Abstract :
[en] The Greenland ice sheet (GrIS) is the world’s second largest ice mass, storing about one tenth of the Earth’s freshwater in the form of snow and ice. If totally melted, global sea level would rise by 7.4 m, affecting low-lying regions worldwide. Since the mid-1990s, increased atmospheric and oceanic temperatures have accelerated GrIS mass loss through increased meltwater runoff and ice discharge across the grounding line of marine-terminating outlet glaciers.
To understand the causes of recent GrIS surface-driven mass loss, we use the Regional Atmospheric Climate Model (RACMO2) to dynamically downscale climate. This meteorological model, coupled to a multi-layer snow model, simulates the evolution of the surface mass balance (SMB), i.e. the difference between snowfall accumulation and ablation from sublimation, drifting snow erosion and meltwater runoff. In this thesis, we show that RACMO2 realistically simulates the extent of the elevated inland accumulation zone, where the GrIS gains mass at the surface as snowfall exceeds sublimation and runoff, as well as the narrow ablation zone along the low-lying margins where mass is lost through runoff of meltwater exceeding snowfall. Separating these two areas is the equilibrium line, where accumulation and ablation cancel (SMB = 0).
In order to cover a large domain at reasonable computational cost, RACMO2 is run at a relatively coarse horizontal resolution of 11 km (1958-2016). At this spatial resolution, the model does not well resolve small glaciated bodies, such as narrow marginal glaciers, typically only a few km wide, and small peripheral ice caps (GICs), detached from the main ice sheet. To address this issue, we developed a statistical downscaling algorithm that reprojects the original RACMO2 output on a 1 km ice mask and topography derived from the high-resolution Greenland Ice Mapping Project (GIMP) Digital Elevation Model (DEM). Correcting for surface elevation and ice albedo biases over the topographically complex GrIS margins, the downscaling procedure accurately reproduces the large ablation rates over narrow ablation zones, marginal outlet glaciers, and GICs, all important contributors to ongoing sea level rise. This downscaled 1 km product proves to be very useful for studies quantifying mass changes of these spatially restricted ice masses.
For instance, using the new 1 km data set, we identify 1997 as a tipping point for the mass balance of Greenland’s GICs. Before 1997, ablation and accumulation were in approximate balance, and these peripheral ice masses remained stable. After 1997, following significant warming, the accumulation zone retreated to the highest sectors of these ice caps and their mass loss accelerated. Although previously reported by satellite altimetry measurements, no clear explanation has been provided for this acceleration so far; with the 1 km SMB product, we can identify the acting surface processes. Greenland’s GICs are located in relatively dry regions where summer melt nominally exceeds winter snowfall. To sustain these ice caps, the refreezing of meltwater in the snow is a key process. The snow acts as a ”sponge” that buffers a large fraction of meltwater, which subsequently refreezes in winter. The remaining meltwater that can’t be buffered runs off to the ocean and directly contributes to mass loss. Until 1997, the snow layer in the interior of these ice caps could compensate for increased melt by refreezing more meltwater. Around 1997, following decades of increased melt, the snow became saturated with refrozen meltwater, so that any additional summer melt was forced to run off to the ocean, tripling the mass loss. We call this a tipping point, as it would take decades to regrow a new, healthy snow layer that could buffer enough summer meltwater. In a warmer climate, this is highly unlikely to happen: rainfall will increase at the expense of snowfall, further preventing the formation of a new snow buffer.
Similar mechanisms are at play in the Canadian Arctic. Here, the downscaled SMB product reveals that while the northern ice caps, that are larger and more elevated, can still efficiently buffer meltwater in their extensive snow covered accumulation zones, the southern smaller and lower-lying ice fields have already lost most of their meltwater retention capacity, causing uninterrupted mass loss during the last six decades. These results suggest that the ice caps of Greenland and Canada will likely undergo irreversible mass loss. In the southern Canadian Arctic, the remaining ice fields may disappear within the next 400 years.
For now, the main Greenland ice sheet is still safe, as porous snow in the extensive accumulation zone, covering about 90% of the GrIS, still buffers most of the summer melt. At the current rate of mass loss, it would still take 10,000 years to melt the GrIS completely. However, the tipping point reached for the peripheral GICs must be regarded as an alarm-signal for the GrIS in the near future, if temperatures continue to increase.
To better resolve surface processes in RACMO2, we regularly update the model physics and evaluate the resulting simulated climate and SMB using a combination of in situ and remote sensing measurements. The latest model version, RACMO2.3p2, generally improves on previous versions. This model version will be used to conduct future climate scenario projections in forthcoming studies, essential to identify processes causing longterm GrIS mass loss and estimate Greenland’s contribution to the projected sea level rise.
