Hunting the supernova of 1181

Strelitzia reginae in its Manchester habitat. Photo by the author

One of my fondest memories of South Africa (apart from being shot at by police) is the bird of paradise. Not a bird: the birds of this name live in a different continent and a very different habitat. In South Africa, birds of paradise are plants with banana-like leaves growing a meter or more tall. Their official name is Strelitzia reginae. The plant is statuesque, but it comes to life in winter or early spring when flowering. The flowers are flamboyant and spectacular: huge and combining bright orange with purple. Although not hardy in the UK, it is popular (and expensive) nonetheless. But the garden centres promise too much: Strelitzia rarely flowers in the UK. Paradise needs sunlight and whilst this is plentiful in South Africa, it is not quite as frequent or powerful in the UK. Manchester is certainly not the ideal place to get a Strelitizia to flower. But we were lucky: after almost a decade, our specimen flowered for the first time last year and it put in a repeat performance this year. We are wondering whether to put it on ebay as a miracle Strelitzia, being Manchester-proof. But we would not be surprised if it will never flower again. This is a plant to keep in the forlorn hope that one day it will suddenly go from statuesque to spectacular.

(You can also buy Strelitzia nicolai which can be a bit cheaper. Be aware that (1) the flowers aren’t as colourful; (2) it doesn’t flower until it reaches full size, after about a decade; (3) that full size is the height of a tree, which is not ideal when kept as a house plant; (4) if you live in a frost-free climate it can be grown outdoors but keep it far away from your house as the roots can be destructive; (5) buy Strelitzia reginae instead.)

Sky gazing

My daytime occupation is astronomy. (You may see a slight contradiction in terms here.) There are types of events in astronomy that are in some ways comparable to this bird of paradise: they are like a star or asteroid suddenly flowering. One such event is a comet. Most of the time it is an icy body in the frozen outback of the solar system. But every now and then one approaches the Sun, and like the plant it becomes spectacular. A tail forms which can be as long as the distance from Earth to Mars. If the comet gets close to Earth, the tail can stretch across our sky. Hale Bopp, the comet of 1997, and Halley in 1987 may be familiar names. As comets go they were rather disappointing. Hyakutake, the Great Comet of 1996, was much more impressive: it came very close to Earth (ok, 15 million kilometers) with a bright tail more than 50 degrees long. I was lucky enough to see it from both hemispheres. By far the best recent one was Comet McNaught, the Great Comet of 2007, especially as seen from the southern hemisphere. One hundred times as bright as Hyakutake, I saw people stopping their car on the motorway (traffic rules being a bit more flexible in South Africa) to gaze at the multi-tailed spectacle.

Comet McNaught over the Pacific Ocean. Image taken from Paranal Observatory in January 2007. Credit: S. Deiries/ESO

There was another comet three years ago, and in fact another one will be faintly visible in the next few weeks. But those are no competition for the Great Comets. Those are the spectacle of a life time, as beautiful, as rare and as unpredictable as a bird of paradise flowering in Manchester. One day we will have another Great Comet. We live in hope.

Comet C/2022 E3, which will be visible with the naked eye over the next few weeks – if you are lucky. Image by Michael Jaeger, 12 January


Comets are not the only astronomical spectacle. Supernovae are, as the name suggests, also superb. A supernova happens when a star is utterly destroyed in a nuclear explosion. There are two different types, which with the nominative creativity commonly seen in astronomy are called type I and type II. Both types end the lives of their star in a thermonuclear catastrophe. The debris of the explosion leaves an expanding cloud of gas which remains visible for thousands of years. This is the type of event that astronomers live for but which few of us will ever witness. It is rarer than a Great Comet, and even rarer than that flowering bird of paradise in Manchester. Bright supernovae are gold dust, approaching the rarity of a VEI-7 eruption. Centuries may pass before one suddenly and entirely unexpectedly brightens our skies.

When they do happen, they get deserved attention. Suddenly a star brightens and brightens. It may become so bright as to be visible in the day time, reconciling my daytime job and occupation. Since the year 1000, there have been six or seven bright supernovae in our skies, in 1006, 1054, 1181, 1572, 1604, possibly 1670, and 1987. The last one was faint and was in a satellite galaxy to ours (the Large Magellanic Cloud) rather than in our own Milky Way, but it is included here because it was visible with the naked eye. It was spotted near-simultaneously by people in three different continents and even made the front page of Newsweek. The one in 1670 (called CK Vul) was originally classified as another type of object but is now thought to have been a supernova.

Supernovae of history

Which are those bright supernovae? Almost all predate the invention of the telescope. They were found by people looking up at the sky and noticing something new. In some countries, this would have been the job of a court astronomer. (The UK still retains the title of Astronomer Royal.) Elsewhere it was unorganized and left to chance, a bit like an NHS appointment in the UK. A star bright enough to rival Venus (ten times brighter than any other celestial object other than the sun or the moon) would draw attention anywhere. One competing only with the brighter stars could easily be missed, or reported only by word of mouth. The advantage of court astronomers was that the court’s chroniclers would make note of any discovery, in documents more likely to have survived.

In May of 1006, such a bright star appeared in the constellation of Lupus. It became the most dazzling astronomical event in written history. The star became more than ten times brighter than Venus and was easily visible in the day time. It was recorded worldwide, possibly even in the ancient rock art of Arizona’s White Tanks regional park. The star remained visible for two years.

1054 was the year of the Crab, or at least the Crab nebula. Again this supernova was exceptionally bright, nearly as bright as Venus. It remained visible (though fading) for more than a year. It appeared in July in the constellation of Taurus and was recorded in Asia and possibly America, but not, strangely, in Europe.

The Crab nebulae. Image from NASA/ESA Hubble Space telescope

In August 1181, a new star appeared in Cassiopeia. It wasn’t quite as bright as the previous ones – probably comparable to Saturn. It remained visible for 6 months but was recorded only in China and Japan.

Almost 400 years later, in November 1572, Tycho Brahe saw a supernova in the same constellation of Cassiopeia. It became as bright as Venus and remained visible for 18 months. It is now known as Tycho’s supernova.

Johannes Kepler, Brahe’s protege, found a supernova in October 1604 in the constellation of Ophiuchus. It was visible for one year. It was almost as bright as Brahe’s supernova. Brahe and Kepler are still the only people to have supernovae named after them.

In the spring of 1670 Johannes Hevelius found a new star. It was fainter than the previous ones, and not as bright as the brightest stars. The new star remained visible (intermittently) for two years. The remnant of this new star was finally found in the 1980s. It was long thought to have been a nova (a much more common but far fainter type of explosion) but recently it was found to have been a supernova, albeit a faint one.

The case of the missing remnant

Supernovae leave a lot of debris. Much of the material that made up the star is blown into space and it takes a long time before they disperse. Astronomy can be like archaeology: sifting through the debris to find out what happened.

There are many nebulosities (clouds) known on the sky which are identified as supernova remnants. They can be especially bright in radio waves, detected with radio telescopes. Many of these remnants are old and large, and come from supernovae that exploded more than 10,000 years ago.

All the bright supernovae known since 1000 should have clear remnants. In fact almost all do. The most famous is the Crab nebula, which is the remnant of the 1054 supernova. We also have one young remnant without a supernova: it is called Cas A and probably dates from the 16th or 17th century but the supernova was not seen, for unknown reason.

The supernova of 1181 is the outlier. It was identified with a bright radio source called 3C 58 (number 58 in the third Cambridge catalogue). However this identification became shaky, as this particular remnant appears to be some 5,000 years old. That age is problematic as the remnant should not be older than the explosion which caused it! All the other post-1000 supernovae have undisputed counterparts – this one doesn’t.

But the case of the missing remnant is now solved. We found it.

SN 1181

The discovery began in 2013 in Los Angeles. The amateur astronomer Dana Patchick (in his daytime job working at the California Veterans Home) was looking through archives of astronomical images, and came across a round nebula which was not in any catalogues. He is quite a prolific discoverer: this was his 30th find. As is common in astronomy, the catalogue he produced was named after him, and the object became known as Pa 30. The nebula was small, round and seen only at infrared wavelengths. Nothing much was visible in the optical.

Pa 30. On the right is the infrared image, on which Dana Patchick discovered the nebula. In the centre, X-ray emission is overplotted which comes from the exceedingly hot star. On the right is a very deep image in the light of oxygen, which just about shows the nebula. The green cross shows the central star. The bright star to the right of it is not part of the nebula but is much closer to us. Image from Ritter et al. 2021

Patchick passed it on to the experts and went on with his work. It turned out, this nebula was anything but normal. The expert was Quentin Parker, and he began his own work and turned telescopes on to this target. There was a faint star in the centre. It turned out to be extraordinary. The star was extremely hot (240 thousand Kelvin, according to the most recent research) and showed no evidence for either hydrogen or helium. The two elements together account for 99% of the mass of stars. They can’t just go missing!

