Tagish Lake

standby for tagish lake photoC2-ungrouped
Fell January 18, 2000
59° 42′ 15.7′ N., 134° 12′ 4.9′ E. Several hundred thousand fragments of this unique carbonaceous chondrite fell at 8:43:43 A.M. onto ice-covered Tagish Lake, located between Atlin, British Columbia and Carcross, Yukon Territory. The luminous meteor (magnitude ~–22), estimated to have been ~4 m in diameter and to have a pre-atmospheric weight of 56,000 kg (Brown et al., 2002), approached at a velocity of ~16 kilometers per second at an angle of 17.8°. Fragmentation models indicate that the object first exploded at an altitude of 34.4 km with an energy equivalent to 1.7 kT of TNT, which caused 88% fragmentation of the object; a total of 33 fragmentation events are thought to have occurred during descent (Ceplecha, 2007). At a height of 29.2 km, the 2,660 kg main mass entered the dark phase of its flight, its velocity now at 13.1 km/second. From models based on a porosity of 37–58% (ave. 40%, Hildebrand et al., 2006), it is estimated that ~1,300 kg of appreciable fragments fell to the ground, corresponding to a total ablation loss of over 97%.



Sequence captured by Ewald Lemke (Atlin, British Columbia) on January 18, 2000.
standby for tagish lake orbit photo
Tagish Lake orbit. Image Credit: University of Calgary

The multiple explosions and accompanying dust cloud were seen and heard by eyewitnesses, while a few others heard hissing associated with electrophonic sound phenomena, or noticed a metallic or sulfurous odor (Hildebrand et al., 2006). The fall was detected by seismic stations and U.S. Department of Defense satellites, providing data that has helped ascertain for the first time an accurate orbit (right) for a carbonaceous chondrite. The meteoroid moved in a direction of 151.5° ± 2° with an elevation above the horizon of 17.8°. After calculating an orbit for this meteoroid, the aphelion was found to lie in the outer asteroid belt, possibly associated with the Apollo asteroid group—in particular, the low-albedo D-type asteroids and metamorphically related asteroids. There is a high probability that transfer to an Earth-crossing orbit occurred by means of the ν6 secular resonance. Excellent pictures of the event taken within two minutes of the explosion and spanning an extended time afterwards can be seen to the left. It shows the expanding smoke train of the Yukon meteor over a 14-minute period. The first frame shows a smoky red vapor trail just 1 minute and 30 seconds after the initial flash.




It was fortuitous that local resident and pilot, Jim L. Brook, had been previously briefed by meteoriticists from the University of Western Ontario on how best to collect meteoritic dust samples. At ~4:00 P.M. on January 25, 2000, while driving on the ice of the Taku Arm of Tagish Lake, Brook spotted black fragments on and within the snow and recognized them as meteorites. Over the next two days he collected several dozen fragments and fine particles totaling ~870 g, which were placed into clean Ziploc freezer baggies and kept frozen until they could be turned over to the scientists. Between April 20 and May 8, 2000, after the heavy snow cover had dissipated, search parties from the University of Calgary and The University of Western Ontario utilizing ATVs and snowmobiles identified and marked 412 additional fragment sites in situ, some situated deeply within the ice. Approximately 200 of these fragments were eventually collected before the ice lost its structural integrity. A few more pieces were recovered by local residents during the summer.

The photos below show various Tagish Lake fragments as they were found embedded in the ice. All photos courtesy of The University of Western Ontario and the University of Calgary.








standby for tagish lake strewnfield photo
Tagish Lake strewnfield. Image Credit: University of Calgary

The Tagish Lake strewnfield (left) covers an area of at least 16 × 4 km. The total recovered weight of this lightweight, friable meteorite is 5 – 10 kg, but only a small fraction of that (~820 g) is being curated at the University of Alberta preserved in a frozen pristine state (Ralchenko et al., 2014).

