After a runup of 4 months, or 10 hours, or 1.5 hours (depending on how you look at it), Kilauea sprung back to life, in its continuing quest to recreate the summit-wide lava lake of a century ago. There is still some way to go, but comparing this morning’s glowing lake to the Halemaumau lava pit of a decade ago shows how far it has already come. Over this decade the lava has grown from a pond 50 meters wide to a lake close to 1 kilometer across. It has been an eventful period.
The pit probably formed when lava traveling underneath the summit to Pu’u’o’o began to slow down and melt through the roof of the pipe. As the the connection between Kilauea and Pu’u’o’o began to clog up, pressure at Pu’u’o’o decreased and that at the summit increased. The lava in the pit slowly rose; eventually it just about overflowed on to the floor. This phase ended with the collapse of the Pu’u’o’o eruption. The lava found a new way out through the damaging Leilani rift eruption. Halemaumau collapsed.
This collapse eventually ended the rift eruption, when the path of least resistance was no longer out but instead went up. A quiet phase followed, until 17 months later the least resistance cracked. The collapsed crater now provided an easy way out for the magma. Since that time, the crater has filled with lava in a near continuous flow, albeit with two lengthy interruptions. Sometimes the lava would flow on top, at other times it would push up the lid. The second interruption came in September last year. Since that time, GPS measurements have shown that the crater was inflating, as magma collected at perhaps 2 km below the surface. Mauna Loa was preparing its eruption at this time as well. Was that related? Does the pressure in one volcano affect the other, even in the absence of a direct magma link? It is too early to tell. Hector and Chad were warning us that Kilauea was preparing to resume hostilities.
The GPS shows the movement of the CRIM GPS (south of the crater) and UWEV (north of the crater) over the past 2 years. CRIM was moving south, showing the centre of inflation was directly north of it (with some wiggling). UWEV was moving more northwest. The crater was expanding, and the inflation was centred in the crater. It may have been more to the northeast edge of Halemaumau. That is quite uncertain though with only two measurements. Both stations had gone up by some 10 cm, accelerating in recent weeks.
This inflation should not be overstated. The movements remain small compared to what the crater can do. During the 2018 eruption, CRIM went down by 2 meters! It has still only recovered 10% of that. UWEV went down by 50 cm: it has recovered about half of the 2018 collapse.
The picture above shows the 10-year curve. There were three phases of inflation, in 2019, 2021 and late 2022. The slopes are quite similar, at around 15 cm/yr. (There was a brief phase of stronger inflation immediately before the 2021 eruption.) This suggests that the three eruptions are similar in magma growth rate and depth. In contrast, the 2018 eruption did not have a strong inflation phase, but then this eruption did not start at Kilauea: it started with a magma chamber failure at Pu’u’o’o.
Once Mauna Loa ended its circus (it is only a month ago!), Kilauea saw its chance for the lime light. At Christmas, a larger earthquake moved the crater rim notably. A new dike started to form in the past few days, as shown by earthquake clusters. This was the state of the volcano only one day ago.
Just after midday (local time) on Jan 4, the tilt measurement at UWEV began changing. It moved up by about 3 microradians. This means there was a bit of uplift to one side, so that the instrument was tilting a bit. It is a very small effect, corresponding to 1 cm uplift at a distance of 3 km. It is a very sensitive instrument! It stabilized again, but began to tilt further in the early hours of Jan 5. Around 3pm it measured a sudden deflation followed by rapid inflation.
The seismographs confirmed the action. After a little tremor in the morning, a tremor wave hit at around 3pm. Magma was on the move: the seismograph detected the gurgling of the fluid, much like gassy plumbing. The tremor calmed down again by 3:30, after two earthquakes. This was the quiet before the storm. It is not uncommon that seismographs go silent in the hours before an eruption. The conduit is fulland is putting pressure on the plug. Nothing can move until the plug gives way: the magma is slowly increasing the pressure and is waiting for failure. It is not actually the magma that is pushing. Gas has come out of the magma and travels ahead. There is now a high pressure gas pocket directly underneath the plug. After a few months of no eruption, the crust is probably a few meters thick.
Failure came at 4:34 pm. The seismograph resumed gurgling, showing magma was on the move. The path was now open. Magma and gas combined to create fountains, reported as up to 50 meters tall. Underneath the fountains, magma flooded out and covered the solid crust in molten rock. HVO noticed the glow almost immediately.
You may want to read Hector’s post on the start of the December 2020 Kilaueau eruption!
The eruption is a very pretty one. The lava pool is big and red and covered with grey pancakes (thinly crusted lava). The views are clear and close and much better than the cloud-affected Mauna Loa eruption. Kilauea’s lava is also safely contained in a big crater. And one feature has really stood out in the early videos. Are you ready to surf the lave waves? (Not recommended!)
A video was published taken by Janice Wei showing lava fountains with spreading lava. It is fascinating to see how the lava moves in a clear wave, much like a wave in water. It even shows the lava wave breaking.