Vasilii Gvaramadze found the object independently and first published it. He argued that the star came from a collision of two stars. Lidia Oskinova subsequently identified the nebula as a supernova remnant. In the mean time, our own team led by Quentin Parker had managed to measure the age of the nebula by determining how fast it was expanding: 990+-250 years.

It was time to put everything together. The lack of hydrogen and helium could only be explained by nuclear fusion. The nebula must have been ejected in a high-energy event. We found that the gas was fleeing the star at over 1000 km/s! That takes quite an explosion. The combination suggested a thermonuclear explosion: a supernova. A type-I supernova, to be precise.

The Cause

So I began to wonder whether this explosion might have been seen. European astronomical records are very patchy for the middle ages. There however records from the Far East: China, Korea and Japan. Korean records are also intermittent, with some large gaps, but the other two are quite complete. Events in the sky were thought to relate to events down here, at least where emperors were involved. Astronomers had an important part to play in the court proceedings, keeping an eye on the sky and observing (and predicting) solar and lunar eclipses. Detailed records were kept (albeit not with very accurate descriptions).

During the lock-downs I found myself with a spare day or two and went busy with the old Chinese catalogues (translated, of course). The goal was to find recorded events in the correct part of the sky.

The Chinese records list the events, the durations and the approximate location on the sky. They are classified into three types: ‘guest stars’, ‘bushy stars’ and ‘broomed stars’. The last two types are mostly comets: a ‘broomed star’ is a comet with a tail, and a ‘bushy star’ is a blob but has no long tail. Guest stars are stars that appear and disappear, and unlike comets do not show a tail or move on the sky. So I was looking for guest stars in the approximate direction of Pa 30. ‘Approximate’ because the records are not particularly good at describing position. They normally give an indication where on the sky it was or on which constellation but not in great detail. They may even just say ‘in the north’ (which can mean circumpolar), but they can also state which other stars it was near to.

I found two possible identifications in the Chinese records. One was in the year 722 and one in the year 1181. The 722 event seemed to be a normal nova and was seen for only a few days. The one in 1181 was different. It had remained visible exceptionally long (compared to other guest stars), at 185 days. And of all the events I had looked at, it was the only one not classified as a nova or comet. It had been a supernova.

It was one I remembered. Years earlier I had looked up publications on this supernova. There was a remnant in the literature but it did not convince me. It seemed to me that the characteristics of the remnant, 3C58, did not fit well with the event. So when this one jumped out from the catalogues, it seemed plausible that there was another remnant. That was true for no other supernova in the list. I had a prime suspect.

We published our result in 2021 (Andreas Ritter et al, There has been a lively discussion since, but the result that we found the remnant of the most recent unidentified galactic supernova has so far survived. Let’s see.


The Chinese constellations are very different from the western ones. Our familiar constellations started about 4500 years ago, probably in the region of Iraq. We know the date fairly well because the location of constellations on the sky change over time. Stars in our constellations are those that were visible from Iraq at that time. The constellation of Taurus (the bull) was the start of the zodiac: at that time it was the constellation that heralded spring. The Chinese constellations are not as well described, and often used fainter stars than we use and separate smaller parts of the sky. The term ‘asterism’ is used for a grouping of stars that is not one of our constellations.

The chart shows the relevant part of the sky. The size of the dots indicates the brightness of each star. Green indicates our constellations: Cassiopeia and Cepheus. Red indicates the Chinese asterisms: Hugai, Chuanshe, Kotao and Wangliang. There are stories behind them.

From Ritter et al 2021, The remnant and origin of the historical supernova 1181 AD.

These particular asterisms are made up largely by quite faint stars. They describe the area of the imperial palace. The palace itself is represented by Ziwei, an asterism just to the top-right of the area shown. Huagai is the canopy of the emperor. (It may seem to be upside-down, but the palace Ziwei is to the upper right from there, so shielded by the canopy). Wangliang represents the charioteer of the emperor. Chuanshe is the guest house of the palace. There is a historical connection here: just before the new star appeared, an ambassador from the rival northern Chinese Jin empire had arrived and was staying in the guest house. Finally, Kotao (‘mountain road’) connects a number of brighter stars. It is unusual in that this asterism crosses another one, Chuanshe.

The exact shape of these asterisms is not known and may have varied over time! We don’t have drawings from those times: the ones we have were made much later. Huagai has another brightish star just above the left end of the canopy: if this was part of the original canopy it would be better aligned with the palace. Kotoa is quite certain because it has brighter stars. Wangliang is also quite clear. But Chuanshe connects very faint stars (barely visible to the naked eye): we know there were nine stars but we don’t know which nine. We know that Chuanshe was between Wangliang and Huagai, but there are several possibilities. The chart shown here depicts two options.


There are 6 separate documents with useful information regarding the location of the guest star, 4 Chinese and 2 Japanese. The longest one is the Wenxian Tongkao, the diaries from the southern Song dynasty. Two of these state that the star appeared in the lunar lodge of Kui. This is not shown on the chart: it covers Huagai, most of Chuanshe and the area below it. One document states it was ‘beside Ziwei’ and one puts it vaguely ‘at the north near Wangliang’. The most specific locations state that it was ‘guarding’ or ‘invading’ Chuanshe. And one document even states that it was guarding the ‘fifth star of Chuanshe’ – sadly we do not know which star this was!

The important statement is that of ‘guarding’ or ‘invading’. This was used in the sense of ‘adjacent’ or ‘very close’. It implies that the guest star was not in the constellation, but very close to it, just like a guard or invader stands just outside.

The combination of the statements implies that the guest star had appeared in between Huagai and Chuanshe, probably closer to the latter.

The previous candidate 3C58 is indeed there, at the left edge of the possible range. Pa 30 is also in the range but at the right edge.

So which one is the correct location? Either fits the descriptions. 3C58 though is in a special location, at the point where Chuanshe and Kotao cross. It is also close to a brighter star. None of the documents mention this. The ‘fifth star’ could point at 3C58, as it is right next to a star that could be the fifth one. But depending on which stars to count, there are ten (!) other possibilities, two of which would agree with Pa 30. The reference to Ziwei favours a location to the right of Huagai. On balance, Pa 30 has a higher likelihood of being the counterpart.

From Bradley Schaefer, 2023, Chinese and Japanese observations of Supernova 1181.

The chart shows the combination of all constraints from those records. The blue region is the area that agrees with all descriptions. Pa 30 is in this area whilst 3C58 is not.


Celestial events were thought to be associated with the divine (at least considered divine) leaders of the empire. That is not unique to the Eastern world. It was prevalent in the near east, probably from the inception of our constellations 4500 years ago. ‘Wise men’ (i.e. astronomers) from the east even brought it into the bible, a book which was otherwise mostly robust against linking celestial events to our lives. If you call yourself the Sun King, a sunspot can be a sign of a disaster. So obviously, a new star on the sky next top the symbol of the imperial palace would have been a concern, especially when an important ambassador from the rival northern empire was just visiting. Whose side was the new star on? Was there a reason it was said to be ‘guarding’ or ‘invading’ the guest house where the ambassador was staying?

The Japanese were equally concerned. The diary of Fukiwara Kanezane, of the imperial court, reads ‘a guest star has been present in the inner sky, a sign of abnormality indicating that at any time we can expect control of the administration to be lost’. This new star was serious business.


The one thing we don’t know well is how bright this star was. It was bright enough that it was noticed easily, and as it remained visible for 185 days, it must have been fairly bright to begin with as these events fade over time. It was also bright enough to be seen as an omen. On the other hand, it was noticed by court astronomers but did not create a great stir elsewhere. That suggests something say as bright as Vega: not quite the brightest star in the sky, but not too far off. That would be quite faint for a supernova in the Milky Way. Even at its distance of 8000-9000 lightyear, it should have been 10 times brighter or more.

There is one comment in a Japanese document that compares it to Saturn: ‘At the hsui hour [7-9pm] a guest star was seen in the north-east. It was like Saturn and its colour was bluish-red and it had rays’. We don’t know what this mean: Saturn was nowhere near the supernova at the time and is not ‘blue-red’. It is often taken as indicating that the new star was as bright as Saturn. But this description was written a century later. It may have confused things: the entry is clearly about the 1181 star, but the description about the colour and rays and the specific time sound more like aurora.


The case that we have found the remnant of the supernova of 1181 seems convincing – at least to me. It fits the location and age of the remnant well, and the lack of hydrogen and helium is strongly suggestive of a thermonuclear explosion – a supernova. But why was it rather faint?

And why is there still a star? Supernovae are destructive events which the star does not survive. You may get a neutron star, a black hole, or there may be nothing left at all. Pa 30 still has a star at the centre. How is that possible?

