Ajoy K. Baksi

40Ar/39Ar Dating of the Brunhes-Matuyama Geomagnetic Field Reversal

Magnetostratigraphic studies are widely used in conjunction with the geomagnetic polarity time scale (GPTS) to date events in the range 0 to 5 million years ago. A critical tie point on the GPTS is the potassium-argon age of the most recent (Brunhes-Matuyama) geomagnetic field reversal. Astronomical values for the forcing frequencies observed in the oxygen isotope record in Ocean Drilling Project site 677 suggest that the age of this last reversal is 780 ka (thousand years ago), whereas the potassium-argon-based estimate is 730 ka. Results from 4039; Ar incremental heating studies on a series of lavas from Maui that straddle the Brunhes-Matuyama reversal give an age of 783 + 11 ka, in agreement with the astronomically derived value. The astronomically based technique appears to be a viable tool for dating young sedimentary sequences.

Intercalibration of 40Ar39Ar dating standards

A number of standards used for KAr and 40Ar39Ar dating have been intercalibrated. Multiple splits (∼ 15–40 mg each) of MMhb-1, SB-3 Biotite, LP-6 Biotite 40–60#, GA1550 Biotite, Fish Canyon Tuff-3 Biotite, Taylor Creek Rhyolite Sanidine, Bern 4 Muscovite, Bern 4 Biotite and Pennsylvania State University Orthoclase-1A were irradiated in capsules, fused and analyzed. In replicate analyses, most of these standards proved to be homogeneous in 40Ar★39ArK ratio (age) K, Ca, and 36Ar contents. MMhb-1 is inhomogeneous in age at the ∼ 15-mg level, indicating it is unsuitable for use as an interlaboratory standard using material taken straight from the bottle; the orthoclase sample is also inhomogeneous at the ∼ 30-mg level. Quoted relative to an age of 162.9 Ma for the SB-3 Biotite standard, total fusion ages are: LP-6 Biotite 40–60# = 128.1 Ma; GA1550 Biotite = 97.8 Ma; Fish Canyon Tuff-3 Biotite = 27.95 Ma; Taylor Creek Rhyolite Sanidine = 28.0 Ma; Bern 4 Muscovite = 18.51 Ma; Bern 4 Biotite = 17.21 Ma. These ages are in good agreement with the (absolute) KAr ages of these minerals. Different preparations of biotite and sanidine standards from a welded tuff such as the Fish Canyon Tuff may not be of identical ages. Tests on submilligram splits reveals that SB-3 Biotite, GA1550 Biotite, Fish Canyon Tuff-3 Biotite and Taylor Creek Rhyolite Sanidine are suitable for use as monitors in laser work; the sanidine specimen appears to be the most suitable for use as the primary monitor for 40Ar39Ar dating.

Geochronological studies on whole-rock basalts, Deccan Traps, India: evaluation of the timing of volcanism relative to the K-T boundary

KAr and 40Ar39Ar incremental heating studies have been carried out on whole-rock basalts spanning different sections of the Deccan Traps. KAr dates typically fall in the range ∼ 55–65 Ma and, for an individual section, are commonly out of stratigraphic order; indicating post-crystallization loss of 40Ar∗ from some samples. Specimens from six lava flows spanning a composite Western Ghats section and another six from sections in the southern, western, and northern areas of the Deccan, were selected for 40Ar39Ar step-heating studies. Many of these specimens yield low temperature steps with apparent ages 70 Ma) older than crystallization values. Such spectra are characteristic of fine-grained basalts exhibiting 39Ar recoil loss and could involve: (1) 39Ar internal redistribution, out of K-rich sites into K-poor phases, and/or (2) 39Ar recoil loss out of the sample. Comparison of total gas ages with KAr dates permitted constraints to be placed on these two mechanisms. For some rocks, plateau ages were recovered from the intermediate-high temperature steps in the range 64–66 Ma, with 1σ errors of 0.5–1.0 m.y. A dyke specimen from the western Deccan, with a KAr date of ∼ 97 Ma, shows a descending staircase type of age spectrum and its estimated crystallization age is ∼ 65 Ma. Two lava flows from the central Deccan with low KAr dates ( ∼ 50–55 Ma) yield crystallization ages of ∼ 65 Ma; KAr dates deviating significantly from ∼ 65 Ma, do not reflect crystallization ages. Integration of the age data with the magnetic polarity of the different sections of lavas suggests the main reversed polarity epoch trapped in the Deccan volcanic province is chron 29R, which includes the Cretaceous-Tertiary boundary (estimated age 64.5 Ma). The plateau ages presented herein tend to be slightly older ( ∼ 1%) than the age of chron 29R; this probably results from minute amounts of 39Ar recoil loss out of the sites degassed in the intermediate-high temperature range. The Western Ghats section of the Deccan Traps, representing > 80% of the exposed material, was extruded in ∼ m.y.

