post-impact heating of a crater lake
TRANSCRIPT
Post-impact heating of a crater lake
I. Gilmour1, D.W. Jolley2, J.S. Watson1, M.A. Gilmour1 and S.P. Kelley1
1. Centre for Earth, Planetary, Space and Astronomical Research, The Open University, Milton Keynes MK7 6AA, UK 2. Department of Geology & Petroleum Geology, University of Aberdeen, Aberdeen AB9 2UE, UK
Impact hydrothermal systems
• Candidate habitats for the origin and evolution of life – Mars exploration focused on cratering environments with
evidence of a long-term history of aqueous processes • Large impact events can generate a hydrothermal
system if the impact occurs on a water- or ice-rich target
• Potential significance of impact-generated hydrothermal systems as habitats dependent on – longevity – continued availability of liquid H2O and nutrients – hospitable environments for life in the form of post-impact
lakes and lacustrine sediments
Terrestrial impact crater lakes
• Numerous terrestrial impact craters contain lakes and lacustrine sediments – Research mainly focused on the paleoclimatic records
• In craters where the lake formed immediately post-impact, basal lacustrine sediments may have been altered by the impact-generated heating or hydrothermal system enabling constraints to be placed on the extent and duration of heating
Boltysh Impact Crater • 25km complex crater • Ukrainian Shield – impact on
land • Ar/Ar age 65.17 ± 0.64 Ma
(Kelley & Gurov, 2002) • Developed a lacustrine
depositional system in 600 m of accommodation space
• Pre-dates Chicxulub by a few ka (Jolley et al. 2010)
• Drilled in the 1960s & 70s • Cores lost
• 596m cored borehole west of central upli@ • >95% recovery
• 596 -‐ 582m – allochthonous impact breccia • 390m Cenzoic crater fill
• 582 – 490m cyclic, fining upwards, poorly sorted sands and sandy muds > turbidity currents
• 490 – 190m finely laminated organic rich shales • Lacustrine, abundant plant macrofossils
• ~300m abundant ostracods and gastropods (in life posi&on), interbedded with gypsum lamellae > shallow evapora&ve lake
(Ames et al., 1998). This is consistent with differences between thelevel of alteration within the Kara, Popigai, and Puchezh-Katunki im-pact structures, Russia (Table 1), where the most intensive impact-generated hydrothermal alteration took place in the craters thatformed in shallow continental shelf or intra-continental shallow ba-sins (e.g., Kara and Puchezh-Katunki) (Naumov, 2002). The differencein the intensity of hydrothermal alteration of crater-fill impactitesbetween the Haughton and Ries impact structures is notable giventheir similar size (23 and 24 km, respectively) and the fact that theyboth occurred in a continental setting. Critically, the crater-fillsuevites at the Ries are overlain by !400 m of lacustrine crater-fillsediments and sedimentation appears to have commenced immedi-ately following impact (Arp, 1995). In contrast, at Haughton, there isno evidence preserved of a crater lake immediately post-impact(Osinski and Lee, 2005). This suggests that the presence/absence ofan overlying crater lake may play a critical role in determining thelevel of hydrothermal alteration of crater-fill impactites.
2.2.2. Interior of central upliftsIn many terrestrial impact structures, erosion has removed the
superficial crater-fill impactites. This is reflected in the largenumber of sites where hydrothermal alteration has only beendocumented in the more deep-seated lithologies of central uplifts(Table 1). As noted above, central uplifts are formed during themodification stage of complex impact crater formation and theiruplifted geotherms can contribute a heat source driving hydrother-mal systems. Central uplifts form in craters >2–4 km diameter onEarth, and >5–10 km on Mars (Melosh, 1989) and are comprisedof fault-bounded blocks of coherent to brecciated bedrock com-monly with injection dykes of impact melt-bearing or melt-freeimpact breccias. Mineralization within such lithologies is typicallydiscrete and restricted to vug and vein filling cavities and alongfractures (Fig. 2b). Deep drilling of a number of structures, suchas the !35 km diameter Manson (McCarville and Crossey, 1996)and !80 km diameter Puchezh-Katunki (Naumov, 2002) impactstructures, reveal a zoned alteration assemblage with inferredhydrothermal mineral crystallization temperatures that increaseboth with depth and towards the crater center (Fig. 1).
2.2.3. Outer margin of central upliftsStructural studies of complex impact structures have shown
that the outer margins of central uplifts are often highly fractured
and faulted because they represent an interference zone where theinwards-collapsing crater walls interact with the outwards-collapsing edge of the central uplift (Kenkmann and von Dalwigk,2000; Osinski and Spray, 2005). Not surprisingly, these zones com-monly represent sites of more intense hydrothermal alteration,particularly the infilling of fractures to form vein networks (Figs. 1and 2c) (Hode et al., 2003; Osinski et al., 2005). Observations fromHaughton suggest that these outer central uplift regions are buriedunder crater-fill impact melt rocks and breccias in fresh craters.
