(1) Overview


Spatial coverage

Description: New Zealand

Northern boundary: –34.1194851

Southern boundary: –48.9015311

Eastern boundary: 179.5011487

Western boundary: 165.1523805

Temporal coverage

Collections include artefacts that may date from the initial settlement by the ancestors of Māori around AD 1280 through the period of regular contact with Europeans after AD 1850.

(2) Methods

Each artefact was assessed mid-z elements (Nb, Rb, Y, Zr, Sr) and other elements (Fe, Mg) using a pXRF and then matched to a geochemically distinct natural source of obsidian. Geographically ambiguous source assignments, such as assignment to a general volcanic zone, were not included. We note that out of the 15 known geochemically distinct natural sources of New Zealand obsidian [12], only 13 have been identified in archaeological collections.


The two university laboratories that produced this data each specified procedures in a series of publications. The specific lab protocols and machines used evolved over this period and so readers are directed to original studies for a fuller description. Examples of methods for Bruker pXRF, a type used in both laboratories, is summarised below.

University of Auckland

XRF analysis was carried out in the Roger Green Archaeological Laboratories, School of Social Sciences, University of Auckland, using either a Innov-X Alpha (see methods described in [12]) or a Bruker Tracer III SD portable X-ray Fluorescence (pXRF) analyser (see methods described in [11]).

The Bruker Tracer III SD instrument employs an X-ray tube with an Rh target and a 10mm2 silicon drift detector (SDD), with a typical resolution of 145eV at 100,000cps. The X-ray tube was operated with a setting of 40 keV at 10.7µA, through a window composed of 12mil Al and 1mil Ti filters (Bruker’s Yellow filter). Samples were analysed in an air path for 60 seconds. Each specimen was analysed twice on a different portion of its surface area to check for consistency and the values were averaged. A total of 12 elements were quantified (K, Ca, Ti, Mn, Fe, Zn, Th, Rb, Sr, Y, Zr, Nb). Concentrations were calculated as parts-per-million (ppm) using Bruker’s S1CalProcess (ver. 2.2.33) software. A set of 24 reference specimens are used for this calibration, including 18 international standards (AGV-2a, ANU-2000, GA, GSP-2, JA-1, JA-2, JA-3, JF-2, JG-1a, JG-2, JG-3, JR-1, JR-2, JR-3, NIM-S, NIM-G, NIST-278, QLO-1b) and six New Zealand obsidian specimens from the Anthropology Laboratory reference collection that were analysed using Wavelength Dispersive X-ray Fluorescence (AU-17.59, AU-29.16, AU-32.1, AU-7.21, AU-9.3, AU-9.5). Reference standard NIST-278 is routinely analysed to check for consistency.

University of Otago

XRF analysis was carried out at in the archaeology laboratories of the University of Otago using a BrukerAXS™ pXRF. All samples were shot using optimal settings for ‘mid-z’ trace elements (Rb, Sr, Y, Zr, and Nb), specifically 40kv and 8 or 25 microamps at a 300 second live time and with a filter (12milAl + 1milTi + 6milCu, or what the manufacture refers to as the ‘green’ filter). To examine lighter elements (Si, Ti, Al, Fe, Mn, Mg, Ca, Na, K), a second protocol was used that engaged the Bruker pXRF’s vacuum, with the beam set to 15kv and 45 microamps, but with no filter. For both settings, laboratory specific quantification protocols were created and applied. Linear regressions were based on nine pelletized international standards each shot three times for each of the two setting. Green filter linear regressions were improved by applying Speakman’s OB40 calibration to raw counts before the lab specific linear regression. A pelletized USGS basalt standard (BHVO-2) was run alongside samples as a quality check of precision and accuracy, with an additional second standard (SRM-278) run on the vacuum setting to check the pXRF performance over a range of values (ppm).

Sampling strategy

See original studies for details on collections and sampling strategies.

