Effects of Subsea Volcanic Heat

Superheated molten lava at West Mata (Pacific). (Image: NSF and NOAA.)

A while back I tried to calculate how submarine volcanic activity around the West Antarctic Peninsula might affect water temperatures, and in consequence land temperatures in the region. What I lacked was a credible estimate of the size of a submarine eruption. I mean I could imagine a small eruption, but when you see pictures of discrete pillow lava formation (e.g. left) there is no imagining the scale as in a subaerial eruption.

Images of the sea floor around the Axial Seamount were reported recently with news of the discovery of a large eruption.

“An undersea volcano has erupted off the coast of Oregon, spewing forth a layer of lava more than 12 feet (4 meters) thick in some places…”

“…producing hardened lakes of blobby lava, in places more than a mile (1.6 km) across”

Data from Wolframalpha gives an ocean depth of ~1-1.5 miles for this location (Juan de Fuca Ridge) so I’ve enough to work with at last.

I’ve gone for a lava layer (assuming uniformity) of 2m depth and 1 sq. km, although undoubtedly that is large.
Total volume = 2 x 10^6 m3
Density = 2.7 x 10^3 kg/m3 [1]
therefore mass (m) = 5.2 x 10^9 kg

Inital temperature = 1200°C
Final temperature =         4°C (for deep ocean)
Specific heat (Csp) of basalt = 1.50 x 10^3 J/kg/°C [2]
But since the lava will solidify, there is a phase change and we must also include the heat released – latent heat (Cla) of crystallisation =  5.00 x 10^5  J/kg [3]

Energy released Q = m x (Csp + Cla)  x ΔT

For cooling, solidification and temperature reduction to 4°C of 5.2 x 10^9 kg lava the calculated total thermal energy released on cooling = 1.24 x 10^16 J

To heat the surrounding water – if we imagine a column of water 2000m above the eruption and an affected area of 1km2 above the lava field..
Volume of water =  2.00 x 10^9 m3
Mass water (m) = 2.00 x 10^12 kg
Specific heat of water (Csp) = 4,186 J/kg/°C

Water temperature increase ΔT = Q / m x Csp = 1.42°C

Perhaps a 1m average depth of lava would be more realistic, but this still gives an average temperature increase of 0.71°C – over the area of a square kilometer.  And in real life you could imagine that, with convection, a warm plume of water would develop increasing local temperatures even more and spreading out at the surface. Of course finding such transient ‘hot spots’ on the ocean surface, if they even occur, would be like finding a needle in a haystack.

Image from http://weather.unisys.com/surface/sst_anom.html embedded in post at WUWT

My original curiosity was sparked by  Undersea Volcanic Eruption In Tonga, although the premise that this was causing the SST anomaly shown above was dismissed by many (poor) calculations and by Bob Tisdale here. But on the other hand jorgekafkazar says in comments:

“…released volcanic heat would rise by convection very rapidly, especially as the pressure fells below the critical point. The entire column of water (n km by n km) wouldn’t be heated, just the portion almost straight above the vent all the way to the surface, maybe as small as 100 meters in diameter.”

…then at the surface the warm pool would spread.

The volcanic area west of Tonga

Ocean depth in this area (right) is up to 1.5 km, not the 6000+km of the Tonga trench, so the calculations above apply, except that we don’t know the magnitude of the eruption or the exact depth of the vent (more about that in a moment).  So, while I defer to Bob Tisdale’s conclusions that the SST anomaly (area ~500 x 500km) is complex “weather noise”, volcanic heat can make a significant localised contribution.

In case you were wondering, like me, why most submarine eruptions seem to be depicted as pillow lava, yet explosive eruptions can and do break the surface, ocean depth is one of the factors.  When Sohn et al [4] found evidence of pyroclastic deposits associated with volcanic activity in the polar Gakkel Ridge they determined that a high CO2 content in the magma would be sufficient to overcome pressure containment at depth (Table 1).