The most likely explanation is the one by Gvaramadze: this was not a normal explosion, but was caused when two dead stars (white dwarfs) collided or merged. Many stars exist in multiple systems – in fact our sun is a bit uncommon in being alone. Once a star has finished its life, other than by supernova, it leaves a white dwarf behind: very compact, and consisting mainly of carbon and oxygen. A double system can leave a double white dwarf, often orbiting very close to each other. Over time the two stars spiral in, and eventually some may merge.

Such a merger would lead to a partial supernova: there is a thermonuclear explosion, but it does not blow up the entire system. In fact, models suggest this leads to a faint supernova. About 10% of the supernovae in the Universe may be of this type.

We had never seen one in our Milky Way. But it fits well with the supernova of 1181, with its brightness, its remnant nebulae and with the surviving, very strange central star. And with 10% of the supernovae being like this, it is not unreasonable that there would be one hiding among the historical supernovae.

The nebula

The nebula itself is very faint in visible light. However, there are ways to bring it out better. Robert Fesen managed to obtain a deep image through a narrow-band filter that isolates light from sulfur. Supernova remnants can be brighter in such a filter. The result was spectacular. It was just published (Fresen et al. 2023,

Fesen et al, 2023

The new image shows a very large number of stripes. These are more normally associated with planetary nebulae and have not before seen in supernova remnants. The stripes are material that is either swept back by the strong wind from the star, or are shadow regions behind a dense globule where they are shielded from the hot radiation from the star. Either case requires that there are dense globules in the remnant, which can not be seen directly.

In fact, every single stripe looks a bit like a comet! Of course they aren’t (they are 10,000 times larger, for one). But what a way to get the two most special astronomical events in one.

This has been an adventure. It started with two Great Comets and one great flower and ended with a Great Explosion, or if not the explosion itself then at least the debris of one that happened more than 800 years ago and which caused great concern at the time.

All that is still needed is a major (but safe) volcanic eruption. Or was that perhaps Hunga Tonga?

Albert, January 2023

Fuzzy tails in their more usual location, around a planetary nebula. This is a detail from a JWST image of the nebula NGC 3132 (

111 thoughts on “Hunting the supernova of 1181

  1. Great writing, again. Question: How can they (supernovas) stay so bright so long, for many months?

  2. I was ever astonished by the lacking of any european report of the 1054 supernova. I have read (the information came from chinese records, i think) that the “new” star was so bright to outclass a full moon for the light it emitted, and should have been visible for many week during the day.
    How on earth no one has registered its appearance ? At least in southern Europe and in the Mediterranean it was so high on the horizon to be clearly seen, I have read. Were the times so obscure anyone had time to think and register so great an event ?

    • I was nowhere near as bright as the moon – that claim is wrong. It was bright enough to see in the day time though, for about a month. It was briefly close to the moon on the sky (on July 5, 1054 to be precise) and the native american rock art depicted this, with the moon in its last quarter. (It is however very hard to prove that the rock drawing represent the supernova and not say a planet or comet.) There are some disputed notes in Italian documents that could relate to the supernova but have a lot of discrepancies. Europe was not in a good state at this time, evidently. If it had happened two hundred years later, it would have been recorded in many places.

    • I should perhaps add that this supernova was visible in the early morning sky. Perhaps the culture in Europe at that time did not involve early rising!

  3. If you do a simple search of how often supernovae are expected in a big galaxy like our own it says about every 50 years. Obviously most of the Milky Way is not really visible, but about 1/3 is, so is it correct to expect a visible supernova about once every 150 years on average?

    Apparently the last one within that visible 3rd was in the early 20th century but was not bright in visible light, and close to the galactic center.

    I guess we will have to wait until Eta Carinae explodes, it is probably the most certain of the known candidates being it is the biggest star in the local part of the galaxy but then even these sorts of short lived O stars change at a timescale that we cant comprehend so it could still be a while… Eta Carinae is a southern circumpolar star though so I will get a perfect view of the thing if it does go 🙂

  4. Bravo !!

    IIRC, one of my former colleagues’ younger sibs discovered that mentioning his occupation to ‘dates’ too-often prompted delighted witterings about her ‘star sign’.Though a reliable diagnostic for ‘ditzy’, he preferred not to squirm.
    So, he learned to say he worked for a well-known Bank, ‘NDA on detail’.
    Jodrell Bank…

    FWIW, I’d really, really like a definitive mass for Tau Ceti ‘f”. Sadly, as non-transiting, Doppler can only give Sin(i), poorly constrained by possibly-skewed asteroid belt to 3~~9 Earth masses, probably ~~5-ish…

    • During the financial crisis, we discussed whether Jodrell Bank could apply for a bank bail-out..

  5. Fascinating article. Thank you! I’d love to see Betelgeuse blow in my life time. But I strongly suspect I won’t!

    • But then Orion will lose his shoulder =(.

      Hmm, when I was in Iceland a couple months ago, on our way up to the north / Akureyri, we stayed an overnight in a very sparsely populated area and lucked into a crystal clear evening with plentiful auroras. Orion was so vivid and spectacular that evening, and an aurora appeared right through the center of the constellation. Was just draw dropping… I have to get around to developing my DSLR shots from that night.

      Thank you too Albert! This was a pleasure to read. Astronomy was the first science I grew an interest in as a child, I remember my parents got me a textbook about our knowledge of the solar system circa mid 1990’s and it kickstarted a lifelong passion in thinking about the natural world. At one point when I was younger, I actually wanted to be an astronomer, but as I got older I realized I lacked the ability for complex mathematics (or likely, the patience).

      Wonderful article, would love more from you in the future that incorporate your “day” job.

      • Dont worry, Betelgeuse isnt considered a supernova candidate now as studies have found it is still in the He burning stage, meaning it has potentially another million years left in it. It is also possible that supernovas dont actually happen from the red giant stage but the star contracts again.
        There are also Wolf Rayet stars, which are basically planetary nebulae except formed by really huge stars, the star blows off its outer layers but then still goes supernova, so it is about the opposite of a red giant when it blows up, it is a 100,000 K dense ball of heavy element plasma, a million times the luminosity of a normal star 🙂
        Because there is also a nebula around the star the supernova lights this up and it glows way brighter than a normal supernova, which is why they are sometimes called a hypernova. Wolf Rayet stars are also believed to be gamma ray burst progenitors, or at least one of the main causes from what I can find.
        There are also even brighter supernovae that are formed from stars that create a huge amount of nickel, which is formed as a short lived radioisotope that decays to iron. So imagine several solar masses of extremely radioactive material suddenly unconfined and allowed to do its thing, it is like a nuke the size of a star 🙂

        Just saying, but if you want to see a hypernova, then wait for Eta Carinae. It is all of the things above, and only 9000 ly away. But its exact details are not well known enough to give an idea of how long it has left to live. but given how short lived these stars are, it is not going to be in much of a different position relative to our solar system when it does go, so those around wil lget one hell of a light show 🙂

        • Thanks for this, great stuff.

          Unsure if you are into PC gaming at all, but I highly recommend Elite: Dangerous. It’s the best space sim I’ve ever played with respect to the feeling of actually being in a reasonable facsimile of space. It simulates a 1:1 representation of the Milky Way with many major nebulae and stars approachable. People have different opinions of the gameplay, but for me it’s an audiovisual masterpiece and stimulates the imagination like nothing else I’ve played in my 30 years of gaming.

  6. A curiosity: an M6 earthquake within 100 meters of the equator

  7. I love that you managed to bring Strelitzia into this article, Albert.
    I gave one to my mother as a present one year. She managed to get it flowering several times over the next 20 years or so.
    But then, she was both South African, and an expert gardener, even by my standards (I’m RHS-trained. Pershore College).

    • As I mentioned my mother…
      On his retirement, my father took up astronomy as a hobby. He never did anything in a half-hearted fashion, so prior to his passing in 2019 he had developed some expertise in building large, backpack-portable telescopes in his “inventing shed”, replete with large mirrors etc. He had also converted a trailer into a mobile observatory. Even living in Malvern, he was always keen to escape light pollution.

    • Thanks Albert for the article. Really interesting story showing how amateurs can do good science, and how much there is still to explore. (My game of choice at the moment is Stellaris, boldly going where no desktop has gone before!)

      The giant Strelitzia has white flowers with the blue flash, and there’s also the totally white S. alba.

      I nearly killed my inherited S. reginae a few years ago, leaving it out in the Sussex weather while the conservatory was rebuilt. It died back to one tiny spike and has recovered to bloom 3 years in a row, a new flower spike is half-grown now.

    • I have one in my drive way about 8′ tall, flowers profusely, Then again i am living in RSA so thats cheating.😀

      • Growing a plant in its natural environment is unfair. It is not the British way..

  8. Been keen on Astronomy and Space since a toddler – I even took my first steps on Apollo 11’s landing (suspect my Dad tried really hard that day) and met my hero Sir Patrick Moore in his eightieth year. So Volcano Cafe is even more a joy to read with added Space too, and it really is a connected topic in many ways (not just Io and Olympus Mons). Thank you Albert. On a tangential note, can’t help feel Taal is up to something surprising soon.

