In a work accepted for publication in Journal of Geophysical Research, they have introduced new and important components in the energy balance of ECHAM5 GCM, which reduce the sea ice albedo (portion of the incoming solar radiation which is reflected by the surface), particularly in summer, and particularly in the Arctic, and thereby affect the sea ice extent and thickness in the Arctic.

Snow and sea-ice covered surfaces have a high albedo and are important for the climate whenever large areas are exposed to significant solar energy. Sea ice is particularly sensitive to moderate temperature changes as a warmer climate may melt the ice and expose larger areas of open water, resulting in most of the incoming sunlight being absorbed, which again leads to further warming. This amplifies the warming and creates a positive feedback. The sea-ice albedo feedback is important for the energy balance in GCMs. Traditionally, GCMs have treated high-latitude cryospheric processes quite crudely. Most schemes are very simplistic, depending only on surface type and temperature. A few schemes include snow depth and ice thickness, and even fewer include spectral and solar angle dependencies. Also, many GCMs use the sea-ice albedo as a tuning parameter. Previous studies have shown that todays GCMs are unable to capture the annual cycle of sea-ice albedo, particularly in summer where the GCMs overestimate the albedo.

A correct representation of sea-ice albedo in GCMs is necessary to incorporate the physical processes involved in the formation and melting of snow and sea ice. It has previously been shown that more advanced schemes allows for larger albedo feedbacks. During the northern hemisphere summer, solar radiation melts the snow and the upper surface of the sea ice. This produces melt water which later transforms into melt ponds on the ice. These melt ponds substantially reduce the surface albedo and absorb two to three times more solar energy compared to thick bare sea ice. The spatial distribution of melt ponds depends on the topography of the snow and sea ice. First year ice (FYI) tends to be smoother than multi year ice (MYI), and melt ponds on FYI are normally less deep, but cover a larger area. On the rougher MYI the melt ponds form in depressions, and tend to be smaller, deeper and more numerous.

Christina Pedersen and her colleagues propose a new, more physically based sea-ice albedo parameterization scheme for the ECHAM5 GCM. The new scheme separates between four surface types (snow covered ice, bare ice, melt ponds and open water) and determines the albedo and fraction of these types separately. This is the first time a physical description of melt pond albedo is included explicitly in any GCM. Despite the relatively short duration of melt ponds, they contribute significantly to the energy balance because the albedo effect on the climate is largest in summer.

Simulations are done by comparing the new albedo scheme, with the original temperature dependent albedo scheme in ECHAM5. Compared to the old albedo scheme (CTL), the new scheme (ALB) was found to reduce the sea-ice albedo both in winter, due to snow aging, and in summer, due to melt ponds. In addition, the spring decay and autumn increase of sea-ice albedo were captured in a more realistic way in the new scheme.
The new albedo parameterization scheme simulated the annual cycle of sea-ice albedo in a realistic way, capturing the important changes that determine the onset of melt, the duration of melt, and the start of the fall freeze-up. The correlation coefficient between one year sea-ice albedo simulations (ALB) and one year observations at the SHEBA site was 0.83, where the ALB albedo was slightly higher than the observed albedo in summer. The SHEBA albedo may, however, represent a lower boundary for albedo during summer.

The new albedo scheme performed well in modeling the coverage of melt ponds both when looking at the temporal evolution and the mean area covered. For the entire NH, the coverage was on the low side compared to observations, but it gave for the first time an estimate of the spatial variability of the melt pond coverage. The reductions in albedo were predominantly in the FYI areas, as melt ponds tend to cover larger areas on FYI. The melt ponds tend to be shallower on FYI, but the many shallow melt ponds on FYI reduced the albedo more than the fewer deeper ponds on MYI. There was large interannual variability in the ECHAM5 model. This implies there were years when hardly any melt ponds formed, while in others the formation of melt ponds started already in early June and reached depths up to 0.5 m in August. The maximum fraction of melt ponds was between 26% and 48%.

The overall effect on the sea-ice albedo was largest in summer, with average reductions of 23%, or 0.14, in the northern hemisphere in August. In the southern hemisphere the overall effect was smaller. The effect of reduced sea-ice albedo was overall reduced sea-ice thickness, concentration, and volume, with large temporal and spatial variations.

This is based on the paper “A new sea ice albedo scheme including melt ponds for ECHAM5 GCM” by Christina A. Pedersen and others, in press (JGR).

Source and contact: Christina Pedersen, NPI (christina.pedersen@npolar.no)

Read in other sources: Dammer kan forklare smelting (NPI nyheter);

<< Return

Melt ponds on the sea ice surface (photo: Sebastian Gerland, NPI).

Arctic sea ice extent from observations (thick red line) and from 13 IPCC 4AR climate models, and multi-model ensamble mean (solid black line) and standard deviation (dotted black line). The inset shows 9 years running mean. From NSIDC and NASA, 2008.

Measurements of albedo over a refrozen melt pond in the Fram Strait in autumn 2007. The melt ponds have a distinct darker color than the surrounding snow, and absorb more of the incoming solar radiation (photo: Stephen R. Hudson, NPI)

Christina Pedersen during albedo measurements on sea ice using a Field Spec Pro spectrometer (350-2500nm) in Fram Strait, spring 2005 (photo: Christina Pedersen, NPI)

50 years model simulations giving the sea ice albedo in August from the new albedo scheme (ALB) to the left and the difference between the new and old albedo scheme (ALB-CTL) to the right.