Arguments
Part Two: How do we measure Antarctic ice changes?
Posted on 15 July 2010 by robert way
Guest post by Robert Way
This post summarizes the three key methods through which ice sheet losses are measured using satellite data. Although there are indeed some on site measurement techniques which are used by glaciologists, it has become the norm to use satellites for large scale measurements because of their cost-efficiency and coverage compared to on the ground measurement techniques. With regards to understanding Antarctic ice changes, these techniques are all the more important because of the continent’s size, climate, and isolation. Of all the different methods out there, it has become clear that there are three predominant techniques for accurate measurement of ice sheet changes. These consist of measuring elevation change, mass change and flux components.
i) Elevation change
Changes in the elevation of ice across the Antarctic ice sheet is measured using both laser and radar altimetry. Sensors such as these measure the distance from the satellite to the ice surface over time. Radar altimeter signals provide very good coverage for ice sheets but older versions of these satellites had issues with too low a resolution (2-5 km) leading to difficulties in measuring sloped/rough terrain. This low resolution also makes it difficult to measure small outlet glaciers where most mass wastage (ice loss) in Antarctica occurs.
Radar altimeter signals also penetrate the near-surface snow and ice layer introducing a potential bias to the data processing (Allison et al. 2009). A new radar satellite "Cryosat-2" has recently been placed into orbit equipped with a much higher resolution radar altimeter that uses complicated radar combination techniques (radar interferometry) to enhance the radar's horizontal resolution to 250 m while allowing for precise measurements of changes as small as a few centimeters. A further benefit of this satellite is that it will allow for the highest coverage available for an altimeter with measurements capable of being taken up to 88° N/S Latitude. Measurements have already been obtained from this satellite and have stunning detail (Figure 1).
Figure 1: New data from the recently launched Cryosat-2 over the Ross Sea
The second method of measuring elevation change is through using laser altimetry which is essentially a laser beam that measures the height from the satellite to ice surface. This measurement technique is extremely accurate and has a higher resolution than even the new radar altimeters but also has a small footprint (10s of meters) and can be affected by cloud cover (Allison et al. 2009). The most commonly used laser altimeter equipped satellite “Icesat” has no longer any working lasers thereby making satellite borne laser altimetry for ice sheets currently unfeasible but still applicable for aerial surveys or when icesat-2 is launched in 2015.
ii) Mass Changes
Gravimetry directly measures the change in regional gravitational fields through accurate measurements of the amount of effect that gravity has on satellites as they cross these regions. As mass is gained or lost it by the ice sheets, regional gravitational forces are affected, thereby making it possible for mass change estimates to be deduced. Ice sheet mass change has been detected using satellites to measure gravitational field changes through precise monitoring of the separation between a pair of satellites in identical orbits (GRACE Mission) (Allison et al. 2009). Although this method can be significantly affected by things such as crustal uplift and Glacial Isostatic Adjustment (GIA, crustal uplift after the depressing weight of an ice sheet/ice mass is removed), it is important to note that accelerations in mass loss can be detected very effectively and cannot be directly caused by GIA (Velicogna, 2009).
iii) Flux Component Method
The flux component method refers to the estimation of the total mass inputs and mass outputs for each individual drainage basin and ultimately the ice sheet as a whole. Estimating the mass inputs requires knowledge of spatial and temporal patterns of snow accumulation in regions that are oftentimes unsurveyed. In order to make an ice sheet wide snow accumulation map, on site observations of mass through meteorological records, snow pits, ice cores and marker poles are used to ground truth regional accumulation models. These models have been shown to have good agreement with measurements in interpolated regions (Allison et al. 2009). The loss by outflow is usually estimated as the ice discharge across the grounding line of the ice sheet, as this is where ice begins to float and hence contribute to sea level rise. Estimates such as these require observations of the speed of ice flow and the thickness of the column ice. Radar Interferometry has allowed for direct all-weather measurements of surface velocities to be conducted through both a differential method and a speckle tracking method (Figure 2). Depth of ice streams has been estimated using varying means from bedrock topography models to ground penetrating radar surveys.
Figure 2: Example of Radar Interferometry Velocity Measurements. Presentation by Eric Rignot at the Alos Palsar Symposium.
Satellite observations have been used to provide wide coverage of ice surface velocities but because the velocity of an ice sheet varies with depth a further level of uncertainty is added to the discharge estimates. For glaciers which move using mostly basal sliding such as outlet glaciers and ice streams with high velocities, this uncertainty is minimized because there is less variation in velocity with depth.
Conclusion
Overall when considering the different measurement techniques, it is important to note that although each individual technique has its own uncertainties, they each can and do effectively measure ice losses in
Part three will consist of a detailed rebuttal of several of Goddard’s recent pieces. This section will use much of the information from Parts One and Two in order to formulate this rebuttal.
I would like to thank Professor Jon Ove Hagen (University of Oslo) for helping to proofread and discuss these postings.
References
Allison,
Velicogna,
0 
0
The downside of this picture is that it becomes apparent that the observed trends are rather short (less than 15 years), that the uncertainty of the amount of mass loss is rather large (it's somewhere between 0 and -200 Gigatonnes per year) and that the error estimates of the different studies (depicted as the vertical extent of the boxes) sometimes do not even overlap. Cleary, more observations and improved data analysis are required for a clearer picture to emerge.
The obvious pitfall of tracking elevation changes is that height needs to be converted to mass. The exchange rate for this conversion is density. Densification of snow via firn to ice is strongly dependent on the accumulation rate and on snow temperature. So if your satellite detects a change in height, it may not mean you lost mass: it could be that the vertical column of snow and ice you're measuring has become a bit denser. It has recently been shown that the effect of decadal accumulation variability has a signature that is as large as the observed trend in elevation itself (Helsen et al. 2009). It means that long records of elevation change and accumulation rates are required to squeeze out a reliable mass change signal, no matter how accurate your laser or radar is.
The flux component method requires model calculations on the surface mass balance (roughly, snowfall minus melt minus sublimation). Direct observations are almost impossible due to the sheer size of the continent. Modeling surface mass balance is an ongoing field of research, and a simple thing like sublimation due to drifting snow or melt water production is surprisingly difficult to model, especially since validation is thin on the ground. Nevertheless, I believe that this method can yield good results. My personal feeling is that gravimetry will in the end be the method that can give the most accurate results, but future developments will show if I'm right on this.
The multiple lines of evidence are slowly converging to a picture in which Antarctica is losing mass, but we will unfortunately be quite a few years away from nailing down the Antarctic mass loss precisely and consistently among the different methods. On the upside, much progress has been booked in estimating the mass budget of the Greenland ice sheet, serving as an example of where constructive skeptical science can be in a few years time.
It is from this rather nice SCAR presentation. I do remember another more comprehensive image like the IPCC one, but can't remember where from!
Comments