  9. Not on topic of article or about volcanoes, but something I found that is pretty incredible for the future, and finally something us Aussies can take pride in for this sector 🙂

    50% increase in energy and 500% increase in power, would allow for serious electric aircraft let alone basically unlimited options for electric cars. As I understand it, the batteries are basically the same as LiFePO4/LFP chemistry, but are more of a solid state design (not truely solid state though). A 500% increase in power though is insane, it allows for a way smaller and lighter battery without compromizing performance. A Model 3 with this chemistry would be able to achieve peak performance with a 100 kg battery pack, and though such a pack would contain half of the energy of the standard one, the car would also be 400 kg lighter, and that is going to improve range a lot, so there might not really be any difference or only a small one. Considering the normal 3 has a range of about 500 km up and people have been hypermiling them to twice the rated range since 2017 (when range was only 300 km), 400km would still be easily achievable I expect, and the car would weigh the same as a Toyota Corolla. I also expect a 5x power output would allow for at the very least a significantly higher charge rate too, so a smaller battery would probably charge in literally 2-3 minutes at a fast charger if the power was high enough.

    I can move this to the bar, it is off topic, but this is something that could really be a game changer and is not just a lab breakthrough like so many battery innovations are. The fact it is also in Australia and not the US or China, to me as an Aussie it given me some hope we can put our mark on the world in a good way and not just send it overseas anymore 🙂

    • Don’t confuse power (watts) with energy (watt-hours). The company claims a “potential” power density of 190 wh/kg. That should be energy density. Current LiFePo4 batteries claim 90-165 wh/kg. If they reach their potential, it will be good, but not as earth-shattering as it sounds. The miss-leading (potential) 500% power spec refers to discharging rate. I can’t imagine a car or airplane ever needing an especially high discharging rate.
      When they bold-face these statements, they are mis-leading the public. Karma will get them.

      • Thanks billg, saved me the trouble. Basically batteries are chemistry and an atom can only gain or lose a few electrons so with lithium (valence 1, at wt 7 is hard to beat. Beryllium valence 2 wt 9, Boron valence 3 wt 11 (good luck getting those electrons off) have ‘difficulties’.

      • I do know the difference, an increase in the discharge per the weight of the battery is the power density, or the power/weight ratio. The energy density is how long that much energy can be used and how much is stored per the weight.

        It should be said, that adding up at the level of a whole car, a well made BEV and an ICE car are basically the same energy density. A 50% increase in this is expected by incremental advancement in a few years, but larger jumps are likely, it only takes one escaping the lab successfully. But today high performance EVs also have large batteries, a design like this article with very low resistance would change that and also address charge time which is for some reason still important.

        It is harder to get high cycle life out of a battery that functions on atoms using more than one valence electron. This goes both for the anode and cathode, it us why Li/O2 batteries are not widely used or long lived, those have an energy density more than 10x higher than anything in a car now but low power and very poor cycle life.

  10. Mauna Loa now has pictures of the 2022 flows on google earth 🙂

    The opening summit stage of the eruption came from a 10 km long unbroken curtain of fire, which seems to have opened entirely within an hour of the start.

    The whole length of the fissure system is 20 km, it is as long as Laki. The main magma chamber of Mauna Loa seems to be the same part of the volcano structure as the south caldera reservoir of Kilauea. Unlike on Kilauea where there is frequent lava lake activity and at least the Halemaumau magma is degassed, Mauna Loa magma sits fresh until the last moment, hence the enormous SO2 flux, which might be the highest ever measured in an effusive eruption to date.

  11. in space no one can hear you scream

    ( Xenomorph emerging out of Alberts kitchen tap )

    • They built a nest under Alberts kitchen drains … time to call Sigourney Weaver for some home help extermination and cleanup

  12. Skyglow causes decreasing sight on stars, standing on earths surface. The skyglow effect provoked by illumination is still increasing in dense populated areas. I expected led technology would have caused a decrease of glow, the light is projected far more downward where it is absorbed. But apparently it is not?

    Someone knows why skyglow is still increasing?


    “The night sky is rapidly getting brighter, eroding star visibility worldwide
    Peer-Reviewed Publication

    … “In many inhabited places on Earth, the night sky never fully grows dark. It instead glows with an artificial twilight caused by the scatter of anthropogenic light in the atmosphere. This type of light pollution, called skyglow, is responsible for the visible brightening of the night sky and the erosion of our ability to see stars. Although the ubiquity and luminance of skyglow have increased exponentially over much of the last century, its global change over time is not well understood. Satellites that can measure global skyglow are limited in resolution and sensitivity and are often blind to the wavelengths of light produced by modern LED lights that have come to dominate lighting over the past decade. To better understand how growing light pollution is affecting our view of the stars, Christopher Kyba and colleagues evaluated 51,351 citizen scientist observations of naked-eye stellar visibility from 2011-2022. To determine the background brightness of the night sky, Kyba et al. asked participants worldwide to compare star maps of the night sky at different levels of light pollution to what they could see with their own eyes through the online “Globe at Night” platform. According to the findings, the night sky has increased in brightness from artificial light by roughly 7 to 10% per year, which is equivalent to a doubling of the night sky’s brightness in less than eight years. This increase is much higher than estimates of the evolution of artificial light emissions (~2% yearly) based on radiance measurements taken by satellites. “Perhaps the most important message that the scientific community should glean from the Kyba et al. study is that light pollution is increasing, notwithstanding the countermeasures purportedly put into operation to limit it,” write Fabio Falchi and Salvador Bará in a related Perspective. “Awareness must greatly increase for artificial light at night to be perceived not as an always-positive thing, but as the pollutant it really is.”” …

  13. Once, in the Oz Snowy Mountains, a frosty night would bring the stars within reach of your touch. They are now in retreat. In the North east a glow reminds one of the manifold menaces of Canberra. A malign influence from which even the heavens are not spared.

  14. Once, in the Oz Snowy Mountains, a frosty night would bring the stars to within your touch. They are now in retrest. To the North east a glow reminds one of the manifold menaces of Canberra. A malign influence from which even the heavens are not spared.

  15. Kilauea lava lake has gone up by 20 meters since it started, or an average of a bit over a meter a day since it began. Considering how big the lake surface is now that is some rapid rise.
    The latest map gives a rise of 9 meters, so there seems to be a discrepancy between the models of the whole crater and wherever the laser is pointing. My take is the laser was pointed at the 2021-2022 lake, which receded, and that the pit was a lot deeper than the crater floor as a whole. When it filled again, the whole crater floor lifted by 9 meters, and the lake within the pit also rose by a similar amount. So half a meter a day, still impressive.

    The lake at the east though seems to have risen more, while the edges of the crater have not risen at all, there are no breakouts only some small overflows of the active surface lakes. I guess the intrusion of last September sagged the lake surface down enough that the eruption now has not offset it yet, or at least not enough to break the surface now after a few months to thicken.

    • Phil Ong of Hawaii Tracker thinks that the rangefinder target has been moved to a new position on the latest lake. There’s a break on 15/1/23 where the level oscillates between 378 and 381 metres, consistent with the instrument alternating between two levels. As the original pond backfilled from the new vent, the level is lower and not as responsive to change as the new perched pond.

      • I wondered about that too, but HVO havent said anything about it so left it an open question.

        I dont think the old lake backfilled from the new eastern lake though. Tbe small pond between the big lake and the old island is the original vent that first opened, on the 6th and was already a pit last year that would glow and stayed glowing until this eruption. The 2022 lake had multiple openings within it, originally it flowed from the side inward and drained towards the east and south, but now it fills directly from the east, reverse to last year. The whole crater is filled with lava that sits probably 10 meters below the surface at most, the vent that opened on the 6th is probably directly under the biggest lake but would really open into the whole crater first, so all the lakes are fed from below not by tubes from each other this time.

    • Looks like the lake has risen up enough to break out through the edges again, the surface lava lake might not overflow anymore but the crater floor will rise up now as a whole. Given that there is at least somewhere on the crater floor that is 20 meters above where it was at the beginning of the year it might be quite fast for the lake to overflow onto the downdropepd blocks now, probably within the next month if not sooner. It was expected this would happen last November before the lake stopped rising in September, so seems like it is playing catch up 🙂

      Even though there isnt a lot of flow activity in the crater the effusion rate seems at least higher than it was on average last year, even now after over two weeks there is still a strong fountain within the lake, which is something that largely stopped by now in the last two eruptions. The GPS has levelled off, the effusion rate now seems to be the same as the rate of resupply, where most of last year there was inflation indicating not everything was reaching the surface, and the same was the case in 2021. Maybe the vent of this eruption opened up right at the bottom of the lake, so in effect now none of it is rootless anymore.

