Variations in magnetic properties of target basalts with the direction of asteroid impact: Example from Lonar crater, India

Authors


  • [This article was corrected on August 7th 2012 after initial online publication. Author name M. Arif was changed to Md. Arif in every instance throughout the text.]

Corresponding author. E-mail: misras@ukzn.ac.za; misrasaumitra@gmail.com

Abstract

Abstract– The Lonar crater in Maharashtra state, India, has been completely excavated on the Deccan Traps basalt (approximately 65 Ma) at approximately 570 ± 47 ka by an oblique impact of a possible chondritic asteroid that struck the preimpact target from the east at an angle of approximately 30–45o to the horizon where the total duration of the shock event was approximately 1 s. It is shown by our early work that the distribution of ejecta and deformation of target rocks around the crater rim are symmetrical to the east–west plane of impact (Misra et al. 2010). The present study shows that some of the rock magnetic properties of these shocked target basalts, e.g., low-field anisotropy of magnetic susceptibility (AMS), natural remanent magnetization (NRM)/bulk susceptibility (χ), and high-coercivity and high-temperature (HC_HT) magnetization component, are also almost symmetrically oriented with reference to the plane of impact. Studies on the relative displacements of K3 (minimum) AMS axes of shocked basalts from around the crater rim and from the adjacent target rocks to the approximately 2–3 km west of the crater center suggest that the impact stress could have branched out into the major southwestward and northwestward components in the downrange direction immediately after the impact. The biaxial distribution of AMS axes in stereographic plots for the unshocked basalts transforms mostly into triaxial distribution for the shocked basalts, although transitional type distribution also exists. The degree of anisotropy (P′) of AMS ellipsoids of the shocked basalts decreases by approximately 2% when compared with those of the unshocked target (approximately 1.03). The NRM/χ (Am−1) values of the shocked basalts on the rim of the Lonar crater do not show much change in the uprange or downrange direction on and close to the east–west plane of impact, and the values are only approximately 1.5 times higher on average over the unshocked basalts around the crater. However, the values become approximately 1.4–16.4 times higher for the shocked basalts on the crater rim, which occur obliquely to the plane of impact. The target basalts at approximately 2–3 km west of the crater center in the downrange also show a significant increase (up to approximately 26 times higher) in NRM/χ. The majority of the shocked basalt samples (approximately 73%) from around the crater rim, in general, show a lowering of REM, except those from approximately 2–3 km west of the crater center in the downrange, where nearly half of the sample population shows a higher REM of approximately 3.63% in average. The shocked target basalts around the Lonar crater also acquired an HC_HT magnetization component due to impact. These HC_HT components are mostly oriented in the uprange direction and are symmetrically disposed about the east–west plane of impact, making an obtuse angle with the direction of impact. The low-coercivity and low-temperature (LC_LT) components of both the unshocked and shocked basalts are statistically identical to the present day field (PDF) direction. This could be chemical and/or viscous remanent magnetization acquired by the target basalts during the last 570 ± 47 ka, subsequent to the formation of the Lonar crater. The shocked Lonar target basalts appear to have remagnetized under high impact shock pressure and at low temperature of approximately 200–300 °C, where Ti-rich titanomagnetite was the main magnetic remanence carrier.

Introduction

The Lonar lake (19°58′ N, 76°31′ E) in Maharashtra state, India (Fig. 1), is always a special attraction to planetary scientists because it is one of the few among approximately 176 known terrestrial asteroid impact craters (http://www.unb.ca/passc/ImpactDatabase) that is completely excavated on the basaltic target rocks and fully accessible (Gilbert 1896; La Fond and Dietz 1964; Nayak 1972; Fredriksson et al. 1973, 1979; Kieffer et al. 1976; Morgan 1978; Stroube et al. 1978; Fudali et al. 1980; Rao and Bhalla 1984; Ghosh 2003; Ghosh and Bhaduri 2003; Hagerty and Newsom 2003; Kumar 2005; Osae et al. 2005; Son and Koeberl 2007). The other one is the Logancha crater, Russia, which could be on the Siberian Trap basalts (Reichow et al. 2002), but little information is available on this crater (Feldman et al. 1983; Masaitis 1999). The most recently known impact structure in a basaltic target is the Vista Alegre crater on the Paraná flood basalt (approximately 133–132 Ma), Brazil (Crósta et al. 2010). The Lonar crater is therefore one of the few known terrestrial analogs available for evaluating the consequences of hypervelocity asteroid impacts on planetary surfaces having basaltic crusts.

Figure 1.

 Sketch maps of (a) India showing the location of Lonar crater and (b) the Lonar crater showing most of the locations of drilling sites (filled black circles) of the present study. DT = Durga Tegri, SWT = Saraswati village, CRW = wall to the crater west, KHN = Khini village, KPD = Kalapani dam.

The most precise radiometric ages (40Ar/39Ar) suggest that the Lonar crater was formed at 570 ± 47 ka ago (Jourdan et al. 2011) by the hypervelocity impact of a chondritic asteroid that struck the preimpact target basalt from the east at an angle of between approximately 30 and 45o to the horizon (Misra et al. 2009, 2010). The total duration of the shock event producing the shock melting at Lonar crater specifically was suggested to be approximately 1 s (Kieffer et al. 1976). The target Deccan basalt of the Lonar crater erupted close to the Cretaceous-Tertiary (KT) boundary at 65 ± 0.9 Ma (Hofmann et al. 2000; also see Courtillot et al. 2000) or close to 67.4 Ma (Pande et al. 2004), although some controversy exists regarding the duration of this flood basalt volcanism (Wignall 2001; Courtillot and Renne 2003). The more recent idea suggests that the first extensive phase of the Deccan volcanism, which might have lasted only a few hundred thousand years, occurred at approximately 67.5 Ma at the northern half of the present Deccan outcrops and after approximately 2.5 Ma of quiescence, the second major phase of volcanism occurred at approximately 65 Ma (Chenet et al. 2007).

Although the Lonar crater is one of the best studied terrestrial impact craters and unique for its target composition, the criteria for identification of distribution of shock fronts around this structure are very limited. Kieffer et al. (1976) identified some petrographic criteria; e.g., well-developed irregular fractures particularly in plagioclase phenocrysts, for identification of shocked basalts around the Lonar crater. More recent studies, however, suggest that apparently unshocked Deccan basalts can also be present on the Lonar crater rim (Osae et al. 2005), and petrographic criteria could not always be sufficient for identification of the shock front around this crater (Fig. 1).

Alternative criteria that can efficiently be used for the identification of shock fronts around the Lonar crater are the rock magnetic properties of the target rocks (Urrutia-Fucugauchi and Velasco-Villareal 2008). Experiments on solid intrusive rocks confirmed that shock pressure in the order of tens of kilobars (kbar) was sufficient to affect the NRM of terrestrial materials; the principal effect of shock was to demagnetize the preexisting remanence, and the acquisition of secondary magnetization in the direction of the ambient field at the time of impact (Cisowski and Fuller 1978). Although longer and more intense shock pulses might affect the higher coercivity fraction, the low-coercivity fraction of the target rocks was most susceptible to magnetic resetting by shock. More recent work on the Vredefort impact crater, South Africa, showed that the random orientations of the low-field anisotropy of magnetic susceptibility (AMS) axes of intrusive target rocks could be taken as a characteristic feature of the asteroid impact event (Carporzen et al. 2005). Experiments on basaltic andesites from central Japan and basalts from the Lonar crater, India, further showed that the orientations of AMS axes could be an indicator of the propagation directions of stress waves generated in rocks at terrestrial impact structures (Nishioka 2007; Nishioka et al. 2007; Nishioka and Funaki 2008). In the high-pressure range (>3 GPa) of experiment, the anisotropy degree was increased, the minimum susceptibility axes (K3) were oriented towards the shock direction, and the average susceptibility was decreased.

