The 21.4 ka (cal. yr) Okareka and 17.7 ka (cal. yr) Rerewhakaaitu eruption episodes represent the first two recognised stages in the evolution of Tarawera Volcano, New Zealand. The deposits of both episodes have been examined to detail the eruption processes that occurred, and to investigate the geochemistry and magmatic processes of the eruptives. The Okareka eruption episode consisted of tephra eruptions that were terminated by the extrusion of lavas. The Okareka Tephra consists of multiple units (A-I), defined from medial­ distal exposures. Basaltic scoria (unit A) of 0.02 km³, was the first eruption deposit, with little or no time break before eruption of biotite and non-biotite bearing rhyolitic tephras (units B-I). Copious evidence exists in eruptives of unit A and B for mixing and mingling of basalt with the initial rhyolite represented by unit B. Unit B tephras (0.21 km³) represent phreatomagmatic eruptions with fluctuating water-magma interaction, shown by the interbedding of fine ash and coarse lapilli bands, and are distinguished by low Sr, Zr geochemistry. Later rhyolitic tephras consist of plinian (unit C, 0.48 km³); ignimbrite and co-ignimbrite ash (unit D); phreatomagmatic (unit F), and further plinian to sub-plinian (units G-I) deposits. The Hawea and Ridge Lavas (extruded at the end of the episode), are presumed to have buried the active vents of the episode. High Sr, Zr geochemistry characterises the unit C-I pumices and Hawea and Ridge lavas. The two geochemically distinct eruptive groups reflect eruption of two separate rhyolite magma bodies, with the low Sr, Zr magma erupted before the larger volume high Sr, Zr magma. Ascent and eruption of basaltic magma, coupled with brief basalt interaction with the low Sr, Zr rhyolite magma at depth, is believed to have aided in triggering the rhyolite eruptions.

The Rerewhakaaitu eruption episode produced rhyolite only and began with the eruption of Rotomahana Dome, followed by tephra eruptions consisting of multiple units (A-N), before the extrusion of more lavas (Southern Dome and Te Puha Lava Flow). Medial-distal exposures show that explosive eruptions began with a short plinian phase (unit A, 0.07 km³) , followed by dominantly phreatomagmatic activity (units B-D, 0.28 km³) and then renewed plinian eruptions (units E-J, 1.18 km³), which represent the bulk of the tephra volume in the study area and were dispersed strongly to the southeast. After a break in activity (unit K), further plinian activity was established (units L-N), but this sequence is poorly defined. Rerewhakaaitu eruptives consist of three rhyolite types, 1) hypersthene rhyolite (low Sr, Zr); 2) biotite-hornblende rhyolite (high Sr, Zr) and 3) hornblende-hypersthene rhyolite (very high Sr, Zr, very low Rb), representing eruption of three separate magma batches. Types 1 and 2 were erupted simultaneously from different vents as pumices representative of both occur in the same units around the volcano. Type 3 magma was erupted as lava only (Western Dome), towards the end of the episode. Mingling of type 1 and 2 lava is observed within Rotomahana Dome, whilst distinct type 1 and 2 lavas occur in Southern Dome, as well as lavas exhibiting slightly mixed geochemical trends. Proximal pyroclastic eruptives also exhibit some mixing. Interaction of magma types 1 and 2 is thought to have contributed significantly to triggering the Rerewhakaaitu eruptions.

Precursor activity to similar future eruptions may vary in duration from weeks to years. Okareka-type eruptions could be preceded by a shorter precursor period (a few weeks), than that estimated for the ~1315 AD Kaharoa episode (~5 years), as basalt was not injected through the rhyolite magma body (though it did interact with a margin of the body), and was erupted discretely. Rerewhakaaitu-type eruptions would probably involve a longer precursory period (months-years) due to rhyolite magma interactions at depth priming the bodies for eruption and the relative absence of any basalt. Tephra fall would be the greatest hazard resulting from eruptions similar to the Okareka or Rerewhakaaitu episodes, with much of the Bay of Plenty buried beneath at least 30 cm of tephra, and small amounts reaching Auckland. Eruption episodes similar to those studied are expected to be of similar duration (i.e. years) to those represented by younger Tarawera eruptions, and have much larger societal impacts than recent relatively short duration, low volume eruptions in New Zealand (e.g. Ruapehu 1995-1996).