Only a very few fireballs are connected with meteorite falls. A meteorite may survive its atmospheric flight and may perhaps then be found if at least part of the body is decelerated from its entry velocity down to its free-fall terminal velocity of about 100 m/s. According to the models in use (see Wetherill and ReVelle (1981) and references therein) and using the ablation coefficient derived from the Príbram, Lost City and Innisfree falls, less than 50% of the remaining mass will be ablated as soon as the velocity falls below 8 km/s.
Another destructive process which may operate against the fall of a meteorite is fragmentation of the meteoroid. The peak pressures these bodies can stand as derived from large bodies such as the recovered meteorites were found to be about 106N/m². This is much less than the crusting stress of about 108N/m² found in laboratory measurements. The reason may be in the earlier history of the meteorite parent bodies. Obviously, a very large fraction of meteoroids undergo a breakup when entering the atmosphere. This was visually observed in the past, but becomes very obvious from fireball photographs such as the Lost City images (McCrosky et al., 1971) and other meteor images (Babadzhanov, 1968)), as well as the video recordings taken of the Peekskill meteorite fall on October 9, 1992 (Brown et al., 1994).
We may then assume that a successful passage of the atmosphere will occur if the object enters at less than 23 km/s and survives to reach a velocity of less than 8 km/s without serious disruption. All of these values were derived from ordinary chondrites however and are likely to be valid only for comparable materials. Even so, the above is only true up to a certain mass. Very large objects of several meters in diameter which may cause meteorite craters are not decelerated very much (Ceplecha 1991).
After the transformation of its kinetic energy and its total deceleration, the surviving meteorite falls only by the Earth's gravitation without the emission of light, a period known as dark flight. During this phase the wind's force and direction have an important influence, and the surviving body may drift away from its original course. The effect can be considerably because both the free-falling meteorite and the wind have comparable velocities. Furthermore the dark flight trajectory is affected by the shape of the meteorite.
All the parameters characterizing a bright fireball -- its entry velocity, initial mass, angle of entry, the final height of its luminous trail and its brightness -- are put together in the PE-criterion (Ceplecha and McCrosky, 1976) or similar criteria (Wetherill and ReVelle, 1981). The PE-criterion is an empirical definition for the meteor's end height, weighted by terms for the initial mass, initial velocity, and direction of flight with respect to the vertical. The PE-criterion is related to different types of material and thus may help decide whether a meteorite fall is probable or not. The types are:
This has been further investigated by Ceplecha (1994). The statistics behind this contains several hundred photographed meteors. The `characteristic orbits' give some idea about the source region of the meteoroid, but are not meant as an average with a known scatter; the inividual orbits may significantly differ from each other.
---------------------------------------------------------------- bulk characteristic orbit assumed Type density for 0.1 ... 1 m size composition rho a e i [g/cm³] [au] [degrees] ---------------------------------------------------------------- I 3.7 2.4 0.68 6 stony II 2.0 2.3 0.61 5 carbonaceous IIIA 0.75 2.4 0.82 4 cometary IIIAi 0.75 infinitely 0.99 any cometary IIIA[CIII] 0.75 2.7 0.67 any cometary IIIB 0.27 3.0 0.70 13 soft cometary ----------------------------------------------------------------
From photographs of the meteorite- producing fireballs of the Innisfree, Lost City, Príbram, and -- with greater uncertainty -- Hohenlangenbeck (occasionally called Salzwedel, Germany) events, end heights for the luminous path of about 21 km are found. This seems to be slightly higher in the case of the very shallow trajectory of the Peekskill meteorite (Brown et al., 1994). The last point measured on the video records corresponds to a height of 33.6 km, but this is not the end of the luminous path.
To reach the ground from this height, a meteorite needs about 3 to 4 minutes in free-fall, so there will normally be a delay of this duration between sightings of the end of the track and reports of impact noises being heard. If there is information about such impact noises, a systematic search for meteorites is recommended. Impact noises were important for the recovery of the Hohenlangenbeck meteorite which fell on 1985 November 14, the Trebbin, Germany (1988 March 1), and the Glatton, UK (1991 May 5), daylight meteorite falls for instance.
The identification of a meteorite is in many cases not indisputable. A freshly-fallen meteorite usually has a black and relatively smooth melted crust a few tenths of a millimeter thick. This crust may crumble away and be lost within a short time-scale, dependent on the material involved. Carbonaceous chondrites are especially susceptible to this effect. The interior may be very different, and unambiguous identification is only possible using analytical methods such as looking at a specially-prepared transparent thin-section of the rock, and searching for chondrules (characteristic inclusions in chondritic meteorites, these are millimeter-sized spherules of fine-grained silicate rock), or etching a cut and polished plane with acid to find the so-called Widmannstätten patterns in iron meteorites. In the case of fresh meteorite falls, an isotope analysis should be made as soon as possible after the event. During recovery searches for meteorites all meteorite-like material should be collected -- it is better to include a lot of terrestrial material than to miss a real meteorite (McSween, 1987).