The existing rock-magnetic and paleomagnetic studies on the Lonar crater did not aim to evaluate the distribution of shock front around this crater. Rao and Bhalla (1984) reported that some magnetic parameters, viz., Jn (NRM), K (susceptibility), Qn (Koenigsberger ratio), and declination, of basaltic samples collected from the inner walls of the Lonar crater showed systematic variations, whereas random variations were observed for the surrounding target rocks. A soft secondary shock component was also identified in their study that was acquired in the Earth’s present magnetic field. More recent work, however, suggested that this low-coercivity and low-temperature (LC_LT) component was not directly related to shock metamorphism, but was replaced by a postimpact component acquired by viscous and/or chemical remanent magnetization (Louzada et al. 2008). Weiss et al. (2007) also reported rock magnetic data of the impact spherules and fläden (0.01–1 cm in size) recovered from the ejecta blanket around the east and west of the Lonar crater rim. The ratios of NRM to saturation isothermal remanent magnetization (SIRM) for the small glasses reported by them were only 0.5–1 × 10−3, while the large glasses had ratios twice as large. These values were nearly an order of magnitude lower than those measured for the nearby Deccan basalts (Louzada et al. 2008). The low NRM/SIRM ratios of the Lonar glasses were interpreted to indicate the absence of any impact-generated paleofields substantially higher than several tens of micro Tesla (μT) at the Lonar crater (also Weiss et al. 2010), and these glasses marginally underestimated the intensity of the field in which they cooled, probably due to the effects of rotation during cooling. A detailed AMS study on the target rocks established the oblique asteroid impact hypothesis from the east for the Lonar crater, which showed that the target basalts from approximately 2 km west-southwest of the crater rim were highly shocked compared with the unshocked basalts at approximately 2 km east-southeast of the crater rim (Misra et al. 2010). However, this study was directed to evaluate neither the distribution of the shock front around this hypervelocity impact crater nor the variation in rock magnetic properties with reference to the plane and direction of asteroid impact. The main objectives of our present study were therefore to investigate a detailed rock magnetic study of the target basalts from around the Lonar crater rim and surroundings to evaluate (1) the possible distribution of impact-induced shock front around the crater rim through AMS, (2) any possible variation in NRM/bulk susceptibility (χ), REM (=NRM/SIRM ratio expressed in percentage), LC_LT and high-coercivity and high-temperature (HC_HT) magnetization components in relation to the direction of asteroid impact and across.

Geological Setting of Lonar Crater

The Lonar crater is a simple, bowl-shaped, near-circular impact crater (Kumar 2005), with N–S and E–W diameters of approximately 1832 and 1790 m, respectively, with a circularity of approximately 0.95 and a depth of approximately 150 m (Fredriksson et al. 1973; Fudali et al. 1980; Misra et al. 2010). All around its circumference, except for a small sector in the NE, there is a continuous rim raised approximately 30 m above the adjacent plains, whereas the crater floor lies approximately 90 m below the preimpact surface. The rim is surrounded in all the directions by a continuous ejecta blanket that extends outward with a gentle slope of 2–6° to an average distance of approximately 700 m from the crater rim, except to the west where it extends for little more than a kilometer (Misra et al. 2010). The interior of the crater is occupied by a shallow saline lake; below the lake water, a sequence of approximately 100 m thick unconsolidated sediment is reported that overlies the base of the crater made up of highly weathered Deccan Trap basalt (Nandy and Deo 1961; Fudali et al. 1980). About 700 m north of the rim, there is another relatively shallow depression known as the Little Lonar (Fig. 1b), which has a diameter of approximately 300 m; however, drilling into the structure revealed no evidence of impact (Fredriksson, personal communication; Maloof et al. 2010).

The target rocks at Lonar crater are subhorizontal Deccan Trap basalt flows that overlie the Precambrian basement with a thickness of >350 m (Kumar 2005). These basalts contain intertrappean sediments of small areal extent of fluviatile and lacustrine origin of varied thickness up to 3 m (Jhingran and Rao 1958; Venkatesh 1967; Krishnan 1968). There are altogether six basalt flows of approximately 8–40 m thickness in and around the Lonar crater, of which the four bottom flows are only exposed along the crater wall (Ghosh and Bhaduri 2003). The two topmost flows occur away from the crater, and do not show any impact-induced deformation. Flows are separated from one another by a discontinuous marker horizon like red and green paleosols, chilled and vesicular margins, and vugs filled with secondary minerals. Fresh basalts occur only in the upper approximately 50 m of the crater wall, whereas below this level, the flows are heavily weathered and friable (Fudali et al. 1980). The preimpact black, sticky, humus-rich soil of approximately 5–90 cm thickness is still preserved at places between flows and overlying ejecta (Ghosh and Bhaduri 2003). All the basalt flows have a common mineralogy and texture except some minor petrographic differences in the abundance of plagioclase phenocrysts, glass, and opaque minerals (Ghosh and Bhaduri 2003; Osae et al. 2005). The basalts contain occasional phenocrysts of plagioclase and rare olivine set in a groundmass of plagioclase, augite, pigeonite, titanomagnetite, palagonite, and secondary minerals, such as calcite, zeolite, chlorite, serpentine, and chlorophaeite (Ghosh and Bhaduri 2003).

Sampling and Experimental Techniques

Oriented drill core samples of approximately 2.5 cm diameter were collected from all around the Lonar crater rim and adjoining areas (Fig. 1b) with a portable gasoline-powered rock drill during three field seasons between the years 2008 and 2010. Each core was cut into two to three specimens of approximately 2.2 cm height in the laboratory. All paleomagnetic and rock magnetic measurements were performed at the Environmental Magnetism Laboratory, Indian Institute of Geomagnetism, Navi Mumbai, India. The details of instrumentation used for our measurements and necessary analytical techniques are given in the Supporting Information.

Rock Magnetic Studies of Lonar Crater

Magnetic Mineralogy of Lonar Basalts

The magnetic carriers of Lonar target basalts were found to be titanomagnetite (Cisowski and Fuller 1978). Micron-sized Ti-rich exsolution lamellae divide ferrimagnetic Ti-poor titanomagnetite grains (tens of microns in size) into interacting single-domain needles of high coercivity, although some grains have poorly developed lamellae and low coercivities.

The Lonar target basalt samples have been studied in the present study using rock magnetic techniques of IRM acquisition and backfield curves, progressive thermal demagnetization of SIRM, low- and high-temperature variation in magnetic susceptibility (between −196 and 700 °C), and hysteresis parameters, viz., saturation remanence (Mrs), saturation magnetization (Ms), remanent coercivity (Hcr), and coercive force (Hc) (see Figs. S1 and S2).

The IRM acquisition curves of unshocked target basalts from Durga Tegri and further east in the uprange, and shocked basalts from around the Lonar crater rim and approximately 2–3 km west of the crater center in the downrange (total number of samples (n) = 33, Fig. 1b) (the shocked and unshocked basalts are defined based on AMS properties of the target rocks as suggested in Misra et al. 2010) saturate at low field of <200 mT (Fig. S1a), indicating low-coercivity magnetic mineral as the main remanence carrier. The IRM backfield curves (Fig. S1a in inset) indicate a broad range of Hcr from 15 to 45 mT, revealing the grain size as single-domain (SD) to pseudosingle-domain (PSD) type (Cisowski 1981; Dankers 1981). The SIRM thermal demagnetization curves show a range of unblocking temperatures (Fig. S1b) corresponding to titanomagnetite (approximately 200 and 300 °C) and magnetite (approximately 580 °C).

