The downside of being a scientist rears it’s ugly head: it’s 5.30 am and I cannot stop thinking about glacier monitoring from space. Argh! This happens ever once in a while. I want to sleep, my body tells me it needs sleep, and my brain goes into overdrive. Wonderful.
Today’s ruminations are on glacier monitoring, in particular Global Land Ice Measurements from Space (GLIMS). If you think meaningful glacier monitor is simple, you would be wrong. Let us take a simplified model of a glacier. Our glacier is 2.5 km long, typical for a Swedish glacier. To keep things simple we have divided the glacier into five mass balance zones: in each zone the glacier accumulates or melts (ablates) a certain volume of snow, measured in metres of water equivalent (m w.e.). Each zone transfers a net mass equal to the net accumulation each year to the zone below to represent glacier flow (see Figure).
In Year 0 the glacier is changing from a long period of stability (steady state) and mass is accumulating at the upper end of the glacier. At the glacier front there is still some, thouh not much, melting.
In Year 1, there is more accumulation over the upper glacier, and accumulation from the previous year starts to travel down glacier increasing the volume and raising the surface (in our simple model negative mass change could also travel down glacier, lowering the surface). But there is melting at the front leading to retreat.
Year 2 is mild and the glacier is in balance (accumulation is equal to ablation), the front is till retreating. But, in Year 3 there is a lot of melting and the net balance is negative. Don’t forget mass is still be transferred down glacier. In Year 4 the mass balance is similar toYear 0 with net accumulation. However, now the mass from Year 0 has reached the front of the glacier and the cumulative balance at the front is in fact +1 as +2 m w.e. have travelled down from the upper glacier from Year 0 (losing 2 m w.e. on the way through ablation). Hence the front advances: despite ablation at the front during four of five years the glacier has retreated during two and advances in Year 4.
When we measure the front then influences how we view the ‘status’ of the glacier. In Year 4 we see an advancing glacier, in year 3 a retreating glacier: over the five year period the glacier has a small net loss at the front. Years 0 and 4 have strongly positive mass change (accumulation), Years 2 and 3 strongly negative mass change (ablation). Overall the glacier has accumulated much more than it has ablated during our five year period. Volumetric measurements (the so called geodetic method) would show this, but measuring the position of the front would not.
So measuring glacier change is not simple. Mass balance measurements (measuring the net surface balance at several points on the glacier) would show change year to year. Surface height/volume change would present a more realistic and complete picture of the changing glacier, and front measurements would show something more complicated and ambiguous.
In practical terms measuring annual mass balance or volume is costly and difficult, requiring field work. These types of measurement are therefore relatively rare: in Sweden fewer than 10% of glaciers are measured this way: in Norway which has many many more glaciers the figure is about 0.1 %. Front measurements by GPS or surveying are quicker and easier: a global database of annual measurements is maintained by the World Glacier Monitoring Service (WGMS) but the number of observations is still very small. GLIMS aims to supplement the WGMS observations with periodic measurements from satellite data covering a large number of glaciers (ideally all glaciers). But, as we have seen, when we measure the front is important if we can’t do it every year.
So what I have been thinking about is how to combine these observations in a GIS to optimize the quality of the observations and spatio-temporal density of observations. The advantage of a GIS approach is that we can combine diverse measurements and analyse the relationship between glacier change and other parameters (climate, topography etc). There is more to say about this: we are working on Swedish glacier data at the moment with several years of satellite observations of all Swedish glaciers and mass balance and front measurements.
And that is why at 5.30 this morning I was awake and ruminating on glacier mass change and how to combine measurements.