Table 1. Variation in pyroclastic jet characteristics with magma chamber depth[4]

Magma chamber roof depth (m)
Minimum CO2 volume fraction in volatile-rich layer
Pyroclastic jet mixture density at vent (kg/m3)
Average jet exit velocity at vent (m/s)
Plume rise height in water column (m)
1,000 0.6443 568 236 544
2,000 0.5647 462 343 956
3,000 0.5026 398 424 1,276
4,000 0.4528 355 490 1,532
5,000 0.412 324 544 1,741
6,000 0.3779 302 591 1,916

We calculate the minimum CO2 volume fraction in a volatile-rich layer accumulating under the roof of a crustal magma chamber required for producing pyroclastic activity on the sea floor at a depth of 4,000 m. Magmas with the gas volume fractions shown in column 2 will fragment just before reaching the sea floor, producing very small deposits. However, if the CO2 volume fraction in a volatile-rich layer is ~0.75, fragmentation occurs at the magma chamber depth, and a much more energetic eruption occurs as the gas accelerates during ascent to the sea floor, producing the approximate conditions shown in columns 3–5.

Large-volume pyroclastic deposits have been found in shallow water (500–1,750m water depth) on the Azores Plateau, but the only previous evidence of pyroclastic material at water depths greater than 3,000m (the critical depth for steam) is limited to small fragments recovered in sediment cores. Hydrostatic pressure inhibits volume expansion, and below the steam threshold any explosive activity must result from magmatic volatiles rather than secondary surface effects. CO2 is the most plausible evolved volatile component for MOR basalts, and at 4,000m water depth a CO2 weight fraction of ~14% (ref. 3) [in the paper] is necessary to achieve the volume fraction of ~75% needed to fragment an erupting magma. This value exceeds the maximum dissolved CO2 concentrations measured in a MOR basalt (~1.4 wt% in a ‘popping rock’) by an order of magnitude.

But I digress…

Since I’ve mentioned the Gakkel Ridge, Tom Moriarty of Climate Sanity, in Volcanos in Gakkel Ridge NOT responsible melting the Arctic ice, used the Mt St Helen’s eruption as a starting point.  According to the U.S. Geological Survey Mount St. Helens released 24 megatons of thermal energy, 7 of which was a direct result of the blast. Tom concluded that an equivalent release of energy by the Gakkel Ridge, would melt just 100 sq km of 3m thick ice, or 300 sq km at 1m thick. That is nothing compared to the seasonal ice melt.  My own calculations agree. The Gakkel Ridge is also in very deep water

So, so far it’s not looking likely that volcanic heat has much, if any, effect beyond localised heating of water in the vicinity.

The track of the echo sounder shows the submarine volcano rising from 1000m water depth to a peak with water depth of 250m only. (photo: Gemma Kirkwood - http://www.ldeo.columbia.edu/news/reports/2004/CORC04/05_10_04.htm)

So going back to the West Antarctic Peninsula, for example Scientists Discover Undersea Volcano Off Antarctica, could this have an effect on local sea temperatures? Sea temperature can and does affect coastal land-based temperatures and we know the WAP is warming.

“Highly sensitive temperature probes moving continuously across the bottom of the volcano revealed signs of geothermal heating of seawater. The heating was noticed especially near the edges of the feature where the freshest rock was observed.

These observations, along with historical reports from mariners of discolored water in the vicinity of the submerged peak, indicate that the volcano has been active recently.”

In such shallow water the effects are interesting (from my calculations and estimation of what would happen in water with convection etc.). A small eruption at 250m of say 10 x 10m and 0.5m thick deposit (50m3), would be sufficient to raise the local water temperature in a similar 100 sq m area by more than 2.8°C (uniformly within the 250m water column, which of course would not happen that way).

I wondered how much lava at this depth there would have to be (making the unlikely assumption it was not explosive) to raise the temperature of an area say 10km2 by >2.0°C. The answer?  4 million cubic metres (0.004 km3). To put that in perspective Kilauea typically produces ~ 0.1 km3/year (100,000,000 m3) of basaltic lava. Yes, local effects are possible if ocean depth is relatively shallow.

Global (magma and lava) production rates are estimated at 3km3 per year for mid-ocean ridge systems and 1km3 per year for continental volcanic systems. For global impact, (mid-ocean ridge (MOR) ~80,000km long; average water depth of 2,500m) there’s enough energy in the volume of lava produced by MOR annually to raise the temperature of 8,000 km3 of seawater by just over 0.5°C – that’s a drop in the ocean (1.3 billion cubic kilometres).