We have observed the variation in magnetic susceptibility with temperature (χ-T) (Fig. S1c–g) on a set of systematically collected target rock samples from around the Lonar crater, which includes a sample of unshocked target basalt (A4/100o-2.93; meaning of the notation: sample site A4 is located at 2.93 km distance from the crater center along a direction of 100o from the geographic north in clockwise direction), two samples of shocked basalts from around the crater rim (L5/092o-0.98, A18/272o-0.92), and an additional two samples of shocked basalts from the downrange direction (A11/251o-2.92, A14/300o-2.10). The χ-T curve of unshocked sample A4 (100°-2.93) has apparently two different thermomagnetic phases during heating (Fig. S1c). The lower Tc (Curie temperature) ranges between approximately 280 and 350 °C, while the higher one is at approximately 580 °C. The cooling curve shows only a single phase, with a Tc close to that of magnetite (approximately 585 °C). Such irreversible χ-T curves could be due to the presence of titanomaghemite, which probably transformed into magnetite; their low-temperature χ-peak at −155 °C reflects the isotropic point of multidomain (MD) magnetite (cf. Radhakrishnamurty et al. 1978). The χ-T curve of shocked basalt sample L5 (092°-0.98) from the crater rim shows a gradual increase in susceptibility with temperature until approximately 235 °C followed by a sharp decrease up to approximately 580 °C with variable slopes (Fig. S1d). An initial Tc between approximately 235 and 325 °C indicates Ti-rich titanomagnetite, followed by a final Tc of magnetite at approximately 580 °C. For this sample, the low-temperature variation in χ (warming from liquid nitrogen to room temperature) shows a smooth increase, implying the presence of high-Ti content in the sample, which shifts the Verwey transition (Carter-Stiglitz et al. 2006). This is also similar to the behavior with SD nature of magnetic carrier (Radhakrishnamurty et al. 1982). The χ-T curve of another shocked basalt sample A18 (272°-0.92) from the crater rim shows an increase in χ at approximately 300 °C and a subsequent decrease at approximately 400 °C, followed by a peak at approximately 500 °C with a sharp drop at approximately 580 °C, which indicate Tc due to magnetite (Fig. S1e). The low Tc (approximately 300 °C) indicates Ti-rich titanomagnetite, titanomaghemite with intermediate Tc (approximately 400 °C), and finally high Tc (approximately 580 °C) of magnetite. The low-temperature susceptibility behavior shows a suppressed χ-peak (−163 °C). The possible presence of pyrrhotite in this sample (A18) could be excluded because in the first three progressive heating–cooling cycles—i.e., upon heating to approximately 300, 350, and 400 °C (Fig. S2a–c)—the χ-T curves show nearly reversible behavior with very small degrees of irreversibility. Upon further heating to 500 °C, larger degrees of irreversibility are observed (Fig. S2d); and at the final heating–cooling cycle to 600 °C, titanomaghemite is completely inverted, probably forming an iron-rich spinel inversion product with Curie temperature (Tc) at approximately 585 °C (Fig. S2e). Irreversibility of χ-T curves suggests the onset of titanomaghemite inversion at or slightly below approximately 400 °C (i.e., in a 325–400 °C temperature range). The magnetic susceptibility remains constant during gradual increase in the field from approximately 2 to 450 A m−1 (Fig. S2f), which indicates the absence of pyrrhotite in sample A18 (cf. Hrouda et al. 2006). The χ-T curves of the shocked basalt samples A11(251°-2.92) and A14 (300°-2.10) in the downrange show a single ferrimagnetic phase with Tc at 580 °C corresponding to that of the Ti-poor titanomagnetite or magnetite (Fig. S1f, g); the low-temperature χ-peak at −155 °C reflects the isotropic point of MD magnetite (cf. Radhakrishnamurty et al. 1978).

The hysteresis data are plotted on the reappraisal of the Day plot of ratios Mrs/Ms against Hcr/Hc (Day et al. 1977; modified after Dunlop 2002) (Fig. S1h). Most of the samples are plotted on a linear trend within the PSD field between the theoretical mixing curves of Dunlop (2002), the linear trend defined by the samples is directed toward the SD grain size. Our experiments, thus, suggest that the Lonar target basalts essentially contain PSD grain size Ti-rich to Ti-poor titanomagnetite and their low-temperature oxidized products as the magnetic carriers (cf. Basavaiah 2011).

Low-Field Anisotropy of Magnetic Susceptibility (AMS) of Lonar Basalts

The basaltic rocks occurring in and around the Lonar crater (Fig. 1b) can apparently be classified into two types depending on our previous observation on impact-induced rock deformational features and AMS studies (Misra et al. 2010), viz., (1) unshocked basalts: those lying away from the crater rim, but close to the crater to the east and (2) shocked basalts: lying along the crater rim and walls, and in the downrange direction to the west. The best exposed, thick, unshocked basalt flow that could be stratigraphically equivalent to the fourth flow along the crater rim (thickness approximately 40 m, Ghosh and Bhaduri 2003) was sampled from the base of hillock Durga Tegri (DT/100°-2.93) (Fig. 1b). Additional samples of the fourth basalt flow were also collected from farther east of Durga Tegri (A5/102°-4.61).

It was observed that the Lonar unshocked basalts (n = 47) show an oblate-shaped susceptibility ellipsoid with degree of anisotropy (P′) values mostly between approximately 1.02 and 1.04, although a few higher values up to approximately 1.06 are also present; the average is close to 1.033 ± 0.008 (Misra et al. 2010) (Fig. 2a). The ellipsoids of shocked basalt samples from around the crater rim and farther west are, however, oblate to prolate in shape with low average P′ value of 1.01 ± 0.006 that mostly lies between approximately 1.00 and 1.03.

Figure 2.

 Degree of anisotropy (P′) versus shape parameter (T) plot for unshocked and shocked basalts from around the Lonar crater rim and at approximately 2–3 km west of the crater center in the downrange direction (for sample locations see Fig. 1b). Note that higher P′ value and oblate shape of susceptibility ellipsoid are characteristics of unshocked target basalts (a), whereas shocked basalts have restricted and lower P′ values with variation in shape of susceptibility ellipsoids from oblate to prolate type (b–k); P′ and T values are computed after Tarling and Hrouda (1993). Note: each sampling site is represented by its distance from the crater center along a direction from geographic north in clockwise direction; Abbreviations as in Fig. 1b; for further details, see text.

A more extensive set of samples of shocked basalts (number in total = 478, which was approximately 320% higher over Misra et al. [2010]) were sampled mostly from the top of the basalt flow exposed on the crater rim, and the horizontal basalt flow at approximately 2–3 km west of the crater center in the downrange direction (Fig. 1b). None of the samples was collected from the possible folded basalt flows on the crater rim in the present study. The shocked basalt samples from the eastern sector of the Lonar crater rim (092°-0.98, n = 30) show that the average P′ value (1.017 ± 0.009) of mostly oblate to prolate susceptibility ellipsoids is lower than earlier shown (approximately 1.02, Misra et al. 2010), although the data show significant variation of P′ between approximately 1.006 and 1.042 (Fig. 2b). The mostly oblate to prolate-shaped susceptibility ellipsoids from the southeastern sector of the crater rim (138°-0.88, n = 54) show average P′ of 1.009 ± 0.004 with a range between approximately 1.004 and 1.025 (Fig. 2c). The oblate to prolate-shaped susceptibility ellipsoids from the southern sector of the crater rim (190°-0.86, n = 40) also have similarly low average P′ of 1.012 ± 0.002 with a restricted range of variation between approximately 1.005 and 1.016 (Fig. 2d). The data from the southwestern sector of the crater rim (229°-0.82, n = 44) show low average P′ of 1.01 ± 0.003 for the oblate to prolate ellipsoids with a restricted range of variation between approximately 1.006 and 1.017 (Fig. 2e). Unlike our previous observation, the extensive data from the western sector (272°-0.92) and northwestern sector (319°-0.92) of the crater rim (n = 60 each) suggest strong variation in the shape of susceptibility ellipsoids from oblate to prolate with a low average P′ of 1.01 ± 0.004 and range of variation between approximately 1.004 and 1.025 (Fig. 2h and 2j). The susceptibility ellipsoids from the northern sector of the crater rim (358°-0.92, n = 34) remain mostly oblate with low average P′ of 1.012 ± 0.003 with a range between approximately 1.007 and 1.02 (Fig. 2k).

