Thursday, July 28, 2011

Katla - perhaps a pause in the development

The recent changes in the earthquake pattern around the caldera of the Katla volcano in Iceland suggest that the eruption that I was anticipating may be somewhat longer in the making than I had stated earlier in the month. The pattern is now diffusing away from the concentrated patterns of quakes that were occurring earlier in the month, although the distribution remains largely close to the initial regions of focus.
Earthquakes around Katla July 13th

Earthquakes around Katla July 18th 2011

Earthquakes around Katla July 23rd 2011

Earthquakes around Katla July 28th 2011

In an earlier post I tried to explain the process that went into the focusing of the quakes in a tighter swarm around a possible eruption site. And for a while this focusing occurred. However, while there was the evidence through the jokulhlaup of some magma penetration that caused glacial melt and the flood that took out the bridge, there has not been any physical evidence on the surface of magma movement since. There have been some harmonic tremors (reported by Jon ) indicative of some magma movement, but no expression of it to the surface.

If we consider that there is a magma chamber under Katla,, Sturkell et al have looked at the evidence of ground swelling to project that this chamber may be around 5 km deep, with up to 700 m of ice in the caldera. They base the location of the chamber on the idea that as the magma flows into the chamber it is under the pressure of the driving fluid that moves it into the chamber, and that this pressure will lift the rock and glacier above the chamber. The pressure is insufficient to break that rock, initially, which is part of the basis for my earlier reasoning that the series of quakes was indicative of the fracturing in the overlying rock, which would weaken the cap and provide passages through it for the magma to flow upwards and thus cause the eruption.

Simplified projected section through Katla after Sturkell et al)

However one thing that does occur as these fractures develop is that some magma may flow into them before they connect a pathway to the surface. In the short term this will lubricate the fractures and ease their growth toward total cap failure. However if the magma is stalled then it is possible that it loses sufficient heat to the surrounding rock that it solidifies. This then becomes the glue that I wrote about in the earlier piece, since it now closes and binds the rock passage walls, so that the cap re-acquires some integrity.

With the possible loss of some pressure with magma escape into the overlying rock, this may then return the situation to a point earlier in fracture development, with the somewhat relieved rock being re-stressed before it will develop a new set of fractures to coalesce, weaken the cap and potentially again raise the risk of eruption. I begin to suspect that we are in this phase at the moment. But, as I said earlier, each case and condition is different and while continued focused quakes in the region of the caldera continue to indicate further weakening of the cap, we do not know how much will be required before an eruption occurs.

Note that the flow of magma under the rock, and the pressure that it exerts, causes the surface of the volcano, and thus of the glacier above it, to move. This movement is monitored using GPS and plots of the data are available.

1 comment:

  1. Wiki says `the melting point of silicon at ambient pressure (0.1 MPa) is 1415 °C, but at pressures in excess of 10 GPa it decreases to 1000 °C', so it could be that pressure effects will be at least as important as temperature when it comes to `glueing' i.e. the nearer the surface, the more glutinous the rock becomes? Of course magma is not silica, so it may not be true for magma. Certainly the free flowing magma we see flowing out of volcanoes would suggest an extremely liquid core at both high temperatures and pressures.