Petrogenesis and timing of volcanism in the Rajmahal flood basalt province, northeastern India

A suite of rocks from the Rajmahal-Bengal-Sylhet Traps of northeastern India has been analyzed to ascertain the timing and duration of volcanism and elucidate their petrogenesis. 40Ar39Ar tep-heating studies identified specimens that suffered post-crystallization loss of 40Ar∗ and indicate the Rajmahal Province was extruded in ∼2 Ma around 117 Ma ago. Trace- and rare-earth-element data suggest the existence of three different types of magmas. Rajmahal quartz tholeiites were formed from primary melts, following considerable gabbroic fractionation. Bengal Trap olivine tholeiites represent lavas formed by large partial melting of mantle material, leaving garnet in the residue. Alkali basalts in the Bengal Traps appear to represent partial melts of mantle containing LILE-enriched sections, rather than very small ( < 2%) melts of a garnet lherzolite source. Whole-rock δ18O-values for slightly altered tholeiites fall in the range +5.9 to +6.6‰, indicating mantle-derived melts that have suffered minor crustal contamination; two alkali basalts, formed following considerable crystal fractionation of primary magmas, yield values of ∼ + 7.2‰. SrNd isotopic analyses show two different contamination trends, overlapping those observed in an earlier study of surface Rajmahal quartz tholeiites, with the most primitive material showing 87Sr86Sr ∼ 0.70400, 143Nd144Nd ∼ 0.51280 at 117 Ma ago. The Bengal Trap olivine tholeiites were formed following assimilation of high-87Sr86Sr (granulitic?) material. The main contamination trend includes quartz tholeiites from the Rajmahal Traps and alkali basalts from the Bengal Traps. Tholeiites, showing considerable isotopic modification, suggest ingestion of a high-Sr component, unlikely to be upper-crustal material; for the alkali basalts, with high Sr (∼ 1000 ppm) and Nd (∼ 55 ppm) contents, incorporation of a few percent of “exotic” material (in the source region?) is indicated. Carbonatite is the probable contaminant, strengthening the postulated link between flood basalt volcanism and carbonatite-lamproites in this area. The occurrence of two lavas with reversed magnetic polarity, in association with the 40Ar39Ar ages reported herein, suggests the ISEA reversed event is displayed in the lavas of the Rajmahal Traps.

Widespread Early Cretaceous flood basalt volcanism in eastern India: geochemical data from the Rajmahal-Bengal-Sylhet Traps

Abstract K-Ar dating of a set of flood basalts from eastern India indicates that volcanism ∼ 115 Ma ago was much more voluminous than heretofore recognised, covering an area of at least 2·105 km2. Separation of India from Antarctica began ∼ 125 Ma ago and in this rift dominated regime, voluminous flood basalts were extruded with the Kerguelen Hotspot perhaps serving as the heat source. We report here on the occurrence of both alkali basalts and olivine tholeiites from the Rajmahal area; preliminary data indicate that more than one type of primary magma was present, one of which resulted from partial melting of a section of the mantle containing metasomatised veins.

A geomagnetic polarity time scale for the period 0–17 Ma, based on 40Ar/39Ar plateau ages for selected field reversals

Utilizing 40Ar/39Ar plateau ages for selected field reversals at ∼ 1, 2, 10, 16 and 34 Ma, a geomagnetic polarity time scale (GPTS) is constructed, following the technique of Cande and Kent [1992]. This differs markedly from previous GPTS, especially in middle-late Miocene time; the ages for the chrons in post-Miocene time are in agreement with values derived by the astrochronological technique. Refinements in this GPTS await acquisition of 40Ar/39Ar plateau ages for field reversals in the time frames 3–8 Ma and 11–15 Ma.