[1] Stephen R. Sparks J. & Huppert H.E. (1984) Density changes during the fractional crystallization of basaltic magmas: fluid dynamic implications. Contrib. Mineral. Petrol. 85:300-309
[2] http://www.engineeringtoolbox.com/specific-heat-solids-d_154.html
[3] Kojitani & Akaogi (1995) Measurement of heat of fusion of model basalt in the system Diopside‐forsterite‐anorthite, Geophys. Res. Lett., 22(17), 2329–2332, doi:10.1029/95GL02064.
[4] Sohn, et. al., (2008) Explosive volcanism on the ultraslow-spreading Gakkel ridge, Arctic Ocean, Nature 453, 1236-1238, doi:10.1038/nature07075 http://www.nature.com/nature/journal/v453/n7199/full/nature07075.html (abstract)

This entry was posted in Volcanoes and tagged , , . Bookmark the permalink.

34 Responses to Effects of Subsea Volcanic Heat

  1. Willis Eschenbach says:

    Fascinating, Verity.

    Let me offer another mechanism. I was the navigator on a tramp steamer in that area near Tonga, and I want to point out another mechanism bringing the heat to the surface.

    This is the effect of the massive quantities of pumice that is released by these Tongan Trench volcanoes. I’ve sailed across huge (a couple km across) patches of pumice in the area. It is quite common to see rafts of pumice drifting onto the Fijian beaches, driven by the ceaseless south-east trade winds.

    In the same area around Tonga I’ve also seen (once) pumice, bubbles, and discolored water rising together to the surface. It put on a very good imitation of an uncharted reef, and gave me some very uneasy minutes double-checking my sextant sights and scrutinizing the charts (which said deep water everywhere) until I realized what it was.

    In any case, the combination of rising bubbles of unknown composition and rising chunks of pumice can entrain a whole bunch of warm water and bring it to the surface.

    Don’t know what that has to do with pillow lava, but that’s what I observed there. Localized heat at the surface.


    • Verity Jones says:

      How interesting! I imagine encountering an actual subsurface eruption as you describe must be a rare occurrence. Actually it reminds me of The Wine Dark Sea by Patrick O’Brian where the protagonists encounter a submarine eruption.

      Regarding the formation of pumice, that is quite easy to imagine. If you think of an eruption below 3,000m depth and therefore not generating steam (as described in the post)… If it exudes as pillow lava it is because it has too little dissolved CO2 and other gasses to overcome the pressure containment at the eruption depth; if there is a pyroclastic eruption then the dissoved gasses are sufficient to overcome the pressure containment. However, there could be situations where there is enough dissolved gas and just the right depth (pressure) that there is a small, but not explosive, expansion of the dissolved gas, that happens simultaneously with cooling and solidification. This is just enough to reduce the density of the rock and give it buoyancy in the water column and hey presto! we have much greater heat brought up by the rocks from the depths in a localised area.

      Of course deep submarine lavas must trap a lot of dissolved CO2 within the rock and ensure it stays sequestered.

  2. Bob Tisdale says:

    Verity: The reason I dismissed Goddard’s post was because he missed the location of the volcano by a couple thousand km.

    That’s from my comment here on that WUWT thread:

    Also, that’s a very common SST anomaly pattern along the South Pacific Convergence Zone.


    • Verity Jones says:

      Thanks Bob. Actually I was intending not to draw attention to his already very public faux pas. I just wanted to make sure that such myths do not perpetuate.

      It’s not impossible of course, but it would take a gargantuan eruption (supervolcano?) and in that case we’d probably be rather thankful that it was submarine.

      • Jim Kelley says:

        It is possible however, that a continuous release of energy for an extended period of time could in fact raise the temperature of a smaller body of water the size of the Arctic Ocean by an amount comparable to that which has been observed.