  16. Its not like what was going on in the first week of the year, but there are still a lot of earthquakes under Kilauea, but these are not shallow they are at 5-10 km deep, they are underneath the major magma chamber that feeds all of the volcano not just Halemaumau. What this means I dont really know but this idnt happen after the 2020 or 2021 eruptions, it did happen in 2018 but then there was a lot going on that year…

    All of the summit GPS are showing downward movement, which would indicate deflation, but they are also not showing movement relative to each other so the deflation seems to be coming from deeper down, not from Halemaumau but the main magma chamber. There are still DI events though, which is interesting. Seems that maybe the whoel summit is sagging down as this eruption goes on, which is not really something that was seen outside of the very intense start of the last two eruptions. But this eruption slowed down faster and seems to have stabilised at a higher rate than the other two.

    The only thing I can think of that would cause such quakes in the walls of the magma chamber though is pressure, which is not exactly what the instruments show. It could be a new magma intrusion below the summit, the eruption could change in the next few weeks.

  17. Not sure if this was already posted, but a recent report on research done by C.I.T. on the ongoing Pahala swarm is concluding that the quakes are indeed magmatic in origin.
    Mostly through the use of a high density array of micro-seismometers and other instruments, they were able to actually track pulses of magma as it moved through a combination of conduits and sills.
    The paper also shows that the main feed at depth is hydraulically connected to Kilauea…while Mauna Loa has a discrete magma supply.

    • There was a pretty big quake down there just today, a 4+. Not huge but it really threw up the cumulative seismic energy on the graph, the last quake like it was the one north of Kilauea a couple days before the eruption resumed that was a 4.

      To make a mag 4 quake that deep in the mantle is crazy, this is some of the hottest mantle on the planet, if it wasnt for there being an island above it the location would be liquid at the temperatures present. And the scale if these sills is huge, there are probably many tens of km3 of magma moving down there, even if only 10% erupts one day at Kilauea that could be a big deal.

    • The link has an extra trailing p in it, so clicking on it does not work. If the p is removed it works. They say the largest sill is 6km by 7km and 300m thick. If we assume an ellipsoid shape, that’s around 50km3 for the largest sill, and there are “more than a dozen” sills down there.

      • 12×50, that is 600 km3. If even 1% of that erupts, that is the same as all of the magma erupted in Hawaii since 1980, from both volcanoes abd including Mauna Loa last year. If this is a sill complex that is a whole different thing to a deep cumulate layer or a batholith, those are plutonic, but sills are not really, sills dont tent to just stay put they try to erupt and big sills feed big volcanoes. There are actually shallow small quakes under 10 km depth above the swarm from the force on the crust these intrusions are generating, says it all really…

        I hope now with this data a timelapse of sorts can be made of all the activity since 2018, are the sills long lived features that have been there for a long time and are part of Kilaueas standard activity, or are these new structures that have only begun to form in the past few years and will potentially drive a major eruption in the near future.

        But consider this, if these structures are new sills… that is a potential supply rate of hundreds of km3 a year of magma to the area. Otherwise the magma was already there but now it has collected and separated from its crystal load. So basically either Hawaii has just become a real LIP or a VEI 7-8 sized magma chamber has just spontaneously formed in the mantle at the source region for Kilauea.

        This… is insane.

        • Wait, I made a silly mistake. Used the diameter instead of radius. Same mistake on 3 axes makes a factor 8 too large value 🙈

          Still quite a lot of magma.

          • The last point still stands, that is a huge volume of magma that has been added in only a few years, especially considering Kilauea has not been less active than normal and Mauna Loa finally erupted. If it is forcefully intruded it is a massive increase in the supply of the hotspot that is trickling through to Kilauea and getting stuck here first.

        • Do you think this may be a step towards the Loa and Kea lines joining together?

          • Something like that probably takes place over longer than a few years 🙂

            But Pahala is between Mauna Loa and Kama’ehuakanaloa, and Kilauea is not visibly affected by Pahala, so if Pahala does feed Kilauea it is possibly from the Loa side, which would make Kilauea sort of a hybrid. There is some Pleistocene rocks from Kilauea which HVO have said are similar to Mauna Loa, so possibly this has happened before.

            But one only needs to imagine what would happen when Pahala makes a break for the surface. If it does go through Kilauea then the caldera probably overflows in a few months and every magma chamber in the volcano erupts.

            If it gles straight up, well then all bets are off. Even if that 600 km3 figure is off by a factor of 8 there is still enough magma to do a Laki sized eruption even accounting that most of it will stay put. These deep chambers seem to be slower to drain, they dont collapse into calderas, just erupt at a high rate for months on end, like at La Palma or Lanzarote.

            Usually, this sort of infrequent deep derived volcanism would be a late stage thing, but what is actually to stop it occurring during shield building too…

          • Looking at this concept of Pahala being a deep magma chamber though, I think that is actually what spawned Laki and Eldgja. Both of those are dfferent to the Veidivotn rifts, neither Katla or Grimsvotn are habitual massive erupters, VEI 3-4 is standard, and the couple of VEI 5s were more like 3-4s that lasted longer. Bardarbunga has a piston caldera that shows what happened at Holuhraun has happened many other times, not so at Katla or Grimsvotn. And though I expected the calderas of those two volcanoes to have formed in the two eruptions respectively the composition is not identical, so they didnt come from shallow. Probably there were intrusions deeper under the volcanoes, that when presented with a chance to access a rift did so, the shallow magma chambers being spared. Bardarbunga is a large single chamber connected to its caldera, Veidivotn eruptions are probably much more intense than Laki or Eldgja were on average.

            Question now is if Kilaueas southwest rift structure goes down deep enough to act as a potential outlet, or if the magma will have to go through Kilauea proper. I doubt it can go straight up but then this scenario is rather unprecedented…

      • Thank you for noting the problem with the link! Much appreciated.
        Maybe a Dragon can modify?

  18. Hi Folks,

    Astrophotography Request.

    Recently took some Astrophotography pics of the Southern Sky using a an SJCAM 10pro (Think GoPro).

    The results were fantastic, going well beyond naked eye level capturing many features .

    Im not familiar with the southern sky and have no ability to map beyond the brighter stars using star maps.

    Is there anyone in this chat who would be willing to have a look and see to what magnitude was captured?


  19. We might have a little deep tremor on the Big Island. Not a clear cut example, since some of the traditional instruments that I look at are not showing a signal, MITD is one. Potential quakes in the time frame below are in the general area.

    2023-01-23 06:40:16
    2023-01-23 06:35:27

  20. Interesting Albert !
    I have been doing astrophotography for a year, and since You write about supernova’s there was one last april, I managed to get a short serie of photos of it april 24th. This one was very faint, so only visible through a telescope.

  21. 🙂 Hopes the pahala structure formed quickly
    That means that I Will get Ionian stuff soon 🙂
    Hopes Kilauea turns into a rouge LIP volcano with unpredicatable nasty behaviour, suddenly puking out a +50 km3 rift flow in just a few months

    Just my goal for Kilauea : D

  22. I feel people are getting a little too excited about Pahala. Yes, it is a big discovery. However, there is quite simply no way 50 km^3 or so (largest sill based on their numbers might be close to 10 km^3, but it is also by far the biggest in the 3D model) has been intruded in a couple years. The effects of that would be a *lot* more noticeable even with the depth. It must be a formation centuries old at absolute minimum, realistically probably thousands (Yes, Hawaii has a world-dominating feed rate, however, Pahala probably isn’t getting the full fire hose all the time, like the past couple years being active aren’t *just* an artifact of better sensor data).

    Yes, there is a wave of earthquakes the past few years, but that just means it has gotten a increase of supply causing its stills to be forced a bit bigger. That may well have some impact for us on the surface eventually depending on to what extent it is connected to surface Kilauea, but not sure why it meaningfully strengthens expectations of a Laki+ beyond that inherently associated with the equivalent of all of Iceland in one volcano. Also, it is one article trying to do some cutting-edge things, so its going to be a while before we can know how strong the inevitable articles saying it is totally wrong and overfitting are.

    • I just like to point out again that the 50km3 figure was a mistake from my side. The actual number based on the volume of an ellipsoid is more like 6.5km3. I used the diameter instead of radius for all three axes, which makes the number 8 times too large. It’s a bit embarrassing, since even a box with those dimensions would be less than 50km3. I should have caught it before I posted the comment.

      I fully agree with you that this is not something that has just appeared in the last few years, but we are just starting to get the instruments and numerical tools that enable us to start seeing what’s really down there.

      • Pahala dyke svarm is just a small relativly shallow part of Kilaueas magma system, the upper shallow system extends from pahala all way out to Puna ridge thats 200 km lateraly. Kilauea stores perhaps many 100 s of km3 in its deep rift system in form of sheets and long melt accumulate areas whos liquid canals are sourrounded by crystal mush.

        Deep under Big Island 130 to 140 km down there is probaly an Impressive melt lens thats formed through decompression melting of the Hotspot and thats probaly many many 1000 s of km3 of melt at over 1600 C

        • The deep rift isnt liquid though, it is olivine crystal mush, so it cant erupt. It is sort of like a mafic batholith.

