Obligatory XKCD…
'For those of us who spent no small part of our childhoods avoiding certain bits of carpet because they were lava….'
For those of us who spent no small part of our childhoods avoiding certain bits of carpet because they were lava, boffins reckon they might have an idea why volcanoes erupt in the first place. Esteemed science journal Nature has published a paper saying that, to everyone's surprise, extreme rainfall may be a significant …
'For those of us who spent no small part of our childhoods avoiding certain bits of carpet because they were lava….'
Hmmm.. if the magma is several km deep, the change in pressure from surface water (rainfall) infiltration is likely to be a tiny fraction of what's already there. This is Hawaii.. an island surrounded by Ocean.. if the rock is porous enough for rainfall falling ON the island to connect that deep, presumably the ocean all around the island connects too.
Perhaps what is actually occurring is that the high rainfall flowing down is filling all the normally open vents that allow gas and steam to escape from depth; under normal rainfall, perhaps some remain open, allowing a pressure relief mechanism, which is blocked under high rainfall events?
Changes. It's about the changes. The ocean water has already infiltrated to the extent that it can (subject to salt actions), so that's (relatively) constant.
Heavy rainfall appears to increase the pressure, and this increase is being transmitted with some efficiency down to the magma layer. As a no-nothing, this is surprising, but what do I know?
True, but how big are those changes resulting from rainfall? They are talking about infiltration causing pressure changes 'several kilometres below the surface' near to where the magma is. The total water pressure at that depth would, if linked to the surface / the sea, already be huge: roughly 10kPa per metre head of water, so 'several kilometres', say 3,000m, maybe 30,000kPa?*
How much rainfall are they talking about? A quick internet search suggests that peak average rainfall is 124mm on average, in November. So extreme rainfall = what? maybe 100mm = 0.1m = +1kPa; even if it were 1m of 'excess' rainfall, that's still only +10kPa.
The change in pressure is tiny relative to the baseline value.
*By the way, is there an approved Reg unit for pressure?
Keep in mind that time is an important factor in rainfall. 124mm of rain over 24 hours is very different to 124mm of rain in 2 hours.
Similarly, 100mm of rain each day for three days straight may have a very different effect to 300mm of rain in a single 24 hour period.
I suspect that there is still quite a lot for them to study to work out how this all fits together.
I live in the tropic's not Hawaii, I have personally measured an average of 1000 mm of rain fall everyday for 3 days, and quite considerable rain fall before and after, this sort of rain fall being collected (burst several of our dams) in the corona say 2 km;s wide is a shitload of water = f#$k load of pressure.
Actually pressure differences are not going to be equivalent to the water column of rainfall. In volcanic systems any water that infiltrates (and doesn't runoff in a river or evaporate) is going in to fractures. Thees fractures are going to be quite small, so adding a metre of water at the top of the system may easily raise the water level in a fracture by 20 metres. Even in the UK some water tables fluctuate more than 50 metres vertically in a season (in extreme cases).
While the pressure change might still be small, it can still be the straw that breaks the camel's back - if you imagine pressure building up in the magma chamber linearly a small change in the system can be a trigger - without the rain an eruption would probably still happen - but maybe not for a few more months or years.
Quite correct - I grossly simplified the calculation to illustrate the core point: the variation in pressure caused by an extreme rainfall event is very small relative to the baseline pressure value at 'several kilometres' depth.
In practice you have to allow for the void ratio of the soil/rock, the permeability of the soil and how this varies with degree of saturation, the response of the soil skeleton to changes in water pressure, the variation in proportion of water that infiltrates relative to the proportion that runs off along the surface, etc. The Reg summary of the article indicates that the rock is permeably, but doesn't indicate whether that is due to a few, relatively large and open fissures in an otherwise low-permeability material, or whether the material itself has a high voids ratio, with good connectivity between the voids. There is also no discussion of what if any variation exists in horizontal permeability relative to vertical permeability.
In practice, high rainfall intensity events on 'low permeability' soils often generate a lot of run-off and relatively little infiltration, whereas longer-duration but lower intensity rainfall can result in more dramatic changes in the groundwater level. In 'high permeability' soils, more water tends to infiltrate and you tend to get less run-off, but the infiltration tends to produce a relatively small increase in ground water level, since it tends then to dissipate laterally: a fully saturated soil is generally more permeable than the same soil when it is only partially saturated.
And of course, this is a discussion of an extreme rainfall event - there is rainfall all the time, so the baseline condition is not necessarily static, but reflects normal seasonal variation; we're probably actually considering the pressure difference between the normal yearly high and an exceptional yearly high, which may be more or less than the variation between the normal yearly high and the normal yearly low.
But I was offered a beer to bugger off, so I will.
Thanks.
My PhD was in modelling infiltration into unsaturated soils (that 'Dr' part of my attribution isn't medical).