The additional shocked basalt samples from distant locations at approximately 2–3 km west of the crater center in the downrange direction (cf. Misra et al. 2010) (Fig. 1b) were also investigated in the present study. Samples collected from the wall to the west of the crater rim (CRW/267°-0.92, n = 38) show oblate to prolate susceptibility ellipsoids with an average P′ of 1.015 ± 0.003 and restricted variation of data between approximately 1.009 and 1.023 (Fig. 2g). The shocked basalts from the Kalapani dam (KPD/233°-2.17, n = 44), close to Khini village (KHN/251°-2.92) (Misra et al. 2010) (Fig. 1b), show mostly oblate to prolate-shaped ellipsoids that have significant variation in P′ between approximately 1.003 and 1.056 with an average P′ of 1.019 ± 0.013 (Fig. 2f). The samples from the Saraswati village (SWT/300°-2.10, n = 74) show an equal distribution of susceptibility ellipsoids in the oblate and prolate fields with an average P′ of 1.013 ± 0.006 (Fig. 2i). Most of the data show a restricted variation in P′ between approximately 1.004 and 1.019, although a few samples have values up to approximately 1.04.

In stereographic plots, the minimum susceptibility axes (K3) of the unshocked Lonar basalts from the base of Durga Tegri (DT/100°-2.93) and farther east (A5/102°-4.61) show clustering of data with moderate to subvertical dips (40–70o) toward the east; the maximum (K1) and intermediate (K2) susceptibility axes are distributed on subhorizontal, west-southwesterly dipping girdle describing a bimodal distribution (cf. Cañón-Tapia et al. 1997) for the Lonar target basalts (Fig. 3a). The average orientation of the K1 (maximum) axes on this girdle defines a flow direction of the Deccan lava at Lonar close toward the west (cf. Cañón-Tapia et al. [1997] and references therein) with a dip approximately 20° (Fig. 3).

Figure 3.

 Stereographic projections of AMS susceptibility axes of (a) unshocked target, and (b–k) shocked basalts from around the rim of the Lonar crater, and at approximately 2–3 km west of the crater center in the downrange direction (for sample locations, see Fig. 1b); Note that the bimodal distribution of the AMS axes is a characteristic of unshocked basalts from Durga Tegri (DT) and farther east (A5); most of the shocked basalts show triaxial distribution of AMS axes, except the samples from the north and west crater rim sectors, which show transitional type distribution.

The shocked basalts from the Lonar crater rim and adjoining area to the west, however, show different distribution of AMS axes in stereographic projections (cf. Misra et al. 2010). The susceptibility axes from the eastern sector of the crater rim (092°-0.98) show more or less a triaxial distribution where the K1 and K2 axes are subhorizontal mostly oriented in E–W and N–S directions, respectively, although a moderate overlap of data exists (Fig. 3b, also see Fig. 7 for displacement directions of K3 axes for all shocked basalts). The K3 axes show vertical to subvertical orientation, and are shifted mostly toward the northwest compared with those of the unshocked Lonar basalts (Fig. 3a). The target basalts from the southeastern sector (138°-0.88) show a strong west to west-southward shift of the K3 axes; the susceptibility axes show a broad triaxial distribution where the orientations of K1 and K3 axes are similar to those of the shocked basalt from the eastern sector (Fig. 3c). The stereographic distribution of AMS axes of the shocked basalts from the southern sector (190°-0.86) shows a similar type of triaxial distribution as observed in the eastern sector of the crater rim, although interchange in position of AMS axes is noticed (Fig. 3d). In this sector, the K1 axes are oriented vertical to subvertical; the K2 and K3 axes show horizontal to subhorizontal orientations mostly in E–W and N–S directions, respectively. In comparison with the unshocked target basalts, the K3 axes of these shocked basalts show south and southwestward shift. The stereographic plots of the AMS axes of shocked target from the southwestern sector of the crater rim (229°-0.82) show a prominent triaxial distribution (Fig. 3e). The K1 axes are horizontal to subhorizontal in an ESE–WNW direction, whereas the K2 axes are vertical to subvertical in orientation. The subhorizontal K3 axes show a southwestward shift in position compared with those of unshocked target basalts at Lonar. The shocked basalts from the western sector (272°-0.92) show significant northwestward shift of K3 axes compared with the unshocked basalts (Fig. 3h). Most of the K1 and K2 susceptibility axes are broadly distributed on a subhorizontal girdle dipping toward the south-southwest, and the bimodal distribution of AMS axes as observed for the unshocked target is partly retained. The AMS axes of shocked basalts from the northwestern sector of the crater rim (319°-0.92) show a triaxial distribution of axes: the K3 axes are vertical to subvertical in orientation and show displacement mostly toward the northwest in comparison with the unshocked basalts; the K1 and K2 susceptibility axes are horizontal to subhorizontal and are mostly directed toward the SE–NW and SW–NE, respectively (Fig. 3j). The K3 susceptibility axes of shocked basalts from the northern sector (358°-0.92) are oriented vertically to subvertically and show a dominant westward shift (Fig. 3k). Most of the K1 and K2 axes are broadly distributed on a southwesterly dipping subhorizontal girdle, and the bimodal distribution of the AMS axes as observed for the unshocked basalts is partly retained.

The distribution of AMS axes of shocked basalts from the village Khini (KHN/233°-2.17) at the southwest of Lonar crater (Fig. 1b) has been described in Misra et al. (2010), which show a major shift in the K3 axes toward the southwest compared with the unshocked target, and most of the K1 and K2 axes describe a subhorizontal northerly dipping girdle. The shocked basalts from the Kalapani dam (KPD/233°-2.17), which is close to the village Khini, also show a southwestward shift of the K3 axes compared with the unshocked Lonar basalts, a broadly defined triaxial distribution of susceptibility axes is observed (Fig. 3f). The target basalts collected from a wall to the west of the Lonar crater rim (CRW/272°-0.92) show a well-defined triaxial distribution of AMS axes (Fig. 3g). The K2 axes are mostly subvertical in orientation, whereas the K1 and K3 axes are subhorizontal and oriented toward the northwest and northeast, respectively. The K3 axes of these shocked basalts are shifted toward the north compared with the Lonar unshocked basalts (Fig. 3a). The shocked target basalts from the village Saraswati at the northwest of the Lonar crater (SWT/300°-0.92) (Fig. 1b) also show a triaxial distribution of data (Fig. 3i). Most of the K2 axes are vertical to subvertical; the K1 and K3 axes are mostly horizontal and oriented toward the north and west, respectively. The K3 axes of these target basalts are shifted toward the west and northwest compared with the Lonar unshocked target.

NRM/χ and REM of Lonar Basalts

Different views exist on the possible relationship between the shock pressure due to asteroid impact and resulting NRM intensities of target rocks. Some workers believed that the NRM of target rocks increased many times due to impact-induced shock pressure (Pesonen et al. 1997; Carporzen et al. 2005; Ugalde et al. 2005); others suggested that NRM decreased with shock pressure (Nishioka 2007; Louzada et al. 2008). In our present study, we carried out measurements of NRM/χ (expressed in units of Am−1) instead of NRM of both the unshocked and shocked basalt samples from around the Lonar crater (Table S1). The detailed variations in NRM/χ within each group of samples are shown graphically in Fig. S3, and a summary of our observation is shown in Fig. 4. Our data show that the NRM/χ values of the unshocked target basalts from Durga Tegri and farther east (sites: DT/100°-2.93 and A5/102°-4.61, Fig. 1b) (n = 44) are ≤330 Am−1 where the majority of samples (approximately 68%) have a restricted range of variation (approximately 89–160 Am−1) with an average of 116 ± 23 Am−1 (Fig. 4a), which is similar to the grand mean of all the unshocked target basalts (Table S1).

Figure 4.

 Bar diagrams showing the variation in NRM/χ (Am−1) (white color bar) and REM (gray color bar) of (a) unshocked basalts from Durga Tegri (DT) and farther east (A5); and (b–k) shocked basalts from around the crater rim and at approximately 2–3 km west of the crater center in the downrange direction. Note that the data are represented in logarithmic scale for NRM/χ and in linear scale for REM. For details of variation in NRM/χ and REM, see text. [Correction added on 03 July 2012 after online publication. In Fig. 4 all instances of NRM/ were corrected to NRM/χ.]

