How important are ice streams in accelerating ice sheet deglaciation?

人们普遍认为地球气候正在变暖，冰川正在失去质量，海平面正在上升。小冰川尤其容易受到影响，但它们只代表了地球上以冰形式储存的淡水的一小部分。相比之下，格陵兰岛和南极洲的巨大冰盖储存了几十米的等效海平面，最近的研究表明，它们也在失去质量，而且似乎正在加速。冰盖通过众多快速流动的冰川（称为“冰流”）将冰转移到海洋中。人们已经发现，冰流可以加速、减速，甚至完全停止；以及转换他们的位置。这些变化可能发生得相对较快（几年内），但尚不清楚它们是冰盖收缩长期趋势的一部分（数百年至数千年），还是仅仅反映了它们的自然变异性。另一种可能性是，最近的加速和变薄是一个大范围质量损失事件的开始或“前兆”，但问题仍然存在：冰流在加速冰盖消融方面有多重要，例如，除了可能由气候强迫单独造成的影响之外？为了评估这些短期变化的重要性，我们需要了解冰流是如何在比当前测量允许的更长时间尺度上运行的，我们还需要将冰流视为冰盖内几千年来演变的综合排水模式。这可以通过对过去冰流行为的调查来实现。过去冰流可以被识别，因为与缓慢流动的冰相比，它们的快速流动在现在暴露的冰盖床上创造了独特的冰川地貌。我们可以很容易地在过去的冰盖床上找到这些冰流“足迹”（例如在英国或北美），然后使用测年技术（例如放射性碳测年）和其他与冰盖流动模式相关的证据来估计它们存在的时间和时间。科学家们已经采用了这种方法，并增加了我们对它们长期行为的理解，但研究往往只关注一条冰流或特定地区。我们真正想要的是关于整个冰盖上大量冰流活动的信息，理想情况下，从尽可能长的时间跨度，即从一次完全的冰川消融，当冰盖从最大范围缩小并完全消失时。因此，该项目的目的是建立一个开创性的数据集，重建北美劳伦泰德冰盖（其大小与南极洲相似）从21000年前最大范围到5000年前几乎消失的每条冰流的时空活动。我们将绘制冰盖床上的所有流动模式，包括冰流，并使用现有的约4,000个放射性碳年代数据库（以及其他已公布的日期）来确定这些数据的日期。最近的一项初步研究表明，我们可以确定单个冰流的持续时间在250-500年之间。这将使我们看到整个冰流排水网络在冰川消融期间是如何演变的，以及它们的联合活动是否导致了重大质量损失的主要事件。它还将揭示冰流活动与过去气候突变和海平面变化的联系程度，例如突然变暖或变冷以及海平面的快速变化。综合起来，这将提供一个坚实的背景，以模拟和预测未来的反应以及现代冰盖可能的变化幅度，例如为IPCC的下一份报告提供参考。

It is widely accepted that the Earth's climate is warming and that glaciers are losing mass and increasing sea level. Small glaciers are particularly susceptible but only represent a fraction of the Earth's freshwater stored as ice. In contrast, the huge ice sheets in Greenland and Antarctica store several 10s of metres of equivalent sea level and recent studies suggest that they, too, are losing mass and that this appears to be accelerating.Ice sheets transfer ice to the oceans via numerous fast flowing glaciers called 'ice streams'. It has been discovered that ice streams can speed up, slow down, and even stop altogether; as well as switch their position. These changes can occur relatively rapidly (over a few years) but it is not clear whether they are part of a long-term trend of ice sheet shrinkage (over centuries to millennia) or simply reflect their natural variability. Another possibility is that the recent acceleration and thinning is the beginning or 'pre-cursor' of an episode of widespread mass loss but the question remains: how important are ice streams in accelerating ice sheet deglaciation, e.g. beyond that which might be excepted from climate forcing alone?In order to assess the significance of these short-term changes, we need to understand how ice streams operate over time-scales longer than current measurements allow and we also need to view ice streams as an integrated pattern of drainage within the ice sheet that evolves over several millennia. This can be achieved through investigation of past ice stream behaviour. Past-ice streams can be identified because, compared to slow-flowing ice, their rapid flow creates distinctive glacial landforms on the now-exposed ice sheet bed. We can locate these ice stream 'footprints' on past ice sheet beds very easily (e.g. in the UK or North America) and then use dating techniques (e.g. radiocarbon dating) and other evidence related to ice sheet flow patterns to estimate when and for how long they existed. This approach has been taken by scientists and has increased our understanding of their behaviour over long time-scales but studies have tended to focus on just one ice stream or specific regions. What we really want is information on the activity of lots of ice streams from across an entire ice sheet and, ideally, from as long a time-span as possible i.e. from a complete deglaciation, when the ice sheet shrinks from its maximum extent and disappears altogether. This aim of this project, therefore, is to produce a ground-breaking dataset that reconstructs the spatial and temporal activity of every ice stream in the North American Laurentide Ice Sheet (which was similar in size to Antarctica) from its maximum extent around 21,000 yrs ago to its near-disappearance around 5,000 yrs ago. We will map all the flow patterns on the ice sheet bed, including ice streams, and date these using an existing database of ~4,000 radiocarbon dates (and other published dates). A recent pilot study shows that we can date the duration of individual ice streams to within 250-500 yrs. This will allow us to see how an entire drainage network of ice streams evolves during deglaciation and whether their combined activity caused major episodes of significant mass loss. It will also reveal the extent to which ice stream activity is linked to abrupt climate and sea level changes in the past, e.g. abrupt warming or cooling and rapid changes in sea level. Taken together, this will provide a firm context with which to model and predict the future response and likely magnitude of changes in modern-day ice sheets, e.g. for the next IPCC Report.