        [Reply – I started out thinking that too – but the volume of magma release needed to have ANY effect on the Arctic Ocean is unbelieveably huge. I fear I’ll have to do another post on it – Verity]

  3. Pascvaks says:

    Would think that both the Americans and Russians would have declassified Cold War submarine data that could shed light on this matter. (Like I said, you’d think so;-(

    Reply: Well I guess OK but would anyone ever dig it up? V.

  4. tonyb says:


    For what its worth, last year I found myself siting next to a volcanologist at a Cambridge University Dinner, who said the latest research showed there were 10,000 times more underwater volcanoes-and vents/fissures- than had hitherto been known about. If so these could have a significant warming effect.


    • Verity Jones says:

      HI Tony,
      What would you call a significant warming effect, assuming it is averaged thoughout the entire volume of the ocean? What increase in temperature would you call significant?
      Just from lava/magma?

  5. tonyb says:


    Depends on whether you are talking in scientific or layman terms. You may have seen my recent article at Judith Currys’ concerning SST’s, as a result of which I am communicating with one of the people who put together the SST data.

    IMHO the data is utterly flawed from 1850 until around 1960 as there are so few reliable data points and the methodology is so haphazard. Consequently we have no idea whatsover as to what degree of variability ocean temperatures have exhibited in the past.

    The Met office would claim one tenth of a degree to be significant, so in their context if lava/magna affected the ocean by this amount that woiuld be highly significant. Personally I think that as we have no idea whatsoever of a starting point a couple of degrees would be more realistic.


    • Verity Jones says:

      OK. Let’s assume that instead of 3km3 of lava produced per year by subsea volcanoes there is 10,000 times that – 30,000km3. That has to heat the entire ocean – all 1.3billion km3 of seawater. By the same calculation method as above, that volume of lava would raise the temperature of the entire volume of the oceans by 0.03 degC.

      In reality, not all the additonal volcanoes would be active all the time. Basically I’m keen to put this notion to bed – I’m satisfied there can be local effects, but that’s about it. And before anyone says “What about all the black smokers and hot springs?” let me just point out the difference in magnitude for the Spcific heat values and the heat released by crystallisation:
      Specific heat (Csp) of basalt = 1.50 x 10^3 J/kg/°C
      Latent heat (Cla) of crystallisation for basalt = 5.00 x 10^5 J/kg
      Specific heat of water (Csp) = 4,186 J/kg/°C (=4.186 x 10^3 J/kg/°C
      Compared to 1200°C lava, for water at say 400°C there would have to be 3×10^2 (300) times as much of it as the lava to have the same effect even without the difference in Cla and Csp.

  6. tonyb says:


    I was being purely hypothetical and not sggesting for a moment that lava/magma could possibly raise temperatures by a few degrees, merely pointing out that our knowledge of ocean temperatures is not good enough to record changes induced by whatever method.

    Where i think it could become intereesting is IF there were sufficient colcanic activty to melt ice periodically.i.e there was a localised rather than global effect.

    We do have historic accounts of oceans warming noticeably through a localised volcanic eruption but whether that has ever had any real impact on ice levels-potentially the most notable effect- is something I have never studied.

    • Verity Jones says:

      Hi Tony,
      Actually I wasn’t responding to you in that way – I mean I know you were – I was quite happy to use that unfeasible multiplier to show just how big the oceans are.

  7. tonyb says:


    Its midnight and I’m tired as you can tell from the terrible spelling mistakes-sorry!


  8. Maybe 3 or 4 months ago, an internet blog displayed a satellite photo, denoted ‘a photo the (some authority) did not want available publicly’ showing a complete ring of erupting volcanoes around the antarctic penninsular.
    I find it hard to believe that I didn’t store it somewhere, but cannot now find it.
    Perhaps some other reader has a better recollection and filing system than I.

    • Thanks Tony,
      That’s a start, but the satellite photographic image I recall showed volcano emiissions circling right around.
      Not only do we have ocean heating sources not being taken into account, but reductions in ice volume obviously occurr. (Without human influence!)