But I might be wrong - if my calculations of baseline water pressure and probably change in water pressure are incorrect, please let me know.
And I admit, I haven't read the nature article referred to.
It's science. Science should be challenged - that's how it's proved.
Also, I did not dispute that heavy rainfall might influence volcanic activity, merely noted that the suggested mechanism seemed unlikely.
(Relevant PhD qualified) Bloke on Internet.
My PhD was in modelling infiltration into unsaturated soils (that 'Dr' part of my attribution isn't medical). ...(Relevant PhD qualified) Bloke on Internet.
Which is probably worth mentioning when initially expressing an opinion, e.g.:
"As someone who has a PhD in modelling infiltration into unsaturated soils, I have the following issues ..."
Otherwise, when a random comment is made, why would anyone take it at more than face value? It is the internet, after all, therefore the safest assumption is to assume everyone is just some random person on the Internet ;)
Hmmm - what makes you say that Romans didn't build on flood areas?. Firstly - they did - much of Rome is prone to flooding, and the Romans expended much energy on flood control in Rome and other cities. Secondly - this falls into the fallacy that 'some old buildings survive therefore all old buildings are durable' - ignoring the fact that Roman towns built on flood plains may have left less archaeological evidence than Roman towns not built on flood plains. Thirdly it ignores basic economic geography - most societies have built cities around watercourses. Rivers provide opportunities for trade, crossing points, fishing opportunities, water supply etc. Floodplains are nice flat bits of ground and its easier to build on flat ground than hills.. Societies and individuals make pragmatic trade-offs between economic advantage and risk. From the perspective of someone who just got flooded it may appear that they got the trade-off wrong - though that is counterbalanced by the prosperity that may well have accrued in all the non flood years.Even in our rather comfortable society people are trading off 'nice view' or 'cheap' against a risk, whether that's flooding or coastal erosion.
I get professionally quite grumpy when people imply we don't know this stuff, or that 'our ancestors' or the Dutch or the man down the pub are somehow more knowledgeable about this than the UK's fairly large cohort of experts and professionals. Still - as we are now 'always guided by the science' I am sure that in future no politician/developer/local council/fly by night will overrule hydrologists and engineers in the future.
OTOH, even 100 years ago many areas would be described as "water meadows", and fully expected to be flooded for large parts of the year. At some point, humanity decided that we can control the flow of water and send it just where we want it to go - prime example of course being the Mississippi River, which the US Army Corps of Engineers have fiddled with for over a century to prevent the river changing course and affecting the industry downstream.
Well the point humanity decided that was probably at least 6000 years ago, and probably much earlier. The only difference between the US Corp of Engineers and the Mesopotamians is that the USCE have bigger shovels. Clearly there is an element of 'what prats' when you see a flooded housing estate with a road sign saying 'Soggy Bottom lane' but again it's economics. 100 years ago communities and supply chains were more local. It probably made economic sense to graze dairy cows on a local water meadow. In the modern world you can still do that, but you may get more return by building a factory/warehouse/shop whatever on the same bit of land.
While there are perfectly valid debates to be had about environment vs development, there is less validity in the idea that we can't technically control rivers. We can, and do (up to a point, obviously). What we also do is allow people to externalise costs (build a new estate and expect the state to pick up the tab for flood control) and we tend to elect tax cutting politicians, who then pare back on investment and maintenance (e.g Katrina - New Orleans).
And going back to the subject of volcanoes, how about Pompeii?
While fossicking around Pompeii a few years ago I kept on remembering that Pompeii was relatively small compared to modern day Naples (pop ~1 million and ISTR that about 3 million live in the area that would have to be evacuated if Vesuvius decided to get tetchy). Yes, they have motor vehicles these days, but the roads aren't going to cope if everybody has to GTFO in a hurry.
I recall an article in your rival, paper-based, publication 'New Scientist' many years ago, which, if I recall correctly, stated that there was a recordable increase in volcanism in South America due to only 10cm more snow fall on the Andes.
My guess (and respects to correspondents EvilDrSmith and eldakka above, my PhD is in Math Logic, so not really relevant, but I've watched a lot of David Attenborough and Prof Iain Stewart on TV), the volcanic systems are like escape valves, it only requires a slight increase in pressure to set them off, once the 'lid' is off, the release in pressure allows dissolved gasses and steam in the magma to erupt from solution and cause a bit of a blow out (I'm not being too technical for you all, am I? ;o). )
Anyway, that is my two ha'pence worth.
(My Dunce's Cap is on order from Amazon at this very moment.)
Osmotic pressure can be dramatic, fresh water soaking one side of porous rock, sea water in another part of it, both already at some depth/pressure. There was an interesting scheme to generate power from estuaries utilising the fresh water/salt water osmotic pressure across a big membrane. Just saying