Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://hbar.phys.msu.ru/gorm/dating/vesuvius2.pdf Äàòà èçìåíåíèÿ: Sun Apr 29 00:00:00 2007 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:03:05 2012 Êîäèðîâêà:

Bull Volcanol (2007) 69: 259­263 DOI 10.1007/s00445-006-0071-8

RES E ARCH AR TIC L E

Marvin Lanphere . Duane Champion . Leone Melluso . Vincenzo Morra . Annamaria Perrotta . Claudio Scarpati . Dario Tedesco . Andrew Calvert

40

Ar/39Ar ages of the AD 79 eruption of Vesuvius, Italy

Received: 23 April 2004 / Accepted: 11 January 2006 / Published online: 13 May 2006 # Springer-Verlag 2006

Abstract The Italian volcano, Vesuvius, erupted explosively in AD 79. Sanidine from pumice collected at Casti Amanti in Pompeii and Villa Poppea in Oplontis yielded a weighted-mean 40Ar/39Ar age of 1925±66 years in 2004 (1 uncertainty) from incremental-heating experiments of eight aliquants of sanidine. This is the calendar age of the eruption. Our results together with the work of Renne et al. (1997) and Renne and Min (1998) demonstrate the validity of the 40Ar/39Ar method to reconstruct the recent eruptive history of young, active volcanoes. Keywords 40Ar/39Ar Geochronology . AD 79 eruption . Historic eruption

Introduction
A well-documented eruption of the volcano Vesuvius, near Naples, Italy occurred 24­25 August in AD 79. Many historical facts of the AD 79 eruption come from two letters written by Pliny the Younger to the historian Cornelius Tacitus (Radice 1971). Archaeologists and geologists have reconstructed the history of this eruption using Pliny's letters as well as observations taken at Pompeii, Oplontis, and Herculaneum, cities that were destroyed by the eruption (Lirer et al. 1973; Sigurdsson et al. 1982; 1985; Cioni et al. 1996; Luongo et al. 2003a,b).

Editorial responsibility R. Cioni M. Lanphere . D. Champion . A. Calvert (*) US Geological Survey, 345 Middlefield Road, Menlo Park, CA 94925, USA e-mail: alder@usgs.go L. Melluso . V. Morra . A. Perrotta . C. Scarpati . D. Tedesco Dipartmento di Geofisica e Vulcanologia, Universita di Napoli Federico II, Napoli, 081253111, Italy

Vesuvius had explosive eruptions prior to AD 79 (at 3,400, 8,000, and 18,000 years). The last major eruption of Vesuvius before AD 79 occurred in 1200 BC (Sigurdsson et al. 1982). A strong local earthquake in AD 62 did extensive damage in Pompeii and in Herculaneum which was still under repair in AD 79. On 20 August in AD 79, tremors again were felt in the region. Earthquakes occurred with increasing frequency during the next 4 days as magma moved up the feeder pipe of the volcano. The eruption started on 24 August when, after minor phreatic eruptions, a high-sustained eruptive column rose convectively into the stratosphere reaching an estimated maximum height of 32 km (Sigurdsson et al. 1985). The plinian phase of the eruption lasted for about 18 h, and during this period, Vesuvius produced about 2.6 km3 of pumice (Sigurdsson et al. 1982). A thick layer of pumice lapilli resulting from the column fallout covered a wide area south of the volcano (Lirer et al. 1973; Fig. 1). This deposit reached its maximum thickness of 280 cm at Pompeii; the deposit totally buried flat roofs and the lower parts of buildings. On the morning of 25 August, the plinian column collapsed and pyroclastic density currents were emplaced which destroyed every settlement within a radius of 10­15 km from the volcano. Luongo et al. (2003a,b) described the effects of the eruption on the inhabitants. The well-documented historical age of this eruption has been used to extend 40Ar/39Ar dating to young events. Renne et al. (1997) first applied 40Ar/39Ar dating to this Vesuvius eruption yielding an age of 1925±94 years (1 uncertainty) for sanidine from Villa Poppea in Oplontis for the eruption (their age is 1932 years ago corrected to the year 2004). They used a series of total fusion analyses on separate aliquants of sanidine, and then pooled these analyses to yield an age. We decided to make incrementalheating analyses of sanidine to see whether this approach would yield results similar to that of Renne et al. (1997) with perhaps better precision. Using incremental-heating we obtained an 40Ar/39Ar age of 1921±66 years in AD 2000 (1925 years corrected to the year 2004), the age of the AD 79 eruption.