The shocked basalts collected from around the Lonar crater rim show a general increase in NRM/χ. The ratios of the number of target basalt samples having NRM/χ higher and lower than the average unshocked target (R1) within each group of target basalts collected from the different sectors of the crater rim vary over a wide range between approximately 3:1 and >54:1 (Table S1). The lowest range of ratios (approximately 3:1) is only observed for the samples collected from the eastern and western sectors of the crater rim. The higher ranges of ratios (approximately 10:1 and >54:1) are only observed for those crater rim target basalts that are situated oblique or perpendicular to the east–west impact plane of the crater (Table S1; Fig. 4).

The NRM/χ values are also found to be low for the majority of samples collected from the eastern and western sectors of the crater rim (Table S1; Fig. 4). The majority of samples (approximately 67%) from the eastern sector (n = 27) of the crater rim (092o-0.98) (Fig. 4b) have average NRM/χ approximately 1.5 times higher over the unshocked target, and the value is very similar when the entire range of dataset for this sector is considered. Most of the samples (approximately 80%) of the shocked basalts collected from the western half of the crater rim including the southwestern (229o-0.82), western (272o-0.92), and northwestern (319o-0.92) sectors (n = 161) show a restricted range of NRM/χ (between approximately 90 and 210 Am−1), and their average (159 ± 28 Am−1) is only approximately 1.4 times higher over the unshocked target (Figs. 4e, h, and j).

The variation in NRM/χ of the shocked basalts from the southeastern sector (n = 54) of crater rim (Fig. 1b) is very interesting (Fig. 4c). The basalts from this sector show three ranges of variation, viz., approximately 50% samples show relatively low and a restricted range of NRM/χ, which is approximately 1.8 times higher over the unshocked target in average; approximately 35% samples show very high NRM/χ, which is approximately 16 times higher over the unshocked target; and a small proportion (approximately 9%) show the highest NRM/χ that are greater than approximately 5000 Am−1. Our experimental data on the shocked basalts from the southern sector (n = 40) of crater rim are also similar type (Fig. 4d); approximately 48% samples show wide variation in NRM/χ and their average value is approximately 5 times higher over the unshocked target; a significant proportion of target samples (approximately 18%) show very high NRM/χ between approximately 1000 and 2000 Am−1, and an equivalent proportion of target (approximately 23%) shows the highest NRM/χ greater than approximately 4000 Am−1. The shocked basalts from the northern sector (n = 34) of crater rim also show a three-fold distribution of NRM/χ, viz., approximately 65% samples show low and restricted range of NRM/χ, which is approximately 1.4 times higher over the unshocked target on average; approximately 29% samples show higher NRM/χ values with a wide range of variation and their average is approximately 4 times higher over the unshocked target; and the rest approximately 6% samples show the highest NRM/χ greater than approximately 4000 Am−1 (Fig. 4k).

The samples studied from the Kalapani dam and Khini village (KDP/233o-2.17, KHN/251o-2.92), wall to the west of the crater rim (CRW/272o-0.92), and village Saraswati (SWT/300o-2.10), which are at approximately 2–3 km west of the crater center in the downrange direction (Fig. 1b), also show both low and high NRM/χ values compared with that of the unshocked target (Table S1). The ratios of the number of target basalt samples with NRM/χ higher and lower than the average unshocked target (R1) in the southwest and northwest directions of the crater rim are lower, and the values are extremely high for the target samples taken from the west of the crater rim. Nearly half of the samples from the Kalapani dam and Khini village (n = 90) southwest of the Lonar crater show low and very restricted NRM/χ (Fig. 4f) and their average value is very close to that of unshocked target; a subordinate proportion (approximately 26%) show higher NRM/χ values with wide range of variation and their average is approximately 7 times higher over the unshocked basalt; the rest (23%) show still higher NRM/χ greater than approximately 1600 Am−1. Nearly 50% of samples collected from the wall to the west of the crater rim (n = 38) show low and restricted range of NRM/χ values (Fig. 4g) and their average is approximately 2.5 times higher over the unshocked target; approximately 37% of samples show higher NRM/χ values that vary over a wide range and their average is approximately 15 times higher over the unshocked basalt; the rest (approximately 13%) show highest NRM/χ greater than approximately 3810 Am−1. The majority (approximately 77%) of shocked basalts from the Saraswati village (n = 74) at the northwest of crater have lower but considerably variable NRM/χ (Fig. 4i), and their average is approximately 1.9 times higher over the unshocked target; approximately 16% of samples have higher NRM/χ with a wide range of variation; and the remaining samples show highest NRM/χ greater than approximately 4090 Am−1.

The REM (NRM/SIRM ratio in percentage) provides an estimate of the paleomagnetic field (Kletetschka et al. 2003) and a ratio of approximately 1.5% indicates an Earth-strength (several tens of μT) field (Gattacceca and Rochette 2004; Kletetschka et al. 2004; Yu 2006). For basaltic lava flow samples, a restricted REM range of approximately 0.5–1.5% corresponds to normal thermoremanent magnetization (TRM) in magnetic fields comparable to the geomagnetic field (Parry 1974). The low-field processes other than TRM (e.g., viscous or chemical remanent magnetization) yield lower REM values for the same paleofield (Fuller et al. 1988), whereas high-field processes (e.g., lightning-induced or artificial IRM or plasma-induced magnetization) yield REM values above 10% (Wasilewski and Dickinson 2000). REM data have also been used for extraterrestrial materials to evaluate the magnetic field intensity of planetary bodies (cf. Yu [2006], and references therein).

The REM ratio of most of the unshocked target basalts (approximately 73% data) from the base of the Durga Tegri (DT/100o-2.93) and farther east (A5/102°-4.61) (Fig. 1b) (n = 26) is low (<1.29%), which is lower than the typical Earth strength field (approximately 1.5%, often considerable up to 3% if NRM is carried by low-Ti titanomagnetite; Yu 2006) and has an average of 0.70 ± 0.24% (Table S1; Fig. 4a); only few samples (approximately 15%) show REM ratio greater than 1.5%.

We have computed the ratios (R2) of number of basalt samples having REM higher and lower than that of the average unshocked basalt (approximately 0.70%) from each sector of the crater rim (Table S1). It is observed that the number of basalt samples with REM higher than the average unshocked basalt is always lower than the number of samples with REM lower than the average unshocked basalt from all the sectors of the crater rim except the southern sector, where the ratio is 2:1. The majority of the crater rim target basalts (approximately 67%) also show REM values lower than that of the average unshocked target basalts from Durga Tegri and farther east (approximately 0.70%) with few exceptions (Table S1; Fig. 4). A subordinate population of target basalts (approximately 26%) from the eastern, southeastern, western, and northwestern sectors has higher REM between approximately 1.8 and 5% in average. However, majority of samples (approximately 64%) from the southern sector of the crater rim only show higher average REM (approximately 3.5%).

Extensive samples of shocked basalt from approximately 2–3 km west of the crater center in the downrange direction from the Kalapani dam (KPD) and Khini village (KHN), wall to the west of crater rim (CRW), and Saraswati village (SWT) (Fig. 1b) (n = 59) have been studied to evaluate shock-induced REM of target basalts. An observation on R2 ratios (Table S1) suggests that the number of shocked basalt samples with REM higher than that of the average unshocked basalt (approximately 0.70%) always constitutes a dominant portion of each sample population, and the R2 ratio is found to be the highest (approximately 3.7:1) for the samples collected from the west of the crater rim (CRW) (Fig. 4). Nearly half of the sample population (approximately 57%) from the west of the crater to the downrange shows REM less than approximately 1.50% with an average comparable to the unshocked target. The rest of the samples, however, show higher REM; approximately 34% of samples show REM (between approximately 1.7 and 7%) with an average that is approximately 5 times higher over the unshocked target, and a few samples (approximately 9%) show the highest REM >12% (Figs. 4f, g, and i).