          Pahala though is a proper eruptable magma body, although where it is it will take some time to erupt yet, probably several more years at this rate. It is interesting what is actually feeding it though because Kilauea is just as active now as before 2018, so it isnt a diversion of Kilaueas magma.
          Maybe it is its own thing, but rather than erupting directly it is trying to reach Kilauea, it does go that direction more than anywhere else.

          • How do you know that the entire Pahala structure is eruptible magma? From what I heard so far we know nothing certain, except that there is an awful lot of hype around this structure (that, as explained further above, is not only here since 2016..).

          • Sills form from melt, old sills in ancient settings that are eroded to the surface form layered intrusions of crystals with more crystal free rock above. Eruptible in this case doesnt mean an eruption is certain to break out of it just means it could, that this is not a plutonic structure like a batholith or mafic equivalent olivine cumulate. It is not certain at all that there is a lot of abundant magma there, but finding that there is not much would suggest that our interpretation of these being sills is probably wrong.

            Pahala is probably overhyped but at the same time it is exiting because it is an unknown and uncharted phenomenon, and especially being in such an easily accessible and well monitored place as Hawaii. Perhaps many such volcanoes have these structures, but it also stands to reason that the biggest volcanoes with the highest supply would have the biggest capacity of magma storage through the whole system, something that the modest size of the calderas in Hawaii hides rather well. Given how large of an area the Pahala area covers even a small melt percentage is a lot of magma, and melt fraction in sills and dikes is typically very high in basaltic magma so if these are sills then it is pretty plain to see.
            And if one is to consider Pahala overhyped then the same needs to be said of most of the Iceland predictions of the past decade. It was fun, and at that time a lot of possibilities existed, but now it has been 10 years. Greip kind of stopped, Holuhraun was not just an entree to a bigger eruption, Grimsvotn is really taking its time. The only prediction that came true (admittedly the least likely sounding one) was that of a Reykjanes eruption.
            In fairness we should have known Bardarbunga was finished in 2015, all of its rifting eruptions are at best followed by minor eruptions, Holuhraun scale eruptions are about once a century or longer, and have long subsequent dormancy. GVP actually had dates for all of the big eruptions which show pretty clearly they space apart well, but that was always brushed off as being incomplete.

      • I’m talking about the whole kit, all 12 sills (and perhaps some smaller bodies), not just the largest one like you were. At least the earthquakes in the 3D model don’t make it really look elliptical, so volume of the big sill is probably somewhere between your 6.5km^3 and the 12.6 km^3 it would be if it were a rectangle.

        • If you plot the map of the sills onto google earth it lets you make polygons that will give you the surface area, then a rough calculation assuming it is a prism is enough considering the low resolution.

    • Even at the very smallest figures these are enormous structures containing many many Holuhrauns for each individual sill structure

  23. Realistically I image the odds are pretty good about this being some fairly normal thing for deepfeed volcanos to have a magma accumulation area at a transition point in the mantle. Over time a complex mass forms, like if you look at the 3D it looks like there are a lot of smaller bodies besides the Big Dozen. For that matter if it really has connections up to Kilauea and most of the way to Mauna Loa, it could have been there a very, very long time indeed. I guess one potential cap on how old it is how long would it take for all those sills to melt themselves into one indistinct mass? I have no idea.

    • Likely there are other such structures at most volcanoes but what is going on under Pahala is pretty extreme, there has been something like 150-200 thousand quakes down there since 2018,

      • Hawaii turning into a full scale LIP? : D

        Kilauea becomming a rouge volcano capable of 100 km3 flows

        • That is a long way away. Like 50km^3 or so is a lot, but all of that won’t erupt and a bunch will be essentially dissipated in forcing the way up (nearly 40kms!). Like Laki is 15km^3, this is probably the volume ballpark you would need for a Laki. If you want 100km^3 and a giant eruption is what it plans, you probably would need to be patient for a few thousand more years.

      • Well real LIP scale flows is over 10 000 km3 for some lava flows

    • People forget. Just because you discover something new doesn’t (in general) mean its just happened. Its likely been there all along ‘undiscovered’.
      Its a bit like discovering microbes and saying everyone will die of microbes imminently (see also PM2.5 which was there before we noticed it).
      So the likely scenario is: everything normal …

      • It isnt exactly that though. That the pahala swarm has been there for longer than 3 years is known but what has happened after 2018 is not normal.

      • The voices of reason.. which I think are correct. Pahala can plausibly form its own volcano in the future. That would most likely be in the ocean though and not on land. It would even help buttress Mauna Loa’s flank

        • And immediately Pahala responds with a swarm off the coast ..

          • At least the authors seem to think it connects to Kilauea and most of the way to Mauna Loa (whether that is a dying or developing connection). So it may not be planning to do anything that grand. Given Hawaii Volcanos like at least 30km distancing and Loihi volcano and an eruption probably wanting to head downslope, not up, the any new volcano from it would probably be about where the ocean section of Mauna Loa’s southwest rift’s eastern slope hits the sea bottom.

            That however, seems a bit off the general track, would tend to expect the next one to be *east* of Loihi. I feel like the order of probabilities is 1) eventually new feed filters up to Kīlauea or *maybe* Loa and increases activity for a while 2) a Laki drainout sometime, could be soonish, could be a few thousand years after many similar swarms come and go 3) way down the list, actually create a new volcano, like that is a once-in-100,000 year’s event, and would be a bit surprising location-wise.

          • Wouldn’t be as surprised like with Giep since Iceland produces an order of magnitude more volcanos, so the odds of a new one being in the process at any given point are quite good.

    • I will re-upload some images I shared in earlier comments about Pahala that might be insightful. This is a map of earthquakes from a catalogue that extends to 2018, so predates the most intense part of the Pahala earthquake swarms that started in 2019. The earthquakes are 42-35 km deep:

      It is not just Pahala that has deep earthquake activity, but rather the entire south coast of the island is underlain by vigorous seismic activity. Because extreme activity has been taking place in Pahala since 2019 we may forget it is part of a larger seismogenic structure. At other moments in time different areas in this structure will have likely experienced activity stronger than Pahala. For example during the 1950s deep powerful earthquake swarms often happened NE of the summit of Kilauea, the last of which was in 1960 or 1961. In the early 1960s, swarms were instead happening directly under the summit, and have not again repeated at that location with such intensity since. Activity likely shifts between different regions over time, and because data only stretches back a couple of decades, then we do not have enough information to understand this process. As far as I know these deep surges in earthquake activity have not shown a very obvious response in the shallow systems of Kilauea and Mauna Loa before.

      Pahala is an special place. It is the junction between the two most important deep seismogenic structures under the island. It is part of the ~40 km deep band of earthquake activity extending under the south coast of the island, but as it gets shallower it shifts to a ~30 km band of earthquakes that extends from Pahala to the summit of Kilauea, the Mantle Fault Zone, Pahala being the point where these two meet.

      This other image shows magnitude of earthquakes versus time in Pahala. Green are the usual volcano-tectonic earthquakes, red are the 41 km deep tremor quakes that usually come within spasmodic tremor episodes:

      Pulses in tremor activity can be seen to precede increases in volcano-tectonic earthquake activity. There have been three pulses in tremor activity, the first in 2014-2015, before 2018. So yes, not all of it postdates 2018, the present Pahala magmato-tectonic crisis was probably already underway in 2014. Volcano-tectonic earthquakes from 2015 onwards started to flare definite structures which had not been distinctly active in earlier historical times. Usually swarms would affect small areas with a very high density of earthquakes, the first of this possibly being the October 2015 swarm. Activity skyrocketed after August 2019 and started happening across different sections of the Pahala Swarm in complex manner. There have also been occasional flurries of earthquakes offshore, south of Naalehu, since at least 2017, as well as the October 2021 M 6.2 earthquake and its aftershocks in this same region. Overall, seismic activity peaked in mid-2021 and has been diminishing since then, it is at present only at about a third of the intensity it had back then, and because tremor activity is down right now, then I expect overall seismic activity will keep decreasing for the time being.

      As for what might happen, I do not know, because I don’t know what is going down there, it could be many things. I’m not so certain the earthquakes are happening in sills though. Earthquakes happen where rock slips past each other in faults. So if you find a plane of earthquake activity, the likeliest option is for it to be a fault. Intrusions are not that good at making earthquakes within themselves, magma doesn’t make earthquakes, you need for it to interact with faults. The Pahala Swarm has also produced 15 earthquakes >M 4, which should require a substantial rock-to-rock slip of more than 1 km2, and probably a few times more in the largest ones that reach up to M 4.6.

