He of 19.6 ± 0.5 km, where the fireball decelerated to 8.7 ± 0.5 km s−1. This height is a common barrier for meteorite-dropping fireballs. The slope of the trajectory was extremely steep—the zenith distance of the radiant was only of about 10° and the difference between the beginning and the terminal height was 18.6 km.

The brightest flare occurred near the beginning of the recorded trajectory at a height Hmax of approximately 35.0 km, where the first breakup must have occurred under an aerodynamic pressure Pdyn of about 1.6 MPa. Although the details may vary, there is a consensus that a 35 km explosion height is typical for meteorite-producing bolides (Foschini 1998; Brown et al. 2000). The second small flare of the fireball occurred at a height of 28 km, where the aerodynamic pressure was equal to 3.8 MPa. The maximal aerodynamic pressure was 4.6 MPa when the body experienced the third final flare. The apparent radiant was in Bootes. The resulting data on the atmospheric trajectory and radiant coordinates are given in Tables 5 and 6.

Table 5. Parameters of the atmospheric trajectory of the bolide.
 BeginningMax. lightTerminal
Velocity (km s−1)14.1 ± 0.513.4 ± 0.58.7 ± 0.5
Height (km)38.2 ± 0.535.0 ± 0.519.6 ± 0.5
Absolute magnitude−20.3
Table 6. Radiant data (J2000.0).
Right ascension (°)221.3 ± 2.1216.2 ± 2.5
Declination (°)32.4 ± 2.130.75 ± 2.56
Initial velocity (km s−1)14.3 ± 0.59.0 ± 0.836.5 ± 0.7

From the calculated radiant and preatmospheric velocity (V = 14.3 km s−1), the heliocentric orbit of the meteoroid was computed by the standard procedure by using the software developed by the Spanish Fireball Network (SPMN) (Madiedo et al. 2011). The derived result suggests that the meteoroid had the orbit with aphelion distance in the Main Belt of asteroids (see Fig. 7) and was of asteroidal origin. The exact time of the fireball's passage is also important for this calculation. The best information comes from the visible-light satellite system, which registered the brightest flare of the fireball at 14:45:25 UT. The resulting elements of the heliocentric orbit for the equinox 2000.0 are given in Table 7. A sufficiently good trajectory and orbit have been derived from the images of the dust trail, although the precision would have been better if direct images of the fireball were available.

Table 7. Computed orbital data (J2000.0).
Orbital elements 
Semimajor axis (AU)2.13 ± 0.25
Eccentricity0.523 ± 0.057
Perihelion distance (AU)1.015 ± 0.001
Aphelion distance (AU)3.24 ± 0.51
Argument of perihelion (°)175.3 ± 1.6
Ascending node (°)120.9882 ± 10−4
Inclination (°)9.57 ± 0.89
Figure 7.

Orbit of the superbolide of July 23, 2008 projected on the ecliptic plane.

As pointed out above, in addition to the satellite data, the Tajikistan bolide was measured with ground-based infrasonic and seismic instruments. These additional data sources provided estimates of the energy released and the preatmospheric mass of the object. Thus, the data of optical detection of the July 23, 2008 fireball by satellite-based light sensors were used in Popova et al. (2011) to estimate the initial kinetic energy of the meteoroid. As a result, an initial kinetic energy of about 0.53 kT TNT was derived, by assuming an integral luminous efficiency of 9.3% based on the calibration performed with infrasound data. Theoretical estimates of the integral luminous efficiency yield an initial kinetic energy of about 0.59 kT TNT (Popova et al. 2011). This initial kinetic energy of the meteoroid (0.53–0.59 kT TNT), coupled with the computed initial velocity 14.3 km s−1, yields an equivalent preatmospheric mass of 20–25 tons.