260 Fig. 1 Map of the Vesuvius region and Bay of Naples (after Sigurdsson et al. 1982), showing the extent of the area affected by pumice and pyroclastic flows during The AD 79 eruption. Broken lines are isopachs of the pumice fall during the Plinian phase (Lirer et al. 1973)
MISENUM

NAPLES

VESUVIUS
HERCULANEUM

OPLONTIS

POMPEII
100 200

PYROCLASTIC FLOW DEPOSITS

0

10 km

200

100

100 ASH-FALL ISOPACHS, CM

10

Previous work
The first good results for a measured isotopic age of the AD 79 eruption of Vesuvius were reported by Renne et al. (1997). They analyzed 12 aliquants of sanidine from one sample by heating each aliquant in 3­5 steps using a CO2 laser. Power settings were from a laser, and they produced 46 separate measurements. The temperatures of their heating steps are not known. They plotted the 40Ar/39Ar ratios of the 46 determinations on a 36Ar/40Ar vs. 39Ar/40Ar diagram after making corrections for non-radiogenic and interfering isotopes. On this diagram, the 46 measurements fit an isochron indicating an age of 1925±94 years (in 1997). Two subsequent irradiations done in the same fashion yielded isochron ages of 1889±132 years (59 measurements) and 1936±207 years (37 measurements; Renne and Min 1998). The number of aliquants measured is unknown. They attributed the larger uncertainties in these ages to larger sample sizes. They said that larger uncertainties resulted from larger corrections for mass discrimination and from non-uniform heating with the laser. They gave no measurements of mass discrimination. Renne and Min (1998) suggested that these uncertainties be ignored and the ages simply be pooled to yield an age of 1917 years with an uncertainty of ±14 years. We believe this to be an
Table 1
40

incorrect use of statistics. The proper way to pool these three isochrons (Renne et al. 1997; Renne and Min 1998)is weighting by the inverse of the variances of the three determinations which yields a weighted mean age of 1,922 years with an uncertainty of ±72 years. The weighted-mean age has an uncertainty of ±72 years which seems more realistic from the internal uncertainties of the three data sets.

Analytical techniques
For the current study, pumice clasts were collected within two buildings: Casti Amanti in Pompeii and Villa Poppea in nearby Oplontis (Fig. 1). The sample from Villa Poppea is different from the sample analyzed by Renne et al. (1997). Both Pompeii and Oplontis were buried by the AD 79 eruption. Oplontis has been excavated inside the current Naples suburb of Torre Annunziata. Sanidine for the current study was separated from pumice using standard magnetic and heavy-liquid methods. One sample each from Casti Amanti and Villa Poppea was processed; each sample was from white pumice and was divided into aliquants for irradiation. Aliquants of sanidine weighing 223 to 271 mg were encapsulated in Cu foil and placed in quartz vials along with fluence monitor

Ar/39Ar incremental-heating ages of sanidine from Casti Amanti, Pompeii and Villa Poppea, Oplontis Plateau age (years) 2061±181 1707±185 2099±185 1821±200 1938±186 2181±182 1667±184 1867±185 Plateau
39

Aliquant number Casti Amanti-1 -2 -4 -5 -6 -7 -8 Villa Poppea-2

Ar (%; steps)

40

Ar/36Ar isochron intercept

99(23 of 25) 87(21 of 26) 86(17 of 26) 86(18 of 25) 100(25 of 25) 100(825 of 26) 100(26 of 26) 96(22 of 26)

