The bed of Lake Taupō is rising and falling as magma moves around in the active volcano lurking below, new research shows.
Researchers from Victoria University of Wellington looked at 42 years of measurements recording minuscule changes in the lakebed.
The peer-reviewed research was published in the New Zealand Journal of Geology and Geophysics on Tuesday. It found the northern end of the lake had sunk by 14cm over the past four decades but this was offset by a 16cm rise within the lake near Horomatangi Reefs.
Co-author of the research, Dr Finnigan Illsley-Kemp, said the change in height could be due to magma moving closer to the surface during periods of unrest.
"Lake Taupō conceals one of the world's most active caldera volcanoes, which last erupted 1800 years ago," Dr Illsley-Kemp said.
"Movements of magma and tectonic faults beneath the surface frequently cause the ground surface above the volcano to uplift and subside.
"In 1979 we began a novel surveying technique which uses the lake surface to detect small changes, with four surveys made every year since.
"In this paper we summarise this 42 years of data to show that within the lake, near Horomatangi Reefs, the volcano has caused 160mm of uplift, whereas north of the lake the tectonic faults have caused 140 mm of subsidence."
Dr Illsely-Kemp said the data shows Taupō is still an active volcano.
"This shows that Taupō is an active and dynamic volcano which is intimately connected with the surrounding tectonics."
The research found that most of the unrest that occurred near the youngest vents was the result of magmatic inflation. Whereas the observed modern subsidence found in the centre of Taupō Fault Belt is likely to be driven by tectonic extension of the rift or even deep magmatic cooling.
The research said highlighting and documenting the changes could help scientists better manage the lake and active volcano below it.
But the study said additional monitoring equipment is needed if scientists wish to detect the precursors to an eruption.
"On the one hand the lake levelling sheds more light on the behaviour of the Taupō Fault Belt, perhaps yielding additional clues on the interaction between its tectonic and magmatic controls," the study noted.
"On the other hand its potential in detecting and quantifying an inflationary episode signalling volcanic unrest potentially leading to eruption is clear, especially if the data from each survey are processed without delay.
"To successfully combine the survey and cGPS and InSAR data it is obvious that installing at least two additional receivers - in the vicinity of Rangatira Point and Scenic Bay – in addition to the one at Horomatangi Reef, would significantly improve the prospects of detecting any eruption precursors in the earliest stages, immediately initiating urgent additional monitoring and resulting in better-informed prediction."
There was a period in 1983 of unrest where an "earthquake swarm" occurred in February and March approximately 3km north of one of the stations on the northern shore of the lake.
A routine lake levelling survey in mid-March showed that the Kinloch region had risen significantly at some stage over the previous five months.
"Frequent repeat surveys found that the uplift was continuing steadily until the Kinloch station suddenly dropped by 50mm following the commencement of a strong swarm reported in the same locality," the report said.
"The Kaiapo fault was displaced, also by 50mm, exactly a week later while the Rangatira Point area continued to rise steadily as the ongoing earthquakes spread southeast."
After the unrest in 1983, the lake levelling was relatively inactive, but it was followed by the uplift of Horomatangi lasting four years and then pausing.
It was a particularly active period from 2000 to 2009 for deformation and seismicity. Both Horomatangi and Rangatira rose 70 to 80mm in a series of steps before stabilising again for the following 10 years, apart from a minor subsidence and recovery of Horomatangi before a 30mm uplift occurred there in 2019.
"The repeated uplift episodes at Horomatangi and Rangatira, and one towards Rotongaio, occurred sometimes independently and at other times in unison," the study said.
"Both Horomatangi and Rangatira had reached a total uplift since 1979 of 165mm by mid-2021. In general, the uplift episodes were in the 25 to 60mm range with each lasting less than 1 to 4 years, resulting in rates of 6 to 75mm/yr, although the majority recorded rates of 13 to 20mm/yr."
All uplifts became permanent although many, especially at Horomatangi, were followed by minor reversals.
A history of the surveys
The lake levelling surveys started as an experimental project by the NZ Geological Survey to use the lake for measuring vertical ground displacements with a portable water level gauge at fixed points on the shoreline, the study said.
The initial aim was to develop a technique that was sensitive enough to monitor even minor vertical deformation across the whole Lake Taupō, which previously had only been found by occasional precise levelling traverses on the eastern shore and monitored by a series of tilt levelling patterns surrounding the lake.
The first six stations of the lake's levelling survey network were established in April 1979. They consisted of fixed reference points, such as nails or screws generally located on jetties and rock faces from where researchers could measure the vertical offset of the gauge suspended from a small hook below it.
Measurements were also made at beaches by suspending the gauge from a tripod set up in shallow water and calibrating it by precise levelling across to a benchmark above the beach.
Eight further stations were added between August 1982 and July 1983, immediately before and during a period of heightened seismic activity, the study said. Six more stations were then added in 1985 as part of a five-year deformation seismic study. The final two stations were added in April 1986, bringing the total to 22 stations.
To make sure station marks weren't lost, all stations had at least one additional mark nearby on a separate structure or rock face, or an extra benchmark of some form.
The lake level was found by looking at the observations taken at the beginning and end of each day's survey at the base station near Rainbow Point on the outskirts of Taupō. Experimental continuous observations were later made using an infrared sensor in a nearby stilling well there and at Kinloch Marina. Records from the automatic lake level gauge at Acacia Bay were also sometimes used to check the lake levels taken from the Rainbow Point observations.
Lastly, each station height was found by adding the average water level reading at the station to the lake level interpolated for that time from the Rainbow Point observations. A correction was then applied for the difference in height between the scale zero and the original station reference mark or benchmark.
After the early experimental surveys, it became more regular and included at least three regular surveys each year, sometimes spread out over a week or more if interrupted by adverse weather, the report said. By the time the network was completed, in April 1986, a standard routine of four surveys per year had become established, although additional observations were made during occasional earthquake swarms in an attempt to see whether there was any correlation.