  9. ArndB says:

    Hi Tony,
    your SST paper is great and I agree with your conclusion: „ Historic Sea Surface Temperatures in particular are highly uncertain and should not be considered as any sort of reliable measure.”, particularly as far as these data are used in climatic research.
    My experience concerning SST measurements, covering the time period from about 1955 to 1964, is that they may have well served preparing weather forecasts for the next few days, and that even when all care and precautions had been taken, the result was at best plus/minus 0.5°C. “Correcting” this data for ‘climate-change research’ is dubious to say the least. That applies particularly for the time period when the measuring methods changed around the mid of the last century. That was the time of WWII, which is subject of two papers I wrote a long time ago:
    ____ (1997); ‘Reliability of sea-surface temperature data taken during wartime in the Pacific’, presented at Symposium on Resource Development, August 8-9, 1997, Hong Kong, in: PACON 97 Proceedings, pp. 240-250. (www.oceanclimate.de, Previous Essays).
    ____ (1998); “How useful are Atlantic sea-surface temperature measurements taken during World War II”, paper submitted at the Oceanology International 1998 Conference, “The Global Ocean”, 10-13 March 1998, Brighton/UK; published in Conference Proceedings Vol. 1, p 121-130. (www.oceanclimate.de, Previous Essays).
    Best regards Arnd

  10. Verity Jones says:

    @Ken McMurtrie
    I think I remember coming across something like that a while ago too and likewise regret not bookmarking it.
    thanks for the link
    That’s just it isn’t it? The accuracy of the measurements taken (sea / air) were (or indeed are) fit for the purpose intended at the time. However, if you apply +/- 0.5 degC accross the board to such measurements and consider the accuracy of any average temperature calculated by such methods, you have no certainty in the average for fractions of a degree.

  11. tonyb says:


    Many Thanks for your kind comments. I am in conversation with the Met Office concerning SSTs. They truly believe that through the use of modern computers/analysis they can take data that is completely hopeless and turn it into a scientifcically valid global data base stretching back to 1850 that is so accurate it can inform policy. Complete madness!

    As Verity says the SSTs were taken in a different age for a different purpose. Your papers are very interesting and with your permission I might like to reference them to the Met office.


  12. ArndB says:

    # Verity Jones says: September 10, 2011 at 11:57 am :
    # by such methods, you have no certainty in the average for fractions of a degree.#

    Maybe I have not fully understood the meaning of the sentence, but it seems curious to me if the precision of sea to air temperatures are “compared on an even level”. In the climatic system the value of one-tenth of a degree in SST is of much more of weight as in air temperature.
    __A three meter thick sea SS-layer has roughly the heat potential as the 10’000 meter air column above.
    All what I want to say with this is that SST with a “0,5° accuracy” is not necessarily accurate and useful for climatic research (beside from a number of other measurement deficiencies).

    Unfortunately, Verity, I have no comment on the interesting subject: Subsea Volcanic Heat, but read it with great interest.

    # tonyb says: September 11, 2011 at 7:18 am #
    Thanks for your interest and kindly make use of the WWII – SST papers as you regard appropriate.

    Regards AB

  13. mizimi says:

    Wikipedia has a page….List of volcanoes in antarctica…which tables around 36 volcanoes , half of which have erupted during the current holocene and about a quarter in the last 1000yrs; seven of them post 1876.
    The page also lists co-ordinates and elevation.

  14. Verity Jones says:

    Thank you – there’s also a good list here with information on the individual volcanos (32 are listed) :http://www.volcano.si.edu/world/region.cfm?rnum=1900

    I’m still keeping an eye out for Ken McMurtrie’s volcano image (comment September 11, 2011 at 1:06 pm). There’s som information on Antarctica on Erik Klemetti’s blog – http://bigthink.com/ideas/23297 (now at Wired: http://www.wired.com/wiredscience/eruptions) where he links (indirectly) to this article on the discovery of a new, ancient subicecap volcano in West Antarctica: http://www.nytimes.com/2008/01/21/world/21volcano.html?_r=1&refer=science There is further discussion here: http://www.tgdaily.com/trendwatch-brief/41171-updated-possible-natural-explanation-found-for-west-antarcticas-warming

  15. Jim Kelley says:

    I think it’s clear that geothermal heat cannot, by itself, account for the change in the average temperature of the Arctic ocean. The problem is, neither can any other isolated mechanism. But any increase in the influx of heat will raise the water temperature and affect the ice melt by some amount, as you have noted. This would represent a change in the energy budget. Any affect on the ice melt (as has been observed) affects the albedo of the ocean surface. Changes in the albedo of the ocean surface create a change in the amount of absorbed Solar energy at the ocean surface. This changes the energy balance even further, and in the same direction. Then, an accompanying change in Solar flux (an increase had been observed until this latest solar cycle) would tend to magnify the effect of the other changes.