297.4±2.8 298.4±3.0 300.1±4.4 299.7±1.0 290.8±3.6 295.9±2.0 294.2±3.4 302.1±5.3

Weighted mean age: 1921±66 years (in year 2000)


261
10

Casti Amanti - 1

10

Casti Amanti - 5

Apparent age, ka

Apparent age, ka

8 6 4 2 0 2061 ± 181 years
1025 750 900 950 975 1075 1125 1100 1050 1150

8 6 4 2 00
900 1050

750

1821 ­ 200 years
950 975 1000 1025 1125 1075 1100 1150

1000

0

10

20 39

30

40

50

60

70

80

90

100

10

Ar released, cumulative percent

20 39

30

40

50

60

70

80

90

100

Ar released, cumulative percent

Fig. 2 40Ar/39Ar age spectrum for sanidine from aliquant Casti Amanti-1. Half of the vertical dimension of the increment boxes is the estimated SD (1) of precision of the age spectrum age. The error in the weighted-mean plateau age is the weighted SD

Fig. 4 40Ar/39Ar age spectrum for sanidine from aliquant Casti Amanti-5. See Fig. 2 for explanations

mineral TCR sanidine (t=28.92 Ma). The quartz vials were then wrapped with Cd foil and were irradiated for 30 min in the US Geological Survey TRIGA reactor in Denver, Colorado. Samples were irradiated in two different groups: Menlo Park irradiations CLIX and CLXIII. Shielding a sample with Cd foil dramatically decreases the production of 40Ar from 40K by thermal neutrons. In the USGS TRIGA reactor, the ratio of 40Ar/40K in K-glass shielded by Cd foil is 0.0000±0.0000. The calculation of ages for these young samples would be very difficult if the correction for reaction with thermal neutrons was not eliminated with Cd shielding. A total of 3­4 grains of the fluence monitor mineral were fused using an argon-ion laser (Dalrymple 1989). For each of the eight aliquants in Table 1, a weightedmean plateau age has been calculated; age-spectrum diagrams are shown for four of these aliquants as Figs. 2, 3, 4, 5. It is important to realize that these age spectra are only pictorial representations to aid in interpretation. Precise correlations of individual gas increments are made analytically. We prefer to make comparisons using the weighted-mean plateau from age spectrum diagrams. The age spectra actually are surprisingly good given that the quantity of 40Arrad in each step is very small, generally in the range 1 to 9â10-16 mol; this small amount of radiogenic Ar is difficult to measure precisely. There are small
10

amounts of extraneous Ar in low-temperature steps and these have been eliminated from the plateau age calculations. Three of the eight determinations (Casti Amanti­6, 7, and 8) do not have any extraneous Ar (Table 1). These three determinations have the lowest 40Ar/36Ar intercepts. The total amount of extraneous Ar is very small and ranges from 6â10-16 mol to 2.7â10-15 mol. This amounts to 1­13% of the measured 39Ar. Samples of irradiated sanidine were heated in a furnace by increments in this study, generally 25 or 26 temperaturecontrolled increments per sample. Backgrounds were measured on every other step, and backgrounds for the intervening steps were determined by interpolation. Heating was done using a resistance-heated furnace connected to an extraction line and a mass spectrometer (Dalrymple 1989). Mass analyses were made using a MAP-216 mass spectrometer. Mass discrimination of the extraction line and mass spectrometer was monitored using a gas pipette from a reservoir filled with atmospheric Ar. The mass discrimination during this study was 1.009145±0.000041 per mass unit (1 uncertainty). The incremental-heating temperatures were carefully controlled with an optical-fiber thermometer. Each aliquant of sanidine in the current study yielded a plateau age with an uncertainty of 8­9%. The percent 40Arrad for most temperature steps would be from