Reevaluation of Plate Motion Models Based on Hotspot Tracks in the Atlantic and Indian Oceans

Plate motion models based on hotspot tracks in the Atlantic and Indian Oceans predict minimal movement (less than a few millimeters per year) between these hotspots and their counterparts in the Pacific Ocean for the past ∼100 m.yr., whereas plate circuit exercises indicate relative motions of ∼20 mm/yr. Hotspot‐based models also suggest that the Rajmahal Traps, India, were located ∼1000 km away from the Kerguelen hotspot at ∼115 Ma, and the Deccan Traps, India, were located a similar distance from the Reunion hotspot at ∼65 Ma; this is at odds with conclusions derived from paleomagnetism, plate circuits, and geochemical parameters that suggest a genetic link between flood basalt provinces in India and hotspots in the Indian Ocean. These divergent views may be explained by plume action ∼1000 km from its center or errors in the hotspot motion models. The latter hypothesis is scrutinized in this article by examination of the radiometric ages for hotspot tracks in the Atlantic and Indian Oceans. The 40Ar/39Ar step‐heating data for rocks defining the tracks of the Reunion and Kerguelen hotspots in the Indian Ocean and the Great Meteor and Tristan da Cunha hotspots in the Atlantic Ocean are critically reexamined. Of ∼35 such ages utilized for deriving plate motion models for the past 130 m.yr., at best, only three (∼32, ∼50, and ∼52 Ma) in the Indian Ocean and one (∼65 Ma) for the Atlantic Ocean may be treated as crystallization ages. Conclusions based on hotspot track modeling for Late Cretaceous to Eocene time are suspect, and those for the Early to Late Cretaceous period are untenable. In the absence of precise age data for the tracks of hotspots in the Atlantic and Indian Oceans, and inconsistent age progressions noted within a single volcanic chain, plate circuit models serve as the superior technique for tracing plate motions over the past ∼100 m.yr. The degree of (absolute) motion for hotspots remains contentious. For most hotspots, models indicating detectable movement (>20 mm/yr) for the past ∼100 m.yr. are favored; the Kerguelen plume was situated close to its current position (49°S) at ∼115 Ma.

40Ar/39 dating of the Siberian Traps, USSR: Evaluation of the ages of the two major extinction events relative to episodes of flood-basalt volcanism in the USSR and the Deccan Traps, India

40Ar/39Ar incremental-heating studies have been carried out on three whole-rock specimens from the Siberian Traps. A basalt lava flow from the lowermost horizon yields an age of 238.4 ±1.4 Ma (1σ error). A second basalt lava flow from the top of the section, ∼800 m above the first specimen, yields an age of 229.9 ±2.3 Ma, indicating that the duration of volcanism was ∼5-10 m.y. A doleritic dike intrusive into the lower parts of the Siberian Traps contains excess argon and yields an isochron age of 234 ±7 Ma. Critical reexamination of relevant radiometric data relating two separate episodes of flood-basalt volcanism to global faunal extinctions suggests the volcanic event forming the most voluminous sections of the Deccan Traps, India, was coincident to within ±1 m.y. with the time of the Cretaceous/Tertiary boundary. However, the onset of volcanism in the Siberian Traps apparently occurred at a time postdating that of the Permian/Triassic boundary.

Critical evaluation of the age of the Deccan Traps, India: Implications for flood-basalt volcanism and faunal extinctions

Attempts to evaluate the possible link between flood-basalt volcanism and the faunal extinctions observed at the K/T boundary must rely on precise knowledge of the amount of such volcanism that took place ∼65 Ma. The current geochronological data do not permit close resolution of either the timing or the duration of Deccan volcanism; the age of the bulk of the material is placed in the range ∼70-60 Ma. Further resolution of this hypothesis awaits acquisition of high-quality 40 Ar- 39 Ar plateau ages on key sections of the Deccan Traps.

Testing the accuracy of the geomagnetic polarity time-scale (GPTS) at 2–5 Ma, utilizing40Ar/39Ar incremental heating data on whole-rock basalts

Abstract Recent40Ar/39Ar dating and astrochronologic calculations indicate that the K-Ar derived GPTS underestimates the ages of geomagnetic field reversals in the interval ∼ 2-0 Ma. We report tests to check the accuracy of K-Ar data utilized to date the Reunion Event (∼ 2.1 Ma) and part of the early Pliocene portion (∼ 4.6 Ma) of the GPTS.40Ar/39Ar incremental heating studies on whole-rock basalts from the island of Reunion yield an age of2.14 ± 0.03 Ma for the Reunion Event, somewhat older than the earlier K-Ar value of2.07 ± 0.02 Ma on the same set of lava flows. Study of a single specimen from the Newer Volcanics, Australia, indicates that the earlier K-Ar date was accurate; but the new40Ar/39Ar results yield a plateau age (4.56 Ma) with a standard error of ∼ 0.6%, as compared to the K-Ar value ∼ 3%. The improved precision and accuracy resulting from such40Ar/39Ar incremental heating studies will play a key role in calibrating more precisely the GPTS for the interval 10-0 Ma.