2.2.4. Ejecta depositsImpact ejecta deposits are characteristic features of fresh im-
pact craters on Earth and other planets. Such deposits are superfi-cial in nature, typically tens of meters thick for craters <100 km indiameter and as a result of erosion are rarely preserved on Earth.An important observation is that ejecta deposits appear to be com-prised of (at least) two distinct facies or layers in many craters onthe terrestrial planets (Osinski et al., 2011). The Ries structure inGermany is an excellent example, where a patchy layer of impactmelt-bearing breccia overlies melt-free lithic breccias (BunteBreccia). Importantly, the Bunte Breccia deposits were emplacedat ambient temperatures and no evidence of hydrothermal alter-ation has been documented (Hörz, 1982). The overlying impactmelt-bearing breccias, in contrast, were emplaced at temperatures>750–900 !C (Osinski et al., 2004).
A range of ‘‘secondary’’ minerals have been documented withinthe impact melt-bearing breccias, with montmorillonite clay andzeolite minerals being the dominant assemblages (Fig. 2d). Compli-cations with the origin of these assemblages arise due to the natureof the groundmass. In particular, there is evidence for two genera-tions of hydrous silicates with an early undetermined groundmass-forming phase, potentially formed through devitrification orautometamorphism of hydrous impact glasses (Osinski et al.,2004), and later cross-cutting veins of platy montmorillonite clay(Osinski, 2005a). Some favor a hydrothermal origin for at least someof these clays (Newsom et al., 1986; Osinski, 2005a), while others,based on stable isotope studies, suggest a low-temperature origin(<20 !C) (Muttik et al., 2010). However, these lower temperaturesare based on studies of bulk samples so that it is unclear what gen-eration of hydrous silicates were analyzed and/or whether this rep-resents a modern-day overprint of an originally higher temperatureassemblage. Regardless of these complications, it appears that the
Fig. 1. Distribution of impact-generated hydrothermal alteration deposits within and around a typical complex impact crater. The six settings are highlighted and numberedin the order in which they are discussed in the text.
G.R. Osinski et al. / Icarus xxx (2012) xxx–xxx 5
Please cite this article in press as: Osinski, G.R., et al. Impact-generated hydrothermal systems on Earth and Mars. Icarus (2012), http://dx.doi.org/10.1016/j.icarus.2012.08.030
Distribu&on of impact-‐generated hydrothermal altera&on deposits within and around a typical complex impact crater (a@er Osinski et al. 2012)
Molecular parameters of cooling (thermal matura&on)
Hopane in sediment (geological configura&on)
ββ22R
βα22R αβ22R αβ22S
x
ββ/(αβ+βα+αβ)
βα/(αβ+βα)
Thermal maturity parameters
0 5 10TOC /%
540
546
552
558
564
570
576
582
Core
Dep
th /m
K/Pg0.1 0.3 0.5 0.7 0.9
βα/(αβ+βα)0 50100150B. Braunii
0.1 0.3 0.5 0.7 0.9ββ/(αβ+βα+ββ)
C 31ββ
C 31αβ22R
C 31βα
C 31ββ
C 31βα
C 31αβ22R
C 31αβ22S
C 31ββ
C 31βα
C 31αβ22R
C 31αβ22S
15 18 21 24 27δ18OVSMOW/‰
Previous estimates of duration of heating
• ~1.5 – 4.5 ka for the 4 km diameter Kärdla crater (Jõeleht et al., 2005) • ~5 ka for 24-km-diameter Haughton crater (Parnell et al. 2005) • 67 ka for 30-km-diameter crater in an early Martian environment (Abramov
and Kring, 2005) • ~600 ka. and ~1.6 Ma for 23-km-diameter Lappajärvi (Schmieder and
Jourdan, 2013)
• In comparing the longevity of the hydrothermal systems developed at the Ries and Haughton impact structures, Osinski (2012) concluded that crater lakes were critical in the development of longer-lived hydrothermal systems
• Continuous sedimentation record at Boltysh in 600 m of accommodation space provide powerful stratigraphic constraints on timescales
Post-impact timescales
• Palynology of early post-impact sediments
• Early-mid successional community of ferns and angiosperms
• Parallels with inter-lava flow durations
• 2 – 5 ka timescale between the basal lake sediments and fern-spike that marks K/Pg boundary (~581.6 m)
• Need to constrain longer timescale
CLIMATIC OSCILLATIONS STALL K/PG RECOVERY
Imp
act
bre
ccia
cra
ter
lake
tu
rbid
ites
c s f m csand
Lith
olog
y
576.5576.7576.9577.1577.3577.5577.7577.9578.1578.3578.5578.7578.9579.1579.3579.5579.7579.9580.1580.3580.5580.7580.9581.1581.3581.5581.7581.9582.1582.3582.5582.7582.9583.1583.3
0 20
Cup
ress
acea
e
0 20
Pinac
eae
Fagac
eae
Thym
elae
acea
e & u
ndiff
Iaca
cina
ceae
Juglan
dace
ae
Myr
icac
eae
Nys
sace
ae
Plata
nace
ae -
Cer
cidiph
yllace
ae
Ros
acea
e
0 20 0 20 0 20 0 20 0 0 0 0 0 20 40 60
Nor
map
olles larg
e
0 20
Nor
map
olles sm
all
0 20 40 60 80 100
Flood
plain
fern
s
(incl. P
olyp
odiace
ae &
Pte
ridac
eae
0 20
Fern
allie
s
0 20
Schizae
acea
e
0 20 40
Palm
ae
Gymnosperms Angiosperms - post Normapolles Normapolles Ferns
Fe
rn S
pik
e
Ph
ase
1P
ha
se 2
Boltyshpost impact
BarrenZone
Pla
tan
ace
ae
-N
orm
ap
olle
s -
Pin
ace
ae
Fig. 2. Frequency plot of selected taxa summed by botanical affinity. These are shown as count data from aliquot mounts of standard dry weight rock, because of the low pollen frequencies in some samples. The significant increase in diversity and influx of Palmae pollen (mainly Arecipites spp.) at the onset of fern spore spike Phase 2 should be noted. The Platanaceae–Normapolles–Pinaceae association that succeeds Phase 2 is defined by the dominance of pollen from these families from 578.1 m.