Tagish Lake is among the most chemically primitive meteorites known, with a significantly higher carbonate abundance than any other carbonaceous chondrite. It has undergone the most pervasive aqueous alteration of any C2 chondrite studied, and contains water and carbon structurally bound in hydrated minerals, adsorbed onto mineral surfaces, and absorbed within the layers of smectite-group phyllosilicate clays, likely zones of prebiotic organic synthesis (Garvie and Buseck, 2007). Bulk density and porosity measurements of numerous Tagish Lake fragments reveal that the meteoroid had both a lower bulk density and a higher porosity than any other meteoroid similarly analyzed, with values very close to those of IDPs (Hildebrand et al., 2006). The composition of Tagish Lake can be separated into two isotopically distinct lithologies—a dominant carbonate-poor lithology, and a subordinate carbonate-rich lithology (Zolensky et al., 2002).


The carbonate-poor lithology of Tagish Lake is composed of phyllosilicate-rich clasts (comprising µm- to sub-µm-sized mineral fragments) together with interspersed, aqueously altered, rimmed chondrules (mainly porphyritic and barred olivine types) enclosing small FeNi-metal inclusions. Other constituents present include anhydrous forsteritic olivine grains (mainly Fo99), lithic fragments of mainly enstatite composition, framboidal magnetite, and sulfides (primarily pyrrhotite and pentlandite), along with minor phosphides, chromite, spinel-rich spherules, and other rare refractory inclusions.


Many of these constituents are enclosed by fine-grained, low-porosity rims composed of an unequilibrated assemblage of phyllosilicates, FeNi-sulfides, FeNi-metal, magnetites, low-Ca pyroxenes, and forsteritic olivines. It was demonstrated by Greshake et al. (2005) that these rims were accreted in the solar nebula. However, Takayama and Tomeoka (2008) have invoked a parent body aqueous alteration process as the formation mechanism of these rims.


On a nm- to µm-scale, hollow, membrane-like globules composed of pristine amorphous carbon are present, attesting to a persistently cold environment since their formation before or during the formation of the Solar System (Nakamura et al., 2003). All of these components are contained within a dense, fine-grained matrix of mostly Mg-rich saponite and serpentine (likely greenalite), along with Fe–Ni sulfides. Only rare carbonates are present within this matrix in the form of polycrystalline calcium carbonate grains. Micropores are partially lined by these carbonates.

In the proposed alteration sequence of Greshake et al. (2005), which occurred on the parent body at very low temperatures (<100°C), olivine and metal were replaced by phyllosilicates and pyrrhotites. Phyllosilicates also replaced Ca-rich phases in the rimmed objects, leading to the transport of Ca out of these objects by aqueous fluids and into the matrix component where carbonates were then precipitated. Subsequent oxidation conditions resulted in the replacement of some pyrrhotite by magnetite. This was followed by the replacement of magnetite by the precipitation of fine-grained sulfides and the growth of carbonates. Thereafter, sulfur was incorporated into organic phases.

An alternative formation sequence was proposed by Takayama and Tomeoka (2012) to have occurred on the parent body. They discovered some unique clasts containing coarse-grained components lacking rims, along with attached matrix material, all of which have similar bulk chemical compositions, textures, and mineralogy. Based on their studies, they concluded that the present rim material that partially surrounds chondrules and coarse-grained components is actually the remnant of a first-generation of matrix material from an earlier assemblage. They argue that, following a period of brecciation, portions of this original assemblage (of chondrules, coarse-grained components, and matrix) were transported to another region and incorporated into a second-generation host matrix having a different bulk chemical composition, texture, and mineralogy. CARBONATE-RICH


The carbonate-rich lithology is composed of very fine-grained phyllosilicates of mostly saponite, with only a very limited amount of fine-grained clasts, CAIs, and magnetite present. Whitish Fe–Mg–Ca–Mn carbonate grains are very abundant, and include calcite, dolomite, and breunnerite (formed in that sequence), while very little calcium carbonate occurs independently. The Fe–Mg–Mn carbonates, called siderite, likely replaced existing Ca carbonate grains through the percolation of fluids following impact fracturing. Micropores are completely lined by these carbonates. The carbonates display a wide range of compositions, some unknown from any other CI and CM chondrites, and exhibit heavy-oxygen isotope enrichment consistent with a high water/rock ratio (~2) more similar to CI than CM (~0.6) chondrites. Still, in comparison to the CI and CM groups, different components in Tagish Lake are not in isotopic equilibrium.