      Sill-like intrusions in Piton de la Fournaise that have been closely observed in past years are largely aseismic, only make earthquakes where they interact with the caldera ring faults of the volcano:

      And I well know Kilauea dikes only make a band of earthquakes at a very particular ~3 km depth, along the base of the intrusion, or just below it. And only if they are large enough. Kilauea intrusions are also good at triggering seismicity in the south flank decollement fault of the volcano, in the 2018 caldera ring faults, in strike-slip faults of the Southwest Rift, and in the weakened and faulted rift conduits, but no so much within themselves.

      • I should say that there is one time in which a deep earthquake swarm had an obvious response from the shallow systems of Kilauea and Mauna Loa. An enormous swarm of long-period earthquakes 45 km deep under the summit of Mauna Loa that took place mostly during late 2004. This swarm was associated to very fast inflation of Mauna Loa in 2004-2005 that came after more than 1 decade of relative quiet. Starting in 2004 carbon dioxide emissions of Kilauea rose gradually, increasing by a factor of 2.5 above normal levels in 2005, and slowly decreasing back to normal levels over the few years that followed. Rapid inflation of Kilauea’s summit area and East Rift Zone lasted from 2004 to mid-2007, while at Pu’u’o’o there were increased effusion rates in 2005-2009. It is the one time an actual magmatic surge went up to the volcanoes. What is happening now is more unclear. Links:

        • If anything, it might be the opposite actually. The peak years of Pahala earthquake activity in 2020-2021 weren’t very lively for either Mauna Loa or Kilauea.

          Mauna Loa experienced multiple deflation events of its shallow magma chamber in 2020-2021 which undid some of the inflation that was taking place. Magma obviously went somewhere, but there was no intrusion in the shallow system, it may have gone into the deep system instead, maybe the Pahala area. After the biggest deflation event, and the last one I think, in March 2021, the supply seemed to have been reduced compared to that of 2019. Looks to me like there was no inflation at all during much of 2021. Supply only seems to have picked up more substantially in mid-2022, eventually leading to the eruption.

          Kilauea had been inflating rapidly since the 2018 eruption, both its summit and East Rift Zone. But the inflation slowed down in 2020, there was one pulse of inflation into the East Rift Zone in mid-2020 and then the inflation in December that led to the eruption, but other than that there where times in 2020 where not much magma seemed to be flowing into Kilauea. Usually there is this continuous flow of magma up into the volcano that seemed to disappear during much of that year. The eruption that started in December 2020 and lasted through early 2021 was very weak mostly, and had eruption rates well below typical eruption rates of Kilauea eruptions. Even as the eruption ended, there was no flow of magma into the ERZ. A more vigorous eruption did start in September 2021 with eruption rates more typical of Kilauea, but it is true that at the same time the East Rift Zone was experiencing substantial deflation and was likely supplying the eruption, so it is unclear to me whether the supply was normal or below normal.

      • The deep quakes north of Kilauea in the 1950s are generally associated with the hot primitive magma at Kilauea Iki and being the surge that triggered the eruption at Kapoho and ultimately the modern era of eruption every year at Kilauea. I guess in the 1960s it was somewhat hard to see a lot of the magma supply given only eruptions at the very start and end of that decade erupted large volumes that exceeded the intrusion volume.
        So there could well have been quite considerable volume underground in the ERZ that was intruded by the many eruptions between 1961 and 1969. For the following I calculated the dike as a rectangle 3 km tall and 1.5 meters wide, the length is my own estimations of the dikes based on where the vents are, and in some cases from other data. So not official but works for this.

        1961 – 94 million m3, 2 million (96 million m3)
        1963 – 60 million m3, 7 million (67 million m3)
        1965 – 55 million m3, 17 million (72 million m3)
        1968 – 36 million m3, 10 milion (43 million m3)
        Total of 0.28 km3

        This is only for eruptions on the Napau/Makaopuhi fissure swarm, I have not done any calculation for the generally smaller but more numerous eruptions on the upper ERZ. Two of those though, in 1968 and 1965, had large intrusions. There was also Mauna Ulu, I kept that separate, but it technically should be added given when it began, it erupted 70 million m3 of lava in 1969.

        If one does include the Iki-Kapoho eruption in this line too, then that on its own is about equivalent to the same volume as the Napau intrusions in total over the next decade, so already at 0.6 km3 total.
        There was also the total of about 100 million m3 of lava erupted in Halemaumau in that time in 1961 and 1967-68, and then up until Mauna Ulu started Kilauea was also inflating from 1961 to 1969 so not all of the magma was erupting to begin with. I dont know the calculations to find that though.
        So at least I think it can be said confidently the supply was not low back then, just not as visible, it was probably the same as it is now. Those deep swarms seem to have been the start of something. Whether Pahala is too and how long that will take we will need to wait and see but there is an argument for both interpretations.

        • Yes, it is quite clear that there was an increase in supply following the 1960 eruption, hard to know if it was as high as today or not, probably at least half the current supply. Because the volume in the 1960s was mainly intrusive it is indeed difficult to guess. Plus there is the deep East Rift Zone spreading of Kilauea, at that time Kilauea’s rift system was spreading much faster than during the Pu’u’o’o eruption, so perhaps more volume was being allocated there too, hard to say.

          The deep swarms north of Kilauea are interesting. They started in 1953, as shown in “Two hundred years of magma transport and storage at Kīlauea Volcano”. They seem to have died out during 1960. So the timing of those swarms may not be coincidental. Certainly the end seems unlikely to be coincidental given that it matches with the increase in magma supply at Kilauea.

          USGS catalogue of small earthquakes goes back to 1959. Here are the deep earthquakes in 1959 that occurred north of Kilauea. They have mostly > M 2 and depths of 40-50 km:

          That area doesn’t seem to have seem much activity at all during the remainder of historical times.

          In 1961, activity in the Mantle Fault Zone (south of Kilauea), skyrocketed. High Mantle Fault Zone activity continued during the 1960s. The IRIS earthquake browser allows to search small earthquakes in Hawaii back to 1970 and animate them showing their frequency through time. When doing this, it shows that activity been slowly decreasing in the Mantle Fault Zone since 1970, and I think this decrease started in the early 1960s not long after the initial flare of activity in 1961. This map shows deep earthquakes at Kilauea during 1961-64, they concentrate within the Mantle Fault Zone.

          So it is probably more complex than magma rising up from depth and making earthquakes I’d say. I think there is probably rifting at those depths, and magma is intruded into dikes making earthquakes on adjacent decollement faults, with the magma probably staying there to fill the rifts rather than coming up to surface.

          • Those 1950s quakes do go a lot closer to Mauna Loa than I was aware of, I know Mauna Loa seems to have a pretty narrow source but the proximity us still there, and so soon after a period of high supply there, which seems to have peaked with a big intrusion just before the 1950 eruption then dropped to basically nothing. So perhaps these quakes are that missing bit of data.

            There was also that 4.0 quake a few weeks ago, shallower but in this same area, it has not been followed up but still this is notable, at least it seems too deep to be directly caused by Mauna Loa and its recent activity while also being too shallow to be a crustal flexure quake.

            I guess perhaps one interpretation of this model, if there is rifting or an equivalent intrusive activity down there, perhaps that is done when the pressure in the deep system is high, and when that pressure is relieved by an open path to the surface the reverse is true. This may explain some of the very rapid response to caldera collapses at the Hawaiian volcanoes, which is a rare thing elsewhere. Pahala is interesting though because it doesnt sit directly adjacent to any of the central volcanoes, it does sit next to the mantle fault quakes you plotted from the 60s though, which also interestingly is the same place that today is a mostly aseismic gap between Kilauea and Pahala.

            It would be great to know if there is spreading that deep down, if there us does it connect to the surface SWRZ. That may provide a path to a future direct eruption. Its a long shot but still a possibility, something like Mayotte (or Laki) that came from a massive dike out of a deep crustal chamber to directly erupt. If these happen a couple times in 10,000 years on Kilauea we would never know with how fast it resurfaces.

          • Regarding the intrusion numbers, using the same estimated width and height dimensions I got some numbers for the rest of the 1960s intrusions, including those that happened in 1960 and 1969.
            For 1960 though I gave the width to be 3 meters, the eruption was an order of magnitude larger in intensity and I just dont see a 1 meter wide crack being sufficient to create what was observed in that eruption, the 500 meter fountains and the 100+ m3/s eruption rate.
            I also put the very lengthy 1968 Hi’iaka dike as 1.5 km depth because it is specifically described as shallow by GVP and seems to have flowed down an existing crack. Seems it was also the same crack the 2014 lava flowed into, possibly the 1968 dike went very far down even to near Leilani Estates but passively, without the quakes it isnt possible to know I guess, but I stopped at 27 km or about 5 km east of the easternmost observed vent.

            I also didnt do calculations for any of the summit eruption intrusions, 1959 was not large it erupted fast and hot, so little underground travel leterally if any, and the other eruptions were entirely in Halemaumau so probably not even derived from a dike per se. There is really only a very small area available, the volumes would always be negligible unless they escaped the caldera directly and didnt erupt after.