    It’s reasonable I think that the observed change in average ocean temperature (which by the way is warmer and exceeds the change in average atmospheric surface temperature over the past 200 years) would be caused by multiple reinforcing factors, one of which appears to be increased geothermal activity – particularly in the Arctic.

    • Verity Jones says:

      the observed change in average ocean temperature (which by the way is warmer and exceeds the change in average atmospheric surface temperature over the past 200 years)
      Interesting – can you point me to a reference for that?

  16. Jim Kelley says:

    Statistically, one might be able to identify a correlated leading trend between average ocean temperature, and average surface air temperature. And it stands to reason, given the direction of net energy flow from the warmer ocean, to the cooler atmosphere.

    [Reply – let’s not go there! I can see all sorts of possible arguments over defining avaerage ocean temperature (how deep do you go; comparing now and Argo data with older data etc.), Verity]

    • Jim Kelley says:

      Water doesn’t get much below zero C, whereas the air covering this planet does (and most of it is). Roughly 1/2 the mass of the atmosphere is at or below zero C. This tends to skew the average in favor of water. It shouldn’t be all that controversial.

  17. Jim Kelley says:

    In comparing the generally accepted average surface air temperature of approximately 14 to 15 degrees C with the generally accepted average surface sea temperature of approximately 17 degrees C, the average temperature at the surface of the ocean is higher than the average surface air temperature. The second law of Thermodynamics tells us what we can generally expect as a result of such a differential. For purposes of clarity we can focus simply on the referenced data, which is largely absent the ambiguity that may have been inferred from the lack of semantic precision in my previous posts.

    I also found this data to be generally supportive of the previously referenced anomaly data:

    The data are indicative of a general increase in Arctic ocean (surface) temperature, albeit over a relatively short time scale. Again, this represents a greater increase than any apparent cotemporal increase in average atmospheric surface temperature. Regardless, I believe it’s constructive to consider and discuss natural energy sources as being among all possible causes of the increase, including geothermal and solar energy.

    The trend I referred to previously is readily apparent in a plot of both the NASA GISS Anomaly Data, and the Global Average Sea Surface Data on the same graph. (Anomaly temperature is converted to absolute temperature by adding back the mean temperature referenced to the data.) I’m simply reporting my observations.

    • Verity Jones says:

      While perhaps I should defer to you as a geologist, I still tend to discount geothermal sources as having any tangible effect in the Arctic for several reasons:
      – the volcanicity in the Gakkel Ridge region is at great depth (4km) (perhaps there are others, shallower, that I’m not aware of); large explosive eruptions, it is thought, may reach up to 2km depth.
      – the temperature increase calculations above still hold but at such depth the volume of water to be heated “on the way up to the surface” from a depth of 2km is still so great that any increase in ocean temperature would be negligible.
      – even in shallow regions the thermocline and halocline prevent mixing of the cold/fresher surface waters with the warmer/saline deeper waters below ~200m

      Overall therefore at the moment I tend to discount geothermal and attribute warming of Arctic waters to heat transport by ocean currents and solar energy. From what I have read I understand the former to be the most important, since solar energy is moderated by cloud.

      • Jim Kelley says:

        I don’t entirely disagree. However the influx of energy into the ocean can do almost nothing other than increase the total heat energy, which would presumably increase the temperature at least locally. And, although the gradient is entirely in the wrong direction for geothermal energy to propagate toward the surface, reducing the differential temperature across the thermocline would tend to decrease the rate at which energy is transferred from the surface to greater depths of the ocean. A reduction in the rate of energy transfer away from the surface would likely result in an increase in surface temperature, which is what has been observed.

  18. Pingback: Pumice raft may indicate new underwater volcano | Digging in the Clay

Comments are closed.