Casti Amanti - 4
Apparent age, ka

10 8 6

Villa Poppea - 2

Apparent age, ka

8 6 4
850 925 950 975 1075 1025 1050 1125

2099 ­ 185 years
1175 1150 1000 1100

1867 ± 185 years
900 1100 975 875 1000 1025 925 950 1050 1075

4 2 0 0 10
850

2 0 0 10 20 39 30 40

50

60

70

80

90

100

Ar released, cumulative percent

20 39

30

40

50

60

70

80

90

100

Ar released, cumulative percent

Fig. 3 40Ar/39Ar age spectrum for sanidine from aliquant Casti Amanti-4. See Fig. 2 for explanations

Fig. 5 40Ar/39Ar age spectrum for sanidine from aliquant Villa Poppea-2. See Fig. 2 for explanations


262

about 1 to 7% with some steps having more than 10% radiogenic Ar. The background in the mass spectrometer typically is 1.5â10-18 mol for m/z=36, 1â10-17 for m/z=37, 2x10-18 mol for m/z=39, and 1.5â10-16 mol for m/z=40. A typical blank, which includes the mass spectrometer background, consists of about 2â10-18 mol of m/z=36, 1â10-17 mol of m/z=37, 4â10-18 mol of m/z=36, 4â10-16 mol of m/z=40 (Dalrymple 1989).

(Table 1) have isochron intercepts that agree with the atmospheric composition within 2 uncertainties. Villa (1991, 1992) pointed out that other young volcanic provinces (such as the Roman Volcanic Province) contain excess Ar in phenocrysts. However, sanidine from the AD 79 eruption of Vesuvius does not contain excess Ar. At present, the reason for the subtle differences in the 40 Ar/36Ar values of isochron intercepts from the Vesuvius eruptions are not known.

Discussion
Eight aliquants of sanidine (Table 1) yielded a weightedmean age of 1921±66 years during experiments performed in the year AD 2000. Other dating methods have been applied to the AD 79 eruption. 14C ages have been measured on a variety of carbonaceous materials caught up in the eruption. Aciego et al. (2003) recently determined a U-Th/He age of 1,885±188 years on garnet from pumice. This age was determined after correcting for diffusive loss of He, alpha ejection, and initial U-series disequilibrium. The sample studied by Renne et al. (1997) and Renne and Min (1998) was collected at Villa Poppea in Oplontis. We determined ages on sanidine from a different sample collected at Villa Poppea and a second sample at Casti Amanti in Pompeii. Most of our measurements were made on the Casti Amanti sanidine because percent 40Arrad was slightly higher on sanidine from Casti Amanti. It is not necessary to make irradiation times extremely short or to use a very young neutron fluence monitor to do 40 Ar/39Ar dating on young samples. Renne et al. (1997) used an irradiation time of 3 min for samples from the AD 79 eruption and their fluence monitor mineral was sanidine from the Alder Creek Rhyolite (t=1.194 Ma; Turrin et al. 1994). Renne and Min (1998) suggest that very short irradiations require a young monitor in order to keep the dynamic range of isotope ratios low. However, it is not necessary to use such short irradiation times as 3 min to accomplish this. In the present study the irradiation time of sanidine aliquants was 30 min. The dynamic range of 40 Ar/39Ar ratios in TCR sanidine was an easily measured 144­192. In addition, using an older fluence monitor mineral such as TCR sanidine allows the J-monitor curve to be determined more precisely than if one uses a young fluence monitor mineral and very short irradiation times. Short irradiation times are necessary if one has to correct for 40ArK (Ar produced from potassium in the sample), but this correction is unnecessary if the sample is shielded with Cd. Renne et al. (1997) pointed out that their samples contained extraneous Ar (excess or inherited Ar) based on an 40Ar/36Ar intercept of 307±1 for the isochron intercept fitted to 46 measurements of 12 aliquants of sanidine. This intercept is significantly different from the atmospheric value of 295.5±1. Renne and Min (1998) also reported 40 Ar/36Ar intercepts of 309±4 and 308±7 for additional aliquants of sanidine from Villa Poppea. In contrast, the eight sanidine aliquants analyzed by us in all cases

