April 2010 LIP of the Month

Short duration for the eruption of the Grande Ronde Basalt lavas of the Columbia River Basalts, USA.

Excerpts modified from paper in press in Lithos, entitled ‘New ⁴⁰Ar/³⁹Ar dating of the Grande Ronde lavas, Columbia River Basalts, USA: implications for duration of flood basalt eruption episodes’.

T.L. Barry¹, S. Self¹, S.P. Kelley¹, S. Reidel², P. Hooper¹, M. Widdowson¹

¹Volcano Dynamics Group, CEPSAR, Open University, Walton Hall, Milton Keynes, MK7 6AA, UK.; t.l.barry@open.ac.uk

² Washington State University, Tri-Cities Campus, Richland, WA 99354, USA.

Summary

Grande Ronde Basalt (GRB) lavas represent the most voluminous eruptive pulse of the Columbia River-Snake River-Yellowstone hotspot volcanism. With an estimated eruptive volume of 150,000 km³, GRB lavas form at least 66% of the total volume of the Columbia River Basalt Group (Fig. 1) [using the latest estimates of total CRB-related lava volume (Camp et al., 2003; Camp & Ross, 2004) compared with 90% of the volume of the CRBG as previously defined by Tolan et al. (1989)]. New ⁴⁰Ar/³⁹Ar dates for GRB lavas reveal they were emplaced within a maximum period of 0.42 ±0.18 My. A well-documented stratigraphy indicates at least 110 GRB flow fields (or individual eruptions), and on this basis suggests an average inter-eruption hiatus of less than 4,000 years. Isotopic age-dating cannot resolve time gaps between GRB eruptions, and it is difficult to otherwise form a picture of the durations of eruptions because of non-uniform weathering in the top of flow fields and a general paucity of sediments between GRB lavas. Where sediment has formed on top of the GRB, it varies in thickness from zero to 20-30 cm of silty to fine-sandy material, with occasional diatomaceous sediment. Individual GRB eruptions varied considerably in volume but many were greater than 1000 km³ in size. Most probably eruptive events were not equally spaced in time; some eruptions may have followed short periods of volcanic repose (perhaps 10² to 10³ of yrs), whilst others could have been considerably longer (many 1000s to >10⁴ yrs). Recent improvements in age-dating for other continental flood basalt (CFB) lava sequences have yielded estimates of total eruptive durations of less than 1 My for high-volume pulses of lava production. The GRB appears to be a similar example, where the main pulse occupied a brief period. Even allowing for moderate to long-duration pahoehoe flow field production, the amount of time the system spends in active lava-producing mode is small – less than c. 2.6% (based on eruption durations of approximately 10,000 yrs, as compared to the duration of the entire eruptive pulse of c. 420,000 yrs). A review of available ⁴⁰Ar/³⁹Ar data for the major voluminous phases of the Columbia River Basalt Group suggests that activity of the Steens Basalt-Imnaha Basalt-GRB may have, at times, been simultaneous, with obvious implications for climatic effects. Resolving intervals between successive eruptions during CFB province construction, and durations of main eruptive pulses, remains vital to determining the environmental impact of these huge eruptions.

Background

The CRBG is the Earth’s youngest continental flood basalt province with volcanism beginning around 17 Ma. Previous age determinations placed the eruption of the whole province between 17 and 6 Ma (Baksi, 1989; Tolan et al., 1989), with the voluminous GRB lavas estimated to have erupted between ~ 16.0 and ~ 15.0 Ma ago (Long and Duncan 1983; Hooper et al., 2002). The GRB lavas overlie the readily identifiable phenocryst-rich Imnaha Basalt lavas in the south of the province, and immediately underlie the Vantage Sandstone Member which forms part of the Ellensburg Formation (Swanson et al., 1979), that is distributed throughout much of the western CRBG on the Columbia Plateau. In many places throughout the province, the lavas of the Wanapum Basalt lie above the GRB, with or without the Vantage sediments between.

Discussion
Duration of the Grande Ronde Basalt
On the basis of our new Ar data, we can assess individual ages for the top and bottom of the GRB lavas and calculate a period in which the GRB erupted. Furthermore, the interval in which the GRB volcanism occurred can be examined by its relationship with the Imnaha Basalt below and the Wanapum Basalt above.

