Accuracy of obsidian hydration dating based on obsidian–radiocarbon association and optical microscopy

Most archaeologists using obsidian hydration dating (OHD) in the United States develop hydration rates from association of obsidian hydration rims with dates based on radiocarbon, subsequently using the rate for chronometric analysis. The overall accuracy of the process has never been quantified. This paper reports an accuracy analysis in which sources of uncertainty are defined and modeled and their effects quantified. A Monte Carlo simulation is used to quantify errors from rate development, while uncertainties in age resulting from chronometric analysis are calculated analytically. For typical ranges of error values, hydration rate errors of ∼5% or less are achievable in the absence of systematic errors, with errors of chronometric age estimates ∼20–30% or less.     

Prehistoric settlement chronology on Rapa Nui,Chile: obsidian hydration dating using infrared photoacoustic spectroscopy

The prehistoric Polynesian inhabitants of Rapa Nui (Easter Island) utilized obsidian for nearly 700 years in many activities connected with daily life. The near ubiquitous occurrence of the natural glass in both domestic residences and religious structures makes the application of obsidian hydration dating highly suitable for the investigation of cultural change. We have applied previously developed calibrations that estimate hydration rates for obsidian based upon the structural water content of the glass as determined by infrared spectroscopy. The archaeological ages estimated by this method were compared with accelerator mass spectrometry (AMS) dates using short-lived woody species endemic to the island. The convergence between the two dating methods is strong and we suggest that obsidian hydration dating may be used on Rapa Nui to reliably date contexts where suitable material for AMS dating may not be available.     

An archaeologically validated protocol for computing obsidian hydration rates from laboratory data

This paper reports the computation of hydration rate for Topaz Mountain obsidian from laboratory data, and a comparison with archaeological data from a well-dated site, Camels Back Cave, in western Utah. Topaz Mountain obsidian is found to be slow-hydrating, with a rate of 0.071 ± 0.021 μ/yr^½ at an effective hydration temperature of 16.01 °C. This rate agrees with a rate developed from archaeological data from Camels Back Cave within ∼6%. Activation energy of Topaz Mountain obsidian is 10370 ± 544 K, and its diffusion constant is (1.87 ± 9.13) × 10¹³ μ²/yr, both of which are independent of temperature. Its intra-source variability in hydration rate is very low (CV < 0.01), implying a low variability in intrinsic water. We present a model of chemical erosion which shows why earlier laboratory-determined rates were incorrect, and discuss the implications of our findings on the determination of experimentally derived rates in obsidian hydration dating.     

Induced hydration of obsidian: a simulation study of accuracy requirements

Determination of the hydration rate constant of obsidian is basic to the use of obsidian for establishing chronologies. The constant can, in principle, be determined in either of two ways: by correlations with archaeological sequences, or by laboratory experiments using induced hydration. Induced hydration holds promise of great accuracy, but results reported to date have been disappointing. This paper is based on the hypothesis that the outcomes are the result of error build-up in the induced hydration protocol, and describes an analysis based on a Monte Carlo simulation of the measurement and analysis process. Data are presented which show that the poor results are due to errors inherent in optical measurement of hydration rim thickness. It is concluded that successful use of induced hydration requires an order of magnitude improvement in accuracy of hydration rim measurement over the accuracies currently claimed for optical microscopy. The results do not affect the validity of hydration dating based on archaeological correlations.     

Effective hydration temperature of obsidian: a diffusion theory analysis of time-dependent hydration rates

An estimate of effective hydration temperature (EHT) is needed for chronological use of obsidian hydration data. This paper describes a method for calculating EHT by the practicing archaeologist, replacing three techniques that are in general use today: estimates based on mean temperature, numerical integration of models of diurnal and annual temperature variations, and use of temperature cells. The hydration (or diffusion) coefficient of obsidian is a function of temperature and thus is time varying, while the classic quadratic law of hydration is not valid for time-varying diffusion coefficients. This paper presents a mathematical solution to the case of a time-varying hydration coefficient, based on diffusion theory, with a concise definition of EHT. It is shown that the results are not affected by concentration dependence in the diffusion coefficient. A computer program to compute the rigorous solution is described, and data are presented to explore the resulting range of variation. That use of the Lee equation to compute EHT is not appropriate for obsidian hydration studies is evident from the data presented. The effects of paleoclimatic variation are estimated, and an algebraic best fit equation and worksheet are provided as practical aids to the archaeologist.     