In the proposed alteration sequence, which occurred on the parent body at higher temperatures (>300°C) than for the carbonate poor lithology, olivine and metal were replaced by phyllosilicates and magnetite, which were subsequently replaced by Ca carbonate. Following impact fracturing, siderite replaced some Ca carbonates, and coarse-grained sulfides were deposited. The estimated water/rock ratio during this phase is ~2. It may be assumed that the carbonate-rich lithology succeeded the carbonate-poor lithology.

Tagish Lake is a heterogeneous accretionary breccia that likely contains additional lithologies than the two outlined above. Newly identified probable lithologies include a carbonate-rich lithology in which siderite dominates over calcite, and an inclusion-poor lithology low in saponite-serpentine (with a corresponding presence of gypsum) and enriched in magnetite and sulfide (Izawa et al., 2010).

X-ray diffraction techniques and Mössbauer spectroscopy have been used by Bland et al. (2004) to determine the modal mineralogy of several carbonaceous chondrites, including Tagish Lake, and to quantify the compositional range of the olivine phases (here its pure forsterite). Ralchenko et al. (2014) determined the bulk and grain density and the porosity of Tagish Lake samples that had been maintained in their frozen pristine state. They utilized 3-D laser imaging (as opposed to the modified Archimedean glass bead method) and helium pycnometry to determine the bulk volume and grain volume, respectively, and from these values they derived the porosity. The modal mineralogy (vol%) and other physical properties of Tagish Lake are as follows:

    • Olivine (forsterite, Fo100) — 7.0
    • Fe-Mg carbonate ————– 8.0
    • Pyrrhotite ———————— 5.3
    • Pentlandite ———————– 0.3
    • Magnetite ———————— 4.5
    • Saponite–serpentine ——— 71.2
    • TOTAL ———————— 100.0


  • bulk density = 1.8 (±0.03) g/cm³
  • grain density = 2.56 g/cm³
  • porosity = 30 vol%

Tagish Lake has a high bulk carbon content of ~5.8 wt%, higher than CI chondrites and much higher than CM chondrites, with ~2.6 wt% of this carbon incorporated in organic components. The water-soluble organic component, which comprises only 0.01 vol% (much less than in CM chondrites), consists of mostly monocarboxylic and dicarboxylic acids, the former including the straight chain compound formic acid as has been found in the C2 chondrite EET 96029; all of these compounds could have formed by parent body or nebular processes. The remainder of the organic component (>99%) is present as insoluble C, which is predominantly composed of two high-molecular weight PAHs having an aromatic structure. This composition is in contrast to the mostly aliphatic-structured compounds of higher molecular weight found in CM chondrites. The insoluble organic matter in Tagish Lake contains a high proportion of diradicaloids among the aromatic moieties, which is considered to be a distinguishing characteristic of an extraterrestrial organic source (Binet et al., 2004). Since the most highly altered CI and CM group members also contain these same diradicaloids, this organic component must have been synthesized in the presolar nebular disk, and further discussions presented by Alexander et al. (2007) and by Quirico et al. (2012) suggest a possible origin in the interstellar medium.

All of the organic compounds present in Tagish Lake, including nitriles with low H content, are consistent with a low degree of aqueous processing following accretion of large interstellar molecules. The lower abundance of heavier organics in Tagish Lake compared to those in Murchison is evidence for lower alteration temperatures and lower degrees of chemical evolution on this asteroid. Consistent with these facts is the discovery of sub-µm-sized hollow spheres (‘nanoglobules’) first identified within the phyllosilicate matrix of Tagish Lake, which are composed of amorphous carbon that contains isotopically anomalous N and H. This component, perhaps associated with sheet silicates, is likely the major carrier of the isotopic anomalies (Zega et al., 2010). Abundances of anomalous N and H in this amorphous component in Tagish Lake are similar to those observed in interplanetary dust particles (IDPs), attesting to a primitive origin from icy dust particles. The presence of these nanoglobules infers that they formed in a cold molecular cloud or in the outer parts of the protosolar disk, or alternatively, they accreted onto the Tagish Lake asteroid and were stored in a very cold environment (–253° to –263°C) over its entire history (see the PSRD article below). Notably, N has been found to be similarly incorporated as aromatic nitriles in Tagish Lake organic matter, in IDPs, and in cometary dust particles collected from comet 81P/Wild 2 (Clemett et al., 2010).

