Met Office Work Experience - Day 5 (for 25-08-2011)

I realise that these posts on what I did at the Met Office are like a week late but I thought that, because I had such a great time and learnt an incredible amount, I would still write them! Well, what did I get up to on my last day with the Met Office?

First up was a very interesting discussion with a researcher about ice sheets. There is currently a lot of uncertainity about ice sheets and their stability, with much of this uncertainity stemming from a lack of understanding of the mechanisms for ice sheet melt. Currently, it is believed that Antarctica is the most likely to experience significant melt and this is because much of it is under the sea. If the proportion that lies below sea-level was to melt, which scientists think could occur over the next few hundred years (although this may sound quite slow, it is considered to be rather rapid!), it would raise global sea-level by 6 metres!!! Greenland melting is an area that often recieves much attention, perhaps becuase we are unsure just how much freshwater could be released or how quicly it would occur, but for now alteast, many feel that a large freshwater input is unlikely. Before this discussion, I had never really given much thought to the influence that the angle of the bedrock, upon which the ice sheet lies, on melting but it does, in fact, seem to play quite a large role. The bedrock that the Greenland ice sheet lies upon slopes away from the sea, whilst that in Antarctica slopes towards the sea and thus making runaway melting possible. When the Laurentide Ice Sheet existed it was sat upon 'flat' bedrock, something that many considered essential for the occurence of Heinrich events. Due to this, depsite the fact that models currently cannot replicate Heinrich events, it is believed that this current climatic state cannot provoke Heinrich events. The fact that surprised me the most was that increased surface air or sea surface temperatures, as a result of global climate change, are not the biggest threat to the stability of ice sheets and would not be responsible for instigating the greatest volume of melt. Instead it is increased wind speeds.....but why?

This is a bit of an exaggeration of what happens but
hopefully you get the general idea....

Well, increased wind speeds would raise the height of local sea-level and increase Ekman Pumping. Increased Ekman Pumping would provoke old warm waters from the deep ocean to be dragged closer to the surface and over the terminal morraine, which marks the end of glaciers or ice sheets, and towards the base of the ice sheet, thus causing melting to occur. This is believed to be responsible for more melting than raised surface air/sea temperatures. One thing that scientists are unsure of though is what happens to the warm water once it passes the terminal morraine. Does it hit the base of the ice sheet and then continually circle, gradually melting away the base or does it bounce back off and return to the rest of the ocean? Understanding this is, again, crucial if predictions of ice sheet melt are to become more cetain.....

Anyway, all of this 'stuff' is important if scientists are to make more certain predictions of the future of the MOC and there is a lot of debate over just how much of an impact melt of Greenland or Antarctica would have. Most focus is placed on Greenland, as a result of its location. Some feel that perhaps, if enough of Greenland melted, it could significantly reduce the MOC intensity whilst others believe that, due to existance of sinking sites either side of Greenland, that Greenland melt could provoke a switch in sinking sites to the western side of Greenland - a switch that could have the potential to actually warm the UK during winter. There are a couple of other quite specific topics that we covered but I think I will leave them for another blog post.

After this I attended a Modelling Team meeting which was quite interesting as it provided an insight into some of the work that researchers at the Met Office are currently doing and some of the problems they are facing at present. Following this I had a chat with someone regarding ENSO, a topic that fascinates me, and as, again, there was lots that we covered,and I am a bit more confident about talking about ENSO, I am going to write another post solely on this. The afternoon was finished off with a chat about the relationship between the ocean and atmosphere and how this relationship is replicated in models. The relationship between the ocean and atmosphere is really really complex and I literally touched the very very basics. The ocean is sort of like the memory of the Earth climate system. The atmosphere cannot store things, like signals or changes in climate, and so instead it passes the signals on to the oceans. The oceans can store this information for hundreds and hundreds of years, whilst it circulates them around the world, and then passes the signal back to the atmosphere where it provokes a short term, but rapid, response. This coupling is crucial for many things such as ENSO. It is tricky to model all of the processes that link the oceans and atmosphere and all of the exchanges that happen between them (I am in the process of writing a post on the real basics of climate modelling as it is like a whole new science).