Obsidian hydration dating on the South Coast of Peru

We compare over 230 obsidian hydration readings from 30 individual site components from the Southern Nasca Region (SNR) with independent age estimates based on radiocarbon dates and temporally diagnostic artifacts. Although there are problems with small sample sizes, and readings must be adjusted for elevation, a very strong relationship accounting for nearly 90% of the total variation in the data set is found. This suggests that obsidian hydration dating (OHD) works in the SNR and is a viable means of independently estimating age. Residual values from our regression suggest that hydration age estimates are usually within 15% of the radiocarbon estimates. Finally, we present an equation other scholars can use to estimate age for Quispisisa obsidian in the SNR.     

Obsidian provenance studies in Colombia and Ecuador: obsidian sources revisited

The field occurrences, elemental compositions and formation ages of Colombian and Ecuadorian obsidians are revisited. It is shown that the regional sources of this raw material are linked to two major volcanic structures: the Chacana and the Paletara calderas, localised on the eastern cordillera of Ecuador and on the central Andean cordillera of south Colombia respectively. Seventy-two samples were analysed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES), inductively coupled plasma-mass spectrometry (ICP-MS) and/or particle induced X-ray emission (PIXE). The same 10 types of elemental compositions were identified independently from ICP and PIXE. Four of these types were previously unknown. The formation ages of these obsidians previously determined by fission tracks dating are in the range 0.17–1.58 Ma at Chacana and 3.46–4.27 Ma at Paletara. Most Colombian and Ecuadorian pre-Hispanic artefacts present elemental compositions compatible with a Chacana- or Paletara-derived origin of the raw material. However, some of them present fission track ages discordant with the present-day known obsidian occurrences, which implies that the regional source inventory is not yet exhaustive.     

Late Pleistocene/Early Holocene seafaring in the Aegean: new obsidian hydration dates with the SIMS-SS method

Archaeological evidence regarding the presence of obsidian in levels that antedate the food production stage could have been the result of usage or intrusion of small obsidian artifacts from overlying Neolithic layers. The new obsidian hydration dates presented below employing the novel SIMS-SS method, offers new results of absolute dating concordant with the excavation data. Our contribution sheds new light on the Late Pleistocene/Early Holocene exploitation of obsidian sources on the island of Melos in the Cyclades reporting dates c. 13th millennium - end of 10th millennium B.P.     

Obsidian hydration at high elevation: Archaic quarrying at the Chivay source,southern Peru

We examine obsidian hydration as a means to date archaeological sites at high elevation in the central Andes, and in particular quarry sites that are difficult to date by radiocarbon means. The Chivay obsidian source lies in a volcanic depression above the Colca Valley in Arequipa, Peru (71.5355° S, 15.6423° E) at 4950 masl. We compare obsidian hydration readings from one quarry and two workshop locations. Ninety-one flakes from the quarry pit, and 61 and 33 flakes from the workshops were analyzed for hydration bands. Of these, 68 from the quarry, and 54 and 33, respectively from the workshops produced at least one culturally meaningful hydration band. As expected, obsidian appears to hydrate slowly at this high elevation. Yet, variation in hydration readings is low within stratigraphic contexts, suggesting relatively narrow windows of knapping activities in each excavation level. A small number of radiocarbon dates allow us to develop a preliminary hydration rate for Chivay obsidian in this high elevation location. Hydration data indicate that intensive quarrying began by 3800 cal. BC and stopped ca. 2300 cal. BC. By contrast, the two workshops appear to have been deposited 2900 and 1200 cal. BC, and 2700 and 2400 cal. BC. The data are consistent with an uptick in obsidian use by at least the Terminal Archaic period.     

