One of the ways GNG researchers are imaging the Earth’s crust to look for supercritical geothermal resources is through the use of magnetotellurics (MT).
Where are the heated fluids and their pathways? At what depths?
MT is a passive geophysical method that measures the natural variations in electrical conductivity (or its inverse, resistivity) in the Earth’s crust. Conductivity and resistivity are fundamental properties of materials. A low resistivity (high conductivity) means the material is a good conductor of electrical currents, and vice versa.
Resistivity is a property that we can measure from the surface. We are not actually that interested in the resistivity of a rock itself, but the Earth’s resistivity is very sensitive to fluids and temperature. This means rocks containing geothermal fluid, or partially melted, are much less resistive (more conductive) than normal rocks.
Under New Zealand’s central North Island, the Pacific Plate is sliding (subducting) underneath the North Island and causing extension in the Taupō Volcanic Zone (TVZ, an active rift). This extension thins the crust (just like stretching a loaf of bread) and brings heat from within the Earth closer to the surface; this heat drives both the geothermal and volcanic processes.
We use MT data to ‘image’ the hot fluids and melt in the Earth’s crust.
MT measurements are made at the surface, and can be used to image structures to depths of tens of kilometres. But these natural signals are very weak – about one million times smaller than an electric fence.
How is MT data recorded?
A changing magnetic field generates an electric current in a wire (the same principal as the pickups on an electric guitar). For MT, we measure the changing magnetic field at the Earth’s surface using three (orthogonal) induction coils, which are just tubes with a coil of wire inside. As the magnetic field at the Earth’s surface changes, tiny currents are induced in these coils, and measuring these currents is one-half of MT data.
The other half of MT data involves simultaneously measuring small electric currents that are induced in the Earth, which are called ‘telluric’ currents (from the Latin ‘tellus’ meaning Earth). These telluric currents are recorded by measuring the potential difference (a.k.a. voltage) between two pairs of perpendicular electrodes, spaced 50-100 metres apart.
That’s it really, the changing magnetic field at the Earth’s surface (the input) induces electric currents in the Earth (the output). Since the electric currents that are induced depend on the Earth’s resistivity structure, by measuring the input and output, we can use some math to figure out the bit in the middle – the Earth’s resistivity structure.
In the GNG research programme, we are not collecting any new MT data. Instead, we are focussing on analysing data and improving the geophysical models, using the abundance of data already available for the Central North Island.
In the 1970’s and 1980’s, the DSIR recorded ~30,000 DC resistivity measurements in the TVZ. These data mapped the resistivity structure of the upper 250 metres and helped to define the shallow boundaries of the geothermal fields.
Between 2002 and 2008, at GNS Science, we collected data from 150 sites, 5-7 kilometres apart to examine the regional TVZ rift structure.
Then our focus shifted (2009 to present) to improving the understanding of the deep (3 to 7 km) geothermal resource potential, and to image the geothermal ‘systems’ from top to bottom (i.e. from the geothermal fields down to their magmatic roots that provide the heat driving convection of fluid in the crust). In the past decade or so, we have collected data at more than 1000 additional sites, two kilometres apart.
The measured data are processed and a computer model of the subsurface resistivity structure is generated, called an inversion model. These inversion models (using data measured at the surface and spaced ~2 kilometres apart) are able to image the resistivity structure in the crust, and that’s kind of amazing!
These models are very useful, but it's important to note that they are aren’t exact representations of the Earth, they look a bit fuzzy and include uncertainty. It’s a bit like an MRI scan or ultrasound at the hospital, both techniques image within the body using data collected from the surface. Obviously very useful imaging techniques, but images are somewhat obscure and require trained professional to assess and interpret.
MT data allows us to map how geothermal systems work from top to bottom, and to look for any hidden deep geothermal resources outside of the known geothermal fields.
One example is shown in the images below for three high temperature geothermal fields: Rotokawa, Ohaaki and Te Kopia. These three geothermal fields have very different basement resistivity structure. Rotokawa has an aproximately vertical low-resistivity ‘plume’ ascending beneath the field, Ohaaki has a similar low-resistivity zone but it is offset to the northwest from the surface expression of the geothermal field, and Te Kopia does not have a strong connection to any low-resistivity zone in the crust.
These differences tell us that tectonic and volcanic structure in the crust influence heat and fluid flow, and influence the locations of the geothermal fields. These differences also support the idea that geothermal fields are not constant in time, and that their heat output may wax and wane as their magmatic heat source cools off or is replenished from below.
In GNG, we are improving our techniques used to model MT data. Specifically, commonly available 3D MT inversion codes approximate the surface of the Earth as a flat plane. While (most of us!) know that the Earth is not flat, this approximation is acceptable for deep imaging in places where topographic variations are small to moderate. However, for shallow imaging and particularly in places with steep topographic changes (e.g. volcanoes), accurately representing the surface topography in the model is important. With colleagues from Japan (University of Tokyo) we are implementing a new finite-element inversion code to improve our ability to generate images of the Earth’s resistivity structure from MT data.
These improved 'next generation’ images will be compared with other geophysical data (e.g. seismology) to better our understanding of how heat and fluid are transported in the crust, and where resources of supercritical fluid may exist.