            11 million left in Kilauea Iki. Intrusion – 103 million, Eruption – 275 million,
            Total 0.39 km3
            Biggest confirmed single Kilauea eruption before 2018 since 1790, outside of sustained multi-decade events. Also (probably) the 2nd biggest eruption in Hawaii in the 20th century, after Mauna Loa in 1950.

            Napau fissure swarm 1961-1968: as above, 0.28 km3

            1962 Intrusion 24 million m3, Eruption ~1 million m3
            1963 Intrusion 4.5 million, Eruption 2 million m3
            1965 Intrusion 77.5 million m3, Eruption 5 million m3
            1968 Intrusion 61 million m3 Eruption basically 0
            1969 Intrusion 78 million m3 Eruption 70 million m3+
            Total: 0.32 km3

            1961 18 million m3
            1967-1968 75 million
            Total: 93 million

            Total for 1960-1970: 1.08 km3
            Erupted total: 0.49 km3, 45.5% eruption rate.

            If 1960 is not included though, the volume is 0.69 km3 with 0.22 km3 erupted, or 31%. This might be a bit closer to a long term average, given eruptions on the LERZ are uncommon and generally very big compared to eruptions elsewhere on Kilauea. At least, in Hawaii the percentage of magma that erupts it seems really is a lot higher than at a plate boundary, although it is very variable on shorter timescales from probably under 10% to nearly 100%

            Obviously not a completely comprehensive analysis but seems that Kilauea became what it is today following the Kilauea Iki and Kapoho eruptions. So there could be a positive correlation between the end of the deep swarms near Kilauea and an increased magma supply, like the swarms were the buildup and found their exit.

            While 1 km3 in a decade is bog standard at Kilauea, even a bit conservative, I was actually surprised still at how much magma could hide underground in these intrusions, I was expecting to need to find some other way to account for the volume like deep storage or that the intrusions were larger but the number I got is close enough to consider margin of error as is.

          • It’s great seeing some volume estimates for the 1960s intrusions. The 1960s was a very dynamic decade for Kilauea, with such a rapid sequence of major dike intrusions involving the Alae-Aloi and Makaopuhi-Napau swarms, and eventually culminating in the birth of Mauna Ulu.

          • I did the same thing for the 70s, found the volume to be a bit lower at 0.73 km3. That being said from this volume 385 million m3 erupted, mostly at Mauna Ulu but all of those small eruptions actually add up to 1/3 of the erupted volume of the decade, they arent negligible although clearly Hawaii grows the most during shield eruptions.

            0.385 km3 is 52% of 0.73 km3, so more lava was erupted out of the total than in the 60s although it was not nearly as much as I was expecting. The total may be higher, because the 730 million includes a massive 150 million m3 intruded into the SWRZ in the early days of 1975, which is likely an overestimate especially if the intrusion is a sill like the 2021 intrusion, and not a dike.

            A massive part of the discrepancy in volume is probably what happened in the 1975 quake. The biggest pre-2018 deflation at the summit of Kilauea in the historical record was from the 1975 quake, and it is pretty telling that from 2018 and 1868 these quakes are associated with major intrusions (and eruptions). I once found the volume of magma that could fit in a dike of the dimensions of the 1975 ‘void’ was 0.4 km3, a dike is probably not the best model but the volume does fit with an expected supply rate of 1 km3 and an observed rate of 0.73 km3.

            I might try to do this for Pu’u O’o, but that is a massive task, and I think unlike in these decades an average year between 1983 and 2018 probably really was basically a 100% eruption ratio, an open path, given there was no net inflation really until near the end.

          • The period 1975-1984 was somewhat particular. The 1975 M 7.4 earthquake drained away a substantial volume of magma from the summit, larger even than the 1961 intrusion, although not as much as 1960, judging from the amount of deflation. During 1975-1984 East Rift Zone spreading was much faster and with each year the south flank decollement was producing more earthquakes than in any year since 1984. Rates of decollement seismicity peaked in 1979, with 1600 located earthquakes, which is 8 times more than the amount of earthquakes being located in the same area during the 2010s. Coincident with the peak in earthquake activity in 1979 which also extended to 1980, there were numerous sill intrusions in the Middle East Rift Zone, probably more than 15 sill intrusions altogether. Judging from the amount of deflation, the cumulative volume of the sills would probably be comparable to the largest 1960s dike intrusions. So volume was going mostly into deep East Rift Zone spreading and intrusive activity in the MERZ during the later half of the 1970s.

          • I did consider only doing 1970-1975 to avoid the quake but went with the full decade to get a fair comparison. I think off by memory though only 0.1 of the 0.7 km3 of my total was after 1975, so about what 5 years would be expected.

            Based on the fact Kilauea still erupted in the late 70s up to 1983, probably nothing really changed, just most of it was going into intrusions, and the one in 1975 itself was probably a large deeper intrusion, maybe similar to what happened in 2018 except without a high magma stand it never got the momentum to drive an eruption lower down.

            There is the point though, the 1975 quake rifted the ERZ the most between Makaopuhi and Kalalua, and this is where most of the intrusions were after, as well as half of the volume of the 1960s intrusions. This is exactly where Pu’u O’o erupted. So seems there was very significant magma storage there.

          • I have read that article, It is where I got some of my numbers although admittedly that is from memory than actually re-reading it. It is a bit dated now at 2014, there has been a lot since then 🙂 but definitely one of my favorites, and a massive help for the older years that have patchy records. It is unfortunate the Hawaiian record is largely lost, back in the 19th century that was considered nonsense but nowdays we recognise it for what it really is, Hawaii could have had a record as long as that of Iceland but now we only have pieces 🙁

            All of my numbers above are pretty loose, but I got them all in the same way unless stated otherwise so they care comparable to each other and at least are within the ballpark of what is known for the last few decdes so good enough for this purpose I think.

          • Just a quick number, but 2022 on Mauna Loa the dike is about 0.1 km3, I assumed a depth of 2.5 km and width of 2 meters, The volume is still being loked at by HVO so not final but last I saw was that it is a bit over 0.14 km3, so 0.25 km3 combined roughly. 1984 was larger, 130 million m3 intrusion using above dimensions, and 0.22 km3 erupted, so 0.35 km3. So it is not quite as big as the 1960 Kilauea eruption but was erupted rather faster.
            1975 is approximately 0.09 km3 intrusion and 0.03 eruption, for 0.12 km3 total, about half of 2022 and 1/3 of 1984.

            I also did 1950. SWRZ intrusions I think are probably deeper than NERZ eruptions ,the eruptions go down to a lower altitude, so rather than 2.5 km they might be more like 3-4 km. For 1950 I went with 3. The total of my numbers is 0.37 km3 intrusions, and the volume has been given as 0.38 km3, so 0.75 km3. All the intrusions were emplaced within the first 4 hours, and only the lowest vent was still erupting strongly after the first day. There is nothing else I can think of to compare this to.

          • And 1950 with 4 km depth dikes is 491 million + 376 million. So 0.867 km3.

            But I dove into the details for that eruption more. I was going to post it here but I think it deserves a whole article on its own, I cant think of an appropriate English word to describe how insane this event was.

            But I will say that all 22 km of the fissure opened in 80 minutes… 🙂

  24. The fountaining shown in the USGS’s Kilauea Live Stream has shifted around a bit. Split a bit since last night.

  25. Can really see the colour of the lava in this one. It is probably a little overexposed but very clear 🙂

    Albert maybe you can do your viscosity calculation on the lava again now, the USGS livestream has very clear view of the waves near the vent. The fountain is about 5-7 meters tall apparently.

    • Wow its super yellow in daylight! Althrough not direct sun since its cloudy

      Is this 1240 C ?

      • No it is a bit overexposed, given that the lava erupting here is potentially 2 years old I think it is probably if anything a little cooler than the lava of 2018 at the summit, which was about 1200 C. This stuff is still hotter than the lava at Pu’u O’o is my guess though.

    • Very very fluid indeed How it ponds itself like water

      I say around 10 PA.s perhaps 20 PA.s
      Thats as runny as really warm bee honey

      But lava is very dense and looks more fluid than it is.. but No more than 100 PA.s

    • Interesting looking seismic sequence:

      Some of the events are emergent, building up gradually, unlike typical fracturing (tectonic or volcano-tectonic) earthquakes that have well defined arrivals. And the clustering is somewhat odd. Some trains of events look faintly rhythmic. They look to me like spasmodic tremors, similar to the tremors that happened in Pahala in early 2019, and in late 2020 to early 2021, which preceded the massive swarms of rock fracturing earthquakes. Probably related to magma movement.

      • With the largest earthquake being a M 2.6 it would among the stronger Pahala tremor events in terms of magnitude of the individual tremor quakes.

    • The spectacular footage of one of the 2022 paroxysms of Etna from that same channel that you told me about. It is a dramatic close-up view of one of the most violent paroxysms it produced that year, it got somewhat dangerous at one particular time too:

Comments are closed.