This video does not show a scale. The wave moves out over about 10 seconds before decaying. We know that the fountains were about 50 meters tall. Using that as a scale, the velocity of the lava wave is about 10 m/s initially, slowing down to 5 m/s later on. These numbers are only approximate! The height of the wave also decays, from about 5 meters initially to around 1 meter later on in the video. Again, take these numbers with a large grain of salt. The wave reaches about 200 meter from the fountain. There is only one main wave, but a small echo follows some 60 meters later, with a much smaller height.
This allows a simple estimate of the viscosity of the lava. The viscosity determines how much energy is lost by internal friction, and this causes the wave to decay at larger distances. Water waves decay slowly due to low viscosity, while honey waves disappear very fast due to high viscosity. A fun trick is to set up a wave in water that is beginning to freeze. You will find that the wave decays quickly: freezing water has a high viscosity.
We can guess the speed of the wave (roughly 10 m/s), the wavelength (say 60 m), the density of the lava (estimated at 2500 kg/m3), and the damping rate (roughly estimated at 0.04/m, based on the decline of wave height). There is now an equation that relates this to the viscosity. It depends on whether the wave is governed by gravity or by surface tension but this only makes a factor of 3 difference. Plugging my numbers into this equation gives me a viscosity of roughly 20,000 Pa s. (Pa s stands for Pascal-second, the units used for viscosity.)
(If you are interested, the equation for the viscosity is above where ρ is the density of lava, v the velocity of the wave, α the damping rate and λ the wavelength of the wave. This assumes a gravity wave: if instead there is high surface tension, the viscosity becomes three times higher.)
For comparison, Einarsson in 1966 obtained a viscosity of 10,000 Pa S for lava waves around the vent of Hekla. At Etna, Pinkerton & Norton in 1995 measured around 13,000 Pa s for Etna, and around 350 Pa s for Kilauea. Shaw obtained the viscosity of lava in the Makaopuhi pit crater in 1965 (using a different method: see below) and obtained 500 to 750 Pa s. The value I find here is normal for gas-poor lava but it is relatively high for Hawai’i. It could be lower if the wavelength is much less. Making it 20 meters would reduce the viscosity by a factor of 10.
The question was raised whether the lava was fountaining from the lava (magma?) sitting below the crust, or from new lava coming up from the deep. The viscosity suggests the ejected lava is not that fresh but has lost gas and may have crystallized a bit. Therefore it is likely that it is lava from the existing Halemaumau lava pool. The new dike increased the pressure and added gas, but did not (yet) contribute directly to the eruption.
The eruption rate has not been published yet. We can do a very rough estimate. The wave velocity decreased by a factor of 2 as the wave moved out. This corresponds to a decrease in depth of the lava of a factor of 4, because the wave speed scales as the square root of the depth of the liquid. Much of this comes from the decrease in the height of the lava wave itself. A quick estimation gives a depth of the lava surface of 1 to a few meters. Assuming this covered an area of 200 meters by 200 meters, and that the video was taken 15 minutes after the onset, the flow rate becomes 100 m3/s. This is a very rough number, and is likely an underestimate: the actual number could easily be several times higher (or more!).
A magma supply rate of 0.1 km3/yr would give a steady state eruption rate of 3 m3/s. That is too low for a stable eruption: it won’t get through the deep lava pond at this rate. This is the reason why the eruption must be intermittent: over time the eruption rate decreases as the fresh magma is depleted, but before it reaches equilibrium with the input rate the eruption stops. Now it has to wait for the pressure below the surface to increase, to allow the eruption to resume. However, this argument does not tell you how the eruption will last. That depends on the size of the available magma reservoir. Pu’u’O’o kept going for 30 years, although it did eventually peter out.
The eruption is likely to continue for months but will not remain as spectacular as it is now. The lava flood has covered much of the surface in fairly hot and shiny material. The rate at which lava is coming out will reduce as the pressure stabilizes. The surface will develop a grey crust and lava will be seen in patches and breaks. This is likely to continue for months, but eventually the eruption will suspend itself again. And the game repeats itself.
Do realize how far is has already come. Below is a USGS depiction from less than a year ago. The lake is now lapping the shores of the down-dropped block.
What is the end game?
Where will it end? The filling of the Halemaumau crater is not yet complete. If the filling continues, it may overflow in a few years. The situation would be similar to that of 1914-1918, when this crater also filled and over four years, overflowed on the floor of the large caldera. A century earlier, the lava lake reached another high point in 1830-1840. It is time for a rerun.
These high stands do not last. The links to the rift zones eventually fail and the crater drains. The drainage can lead to powerful explosions which can be dangerous to tourists. The rifts can then in turn feed fissure eruptions, as in 1840, but this may not always happen.
So we may have some years or a decade of more fun with Kilauea. Enjoy – but don’t be fooled. This volcano can have a temper when filled up too far.
Albert, January 2023