Open Mind

Greenland Tremors

March 30th, 2007 · 8 Comments

In case you’re wondering, I’ve been travelling for the last two weeks, and haven’t been able to update my blog. But I’m back home now, so here’s the latest…

In a fascinating paper published last year (Ekstrom et al. 2006, Science, 311, 1756), geophysicists at Harvard investigated a new class of seismic events: glacial earthquakes. These were identified only as recently as 2003 (Ekstrom et al. 2003, Science, 302, 622), and are dramatically different from other types of earthquakes. For one thing, their “characteristic period” of vibration is not less than one second, as is typical of earthquakes, but very long — on the order of 30 to 60 seconds. For another thing, all detected events of this type are associated with mountain glaciers in Alaska or with glaciers and ice streams along the edges of the Antarctic and Greenland ice sheets.

In fact they appear not to be earthquakes of the usual type, but large and sudden glacial-sliding motions. In other words, they are sudden and massive glacial movements, not unlike a landslide, which generate elastic waves in the earth that are detected by seismometers.

Studying the period January 1993 to October 2005, the investigators identified 182 such events on Greenland. Study of their occurence times yields some interesting results. For one thing, they show a strong seasonality, mostly occuring during summer rather than winter. Studying non-glacial earthquakes during a similar time period, they note that there is no such seasonal pattern:


This indicates that temperature is a factor in glacial quakes, almost certainly because seasonal ice-melt either weakens the structure of the glacier, or because ice-melt water lubricates the base sufficiently to let them slide, or both.

Another very disturbing result of the analysis is that glacial quakes in Greenland have been growing far more numerous recently. The annual number of Greenland glacial quakes has more than doubled since 2002:


In fact the data for 2005 (the last year included in the study) indicate more than twice as many glacial quakes as any year prior to 2002 — despite the fact that the 2005 data include only the first ten months of the year, not the entire year.

The authors hypothesize that increased temperatures in summer have controlled the annual timing of these events, but increased temperatures overall are the reason for their increased number. This, they say, it because recent warming is altering the behavior of the entire Greenland ice sheet:

Recent evidence suggests that ice sheets and their outlet glaciers can respond very quickly to changes in climate, primarily through dynamic mechanisms affecting glacier flow. The seasonal signal and temporal increase apparent in our results are consistent with a dynamic response to climate warming driven by an increase in surface melting and the supply of meltwater to the glacier base.

All this bodes ill for the Greenland ice sheet. The most recent evidence suggests it is deteriorating faster than before, and one of the strongest criticisms of the recent IPCC report is that it underestimates the risk of rapid wasting of the Greenland ice sheet, and the impact of this on sea level rise.

Tags: Global Warming · climate change

8 responses so far ↓

  • Brian // Mar 31st 2007 at 3:13 am

    Might the observed increase in quake frequency be reflecting a feedback and/or cumulative process within the system? That is…the system is not allowed to “heal” each season because each season is incrementally warmer?

    There are so many mysterious and dynamic processes related to glacial flow.

  • Andrew Dodds // Apr 2nd 2007 at 8:53 am

    It would be nice to see - for comparison - a graph of temperature for the region for the same timeline..

    Found the melt extent:

    The lack of correlation seems quite strange - the implication being that what we see in the glacil earthquakes is a very smoothed version, or has a long lag time.

  • Hank Roberts // Apr 2nd 2007 at 11:38 pm

    I don’t think I’d call that a ‘lack of correlation’ — but that’s just eyeballing it. Anyone actually done an analysis of covariance?

  • Gareth // Apr 3rd 2007 at 5:38 am

    It could be a sort of “weakened buttress” effect, similar to the way that the break up of an ice shelf allows inland glaciers to accelerate towards the sea. In this case, snow accumulation at high elevations in central Greenland (which still has a positive mass balance, I believe) pushes the ice down and out towards the edges. But the edges have now been thinning steadily for years, and so can’t resist the pressure as well as in the early 90s. Therefore the ice moves more, and the rate of glacial quakes continues to rise. To slow this process down will require the edges to stop melting, allowing the buttresses to strengthen, which - given the trends - doesn’t seem likely to happen any time soon.

    If this idea is anywhere near the truth, it might suggest that we passed the tipping point for a substantial reduction in ice sheet size sometime in the late 90s.

  • Alexander Ac // Apr 3rd 2007 at 1:21 pm

    Nice post ;-) I was inspired to use it in my blog. Thank’s.