Alternating Field (AF) and Thermal Demagnetization of NRM of Lonar Basalts

Difference in opinion also exists on the status of low-coercivity and low-temperature (LC_LT) magnetization component of shocked basalts from the Lonar crater (Rao and Bhalla 1984; Louzada et al. 2008). To examine any possible relationship between high coercivity and high temperature (HC_HT) and LC_LT directions of shocked basalts with reference to the trajectory of asteroid impact, we have investigated the AF and thermal demagnetization behavior of NRM of unshocked and shocked Lonar basalts (Table S2).

Our observations show that the LC_LT component in both the unshocked and shocked Lonar basalts is easily erased by peak AF demagnetization between approximately 10 and 25 mT or by thermal demagnetization to less than approximately 300 °C (Fig. 5). The LC_LT direction of the unshocked basalts from the base of Durga Tegri (DT/100°-2.93), and the east of this location (A5/102°-4.61)*, and the apparently unshocked basalts from the far south of the Lonar crater (A1/188°-4.04) (not shown in figure) (n = 25, number of samples is approximately 200% higher over Arif et al. 2011) shows high variation in declinations mostly within NW and NE, and inclinations of both reversed and normal polarities with a site mean of D = 348.6°, I = +56.2°, (k = 3.5, α95 = 32.8°) (Fig. 6b). The LC_LT direction of shocked basalts from all around the crater rim and at approximately 2–3 km west of the crater center in the downrange direction from the Kalapani dam (KPD/233°-2.17), Khini village (KHN/251°-2.92), wall to the west of crater rim (CRW/267°-2.07), and Saraswati village (SWT/300°-2.10) (Fig. 1b) (n = 109, number of samples is approximately 110% higher over Arif et al. 2011) similarly shows wide variation in declination between NW and NE, and inclinations of both normal and reverse polarities with a site-mean of D = 357.7°, I = +49.7° (k = 3.2, α95 = 10.7°) (Fig. 6h). The LC_LT components of both the unshocked and shocked target basalts are statistically identical to present day field (PDF) direction (Dec = −0.6°, Inc = +28.8°; IGRF, International Geomagnetic Reference Field 2011), with approximately 20° inclination bias that is almost the size of α95 (Figs. 5 and 6).

Figure 5.

 Representative Zijderveld plots of alternating field (AF) and thermal demagnetization data of (a) unshocked target basalts from Durga Tegri (DT), and (b–k) shocked basalts from around the Lonar crater rim and at approximately 2–3 km west of the crater center in the downrange direction; open and closed circle symbols represent projections in vertical and horizontal planes, respectively. Note that the LC_LT component for all the samples erased between 10 and 25 mT alternating field or approximately 200–300 °C unblocking temperature.

Figure 6.

 Equal area stereographic plots of unshocked (a, b) and shocked Lonar basalts (c–h) from the Lonar crater: (a) high-coercivity and high-temperature (HC_HT) component of the unshocked basalts from the far south of Lonar crater (A1/188°-4.04, cluster 1), and from Durga Tegri and its adjacent east (DT/100°-2.93, A5/102°-4.61, cluster 2); (b) low-coercivity and low-temperature (LC_LT) magnetization component of both clusters 1 and 2. Other figures show HC_HT magnetization components of shocked target basalts from (c) most of the crater rim, (d) the southeastern sector crater rim, (e) the southern sector crater rim, (f) the southwestern sector crater rim, and (g) at approximately 2–3 km west of crater center in the downrange direction. (h) LC_LT component of all shocked basalts from c to g; and (i) a summary diagram on figures c–g showing symmetrical disposition of HC_HT component of shocked basalts, with reference to the east–west plane of impact and impact direction; for sample locations, see Fig. 1b. Note that data with positive inclinations are shown in solid circles or solid squares, and those with negative inclinations in open circles.

The main difficulty in evaluating any possible variation in HC_HT component of Lonar shocked basalts due to asteroid impact is the lack of any consistent orientation of HC_HT component of unshocked target basalts that can be used as a reference (cf. Louzada et al. 2008). To get a better solution, the paleomagnetic data of unshocked target basalts from the Durga Tegri (DT/100°-2.93) and the east of this location (A5/102°-4.61), and apparently unshocked basalt samples from the far south of the Lonar crater (A1/188°-4.04) (not shown in figure) are analyzed in the present study. Interestingly, the unshocked basalts from the south of the Lonar crater (n = 6) yield a point concentration of data with a site-mean HC_HT direction of D = 108°, I = +47.4° (k = 154.6, α95 = 5.9°) (cluster 1 in Fig. 6a; Table S2), which is different from the mean paleo-Deccan magnetization direction (D = 157.6°, I = +47.4°, α95 = 1.9°, Vandamme et al. 1991; also see Pal and Bhimasankaram 1971; Athavale and Anjaneyulu 1972). The virtual geomagnetic pole (VGP) of these unshocked basalts lies at 5.3°S, 133.5°E (A95 = 6.2°). The unshocked target basalts from the Durga Tegri (DT) and its adjacent east (A5) (n = 19) also show point concentration of data, however, with a different site-mean HC_HT direction of D = 196.1°, I = +68.9° (k = 702.4, α95 = 13.4°) (VGP: 16.2°S, 66.4°E [A95 = 19.7°]) (cluster 2 in Fig. 6a).

The HC_HT component of most of the shocked basalts (n = 43) from around the crater rim, except those from the southeastern (138°-0.88), southern (190°-0.86), and southwestern sectors (229°-0.82) (Fig. 1b), is oriented mostly toward the ESE to SSE with a site-mean of D = 120.5o, I = +34.2o (k = 106.7, α95 = 10.3°) (Fig. 6c). The HC_HT component of shocked basalts from southeastern sector (138°-0.88) (n = 8) has a site-mean of D = 88.8°, I = +66.6° (k = 106.0, α95 = 5.8o) (Fig. 6d); those from the southern sector (190°-0.86) (n = 9) show a different site-mean of D = 51o, I = +15.6o (k = 12.5, α95 = 16.1°) (Fig. 6e), and all those from the southwestern sector (229°-0.82) (n = 10) yield reverse polarity inclination with a site-mean of D = 117.3°, I = −27.4° (k = 52.1, α95 = 7.1°) (Fig. 6f). We exclude calculations on VGP on our studied shocked basalts in the present study because magnetization history of these samples could be different from other basaltic rocks that quenched under normal Earth’s magnetic field.

We have also examined the HC_HT component of shocked basalts from approximately 2–3 km west of the crater center in the downrange direction (n = 39) from Kalapani dam (KPD/233°-2.17), village Khini (KHN/251°-2.92), wall to the west of the crater rim (CRW/267°-2.07), and from the village Saraswati (SWT/300°-2.10) (Fig. 1b). Like most of the shocked basalts from around the Lonar crater rim (Fig. 6c), these target basalts show a site-mean of D = 120.0°, I = +38.0° (k = 165.4, α95 = 8.3°) (Fig. 6g).

The target basalt samples, occurring at approximately 2–3 km west of the crater center in the downrange and having high NRM/χ or REM values compared with those of the unshocked target, yield an LC_LT component and an HC_HT component (Fig. 5f and 5g). The HC_HT components of these shocked basalts are different from the characteristic paleo-Deccan direction at Lonar (cf. Vandamme et al. 1991), but are similar to those obtained for the shocked basalts from most of the crater rim having normal NRM/χ or REM values comparable to those of the unshocked target. Some of these shocked basalts from downrange only yield single magnetization HC_HT component (Figs. 5h and j). On the other hand, the shocked basalts from the southeastern, southern, southwestern crater rim sectors with both normal and high NRM/χ or REM values yield different HC_HT directions in comparison with those shocked basalts from most of the crater rim and approximately 2–3 km west of the crater center in the downrange (Figs. 5c–e).

Discussion

Our previous studies on the distribution of ejecta and structural deformations of basalt flows around the crater rim, together with AMS of unshocked and shocked basalts from around the Lonar crater, suggest that the impactor asteroid most probably struck the preimpact surface from the east at an angle of between approximately 30 and 45o (Misra et al. 2010). Our present investigation suggests that the distribution of K3 susceptibility axes, NRM/χ, REM, and HC_HT magnetization component directions of shocked target basalts from around the crater rim and to the west of the crater in the downrange are almost symmetrically distributed with reference to the east–west plane of impact and direction of impact.

