Why should we reconstruct past ice sheets? | The past is the key to the future | How do we reconstruct past ice-sheet change? | How do we relate ice-sheet change to climate? | Further Reading | Comments |
Why should we reconstruct past ice sheets?
Glacial geologists love to go out into the field, collecting rock samples and bags of gravel. They spend hours mapping a single moraine in detail. They hammer tubes into beds of sand and cover themselves, and the sand, in black bin liners. But what is the focus of all this effort? Why do we care how big or how thick past ice sheets were?
Scientists reconstruction past ice sheets because they want to know how glaciers and ice sheets interact with climate and with the ocean. We can observe modern glaciers melting; we can look at the surface of ice sheets with satellite images and calculate changes in mass balance and we can map mountain glacier recession. It’s pretty obvious that glaciers are shrinking and melting worldwide. It’s also clear that air temperatures are increasing, ocean currents are penetrating deeper onto the Antarctic continental shelf, and precipitation patters are changing. What does this mean for our mountain glaciers and ice sheets? We know that they are already contributing to sea level rise, but by how much? And by how much will they change in the future?
It is the role of palaeo ice sheet reconstruction, or palaeoglaciology, to answer these questions. We investigate past ice-sheet and glacier response to climate change to understand:
- Mechanisms and processes of change
- Thresholds and tipping points
- Processes of glacier flow
- Non-linear responses such as ice shelf collapse, marine ice sheet instability and ice steam fluctuations
- Magnitudes and rates of change under different environmental scenarios
The past is the key to the future
By examining how ice sheets responded to change in the past, glaciologists hope to uncover details that will help them understand how they are likely to change in the future. By looking at how ice sheets changed over long timescales, they can extend the short observation period over the Antarctic Ice Sheet. For example, ice streams in Antarctica have been observed to change, accelerate, switch off and recede. Is this normal behaviour? By looking at past ice stream change, for example in the last British Ice Sheet, we can gain a far better understanding of how ice streams evolve over far longer timescales.
Thresholds and tipping points are crucial. We worry that the West Antarctic Ice Sheet is unstable and could catastrophically collapse. Has this happened before? Will it pass a threshold and then collapse quickly, or slowly melt away? What is this threshold? We can only know by looking at the palaeo record.
Scientists also hope to better understand processes of change. The bed of the Greenland and Antarctic ice sheet is difficult to access; far better to examine the exposed bed of ancient ice sheets, like the last British-Irish Ice Sheet!
The ultimate goal of palaeo ice sheet reconstruction is gain a better understanding of how ice sheets and glaciers responded to change in the past. This will enable us to predict how they will respond to change in the future, and this will mean that we can give more precise and accurate estimates of future sea level rise.
How do we reconstruct past ice-sheet change?
Fortunately, we have many tools at our disposal to reconstruct past ice sheets. Glacial geologists usually have three main objectives:
- Reconstructing past ice extent and thickness
- Reconstructing past ice flow
- Dating past ice fluctuations
Past ice extent and thickness
Reconstructing past ice extent means finding past moraines and glaciated terrain. Glacial geologists compile databases and maps of these moraines, interpolating between them and creating isochrones of past ice-sheet extent.
Glacial geologists can reconstruct past ice-sheet thickness in a number of ways. They can use trimlines to demarcate the height of the former ice surface. They can use equations that predict ice thickness by assuming a certain surface slope to reach a certain extent. And they can use measurements of the Earth’s isostatic rebound to calculate the past volume of ice.
Examining past sea level rise in far-field locations, such as the Bahamas, gives us a record of changing global ice volume over long timescales. Examining local sea level rise in near-field locations (where near-field means close to a present or former ice sheet) helps us understand rates of isostatic adjustment, and hence local ice volume.
Past ice flow
Recent research shows that past ice flow was not isochronous; in fact, the last British Ice Sheet had numerous complex phases of ice flow. Ice streams flowed in different places at different times as ice divides changed in response to changing ice sheet thickness. Mapping landforms such as mega scale glacial lineations, drumlins, roche moutonnées and streamlined, striated bedrock helps glacial geologists to reconstruct these changing ice flow pathways and dynamic ice stream behaviour.