Quality Control

The two university laboratories produced quality control data based on international geological standards. The specific quality evaluations vary from study to study and so readers are directed to original studies for a fuller description. Examples of quality check results for Bruker pXRF, a type used in both laboratories, is summarised below:

University of Auckland: Comparison of calibrated results to accepted values for reference Standard NIST-278. All values are in parts-per-million (ppm). [4]

K Ca Ti Mn Fe Zn Th Rb Sr Y Zr Nb

Accepted value 34534 7004 1499 403 14269 54 12 128 64 39 290 18
pXRF Mean (n = 10) 34898 7221 1507 399 15113 53 13 131 65 40 288 17
Standard deviation 633 196 85 14 113 4.1 0.3 2.4 2 0.2 1.1 1.1

University of Otago: Comparison of calibrated results to accepted values for several reference samples. All values are in parts-per-million (ppm). Overall, values come close to published values and results for each setting produce comparable results. The exception is Th, which for this standard is near the limit of detection (LOD).

BHVO-2 Mn Fe Zn Th Rb Sr Y Zr Nb

USGS recommended 1290 78144 103 1.2 9.8 389 26 172 18
Otago n = 43 (8 microamps) 1270 68332 116 2.2 10 378 24 159 16
sd 174 1599 41 3.2 2 18 2 7 2
Otago n = 7 (25 microamps) 1280 77388 101 0.6 12 371 25 164 17
sd 101 6117 11 2.2 2 28 2 11 2
NIST SRM-278 Mn Fe Zn Th Rb Sr Y Zr Nb

NIST recommended 403 14269 n.r. 12.4 127.5 63.5 n.r. n.r. n.r.
sd 2 140 0.3 0.3 0.1
Otago Lab, N = 3 (25 microamps) 419 13570 49 11.7 131 73 39 284 18
sd 276 190 7 0.4 4 1 1 2 1

As a check on accuracy on obsidians, an Otago lab standard Mayor Island Obsidian (AL380) is shown below as compared with published results from University of Auckland:

Mayor Island Obsidian Mn Fe Zn Th Rb Sr Y Zr Nb

Sheppard et al. (2011), pXRF 871 36942 194.2 n.r. 146.5 LOD n.r. 1146.9 n.r.
sd 113.6 2949.8 24 16.9 131.3
AL380 Otago, XRF (25 microamps, solid) 718 30578 164 19.9 149 LOD 119 1184 101
sd 61 353 11 2.3 2 5 7 3



(3) Dataset description

Object name

Sourced Obsidian Artefacts, New Zealand, 2011–18. There are two files in the Digital Archaeological Record, (1) Master_list_NZ_obsidians _2011_2018.csv and (2) Supporting_Data_NZ_obsidians_2011_2018.zip. The master list gives source assignments by site. The supporting data lists any ppm and/or lithic technology information available and includes 28 files.

Data type

Most are secondary data; previously unpublished primary data referenced as “this study.”

Format names and versions

csv, pdf

Creation dates

The data was created between 01/01/2011 and 09/18/2018.

Dataset Creators

Mark D. McCoy, Southern Methodist University

Ladefoged, Thegn N., University of Auckland




CC-BY 2.0

Repository location

Geochemical Sourcing of New Zealand Obsidians by Portable X-Ray Fluorescence from 2011 to 2018. Mark McCoy. 2018 (tDAR id: 446766); DOI: http://doi.org/10.6067/XCV8C250CS.

Geochemical Sourcing of New Zealand Obsidians by Portable X-Ray Fluorescence from 2011 to 2018 (ppm/tech). Mark McCoy. 2018 (tDAR id: 446767); DOI: http://doi.org/10.6067/XCV87947N7.

Publication date

The dataset is available on the Digital Archaeological Record (tDAR) repository.

(4) Reuse potential

Obsidian sourcing studies have become more commonplace since the widespread adoption of pXRF in archaeology. This dataset is the first national scale summary of artefacts definitively matched to source. It has excellent potential for future study regarding Māori culture history, social network analyses, and methodological development in the application of pXRF.