The anhydrous component of Tagish Lake was studied and compared to that of the CM and CI chondrites (Simon and Grossman, 2003). Tagish Lake contains the silicate andradite, which is absent from CM meteorites, but it contains no tochilinite, which is abundant in CM meteorites. Tagish Lake contains primitive refractory forsterite grains, rare Fo-rich chondrules, and isolated olivine grains, the latter probably reflecting crystallization within host chondrules and their subsequent fragmentation (Russell et al., 2010). This anhydrous component in Tagish Lake (~13 wt%) is much lower than that in CM chondrites (ave. 48 wt%), but higher than that in CI chondrites. Other features that were identified in both Tagish Lake and the CM group, but not in CI members, include hibonite-bearing and spinel-rich refractory inclusions, CAIs, and Cr- and Al-rich spinel. Furthermore, thick, fine-grained, accretionary dust mantles surrounding clasts and inclusions are found in Tagish Lake and the CM group, but have not been identified in any CI meteorites.

Phyllosilicates in Tagish Lake matrix and rims are predominantly Mg-rich saponite in contrast to the Fe-rich serpentines found in the CM group. This Mg-rich composition is likely due to the transfer of phyllosilicate Fe into magnetite. Notably, the unique chondrites MET 00432 (CM2), WIS 91600 (C2), and Bells (C2) have similar mineralogies to those of Tagish Lake (e.g., saponite and magnetite abundances), as well as similarities in O-isotopic composition and reflectance spectra; however, these meteorites have features suggesting that they experienced different degrees of aqueous alteration. It has been conjectured that these four carbonaceous chondrites may constitute a grouplet (Nakamura et al., 2009, 2013). In addition, the heated CM chondrite PCA 91008 shows similarities to this potential grouplet. In their organic chemistry analyses, Yabuta et al. (2010) found that Tagish Lake and WIS 91600 have similar IOM contents and O-isotopic compositions, as well as certain spectral similarities, and it was suggested that these two meteorites may be genetically related. WIS 91600 shows evidence supporting a higher degree of aqueous alteration than in Tagish Lake, and unlike Tagish Lake it experienced short duration impact heating to <500°C.

Interestingly, Vernazza et al. (2010) report that both WIS 91600 and Tagish Lake have spectra that match the blue spectra of the bright Stickney crater on Mars’ moon Phobos, but which do not match the remaining space-weathered (more red spectrum) regions of that moon. They raise the hypothesis that WIS 91600 and/or Tagish Lake may originate from the deeper and fresher material of that captured moon. Mid-infrared spectra of the martian moon Phobos acquired by the Mars Global Surveyor spacecraft indicate that both carbonates and desiccated phyllosilicates are present on its surface, consistent with the mineralogy of D-type asteroids and the Tagish Lake meteorite (Glotch et al., 2015).

Sulfides are present in greater abundance in Tagish Lake than in either the CI or CM group, although the range in sulfide phases is more similar to that in CM meteorites; still, a rosette morphology is found exclusively in Tagish Lake. The lower ratio of pyrrhotite to intermediate sulfides in Tagish Lake reflects a higher degree of aqueous alteration than that present in CM chondrites. In further contrast, no sulfate salts occur in Tagish Lake as they commonly do in CI chondrites. The magnetite abundance in Tagish Lake is similar to that in CI meteorites, but much higher than in CM members. The nitrogen content of Tagish Lake was determined to be 0.12 wt%.