A calculated weighted mean average age for all the previous data (with no exclusions) for the timing of the Imnaha Basalt is 16.61 ±0.28 Ma (MSWD=11.9; Fig. 2). [Weighted mean average ages are calculated using ISOPLOT and include standard error propagation]. The same calculation for our Imnaha Basalt data, excluding the poor analysis of MLC-2 (Table 1), gives a result of 15.99 ±0.20 Ma (MSWD=0.5; Fig. 2). This latter date is outside of error of the weighted mean average age for all published data for Imnaha Basalt, but is our best estimate for the top of the lava pile. The wide scatter in published ages for the Imnaha Basalt (15.30 ±5.60 to 17.67 ±0.32 Ma; Fig. 2) may be the result of samples representing different parts of the Imnaha lava pile and possible diachroneity within the Imnaha Basalt, as a result of complex volcanic architecture such as overlapping lava lobes of differing ages, rather than a single horizon. In addition, the variation in the results may reflect variable degrees of sample alteration or excess argon in some samples. Regardless of such complexities, we suggest that Imnaha Basalt volcanism in southern Washington State - where GRB was subsequently emplaced - finished at around 16.00 Ma. In the absence of any evidence for a significant hiatus between the Imnaha Basalt and the GRB, it seems likely that the climactic GRB volcanism began shortly thereafter. No field-based evidence for interstratification of Imnaha and GRB lavas has yet been found.

A calculation of all published data (Table 1) for N₂ lavas at the top of the GRB succession reveals a weighted mean average age of 15.62 ±0.37 Ma (MSWD = 1.7; Fig. 2). The same calculation for our new data alone, excluding sample CRB05-106 for reasons given in the Results section, gives a date of 15.57 ±0.15 Ma (MSWD = 0.92; Fig. 2). Clearly, these results are well within error of each other and fit with data for the overlying Wanapum Basalt, e.g. Eckler Mountain Member (15.70 ±0.34 Ma) and basalt of Sand Hollow (15.12 ±0.38 Ma). Between the top of the GRB succession and the Wanapum Basalt are the Vantage sediments. It was once considered that the Vantage sediments were deposited in a 250, 000 year period, but it is likely that they formed in a much shorter interval, given constraints from the new argon data, the lack of any sediment in places, and the lack of evidence of erosion on the top surface of the GRB.

On the basis of this new evidence it would appear that the GRB lava succession started erupting shortly after 15.99 ±0.20 Ma and finished around 15.57 ±0.15 Ma. This suggests a total duration of 0.42 ±0.18 (1s) My, or approximately 420,000 years. Given uncertainties in the timing of the end of the Imnaha volcanism and that dates younger than 16 Ma are reported for the Imnaha (Table 1), the duration of the Grande Ronde volcanism could be considerably less than our estimate.

Absolute age of the Grande Ronde Basalt
As described by Swanson et al. (1979), the GRB lava pile contains flows that have been demonstrated to have two normal and two reverse magnetizations (R₁-N₁-R₂-N₂). Two full cycles of polarity overturn within 420,000 years is more rapid than generally observed in the geomagnetic record, although similarly rapid overturns are recorded in detailed studies of the Oligocene/Miocene boundary (e.g., Billups et al., 2004).

Using the geomagnetic polarity timescale and Ar data for Steens Basalt and one Imnaha Basalt sample (Table 1), Jarboe et al. (2006) similarly suggest a rapid succession of eruption for these lava formations; they propose that the bulk of the Imnaha and GRB lavas erupted within a 0.75 My period. However, on the basis of Ar data and paleomagnetic results they suggest that the Steens Basalts erupted in geomagnetic chron C5Cr (Fig. 2; 17.235 to 16.721 Ma; Gradstein et al., 2004), and that Imnaha erupted within the chron C5Cn.3n (16.721 to 16.543 Ma; Gradstein et al., 2004). As a consequence of this and the geomagnetic timescale, Jarboe et al. (2008) neatly suggest that rapid reversals C5Cn.2r (16.543 to 16.472 Ma), C5Cn.2n (16.472 to 16.303 Ma), C5Cn.1r (16.303 to 16.268 Ma) and C5Cn.1n (16.268 to 15.974 Ma) are GRB R₁-N₁-R₂-N₂, respectively (Fig. 2). However, they do not have age data directly from the GRB succession to support this.

