Info

Fig. 15.8 HadCM3 projected Arctic sea ice extent for the 21st century under various IPCC scenarios. The model predicts a continuous decline in sea ice cover towards the end of this century and by 2080 the Arctic will become ice free in the autumn under certain IPCC scenarios

15.5 Greenland Ice Sheet

Ice sheets play a dynamic role in Earth's climate system, influencing regional climate and global sea level and responding to climate change on time scales of millennia. The Greenland ice sheet holds enough freshwater to raise the oceans seven metres if it all melts. An imbalance between new ice formed from falling snow and melting will have an important impact on the Arctic hydrological budget.

Climate models forced with IPCC projected greenhouse gas concentrations predict that the Greenland ice sheet is likely to lose its stable state, although a warmer climate comes with more precipitation. For an annual average global warming of more than 3.1 ± 0.8 K and 4.5 ± 0.9 K regional warming over Greenland, the net surface mass balance of the Greenland ice sheet becomes negative in the AR4-scenarios (Gregory and Huybrechts 2006). Greenhouse gas concentrations will probably have reached levels before the year 2100 that are sufficient to raise the temperature past this warming threshold (Gregory et al. 2004). The most extreme scenario considered in the third assessment report (TAR) of IPCC involves a warming of 8 °C over Greenland, in which case most of the ice sheet will be eliminated within the next 1,000 years (IPCC 2001). Toniazzo et al. (2004) have shown that the loss of Greenland ice sheet is possibly irreversible. By simulating the pre-industrial climate without the Greenland topography, which is equivalent to a cut of greenhouse gas concentrations to the pre-industrial level after the elimination of the Greenland ice sheet, they are not able to generate a long-term snow accumulation over Greenland. On the other hand, Lunt et al. (2004) using a higher-resolution model suggest that reglaciation may also be possible.

Many recent observations indicate increased ablation of the Greenland ice sheet. Repeat-pass airborne laser altimetry measurements indicate that Greenland is losing ice at a rate of -80 ± 12 km3/year during the period 1997-2003, mostly from the periphery (Krabill et al. 2004). A more recent study based on satellite interferometry (Rignot and Kanagaratnam 2006) suggests that the tide-water glaciers (those grounded below sea level) are accelerating and the resulting dynamical loss may have increased the mass imbalance to -224 ± 41 km3/year in 2005 compared to an estimate of -91 ± 31 km3/year in 1996 using the same method. Based on gravity measurements from the GRACE satellite mission, Chen et al. (2006) have estimated a total ice melting rate of -239 ± 23 km3/year for the period from April 2002 to November 2005. While other studies have suggested that the interior may be stable, Johannessen et al. (2005) have shown an increase of 6.4 ± 0.2 cm/year in the vast interior areas above 1,500 m, using altimeter height data from European Remote Sensing satellites (ERS-1 and ERS-2) for the period 1992-2003. They have also confirmed the thinning of ice sheet margins below 1,500 m (-2.0 ± 0.9 cm/ year), but on average they find 60 cm increase over 11 years for the area studied. Because the measurement does not completely cover the marginal areas, it is not possible to know the integral change. The short observational records of some satellite measurements must be treated with caution as they are unable to account for the considerable natural variability in surface mass balance (Hanna et al. 2002).

Characterising the response of ice sheets, to various degrees of climatic forcing, is usually conducted using an off-line model with idealised temperature and precipitation. Such models are used because of the need for high spatial resolution at the steep ice sheet margins, a resolution not available from the coarse resolution climate model grids. A significant step forward has been to implement a fully coupled high resolution (20 km) Greenland ice sheet model within the Hadley Centre climate model HadCM3 (Ridley et al. 2005). Figure 15.9 shows the state of the Greenland ice sheet at various stages in a stabilisation experiment at 4 x CO2. The Greenland ice sheet would be half its size in 850 years (3.5 m of sea level rise) and reduce to a small ice cap in the eastern mountains within 3,000 years. Local climate feedbacks indicate that the ice sheet will melt slower than expected by offline simulations with the same forcing.