Conclusions
Sanidine from the explosive eruption of Mount Vesuvius in AD 79 has been dated by incremental-heating of 8 aliquants. These aliquants of sanidine phenocrysts from two samples of white pumice yield an 40Ar/39Ar age of 1925±66 years (in 2004), the calendar age of the eruption. Averaging the 7 aliquants from Casti Amanti yields an age of 1925±69 years (in 2004). These results show that the 40 Ar/39Ar method can be used to reconstruct the recent eruptive history of young volcanoes.
Acknowledgements Forrest McFarland made the sanidine separations from pumice samples from Casti Amanti and Villa Poppea. James Saburomaru assisted with the Ar determinations. Neutrons were provided by Tim Debey and staff of the US Geological Survey reactor center in Denver, Colorado. We are grateful to Professor Guzzo and Dr. A. Ciarallo of the Soprintendenza Archeologica di Pompeii who allowed our field work in archaeological sites. Julie Donnelly-Nolan and Tom Sisson reviewed the manuscript and offered valuable comments. We thank Professor. R. Cioni, and Drs. Villa and Scaillet for valuable comments on the original manuscript.

References
Aciego S, Kennedy BM, DePaolo DJ, Christensen JN, Hutcheon L, (2003) U-Th/He age of phenocrysts of garnet from the 79 AD eruption of Mt. Vesuvius. Earth Planet Sci Lett 219:209­219 Cioni R, Sbrana A, Gurioli, L (1996) The deposits of AD 79 eruption: CEV-CMVD workshop on Vesuvius decade volcano, September 1996, Naples, pp 17­23 Dalrymple GB (1989) The GLM continuous laser system for 40 Ar/39Ar dating: description and performance characteristics. US Geol Surv Bull 1890:89­96 Lirer L, Pescatore T, Booth B, Walker, GPL (1973) Two Plinian pumice fall deposits from Somma-Vesuvius, Italy. Geol Soc Am Bull 84:759­772 Luongo G, Perrotta A, Scarpati C (2003a) Impact of the AD 79 explosive eruption on Pompeii, I. Relations amongst the depositional mechanisms of the pyroclastic products, the framework of buildings and the associated destructive events. J Volcanol Geotherm Res 126:201­223 Luongo G, Perrotta A, Scarpati C, DeCarlis E, Patricelli G, Ciarallo A (2003b) Impact of the AD 79 explosive eruption on Pompeii, II. Causes of death of the inhabitants inferred by stratigraphic analysis and areal distribution of the human casualties. J Volcanol Geotherm Res 126:169­200 Radice B (1971) The Letters of the Younger Pliny. Penguin, London Renne PR, Min K (1998) 40Ar/39Ar dating of the 79 AD eruption of Vesuvius: An ab initio basis for improving accuracy in 40 Ar/39Ar geochronology. Min Mag 62A:1255­1256


263 Renne PR, Sharp WD, Deino AL, Orsi G, Civetta L (1997) 40 Ar/39Ar dating into the historical realm: calibration against Pliny the Younger. Sci 277:1279­1280 Sigurdsson H, Cashdollar S, Sparks SRJ (1982) The eruption of Vesuvius in AD 79: reconstruction and volcanological evidence. Am J Archaeol 86:39­51 Sigurdsson H, Carey S, Cornell W, Pescatore T (1985) The eruption of Vesuvius in AD 79. Natl Geog Res 1:332­387 Turrin BD, Donnelly-Nolan JM, Hearn BC (1994) 40Ar/39Ar ages from the rhyolite of Alder Creek, California: age of the Cobb Mountain normal-polarity subchron revisited. Geol 22:251­254 Villa I (1991) Excess Ar geochemistry in potassic volcanites. Schweiz Mineral Petro Mitt 71:205­219 Villa I (1992) Datability of Quaternary volcanic rocks: an 40Ar/39Ar perspective on age conflicts in lavas from the Alban Hills, Italy. Eur J Mineral 4:369­383