Using Palaeomagnetic Techniques to Uncover the History of an Archaeological Site in Napier, Hawkes Bay
Fired archaeological features and artefacts may carry records of the geomagnetic field at the time they were last used. During the 2020 demolition of early-mid 20th-century residential properties in central Napier/Ahuriri, an archaeological site was exposed. Six features, including two well-preserved hāngī (NP1 and NP2), two shallow fire-scoops (NP4 and NP6), baked floor of pumice-lined hearth (NP5) and a reddened pumice hearthstone (NP3) were excavated from beneath buried surface covered with shells and pumice. Thirty oriented fire-cracked stones were collected from four features: NP1, NP2, NP4 and NP6. In addition, eight oriented specimens were taken from the baked floor of a hearth (NP5) and a reddened pumice hearthstone (NP3) was also retrieved for palaeomagnetic study and potential dating. The natural remanent magnetization varies between 0.04 and 13 A/m while magnetic susceptibilities range between 0.1 and 4.9 x 10^{-6} m^{3}kg^{-1}.
Petrographic analysis reveals that the stones used in the hāngī are greywacke, commonly found in North Island ranges. Thermomagnetic experiments were conducted to understand the magnetic mineralogy and investigate magnetomineralogical alterations. Principal component analysis of alternating field and thermal demagnetization data yielded site-mean directions (declinations between 7° W and 21.3° E and inclinations between -47.7° and -63.1°) across five sampled features. NP3 was not included in any archaeomagnetic experiments due to time constraints. Some stones show evidence of disturbance during the cooling process. Palaeodirections from these stones were calculated using the low blocking temperature component of magnetization. The feature-mean directions were compared with the regional palaeosecular variation record, NZPSV1k to estimate archaeomagnetic dates for the sites. NP1, NP2 and NP5 provide good estimates while NP4 and NP6 have large uncertainities in their estimated ages. Palaeointensities were estimated using the Thellier-IZZI technique with pTRM and pTRM tail checks. The selection criteria were modified to include high quality data and enhance the acceptance rate of accurate results. The estimated palaeointensities for sites NP1, NP2, NP4 and NP6 range between 43.6 and 51.3 µT. These were compared to other archaeointensity studies from New Zealand, Australia and SW Pacific.
The archaeomagnetic results suggest an extended period of occupation, when prior to drainage and development ca. 1900 A.D., the area was predominantly wetland, and the site was surrounded by swamps and shallow tidal lagoons. There may be several other archaeological features associated with the site and extending beyond the current site, that now lies beneath the post-1931 structural development. Further archaeomagnetic studies on these features, if uncovered, may provide crucial information about the variation of the local geomagnetic field, and of history of occupation of the area.