Analysis of the Dark Flight

The dark flight was simulated by using the standard procedure described in Ceplecha (1987). The deceleration at the terminal point of the trajectory obtained from the estimated values of fireball velocity versus height was −6.5 km s−2, with a trajectory inclination of about 80°. The simulation was performed by taking into account the modeled atmospheric conditions provided by the British Atmospheric Data Center (Swinbank and O'Neill 1994) with a software package developed by the Spanish Meteor Network (SPMN) (Madiedo et al. 2011), which uses a standard Runge–Kutta calculation procedure. Spherical shape was assumed for the particle and a value of the drag factor at the terminal point of 0.58 was used. Under these conditions, the terminal mass of the meteoroid, which depends on its density, would vary from 34 ± 9 kg (d = 3700 kg m−3) to 82 ± 9 kg (d = 2200 kg m−3). Most of the region around the predictedstop impact area (38.358°N, 68.074°E) is covered by agricultural fields, without stones or rocks that would complicate meteorite searches, as the soil consists basically of clay. Thus, the favorable circumstances of an almost vertical fireball trajectory and favorable countryside give hope for successful search for the meteorite. At present, the search of additional and more detailed data is in progress and expeditions to the impact area will be organized to try to find the meteorite.


Our study of the July 23, 2008 superbolide shows that the meteoroid, with a kinetic energy of the order of 0.53–0.59 kt (Popova et al. 2011), entered the Earth's atmosphere with an initial velocity of about 14.3 km s−1. This yields an equivalent preatmospheric mass of 20–25 tons and thus the meteoroid diameter would be of about 2.3 ± 0.1 m assuming a spherical shape and the typical density of ordinary chondrites (3700 kg m−3) (Britt and Consolmagno 2003; Macke et al. 2011). As the meteoroid penetrated deeply into the atmosphere, the aerodynamic pressure increased progressively on the front part of the body. At a height of 35 km, the pressure on the surface of the particle reached 1.6 MPa, causing its breakup. This agrees with the results derived by Popova et al. (2011), who determined that this meteoroid was progressively fragmented at altitudes of about 45–25 km. Among the relatively few documented cases of collisions with bodies of such a large mass, this is a remarkable event because the meteoroid penetrated deeply into the atmosphere and probably dropped a meteorite. We hope to have the chance to recover meteorites from this event, which occurred over an inhabited region.

To obtain an accurate orbit, high-quality multistation sequential images of the fireball are needed. Nevertheless, for the July 23, 2008 bolide studied here, despite there being available just double-station recordings of the dust trail together with the satellite data, we have demonstrated that is possible to obtain reliable results. Trajectory and orbital data are accurate enough, although this accuracy is lower than that obtained for the best measured meteorite-producing fireballs with available optical observations. However, our derived optical trajectory is far more exact than those events analyzed on the basis of eyewitness data only. The multi-instrumental aspect of the collected data has enabled us to study in detail the passage of a meter-sized meteoroid through the Earth's atmosphere. In conclusion, we think that the present case deserves a future ground expedition once trajectory optical data will be combined with the interpretation of the available seismic data that are being reduced. The results of this study will be published in a second paper. In conclusion, we think that this study exemplifies not only the intrinsic difficulties in reconstructing accurately fireball trajectories but also the way to combine results from different techniques to obtain valuable orbital information. By using these new clues, we will continue learning about the origin of meteorite-dropping bolides reaching our planet.


The authors are grateful to all witnesses of the fireball of July 23, 2008 who volunteered their observations, and especially to those who offered photographs of the bolide. The second author acknowledges support from the Spanish Ministry of Science (project AYA2009-13227) and Junta de Andalucia (project P09-FQM-4555). JMTR thanks Spanish Ministry of Science and Consejo Superior de Investigaciones Científicas for their respective grants AYA2011-26522 and CSIC#201050I043. We are very grateful to Dr. J. Borovička and to an anonymous reviewer for the constructive and careful review of our manuscript and comments, which helped to improve the paper significantly.

Editorial Handling

Dr. Donald Brownlee