In contradiction to the many CM-like primary features, Tagish Lake exhibits evidence of an alteration history more similar to that of the CI group (Russell et al., 2010). Both the bulk rock and the matrix O-isotopic composition of Tagish Lake are 16O-poor, with a plot located very close to the CI field following the Orgueil trend line above the TFL. The phyllosilicate O- and H-isotopic compositions of Tagish Lake are also similar to those of CI members, indicating similarities in both water composition and alteration temperature. Still, the anhydrous minerals O-isotopic composition range overlaps that of all carbonaceous chondrite groups. Moreover, the chondrule O-isotopic composition is most similar to the CM and CV groups, and shows a positive correlation between 16O and Fo contents. Beyond that, the bulk H-isotopic composition of Tagish Lake is more like that of CR chondrites, with a much higher deuterium content than that measured in either CI or CM group members.

Neon- and N-isotopic studies of Tagish Lake indicate that interstellar grains (nanodiamonds, SiC, graphite) are present in greater abundance than in other C2 chondrites. Isotopically anomalous Os has been identified in the carrier phase SiC in Tagish Lake, the first such finding ever reported (Humayun et al., 2005). This Os consists of a greater proportion of r-process (nucleosynthesis by rapid neutron captures, as in supernova) over s-process (nucleosynthesis by slow neutron captures, as in AGB stars) osmium, and may reflect the preservation of presolar stardust that is typically heated and destroyed. In another case, primordial ratios of planetary noble gases such as helium and argon have been found encapsulated within the three-dimensional form of carbon known as fullerene.

Trace element data, especially observed in a plot of Zn/Mn vs. Sc/Mn, indicate that Tagish Lake is a carbonaceous chondrite distinct from all others. Furthermore, both Raman and Fourier Transform Infrared Spectroscopic analyses of the carbonate-rich lithology have revealed that the carbonaceous component of Tagish Lake is unique from that of other carbonaceous chondrite groups (Djouadi et al., 2003). Results of a K–Ar age study conducted on bulk matrix samples of Tagish Lake by Turrin et al. (2014) indicate an isochron age of 2.51 (±0.03) b.y., or a corrected isochron age of 2.20 (±0.04) b.y. These investigators suggest that this young age may represent the time of arrival to the inner solar system where solar warming reset the Ar-based chronometer. A cosmogenic nuclide analysis indicates that the Tagish Lake meteoroid had a relatively low CRE age of ~7.8 m.y. It experienced only low shock effects (S1) consistent with other C chondrites.

Based on mineralogy, petrology, isotopic composition, bulk chemical composition, and organic chemistry, Tagish Lake shares some similarities with, but is unique from, both the CM and CI chondrite groups. Based on Re–Os systematics and highly siderophile element concentrations, an interesting scenario has been proposed by Brandon et al. (2005) in which Tagish Lake could have originated from a region overlying CI chondrites and underlying CM chondrites on a common parent body, with aqueous alteration increasing towards the interior. Others have suggested Tagish Lake may represent CI precursor material. On the other hand, studies of the large diversity of clasts present in the Kaidun chondrite breccia led investigators (MacPherson et al., 2009) to the conclusion that a near-continuum of carbonaceous chondrite objects exists, encompassing CR, CM, CI, and other similar precursor objects that experienced unique alteration histories—Tagish Lake is seen as the first representative of a new, evolutionarily distinct C2 carbonaceous chondrite object.

Tagish Lake is spectrally most similar to the D-class asteroids, as well as to the metamorphically related T- and P-type asteroids, the first such match ever made to these asteroid classes (although the Antarctic find WIS 91600 also shares these spectrographic characteristics). The D/T/P-class asteroids are mostly located in the outer belt region, consistent with the calculated orbital aphelion for Tagish Lake of ~3.3 AU. One asteroid in particular, 308 Polyxo, exhibits almost identical visible and near-infrared reflectance spectra to that of Tagish Lake, and it also has a distinct 3-µm absorption band like that in Tagish Lake (and in WIS 91600) indicative of the presence of hydrated minerals (T. Hiroi and S. Hasegawa, 2003). However, it was shown that there are other D-class asteroids besides 308 Polyxo located even closer to one of the Kirkwood Gaps, from which an efficient transfer to Earth’s orbit could be executed. Nevertheless, the totality of the evidence raises the possibility that Tagish Lake could be derived from the D-class asteroid 308 Polyxo.

