Distribution of geothermal resources in New Zealand: insights from geophysics and numerical modelling

New Zealand has abundant geothermal resources which produce ~20% of the country’s electricity. Most geothermal systems are located in the geologically complex rifting arc of the Taupō Volcanic Zone (TVZ), along the Australian/Pacific plate boundary. The TVZ is one of the most geothermally active regions on Earth, discharging ~4200 GW of heat through 23 high-temperature systems. Understanding the regional-scale factors influencing the locations of the geothermal systems is important for exploration and sustainable exploitation.

There are several intriguing correlations between TVZ geology and geothermal system locations. Over two-thirds of geothermal systems occur near inferred caldera margins. Many geothermal systems rise to the surface beneath topographic lows. Geothermal activity forms two NE-SW trending zones of low resistivity which are separated by the densely faulted and geothermally quiescent Taupo Fault Belt; most of the geothermal systems are not associated with any known major faults. We can gain new insights into these potential relationships by merging increasingly extensive geophysical and geological surveys of the central North Island of New Zealand into numerical flow models.

We created generalised numerical models of heat and fluid flow using TOUGH2 software to explore large-scale influences on geothermal circulation in the TVZ. Locations of over half the modelled geothermal systems can be broadly explained by the effects of topographic loading due to water table variations. Locations are further improved when topographic effects are combined with localised heat sources at depth inferred from magnetotelluric models. At three geothermal systems, influences such as more permeable volcano-sedimentary cover or a region of intense faulting that acts as a recharge zone for cold downwelling fluid also seem to be important. Three systems (Ohaaki, Te Kopia and Orakei Korako) cannot be explained with any of our models but are known to be in areas with significant local geological structures, making them interesting targets for future studies.