The present observation on elaborate AMS data confirms our early findings that the degree of anisotropy (P′) of the Lonar shocked basalts (collected from the crater rim and from approximately 2–3 km west of the crater center in the downrange direction) has reduced by approximately 2% on average when compared with the value for unshocked target basalt (approximately 1.03); the shape of AMS ellipsoids also varies from oblate to prolate shape in comparison with the oblate shape for the unshocked target basalt (Fig. 2).

In stereographic projections, the distribution of AMS axes (K1, K2, K3) mostly changes from a biaxial distribution for the unshocked target basalt to triaxial for the shocked basalt, although scatter in data exists (Fig. 3). In a few cases, e.g., from the northern (358°-0.92) and western (272°-0.92) sectors of the crater rim (Fig. 3k and 3h), a transitional type distribution of AMS axes between biaxial and triaxial types is also noticed. Besides the general westward shift, the K3 axes also show either southwestward or northwestward shift for the shocked basalts lying mostly to the south and north of the east–west plane of impact, respectively (Figs. 3 and 7). As the displacement of K3 axes in the shocked basalts is sensitive to propagation of high impact stress (>3 GPa) (Nishioka and Funaki 2008), it can be concluded that the impact stress could have branched out in the downrange direction into major southwest and northwest components from the crater’s center immediately after the impact, making acute angle to the impact direction (Fig. 7).

Figure 7.

 A schematic diagram showing probable direction of shock pressure propagation just after asteroid impact at the Lonar crater, India. E, SE, S, SW, W, NW, and N represent eastern, southeastern, southern, southwestern, western, northwestern, and northern sectors of crater rim, respectively; abbreviations are as in Fig. 1b.

Our observation on the distribution of K3 axes around the Lonar crater is perhaps very similar to the distribution of ejecta around the crater rim in oblique impact experiments. Gault and Wedekind (1978) showed that during oblique impact less than 45°, a forbidden zone in the ejecta distribution first appeared uprange from the crater, and then subsequently at shallower incidences, a second zone appeared downrange, both extending from the crater rim with bilateral symmetry about the path of the projectile trajectory. If this ejecta distribution in the oblique impact experiment reflects the distribution of impact-generated shock pressure around the crater, our AMS data from the Lonar crater are also in accordance with the bilateral branching of the impact-induced shock pressure at least in the downrange direction just after the low-angle asteroid impact (Fig. 7).

The shocked basalts, in general, show an increase in NRM/χ in variable proportions compared with the unshocked basalts in the Lonar crater (Fig. 4). The distribution of NRM/χ in shocked basalts around the crater rim appears to be systematic with reference to the east–west plane of impact. The shocked basalts in the eastern sector of the crater rim in the uprange direction (Fig. 4b) and in the western half of the crater rim including the southwestern, western, and northwestern sectors in the downrange (Figs. 4e, h, and j) show minimum increase in average NRM/χ (approximately 1.5 times) compared with the unshocked target. These two zones on the crater rim in fact represent the forbidden zones of ejecta distribution around the crater rim in oblique impact experiments (Gault and Wedekind 1978), and most likely represent the zones of low shock pressure. Our present observation on AMS also confirms major branching of the impact stress into the southwest and northwest components (Figs. 3 and 7) leaving the western sector crater rim in between these two major shock pressure components in the downrange as a low shock pressure zone during an oblique impact from the east. Although variable, the average NRM/χ of target basalts in the southeastern, southern, and northern sectors of the crater rim, which are oriented oblique to the east–west plane of impact and have experienced high shock pressure (Fig. 7), shows a wide increment (approximately 1.4–16.4 times) in comparison with the unshocked target (Figs. 4c, d, and k). The distant target rocks in the downrange direction at approximately 2–3 km west of the crater center, which could have experienced high shock pressure during oblique impact (cf. Pierazzo and Melosh 2000), also show increase in NRM/χ (approximately 1.1–25.8 times) over the unshocked target (Figs. 4f, g, and i). So it could be concluded that the NRM/χ of the target rocks increases with shock pressure during impact, and its variations around the crater rim and in downrange are dependent on the distribution and variation in the intensity of the shock pressure during an oblique impact.

It is observed in our present study that the REM values for most of the unshocked basalt samples are lower than that of the typical Earth (approximately 1.5%). The exact reason for this observation is not clear at our present stage of knowledge, but it appears that it could be related to the complex evolution of continental flood basalts that is characterized by multiple eruptions and reheating of the stratigraphically older flows (cf. Vandamme et al. 1991). It was also argued that the high-field processes like asteroid impact could have resulted in high REM ratio of the target rocks (Wasilewski and Dickinson 2000; Gattacceca and Rochette 2004); our observation on the Lonar crater is, however, different (Fig. 4; Table S1). Most of the shocked target basalt samples from around the crater rim (approximately 67% of 99 total samples), where the important magnetization factor besides the variable shock pressure could be the impact-generated magnetic field (Weiss et al. 2010), show average REM (approximately 0.37%) that is nearly half of the unshocked target. Although a subordinate proportion of samples (approximately 31%) have REM (average approximately 2.7%) approximately 4 times higher than that of the average unshocked target basalt, they do not show any systematic distribution around the crater rim with reference to the east–west plane of impact at Lonar. Weiss et al. (2010) reported REM of Lonar impact spherules and impact-melts, which are mostly less than approximately 1%, and these low REM values could be attributed to the weak impact-generated magnetic field (<100 μT) that existed during the formation of the Lonar crater. So, it appears that the weak impact-generated magnetic field could be one of the factors besides the variable impact stress around the rim of the Lonar crater that resulted in overall low REM of the target basalts from around the crater rim; however, this aspect of magnetization of Lonar target rocks needs further investigation.

On the other hand, nearly half of the target basalt samples from approximately 2–3 km west of the crater center in the downrange, where the impact-generated stress dominates in oblique impact (Pierazzo and Melosh 2000) and is perhaps the only important factor controlling the impact-induced magnetization, have REMs approximately 5 times higher over the unshocked target basalt. So, it can be concluded that the high impact stress (>3 GPa) could have also increased REM of the target basalt; however, the variable distribution of shock pressure around the rim of the Lonar crater aided with some unknown impact-induced magnetic field that existed on the Lonar crater could have resulted in low REM of the shocked basalts on the crater rim.

Observations on circular/semicircular gravity and magnetic anomalies (approximately 2.5 mGal and 550 nT, respectively) over the Lonar lake suggest that the asteroid impact could have modified the magnetization vector and density of the target (Deccan Traps) up to a depth of approximately 500–600 m below the surface (Rajasekhar and Mishra 2005). The HC_HT component of shocked basalts of Louzada et al. (2008) perhaps supports this idea because the topmost (fourth) basalt flow on the Lonar crater wall, which experienced the highest stress during the oblique asteroid impact, shows a sudden change in HC_HT direction. While the second, third, and fifth flows (the flows exposed on the Lonar crater wall are numbered from the base of the crater) show HC_HT direction close to the paleo-Deccan direction at Lonar (D = 157.6o, I = +47.4o, α95 = 1.9o, Vandamme et al. 1991; also see Pal and Bhimasankaram 1971; Athavale and Anjaneyulu 1972), the fourth basalt flow has acquired a different orientation of HC_HT magnetization of D = 126.4o, I = +44.7o, α95 = 4.1o (Louzada et al. 2008). This sudden change in HC_HT direction between the second and fourth flows of Lonar target was also observed by Rao and Bhalla (1984). Our present strength of data on the fourth basalt flow from the crater rim, except the southeastern, southern, and southwestern sectors, also shows a similar site-mean of D = 117.9o, I = +39.6o, α95 = 4.6o (Fig. 6c). The average HC_HT direction in the shocked fourth basalt flow from around the crater rim is although similar to one of the two relatively less prominent orientations of unshocked basalt at Lonar (cluster 1 in Fig. 6a), considerable scatter in orientations of HC_HT component for this shocked fourth flow is seen (Fig. 6c). Additionally, the fourth flow on the crater rim from the southeastern sector has easterly directed HC_HT direction (D = 88.8°, I = +66.6°, α95 = 5.8o) (Fig. 6d), southern sector has a northeasterly directed HC_HT component (D = 51o, I = +15.6o, α95 = 16.1o) (Fig. 6e), and those from the southwestern sector show a negative inclination of HC_HT direction (D = 117.3o, I = −27.4o, α95 = 7.1o) (Fig. 6f), which are absent in the unshocked target basalt populations and could be acquired due to asteroid impact. The HC_HT direction of unshocked Deccan Trap flows at Lonar and nearby areas is variable (Fig. 6a) and is different from the mean Deccan direction (D = 157.6°, I = +47.4°, α95 = 1.9°, cf. Vandamme et al. 1991), and it is argued that significant scatter between site mean directions within a single Deccan flow may be due to remaining early or late overprints from later reheating. However, our further observations on the AMS data show that unlike the unshocked target basalt samples from the Durga Tegri (DT/100o-2.93) and to its east (A5/102o-4.61), the apparently unshocked samples from the far south of the Lonar crater (A1/188o-4.04) show triaxial distribution of AMS axes, and the AMS ellipsoids vary in shape from oblate to prolate type (Fig. S4), which indicate that the samples were shocked by some unknown process and can safely be excluded from the present discussion.