Obsidian hydration dating by infrared spectroscopy: method and calibration

Low temperature (90–190 °C) hydrothermal experiments have been conducted on seven obsidians where composition of the glass varies significantly in the concentration of structural water within the unhydrated bulk material. Infrared transmission spectroscopy was used to track the diffusion of molecular water into the glass surface as a function of time and temperature. Long-term (60–360 days) hydration sequences at 90 °C show a t^0.6 time dependence for the mass uptake of molecular water that forms the hydration layer. The structural water concentration of the unhydrated bulk obsidian is highly correlated with the pre-exponential and activation energy and may be used to estimate the Arrhenius constants. In addition, secondary ion mass spectrometry (SIMS) hydrogen profiling of Napa Glass Mountain obsidian hydrated at 90 °C reveals that the early stages of diffusion exhibit a dynamic behavior that includes a fluctuating hydrogen concentration and a changing diffusion coefficient that slows with time.     

Strofilas (Andros Island,Greece): new evidence for the cycladic final neolithic period through novel dating methods using luminescence and obsidian hydration

The recently excavated coastal prehistoric settlement of Strofilas on Andros Island (Cyclades, Greece) in the Aegean sheds new light on the transitional phase from the Final Neolithic to Early Cycladic period regarding masonry, fortification, and richly engraved rock art. The fortification possesses early evidence of preserved defensive architecture, as evidenced from the plethora of scattered finds from within and around the settlement. Important features are carvings on rock walls which mainly depict ships, animals, and fish. Initial archaeometric dating via the application of luminescence dating of two samples from the fortified wall bearing engraved ships, and by obsidian hydration of two blades employing the new SIMS-SS method (secondary ion mass spectrometry via surface saturation), has been undertaken to determine the site's chronology. The former yields an average date of 3520 (±540) BC and the latter an average date of 3400 (±200) years BC, both of which, within overlapping errors, suggest the main settlement occurred during the Final Neolithic.     

The source of obsidian artefacts found at East Chia Sabz,Western Iran

Excavations at the site of East Chia Sabz in western Iran uncovered deposits dating from the 9th through 7th millennium BC showing evidence of obsidian use. A total of twenty obsidian artefacts was found at the site and they were analyzed by X-ray fluorescence and neutron activation analysis. The results show all of the obsidian found at East Chia Sabz came from the Nemrut Da? source in southeastern Turkey located a distance of almost 750 km from East Chia Sabz. The results and their archaeological implications are discussed.     

Field data validation of an algorithm for computing obsidian effective hydration temperature

This paper describes an analysis to validate the effective hydration temperature (EHT) algorithm for obsidian [Rogers, A.K., 2007. Effective hydration temperature of obsidian: a diffusion-theory analysis of time-dependent hydration rates. J. Arch. Sci. 34, 656-665], using hourly temperature data from the Amargosa Desert Research Site near Beatty, NV. These data were used as input to a numerical model of the temperature-dependent diffusion process, and EHT was calculated yearly and for the aggregate. The same temperature data were processed to extract input parameters for the climatic model reported and used as input for EHT computation. EHT as computed from recorded hourly data was found to be within 1 °C of that computed from the climatic model. It is also shown that computed EHT differences between sites are not sensitive to whether air or surface temperature data are used, as long as they are used consistently; however, surface temperature data must be used if depth corrections are to be made.     

Patterning in procurement of obsidian in Chaco Canyon and in Chaco-era communities in New Mexico as revealed by X-ray fluorescence