Chron C5Cn.3n ends at 16.543 Ma, and following Jarboe et al. would imply that Imnaha magmatism did not occur after this time. Yet there are many Ar dates suggesting younger ages than 16.50 Ma for Imnaha Basalt (Table 1; Fig. 2). Similarly with the GRB, chrons C5Cn.2r to C5Cn.1n are between approximately 16.50 and 16.00 Ma, yet almost all available Ar-age data for the GRB are younger than 16.00 Ma (Fig. 2). The spread of available age data for Steens Basalt, Imnaha Basalt and GRB make it difficult to be certain when a particular volcanic phase ended and suggests that activity may have even overlapped at times. However, the overriding evidence from all the available age data is that the GRB erupted after 16 Ma and therefore cannot be constrained to the C5Cn.2r to C5Cn.1n chrons, and that Imnaha Basalt eruption cannot be restricted to have occurred only in chron C5Cn.3n (Fig. 2). It is worth noting that very little paleomagnetic work has been carried out on the Imnaha Basalt, and it may yet prove to be more complex than initially thought (V.E. Camp, pers. comm., 2010). The age data even suggest that Imnaha Basalt volcanism could have been occurring for some time after GRB volcanism had begun. Potential evidence supporting the eruption of highly voluminous lavas such as Imnaha Basalt and GRB at around 16.0 Ma and 15.5 Ma, respectively, are dissolution events recorded in sediments off the coast of Africa. These dissolution events are thought to record climatic disturbances caused by changes in the oxygen and carbon dioxide composition of the atmosphere and would fit well with voluminous basaltic eruptions (Kender et al., 2009). In summary, with the lack of knowledge of the detailed stratigraphy and paleomagnetic signatures of the Imnaha lavas, it is unwise to speculate further on the exact absolute age of the GRB lava succession from their relationship with Imnaha Basalts.

Our age data, along with other published information suggest that GRB lavas most likely erupted within the younger geomagnetic chron C5Br (15.974 to 15.160 Ma; Gradstein et al., 2004; Fig. 2). We recognise that this poses a problem for the significance of the reversals within the GRB; there appears to be a lack of rapid magnetic reversals during C5Br (e.g. Gradstein et al., 2004). We have no solution to this dilemma at present, and can only suggest that either (a) the reversals measured in the GRB are short excursions within the geomagnetic cycle, or b) that rapid reversals C5Cn.2r to C5Cn.1n should be younger than 16 Ma. In support of younger geomagnetic chrons for Imnaha and GRB lavas than suggested by Jarboe et al. (2008) are ⁴⁰Ar/³⁹Ar dates for Steens Basalts that are younger than 16.60 Ma (Fig. 2).

Average estimates of eruption frequency
The stratigraphy of the 150,000 km³ GRB lavas has been well constrained and mapped in previous studies (e.g. Reidel et al., 1989). The GRB lavas built up from at least 110 individual eruptions, which over a 420,000 year time span, suggest an average periodicity of one eruption every ~4200 years. Assuming an eruption duration of 10 to 100 years per eruption, that would add up to between 1100 and 11,000 years total eruption time, within the 420,000 year time span. This would suggest that the total amount of time that the system was actively erupting, was no more than 2.6% of the total duration time of formation of the GRB, and quite probably much less than 1% of the time.

In more detail, the individual GRB eruptions have estimated eruption volumes between 90 and as much as 10,000 km³ (Reidel et al., 1989) – the larger estimates rival the outputs of even the greatest volume explosive rhyolitic super-eruption deposits such as at Yellowstone (c. 2500 km³; Mason et al., 2004). In reality, the eruptions would most likely have been irregularly distributed through time, possibly in pulses with some eruptions following another after a few 1000 years, while others may have erupted after, possibly, up to 10,000 years of quiescence. With eruptions lasting decades to perhaps centuries, the chances of distinguishing a hiatus that is intra-eruption versus inter-eruption becomes very difficult.

Conclusions

Our new ⁴⁰Ar/³⁹Ar data provide constraints for the duration of time in which the Grande Ronde Basalt, the climactic phase of the Columbia River Basalt Province, erupted. The Grande Ronde Basalt is bracketed by an age of 15.99 ±0.20 Ma for the top of the underlying Imnaha Basalt and by an age of 15.57 ±0.15 Ma for the end of Grande Ronde Basalt volcanism. A duration of approximately 420,000 years implies that, on average, there may have been an eruption every 4000 years. Based on published estimates for single eruption durations, we find that within a 420,000 year period less than 2.6%, and quite possibly less than 1% of that time, would have been volcanically active. A review of all the available ⁴⁰Ar/³⁹Ar data for the Steens Basalt-Imnaha Basalt-GRB suggests that activity of these petrochemically distinct groups may have, at times, been simultaneous, and if confirmed would have significant implications for potential environmental effects.

References