Superimposed on a long-term reduction in moisture availability marked by increasing Normapolles dominance, declining ferns and Cupressaceae are shorter-term oscillations (Figs 3 and 4). Periods of higher moisture availability occur early in each oscillation (O1–O4, Figs 3 and 4), and correspond to low-diversity palynofloras. These pass into intervals of higher palynofloral diversity in the succeeding drier period. In the early, wetter period of each oscillation, diversity may have been suppressed by competition from Cupressaceae as the dominant plant in the true swamp community. In the succeeding drier periods lower moisture availability would have reduced the ecospace available to Cupressaceae, resulting in an open mosaic of plant communities with correspondingly greater diversity. This ecosystem fragmentation is evident not only from the response of the broadly thermophyllic Normapolles group (Batten 1981), but also in the post-Normapolles angiosperms and ferns. Ferns main-tained higher taxonomic diversities in the later, drier stages of oscillations O1, O3 and O4 (Fig. 4) while undergoing a significant fall in overall spore abundance (Fig. 3).
The K/Pg event occurred early in oscillation O1, amplifying the moisture availability and temperature (Tschudy & Tschudy 1986) and enhancing conditions for the dominance of ferns in the primary recovery vegetation. The subsequent reduction in moisture availabil-ity in the upper part of this oscillation modified the plant successional pathway. Transition to the higher diversity, angiosperm-dominated
mid-successional communities during periods of reduced moisture availability would have retarded the spread of true swamp late suc-cessional vegetation. In consequence, the apparent rate of recovery from the K/Pg event of Boltysh terrestrial ecosystems would have been retarded by oscillations unconnected with the K/Pg.
Duration of biotic reassembly
Constraining the duration of post K/Pg biotic reassembly in the Boltysh area is facilitated by the occurrence of the Dan-C2 negative CIE in sediments overlying the upper limit of oscillation O4 (Gilmour et al. 2012). Onset of the Dan-C2 negative CIE occurred at 65.98 Ma, between 0.20 and 0.32 myr after the K/Pg boundary (66.24 ± 0.06 Ma). The upper limit of oscillation O4 occurs 8 m beneath the base of the Dan-C2. The thickness of each successive oscillation was controlled by the efficiency of clastic sediment input via the marginal fluvio-deltaic system and by ejecta availability with increasing blanket denudation. These factors were probably compounded by decreasing fluvial discharge consequent on the overall warming and drying cli-mate evident prior to the Dan-C2 (Fig. 3, and Gilmour et al. 2012). Variability in clastic sediment input to the crater is evident from the crater fill facies distribution derived from the Russian drilling pro-gramme (Fig. 1), which identified turbidite fan retreat towards the crater margins prior to the Dan-C2. From the K/Pg boundary to the
at Open University on April 24, 2013http://jgs.lyellcollection.org/Downloaded from
5m
Early Successional
Mid Successional
Late Successional
-35 -30 -25 -20δ13CVPDB/‰
-50
0
50
100
150
200
250
300
350
400
Age
rela
tive
to o
nset
of C
IE /k
a
-1 0 1 2 3 4δ13CVPDB/‰
DSDP 527Boltysh
0 1 2 3δ13CVPDB/‰
ODP 1049C
0 1 2δ13CVPDB/‰
DSDP 528
-35 -30 -25 -20 -15δ13CVPDB /‰
Precession Obliquity
Comparison with marine Dan-C2 hyperthermal record
Astrobiological Significance
• Impact hydrothermal environments – Spatial extent – Continued availability of liquid H2O, energy and nutrients over
extended periods of time • Majority of Martian impact craters have diameters of less
than 50 km – many contain well-preserved lacustrine sedimentary deposits
• Boltysh – no evidence for timescales of heating as long as 600 ka – estimate of ~30 – 40 ka. is longer than Haughton crater – Haughton, no evidence preserved for a crater lake forming post-
impact – Supports the suggestion that the presence an intra-crater lake
may play a crucial role in determining the extent and duration of impact-induced hydrothermal systems