X-ray fluorescence analysis of obsidian artifacts from sites located in Chaco Canyon and from three Chaco-era communities in New Mexico permits determination of their geological origin. These source data are used to describe patterning in obsidian procurement in sites located in Chaco Canyon dating from A.D. 500–1150, and in a three non-Canyon communities occupied during the period of Chaco Canyon's regional prominence (ca. A.D. 875–1150). These data demonstrate that the most proximate sources generally dominate the sourced obsidian assemblages from sites of all periods, but also suggest differences in procurement patterning both over time and across space. Within Chaco Canyon, there is a notable shift from Mount Taylor obsidian to use of Jemez Mountains sources over time. These data also suggest that earlier analyses of obsidian from sites in Chaco Canyon misidentified some obsidian artifact sources; these new data indicate the central areas of disagreement and provide a revision of procurement patterning. In the Chaco-era communities located outside Chaco Canyon, procurement patterning diverges. The Blue J community shows an increase in use of the nearby Mount Taylor source over time. Two communities located toward the southern extent of the Chaco great house distribution reveal a markedly distinct procurement pattern, obtaining nearly all of their obsidian from southern sources largely unrepresented at Chaco Canyon. Combined, these data provide new insights into raw material procurement and artifact production at sites in Chaco Canyon, and in communities occupied during the Chaco Phenomenon, the period of the Canyon's greatest regional influence.     

Obsidian source characterization in the Cordillera Real and eastern piedmont of the north Ecuadorian Andes

Recent research in the Quijos and Cosanga valleys of the eastern piedmont of Ecuador’s Cordillera Real has revealed and substantiated previous knowledge of obsidian sources that are unrelated to obsidian flow systems in the Sierra de Guamaní, Ecuador. Neutron Activation Analysis (NAA) and X-ray Fluorescence (XRF) were carried out on 47 obsidian source samples collected from several contexts in and adjacent to the study area. From samples within the study area three distinct obsidians were characterized: Cosanga A, Cosanga B, and Bermejo. These obsidians originate from a number of obsidian-bearing rhyolitic domes recently identified in the hills west of the Río Cosanga. Extensive survey of these dome localities has identified obsidian cobbles large enough for formal and informal tool manufacture. Beyond the study area, samples were collected and analyzed from the El Tablón source in the Sierra de Guamaní, providing much needed data on this poorly understood source. In addition, a sample from the newly identified Conda Dome source, near the Cotopaxi volcano, was characterized with XRF. All samples were then compared to 57 pre-existing samples from the Mullumica–Callejones, Yanaurco–Quiscatola and Carboncillo sources in the Ecuadoran Cordillera Real, as well as to artifacts from the Sumaco area in the Ecuadorian Amazon. Results of the elemental characterization indicate that the Cosanga Valley, El Tablón and Conda Dome obsidians are chemically distinct. Further, visual characteristics of Cosanga Valley obsidian types are useful in source attribution for the large artifact samples from the region. Finally, obsidian collected from the El Tablón flow suggests that this source may have produced obsidian suitable for tool manufacture.     

Possible obsidian sources for artifacts from Timor: narrowing the options using chemical data

Measurements made at the Australian National University using laser ablation ICPMS show that none of the 88 analyzed obsidian artifacts from East Timor match either the known Papua New Guinea or the five Island SE Asian source samples in our ANU collections. There is a coastal journey of more than 3000 km between the occurrence of obsidians from the Bismarck Archipelago volcanic province of Papua New Guinea and the Sunda-Banda Arc volcanic chain, yet obsidian artifacts from the two important PNG sources of Talasea and Lou Island are found at coastal Bukit Tengkorak in eastern Sabah at a similar distance along with material that has no known source. Timor lies south of the eastern section of the active volcanic Banda Arc island chain but it is within range of possible rhyolite sources from there. Although there is a continuous chain of around 60 active volcanoes stretching from west Sumatra to the Moluccas most are basaltic to andesitic with few areas likely to produce high silica dacite–rhyolite deposits. This does not exclude the possibility that the volcanic landscapes may contain obsidian, but without detailed survey and chemical analysis of sources from the Sunda-Banda Arc the attribution of the Timor obsidian artifacts remains to be demonstrated. Timor may seem to be an unlikely source for the presence of obsidians as it lacks reports of the silica-rich rhyolite volcanic centers necessary to produce this material. Despite the absence of detailed survey and analysis of Indonesian obsidian sources, especially from the volcanically active Banda Arc, this paper presents evidence that one of two obsidian sources is clearly from Timor while the other, with less certainty, is also from an unknown local source.     