I apologise as I realise that all of my posts regarding my time at the Met Office have been a bit all over the place but I honestly learnt so much and I am not that great at explaining things. Despite this, I still hope they have been interesting to read and have given you a bit of an insight into the work done at the Met Office and some of the things I was fortunate to do whilst up there. I really cannot thank the people who made this whole experience possible enough - I learnt an unbelievable amount, gained some invaluable advice universities courses and careers etc, got to meet some great and highly intelligent people and simply had just an amazing time!!!

Heinrich Events

Heinrich events were first described in 1988 by the German marine geologist, Hartmut Heinrich, whose study demonstrated that six times during the last glacial huge iceberg armadas discharged from Canada into the North Atlantic, depositing, as they melted, lithic fragments. Peaks in the abundance of these lithic fragments far from land can only be attributed to icebergs melting as sea ice only transports dust and to further support this idea, much of the ice-rafted debris found consists of limestone similar to that exposed in much of Canada today (although lithic fragments from across the North Atlantic do feature). The lithic peaks form layers in sediment cores, known as Heinrich layers, which extend some 3000km across the North Atlantic, and seem to reflect an episodic nature in Heinrich events, with each one lasting approximately 1ka and occurring at intervals of 7-13ka (over the last 100ka).

Laurentide ice sheet during Last Glacial Maximum (LGM)

The six events described by Heinrich are not unique as there are many lesser peaks in the lithic content of sediment cores. They are, however, the largest of the now recognised events in high-resolution core logs. In every event, icebergs were released when the surrounding surface water was cold but then abruptly warmed. All of this can be said with confidence but the exact cause of the periodic surges in the flow of the Laurentide ice-sheet (covered eastern Canada) provides an air of uncertainity as explaining the cause of the characteristic iceberg armadas is slightly harder than identifying them in sediment cores. There are, though, two fundamentally differing models that try to do this.

Denton model for iceberg armadas

The Denton model is essentially climatically driven as it based on global cooling leading to greater snowfall and thus a rapid expansion of ice-sheets/shelves and then marine ablation calving the icebergs from the expanding ice-shelf. Fundamentally the Denton model lays blame with the external cause of ice-sheet/shelf expansion as a result of global cooling as the forcing factor.

Advantages:-

- Evidence supports cooling in the North Atlantic before the Heinrich events

- The model explains the surging of South American glaciers and other global responses without the need for complicated teleconnections, during the same period of time

- Around 1.5ka climate cycles are known to have occured during the last 10ka and, despite them being an order of magnitude smaller than Dansgaard-Oeschger events, they too suggest external forcing over internal forcing

Disadvantages:-

- The model cannot explain the rapidity of events

- Heinrich events do not have consistent cyclicity as early on in the last glacial period they occured every 13ka but then later on they occured every 7ka

The MacAyeal 'binge-purge' model for iceberg armadas

The MacAyeal ‘binge-purge’ model blames an internal cause (ice sheet failure) due to geothermal and frictional heat periodically building up and getting trapped beneath the ice-sheet, thus melting the base and provoking the catastrophic failure of the ice-sheet. So, instead of being climactically driven, it’s reliant on the mechanical failure of ice-sheets due to thermal modification at its base.  MacAyeal suggested that the Laurentide ice-sheet grew whilst its base was frozen solid to both the crystalline rock (currently exposed on land in Canada) and the softer sediments found beneath Hudson Bay, to the south of the ice-sheet. When the base of the ice-sheet was heated enough, by geothermal heat, to cause the sediments to thaw, the rapid ice movement of the purge was initiated. Frictional heating then further increased the temperature, provoking a positive feedback as the ice movement was accelerated further. Despite this, the greater friction at the ice/crystalline rock meeting point prevented the total collapse of the ice sheet. This model highlights the fact that natural systems can display abrupt changes caused by a forcing factor that fails to change with time and so can occur independently of the external forcing factor; one of the reasons for the difficultly in making predictions about the future impacts of climatically driven forcing factors.