    Indeed, it seems, that glaciers are starting to respond in a non-linear fashion…. but there is a small chance, that the trend in increased number of earthquakes will at least slow down…

  • Hank Roberts // Apr 4th 2007 at 12:30 am

    Hmmm. I wonder if it’s coincidental that we had a low hurricane season and a break in the glacial erthquake trend in 2006.

    N: C51C-03 INVITED
    TI: Glacial Earthquakes, three years on
    AU: * Ekström, G
    AF: Lamont-Doherty Earth Observatory of Columbia University, 61 Route 9W, Palisades, NY 10964 United States
    AU: Nettles, M
    AF: Lamont-Doherty Earth Observatory of Columbia University, 61 Route 9W, Palisades, NY 10964 United States
    AB: Three years ago we reported on the discovery of glacial earthquakes, a new class of earthquakes associated with fast-moving glaciers. These seismic events generate long-period (25–100 sec) Rayleigh and Love waves that in amplitude are equivalent to those generated by a traditional magnitude-5 earthquake, but which produce little high-frequency energy. We have detected a small number of events of this type in Alaska and Antarctica, but more than 95% of the located events (some 200 for the period 1993–2005) are associated with outlet glaciers on Greenland. A strong seasonal signal is evident in the frequency of glacial earthquakes on Greenland, with most events occurring during August and September, coincident with the period of maximum melting.

    A clear trend in the frequency of events was identified for the period 2001–2005, with each year generating a larger number of events. This trend appears to have been broken in 2006.

    The seismic waveforms for glacial earthquakes are well explained by a mass-sliding mechanism, and are not consistent with a tectonic stress-release mechanism, leading us to our preferred interpretation that the events are associated with a sudden melt-water-facilitated downhill sliding motion of a large portion of the glacier. Force geometries determined for the events are in general consistent with this interpretation, but ambiguities remain. Alternative explanations, such as large-scale sudden internal deformation, should also be considered, as these events occur where glacial strain rates are very high. Near-field geodetic and seismic observations will be required to further constrain and elucidate the nature of these events, and a deployment of 20 continuously recording GPS instruments on Helheim glacier during the summer of 2006 may have provided new observations to address this phenomenon.

  • Hank Roberts // Apr 4th 2007 at 12:35 am

    Have y’all tried “glacial earthquakes” in Google Scholar and limited it to since say 2006?

    Amazing stuff:

    “… Glacier speeds of c. 0.3 m s–1 would have induced peak flash temperatures sufficient (>1130 °C) to cause partial frictional welding of the deformed quartzitic bedrock beneath an ice-loaded clast. These results support suggestions that glacial seismicity is due to episodic very rapid shifts of large ice masses over the substrate (stick–slip movement). They also indicate that significant effective normal stresses occur at the base of glaciers.”

    Yes, you read that temperature right.


    “…. Here we report on the observation of continuous local seismic activity in the area of
    the David outlet glacier in Victoria Land. We monitored the area with a temporary
    local seismographic array, that allowed us to detect more than 6,000 events, most of
    which originate by ice cracking (and are often dubbed ’icequakes’). Besides, we also
    recorded and located about 120 low-magnitude events, with distinctly different characteristics,
    that do not fit either in the descriptions of icequakes, nor of usual earthquakes
    due to brittle fracture of rock. We interpret them instead as due to stick-slip behaviour
    at the ice-rock interface, presumably due to a single asperity mostly slipping aseismically
    and episodically breaking in a brittle failure, perhaps in connection with a small
    surge in glacier motion. Larger magnitude events, whose signals are recorded at planetary
    scale, have been reported as due to sudden slip of large ice masses in Greenland
    and Antarctica. We may be looking here at a smaller-scale instance of the same phenomenon.
    Further constraints are expected from modelling of fracture dynamics, and
    from geodetic observation of glacier flow.”

    Geophysical Research Abstracts, Vol. 8, 08827, 2006
    SRef-ID: 1607-7962/gra/EGU06-A-08827
    © European Geosciences Union 2006
    Peculiar seismicity in Antarctica: swarms of glacial
    earthquakes with a recurrent magnitude under David
    Glacier, Victoria Land
    S. Danesi (1), A. Morelli (1), S. Bannister (2)
    (1) INGV Bologna, Italy (2) GNS Wellington, New Zealand ( / +39065041181)

  • Hank Roberts // Apr 5th 2007 at 3:35 pm

    Nobody know more? Can anyone point me to more info/discussion elsewhere? The Scholar search is great for papers like those I excerpted; Google isn’t finding any discussion of them with my usual approach of trying out likely keywords.

Leave a Comment