Obsidian types from Holocene sites around Lake Turkana,and other localities in northern Kenya

Obsidian has been widely used by early Holocene hunter-gatherers and succeeding Pastoral Neolithic peoples in northern Kenya. Here we report the results of over 2000 electron microprobe analyses of artifactual and non-artifactual obsidian from the greater Lake Turkana region. Of the 15 compositional types of obsidian observed, a preponderant type is widespread across the region from the Barrier in the south to Ileret in the north and east as far as Kargi. This obsidian is the principal type at Lowasera and most Pastoral Neolithic sites, including the Jarigole Pillar site and Dongodien (GaJi4). The source of this obsidian is not known, but based on its distribution the source may be located on the Barrier or in the Suguta Valley immediately to the south of Lake Turkana. Although there are several possible sources of local obsidian identified for minor types, in stark contrast to the central part of the Kenyan Rift, major sources of obsidian available for artifact manufacture are not known in the Lake Turkana region. The lack of obsidian from demonstrable Ethiopian Rift and central Kenyan Rift sources, and the absence of obsidian with compositions found at the Turkana area sites in assemblages in the central part of the Kenyan Rift suggests that the earlier Pastoral Neolithic peoples around Lake Turkana interacted with each other, but perhaps not as strongly with people farther south along the Rift Valley, even as herding practices were expanding to the southward into central Kenya.     

Multiple origins of Bondi Cave and Ortvale Klde (NW Georgia) obsidians and human mobility in Transcaucasia during the Middle and Upper Palaeolithic

Using PIXE four types of elemental compositions were found among obsidian artefacts from the Bondi Cave and Ortvale Klde, Middle to Upper Palaeolithic sites in NW Georgia. One of those types corresponds to obsidians from the Chikiani source, whose compositions were determined with a very good agreement by PIXE and ICP-AES/MS. The composition of Chikiani obsidians is remarkably constant despite K–Ar and ³⁹Ar/⁴⁰Ar extrusion ages from ca 2.4 and 2.8 Ma. The compositions of two other groups of obsidian artefacts are similar to source materials from eastern Anatolia and Armenia, in particular Ikisdere, Sarikamis, Gutansar, and Hatis. Obsidian is only a minority component in the lithic assemblages at the Bondi Cave and Ortvale Klde. Both Neanderthal and Modern Human populations used obsidian in particular from Chikiani. Considering that the shortest walking distance to this nearest source is at minimum of about 180 km, and to other potential sources of more than 350 km it is suggested that this material reached these two sites mostly, if not exclusively, by a series of ‘down the line’ exchanges.     

Geochemical composition of source obsidians from Kenya

Here we provide a reference resource to archaeologists interested in the sources of obsidian in Kenya, through electron microprobe analyses of 194 obsidian samples from 90 localities. Averaged analyses of each sample and eleven published analyses are categorized into 84 compositional groups of which only about 21 are known to have been used to produce artifacts, possibly because studies of artifactual material in the region are lacking. We also provide trace element analyses determined by XRF and LA-ICP-MS for these same obsidians. In northern Kenya 27 distinct compositions of obsidian have been found, including some of Miocene age, but the source of the most abundant obsidian found in archaeological sites in this part of Kenya remains obscure. The Baringo region contains at least 13 varieties of low-silica obsidian. The Naivasha–Nakuru region contains an abundance of obsidian with 38 compositional types recognized, and is the only region in Kenya apart from the Suregei (northern Kenya) that contains rhyolitic obsidian. Nine compositionally distinct types of obsidian are known from southern Kenya. Although Kenyan obsidians span the compositional range from phonolite to rhyolite, low-silica, nepheline-normative obsidians occur only south of 1°N latitude. One obsidian type, the Lukenya Hill Group, appears to have been derived from a regionally extensive ash flow tuff with a distribution of over 8000 km². From previous studies it is known that obsidians of lowest (Mundui) and highest iron content were used for tool manufacture, as were some obsidians (e.g., Kisanana) with the highest alkali content, and obsidians with both high (Njorowa) and low (Kisanana) silica content.