In a Cr-isotopic study of Tagish Lake, Luu et al. (2009) have found that it contains the highest 54Cr excess ever measured in a silicate fraction of a meteorite. This led them to describe Tagish Lake as the most pristine meteorite yet studied, even more so than CI chondrites, and they consider it likely that it was derived from either a comet or a D-class asteroid residing in the cold outer asteroid belt. A cometary origin for Tagish Lake has also been postulated by Bryson et al. (2018) based on paleomagnetic data, which they suggest indicates the acquisition by magnetite of a very low remanence (0.15 µT). This paleointensity is consistent with that estimated to have been present at a heliocentric distance of ~20 AU, and is similar to that measured for comet 67P. After formation, the Tagish Lake source body would have migrated to a storage orbit in the outer asteroid belt, possibly as a result of scattering during the final outward migration of the gas giants (‘Nice Model’).

Among D-class asteroids with similar reflectance spectra to Tagish Lake, 773 Irmintraud and 368 Haidea provide the best matches in the visible–near-infrared range. The fact that 773 Irmintraud is located closer to Jupiter’s 7:3 mean motion resonance than 368 Haidea is located to Jupiter’s 2:1 resonance may have a bearing on any associations. The D/T-class asteroid 308 Polyxo also has similar reflectance spectra to Tagish Lake in the visible–near-infrared range. Also noteworthy is the finding that 773 Irmintraud, 308 Polyxo, and the C-class asteroid 511 Davida all contain hydrated materials like those present in Tagish Lake. Data for this 3-µm hydration band indicate that the closest match to Tagish Lake is 511 Davida (Hiroi et al., 2003). Further studies will be necessary to differentiate between these possible asteroid candidates.

Notably, parts of the martian moons Phobos and Deimos (trailing edge and leading edge, respectively) are spectrally D-class objects. The low bulk densities calculated for these two moons are similar to that of Tagish Lake. The low densities determined for the martian moons are thought to represent macroporosity effects, calculated to be 25–45% for Phobos (Rosenblatt, 2011), which is consistent with the ~40% porosity calculated for Tagish Lake (Brown et al., 2001). In spite of this possibility, an origin for Tagish Lake in the outer asteroid belt remains a strong probability. Moreover, high-quality orbital Thermal Emission Spectrometer (TES) data obtained by the Mars Global Surveyor for Phobos demonstrate that mid-IR spectral features are inconsistent with the composition of Tagish Lake (Glotch et al., 2018). The TES data for Phobos does indicate the potential presence of a fine particulate basaltic component, which would support the hypothesis of a Mars impact origin for the two moons.

A small ‘CI1’ lithology previously found in Kaidun (a clast-rich carbonaceous chondrite) is reported to have an O-isotopic composition identical to that of Tagish Lake, and with a similar petrology as well. Besides that example, an unusual clast found in Tagish Lake is similar to another clast in Kaidun, one that was identified as a CM1. Furthermore, a number of phyllosilicate-rich micrometeorites that were recovered in Antarctica have mineralogies similar to the Tagish Lake carbonate-poor lithology. It has been argued by Izawa et al. (2010) that a continuum may exist, reflecting similarities in material components and variabilities in mineral modes, connecting the carbonaceous asteroids like Tagish Lake and comets. All of these independent clasts and micrometeorites present a valuable opportunity to increase our understanding of carbonaceous chondrite groups. The specimen of Tagish Lake shown above is a 0.24 g specimen showing the thin fusion crust, and was exported from Canada under Cultural Property Export Permit #65152, August 3, 2000.

For additional information on Tagish Lake, read the PSRD article by David W. Mittlefehldt: ‘Tagish Lake—A Meteorite from the Far Reaches of the Asteroid Belt‘, Dec 2002.

Another PSRD article by G. Jeffrey Taylor, ‘Organic Globules from the Cold Far Reaches of the Proto-Solar Disk‘, Jan 2007, examines the Taglish Lake meteorite and its distribution of sub-µm-sized, organic hollow globules.

Leave a Reply