Looking in detail at the sediments laid down at the ice-bed interface allows glacial geologists to understand in detail the processes of past ice flow. How quickly did the ice flow? Was it cold-based or wet-based? Did the ice flow due to deformation of its bed, slipping, or deformation of the ice? By looking at sections of sediment under the microscope, we can infer detailed information about the processes of sediment deposition.
Looking at sediments and landforms together as a whole allows glacial geologists to reconstruct a Glacial Landsystem that they can use to understand the style of glaciation and take a broader view of processes and external influences.
Dating past ice fluctuations
Of course, it is all very well understand how big an ice sheet was and what direction the ice flowed in. But this is not useful for understanding rates and magnitudes of change if we cannot put it into some kind of time-scale context. Fortunately, we have many techniques at our disposal.
Cosmogenic nuclide dating is extensively used to date the formation of moraines and the speed of glacier recession. This technique gives an exposure age, that is, the time since the boulder was exposed at the Earth’s surface. Cosmogenic nuclide dating in this context gives a maximum age for moraine formation; the moraine must be younger than the exposure age (unless your boulder has rolled, moved, has an inheritance, or any of the other multitude of factors that may result in an inaccurate age). Cosmogenic nuclide dating can also be used to date trimlines and constrain past ice thickness.
Radiocarbon dating is an essential part of the glacial geologists’ toolkit. This technique relies on organic remains. One example of an application may be for a lake dammed by a moraine. A radiocarbon age from the base of the sediment core gives a minimum age for moraine formation; the moraine must be older than the radiocarbon age. Radiocarbon ages are also extensively used from marine sediment cores around the margins of ice sheets, such as the Antarctic Peninsula Ice Sheet. Transitional Glaciomarine Sediments, those glaciomarine sediments laid down immediately after ice sheet recession and that overlie subglacial tills, provide a minimum age for ice-sheet extent at the core’s position; the ice margin was at this position before the radiocarbon age.
Optically stimulated luminescence is used to date beds of sand that accumulated in front of the ice margin by proglacial rivers. Like any other technique, dating glaciofluvial sands has its own challenges, but it can be used to provide limiting ages on ice-sheet extent.
How do we relate ice-sheet change to climate?
With all the data that has been collected mapping out former ice sheet extent and thickness, changes in its flow regime and rates of thinning and recession, we have a good idea of how the ice sheet evolved through time. The next stage is to examine the external forcings that drove this change. There are several ways of doing this:
- By using ice cores and other proxies to reconstruct past climate
- By relating cryospheric change to climate through a numerical model
Reconstructing past climate
Ice sheets and glaciers are an integral part of the atmosphere-ocean-cryosphere system. Change in one inevitably leads to change in all others. So if we want to understand what may have driven past ice sheet change, we need to understand past climate. Fortunately, there are many ways in which we can do this.
Ice cores preserve the composition of the ancient atmosphere. We have detailed accounts of past air temperature over the last 800,000 years. Cores from the ocean also tell us how air temperatures and global ice volume has changed over millions of years. We have detailed records of past environmental change from proxy records from pollen, beetles, insects, algae and in fact any organism that lives within limited environmental conditions. Together, all these data build up a picture of global to local environmental change over short (decadal to centennial) to longer (millennial to epoch) timescales.
Numerical ice sheet modelling
The equations that govern ice flow and mass balance (the balance between snow gained and snow melted) are complex. They depend on velocity, thickness, thermal gradients, ice temperature, air temperature lapse rates, and many other factors.
Fortunately, these physical relationships are reasonably well understood and can be fed into a computer model. These computer models can take in input data (such as past climate from an ice core) and drive a numerical ice sheet model, allowing scientists to investigate detailed processes and these key questions of past rates and magnitudes of change. Models that provide a good fit to the geological record have a high confidence, and allow scientists to investigate atmosphere-ocean-cryosphere relationships. Ultimately, all these data together help scientists to answer the critical question of how the cryosphere will change over coming centuries, and how much global sea levels will rise.
As an example, I finish with this simulation of the glaciers in Yosemite National Park responding to climate change during the Last Glacial Maximum, 21,000 years ago.