Year 0 Year 270 Year 710 Year 1130 Year 1760

Volume 100% Volume 80% Volume 60% Volume 40% Volume 20%

Year 0 Year 270 Year 710 Year 1130 Year 1760

Volume 100% Volume 80% Volume 60% Volume 40% Volume 20%

0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 3000 Bedrock altitude (m) Ice thickness (m)

Fig. 15.9 The state of the Greenland ice sheet during various stages of its decline under 4xCO2 simulated by the HadCM3 model coupled with a high-resolution ice sheet model (After Ridley et al. 2005; figure courtesy of Jonathan Gregory)

0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 3000 Bedrock altitude (m) Ice thickness (m)

Fig. 15.9 The state of the Greenland ice sheet during various stages of its decline under 4xCO2 simulated by the HadCM3 model coupled with a high-resolution ice sheet model (After Ridley et al. 2005; figure courtesy of Jonathan Gregory)

The reversibility of the decline has been further investigated by reverting to pre-industrial levels of CO2 during various stages of the ice sheet decline. These experiments, conducted off-line, have indicated that the ice sheet can return (over ~10,000 years) to its current state if it has not lost more than 30% of its mass. It is unlikely to recover, unless temperatures are colder than under pre-industrial concentrations of CO2, if 70% or more of the ice sheet has been lost. Stable intermediate ice sheets may form, during the climatic cooling, for ice sheets between 30% and 70% of the current mass.

Increasing meltwater flux from the Greenland ice sheet could potentially accelerate the weakening of the THC under global warming conditions. Depending on the scenario, model studies show flux rates between 0.01 Sv (Fichefet et al. 2003) and 0.1 Sv (Ridley et al. 2005) that would dilute the North Atlantic surface waters and potentially suppress deep water formation. However, the simulated THC weakening appears to be highly model-dependent. Fichefet et al. (2003) found a strong and abrupt weakening of the AMOC at the end of the 21st century. In contrast, Ridley et al. (2005) analyzed a climate with four times the pre-industrial CO2 level and found relatively minor changes in the THC. Jungclaus et al. (2006b) calculated meltwater flux rates from IPCC A1B scenarios and found a flux of about 0.03 Sv by 2100. In a sensitivity experiment this additional meltwater input caused only a slight acceleration of the THC weakening.

15.6 Freshwater Content Changes

Observational estimates indicate an average Arctic Ocean fresh water storage of 100,000 km3 with respect to an Arctic mean reference salinity of 34.8 psu. Most of this is stored in liquid form. Solid sea ice is thought to contribute roughly in the order of 30,000 km3. The largest source comes from the combination of surface flux and river discharge, followed by an inflow of relatively low salinity water through the Bering Strait (see Fig. 15.1). The largest individual sink is Fram Strait sea ice export. Recent studies, however, indicate that the CAA outflow might be of the same order (Prinsenberg and Hamilton 2004). From time to time, large pulses of freshwater (in the order of several thousand km3/year) move from the Arctic to the North Atlantic in the form of the so-called Great Salinity Anomalies (GSA, Dickson et al. 1988; Belkin et al. 1998), leading to substantial changes in regional surface salinity and freshwater storage (Curry and Mauritzen 2005; Peterson et al. 2006). There have been various modelling studies using ocean-only or ocean-sea-ice models forced with atmospheric reanalysis data to investigate driven mechanisms of Arctic freshwater content changes (e.g. Zhang and Zhang 2001; Zhang et al. 2003; Hakkinen and Proshutinsky 2004 and Karcher et al. 2005), but here we mainly focus on modelling efforts with fully coupled climate models.

Table 15.1 shows the Arctic mean freshwater budgets and Table 15.2 the volume transports for the 20th, 21st and 22nd centuries simulated by the ECHAM5/MPIOM model (Haak et al. 2005; Koenigk et al. 2007). Terms are averages for each time window and error bars are standard deviations of the annual means. Percentages

Period

Was this article helpful?

0 0

Post a comment