 Advantages:-

- The model is able to explain the rapid initiation and termination of Heinrich events - something that the Denton model fails to do

- Due to its dependency on the size of the ice sheet, it can explain the irregular cyclicity of Heinrich events

- Provides an explanation of the large amount of ice-rafted debris, which forms the Heinrich layer, found in the North Atlantic

Disadvantages:-

- Cannot explain the cooling known to have occured before each Heinrich event

- It needs a mechanism, such as the NADW, to transport the 'signal' around the world

Both models have their advantages and disadvantages and as a result there is a lack of consistency in which is favoured by scientists. For some time though, the 'binge-purge' model was favoured by most but then it was revealed that some of the Heinrich layers contained material that could have only came from other ice sheets, other than Laurentide. Attempts to generate a combined model have failed to paint a clearer picture of the mechanism provoking iceberg armadas but have led to further research into the conditions surrounding such events. Some of the sediment in the Heinrich layers has been linked to areas other than those covered by the Laurentide ice-sheet, such as Iceland due to basaltic glass fragments. This suggests that separate ice sheets surged simultaneously, something unlikely unless climatically driven or as a result of increased marine ablation due to eustatic sea-level rise thanks to Laurentide ice-sheet melt. The westward thickening of Heinrich layers, across the Atlantic, and its continuation towards Hudson Bay, point to the latter being correct and it is possible that its break-up triggered a response in other ice-sheets. Evidence insinuates that three gradual advances and rapid retreats of the Laurentide ice-sheet occurred towards the end of the last glacial, with glacial advances culminating before Heinrich events; thus provoking rapid ice discharge into the Atlantic, reducing southward ice flow and resulting in rapid retreat of LIS. There also appears to be synchronicity between the North Atlantic ice-rafting events and ice-sheet growth/ collapse in the Andes and New Zealand; something which supports the idea of strong inter-hemispheric coupling of changes in temperature  and therefore global forcing of climate change. Research into the abundance of left-hand coiling in foraminiferid populations in ocean floor sediment cores, accompanied with studies into the origins of lithic sediments, have indicated that iceberg-calving events have occurred more frequently than first believed (intervals of 2-3ka), albeit on a smaller scale than the six originally identified Heinrich events. Of greater importance, is that many of the fragment peaks coincided with >90% proportions of left-hand coiling foraminiferid, thus revealing that the launching of iceberg armadas corresponded with low North Atlantic SST’s, symbolising stadial periods followed by the prompt warming leading into an interstadials. The use of d¹⁸O variations as a proxy record for eustatic sea-level rise and a lack of coherence between evidence of temperature rises in ice-cores from Greenland and Antarctica imply that meltwater discharge pulses, during the above events, originate from the Northern Hemisphere; an inference only endorsed by changes in salinity that are known to have occurred in the North Atlantic. As a result it can be said with confidence that Heinrich events influenced alterations in ocean circulation. 

There exists conflicting views regarding the influence of Heinrich events on THC. Some believe that the AMOC collapsed over the course of the six known Heinrich events, as a result of the influx of freshwater from glaciers interrupting the ‘normal’ circulation. Others suggest that NADW cessed as a result of each Heinrich event as the iceberg armadas placed a freshwater ‘lid’ over the northern North Atlantic. A Heinrich event is estimated to have provided a freshwater input in the order of 0.1Sv to the Atlantic, a volume believed to be sufficient to halt NADW formation, whose magnitude dictates the deep-ocean THC, and thereby explain the cooling observed in proxy data from the mid-latitude Atlantic. Some believe that Heinrich events were actually triggered by a reduction in NADW formation, due to freshwater fluxes to the North Atlantic as a result of the early deglaciation of the Fennoscandian ice-sheet. The reduced NADW formation would generate warmer SST’s, therefore perturb the ice-shelves, thus triggering iceberg armadas, allowing for the additional freshwater to further weaken, ultimately leading to the cessation of the MOC. Alternatively, several think that the collapse is relatively independent of the magnitude and origin of the freshwater input produced by Heinrich events, as long as it is transferred to North Atlantic convection sites; whilst events restricted to the Nordic Seas reduced NADW formation but didn’t provoke the cessation of the global conveyor; thus perhaps partially explaining the cause of Dansgaard-Oeschger events.

This might seem a bit random, but I have just written up a section on the past changes that occured to the ocean circulation across glacial/interglacial and stadial/interstadial and so have mentioned Heinrich events, Dansgaard-Oeschger events and Bond cycles, amongst other things; unfortunately it is beyond the scope of my project to go into any real detail into the models used to suggest the causes of iceberg armadas and so, instead, I thought I would mention them on here. I have one question though, that I would be quite interested in knowing the answer to if anyone knows, what excatly caused the frequency of Heinrich events to change, from 13ka to 7ka, over the course of the last glacial?