There could be two possible explanations on the above-mentioned observations. The fourth target basalt flow on the Lonar crater rim did not remagnetize during the impact and could represent the primary Deccan magnetization component. This possibility was suggested by Louzada et al. (2008), although their data set was spatially and temporally too limited to average out secular variation. Rao and Bhalla (1984) also estimated the HC_HT component of Lonar target basalt samples collected from the inner walls of the crater along two profiles in ENE–WSW direction (ten sites: six from eastern profile, four from western profile). Their site-mean HC_HT direction (D = 136o, I = +42o, K = 71.4, α95 = 5.2o), which was computed on data collected mostly from the fourth and second basalt flows, was also different from the mean paleo-Deccan direction (Vandamme et al. 1991) and the average orientation observed for the unshocked target basalts from the east of the Lonar crater (Fig. 6a). So, the existing idea that the observed HC_HT component of the fourth basalt flow around the Lonar crater rim could represent the primary Deccan magnetization component is not well supported by data.

Alternatively, a more favorable idea is that the fourth basalt flow exposed on the Lonar crater rim could have acquired an HC_HT magnetization component during the impact. This is because (1) the HC_HT magnetization component of the fourth basalt flow is different from the average Deccan direction at Lonar (Vandamme et al. 1991) and that of the unshocked target from the east of the Lonar crater (Fig. 6a, cluster 2); (2) these are symmetrically disposed with reference to the east–west plane of impact; and (3) these are directed to the uprange direction and make obtuse angle with the direction of impact (Fig. 6i). Observation on the systematic displacement of the K3 susceptibility axes (Fig. 3) suggests that the Lonar target basalt around the crater rim could have experienced a shock pressure >3 GPa particularly in the northern and southern sectors of the crater rim (cf. Nishioka 2007; Nishioka et al. 2007; Nishioka and Funaki 2008). This intense shock pressure was perhaps sufficient to remagnetize the HC_HT component of the Lonar target basalt within an impact-induced magnetic field (Cisowski and Fuller 1978).

The LC_LT magnetization components of both the unshocked and shocked target basalts are statistically identical to PDF direction (Table S2), and it could be the chemical (and/or viscous) remanent magnetization acquired during the last 570 ± 47 ka subsequent to crater formation (cf. Louzada et al. 2008).

It is understood that the shock heating of target basalts and ejecta blocks in and around the Lonar crater did not exceed approximately 200 °C (Louzada et al. 2008). Our experiments on variation in magnetic susceptibility with temperature (Fig. S1d) and unblocking low-temperature component in Zijderveld plots (Figs. 5b, f, and k) also suggest that the shocked basalts around the crater rim and downrange to the west had remagnetized at around 200–300 °C. The mineral that mostly responds to remagnetization at low temperature is Ti-rich titanomagnetite with its low Curie point at around 235 °C (Figs. S1d). Also, our low-temperature susceptibility observations have indicated that χ-peak can occur anywhere in the temperature range of −163 to −155 °C (Figs. S1c, e, f, and g), possibly arising from shifts in the isotropic points of magnetite grains because of low concentrations of Ti in them. All such cases represent the presence of MD grains of magnetite. The χ-peak susceptibility shown by MD grains is suppressed for the SD states due to shape anisotropy of the SD grains that possibly occurred during remagnetization. Hence, our present observation on AMS, NRM/χ, partly on REM and paleomagnetism of target basalts, suggests that the impact-induced remagnetization of Lonar shocked basalts must have taken place under low-temperature–high-impact shock pressure conditions, where predominantly PSD and SD Ti-rich titanomagnetites perhaps were the magnetic remanence carriers.

Conclusions

(1) The degree of anisotropy (P′) of the Lonar basaltic target rocks is decreased by approximately 2% due to shock pressure generated by asteroid impact; the shapes of the susceptibility ellipsoids include both oblate and prolate types in contrast to the oblate shape of the unshocked target.

(2) The distribution of AMS axes in stereographic projections changes from biaxial type for unshocked basalts to mostly triaxial type in shocked basalts, although scatter in data and transitional type distribution also exists. Besides the major westward shift of the K3 axes, the shocked basalts also show major southwestward and northwestward orientations of K3 axes for those members lying to the south and north of the east–west plane of impact, respectively. This distribution of K3 axes is symmetrical about the east–west plane of impact and makes acute angle to the direction of impact in the downrange direction; it suggests a major branching of impact stress toward the SW and NW from the point of impact just after the impact of asteroid on Lonar target basalt.

(3) The NRM/χ of the target basalts is proportionally increased with the impact shock pressure. The target basalts from the relatively low shocked eastern sector of the crater rim and the western half of the crater rim including the southwestern, western, and northwestern sectors show only approximately 1.5 times increment of NRM/χ over that of the average unshocked target basalt; a significant increase in NRM/χ of approximately 1.4–16.4 times is seen for the target that are situated oblique to the direction of impact within the high shocked part of the crater rim. The shocked targets from approximately 2–3 km west of the crater center in the downrange also show an increase in NRM/χ of approximately 1.1–25.8 times over the unshocked target.

(4) Our study does not show any systematic increase in REM of the target basalts from around the crater rim due to asteroid impact. The shocked targets away from the crater center at approximately 2–3 km west of the crater center in the downrange direction show approximately 5 times increase in REM (average approximately 3.63%) over the unshocked target.

(5) The topmost basalt flow (fourth flow) around the Lonar crater rim, which experienced the highest shock due to asteroid impact, has acquired an HC_HT component that is different from the mean paleo-Deccan direction (C29R, Vandamme et al. 1991) at Lonar. This component is systematically oriented with reference to the east–west plane of impact, and directed in the uprange direction, making an obtuse angle with the impact direction. The LC_LT components in both the shocked and unshocked basalts are statistically identical to PDF direction and it could be chemical and/or viscous remanent magnetization acquired during the last approximately 570 ± 47 ka after the impact.

Footnotes

  • *

    We have to ignore one more location (A6: 113°-7.30) from this population because we came to know that this site was affected by dynamite blast during a dam construction, and this shock effect is found to have affected the AMS properties of the target from this site.

Acknowledgments— We are grateful to W. Hastie, J. Spray, S. Gilder, J. Plado, and L. Carporzen for their valuable comments on the early version of the manuscript. We thank Director, Indian Institute of Geomagnetism, Navi Mumbai, India, for funding this research project. S. M. wishes to acknowledge the Department of Geological Sciences, SAEES, University of KwaZulu-Natal, Durban, South Africa, for providing infrastructural facilities in completing this research.

Editorial Handling— Dr. John Spray

Ancillary