Snowclimate feedback

The impact of snow cover on surface air temperatures is well established. Snow cover is thought to influence surface temperatures in several ways: the high reflectivity of fresh snow cover can increase the surface albedo by 30-50%; fresh snow cover acts as a thermal insulator due to its low thermal conductivity; and melting snow is also a sink for latent heat since large amounts of energy are required for melt (Cohen, 1994).

The most widely studied climate impact of snow cover is the reduction of surface air temperatures mostly due to the high albedo of snow cover (Namias, 1960, 1962, 1985; Wagner, 1973; Dewey, 1977; Walsh et al., 1982, 1985; Klein, 1983, 1985; Klein and Walsh, 1983). Early empirical studies found that anomalous snow cover can cool surface temperatures by about 5 °C. Later modeling results confirmed empirical calculations that the presence of snow-cover reduced surface temperatures; however, the amount of cooling varies greatly in those studies. A modeling study by Walsh and Ross (1988) concluded that cooling by anomalous snow cover can be as large as 10 °C while Cohen and Rind (1991) found cooling by snow cover to be only on the order of 1-2 °C. They attributed the reduced cooling to the inclusion of net warming from sensible and latent heat fluxes in the surface energy balance. For the most part snow-cover parameterizations have concentrated on realistically capturing the high albedo of snow cover; nevertheless, other important thermodynamic properties of snow cover may not be well represented in current GCMs. These properties include the high thermal emissivity and low thermal conductivity of snow cover and processes associated with melting snow, including evaporation and runoff.

By computing and comparing the snow feedback parameter in 17 different GCMs, Cess et al. (1991), showed that the snow-climate feedback can vary significantly among numerical models. The snow feedback parameter is a proxy to measure the impact of snow cover in a climate change scenario, where climate change was defined as a warming of SSTs by 4 °C uniformly across the globe. All GCMs in the experiment ran two climate change scenarios: one where snow cover remained fixed at the coldest SST run, and the second where snow cover was allowed to respond freely to the SST warming of 4 °C.

The response of the GCMs to snow-climate feedback varied considerably from strongly positive (where decreased snow cover increased global warming) to slightly negative (where decreased snow cover modified or damped global warming). However when only clear sky regions of the GCM were included in the snow-climate feedback, all GCMs had at least a small positive feedback (see Fig. 4.17). The authors concluded that the direct impact of snow cover on the incoming solar radiation is to reduce the amount absorbed by the earth and is therefore a net cooling forcing on the climate. However cloud feedbacks and changes in outgoing longwave radiation could enhance or moderate the cooling effect of the high albedo of snow cover. Randall et al. (1994) confirmed the results of Cess et al. (1991) by looking at the snow forced response of 14 atmospheric GCMs; here too a wide variation of responses occurred. Randall et al. (1994) found, however, that two GCMs had a negative feedback even under clear sky conditions. Both studies demonstrate that as more GCM groups examine the impact of a retreating snow cover on the future climate in a global warming scenario, the results will be highly varied (but the

Figure 4.17. (a) The snow feedback parameter X/Xs for 17 GCMs for both global (•) and clear (o) designations. (b) Values of SRR/G (a measure of the snow radiative response) from the 17 GCMs and for both global and clear designations. Taken from Cess et al. (1991).

predominant response should be an enhanced warming with high-latitude or polar amplification).

4.5.3 Climate change scenarios

The consensus among scientists is that global surface air temperatures have been increasing since the beginning of the industrial revolution (Folland et al. in IPCC 2001). The warming has not been uniform though, with the greatest warming occurring in the mid- to high-latitudes of the Northern Hemisphere. Also the warming has been largest in the winter and spring and least in the fall (Easterling et al., 2000). What has been the role of snow cover on the observed warming and what may be its future role in a rapidly warming climate?

The easiest method of predicting future snow cover and its influence on a warming climate is to extrapolate from the most recent climate records. Accurate observations of snow cover from satellites have occurred over the past 35 years. Over that period the annual mean snow-cover extent has decreased. However, little trend or no trend is observed in the fall and winter with most of the decrease in observed snow cover occurring in the spring and summer. Groisman et al. (1994a,b) have attributed at least some of the spring warming in surface air temperatures to the decrease in

—I—I—I—I—I—I—I—I—I—I—I—I—I——I—r

Figure 4.17. (a) The snow feedback parameter X/Xs for 17 GCMs for both global (•) and clear (o) designations. (b) Values of SRR/G (a measure of the snow radiative response) from the 17 GCMs and for both global and clear designations. Taken from Cess et al. (1991).

Model number

Model number spring snow cover. But winter warming has also been large and yet observed trends in winter snow cover have not been significant even at the snow margins and low latitudes. Therefore interpreting the role of snow cover on surface temperatures from the observations is not straightforward and ultimately GCMs will need to be relied upon to predict the future role of snow cover in a doubled CO2 climate.

Many numerical experiments pertaining to global climate change due to greenhouse gases have been conducted and documented; however, the impact on snow cover has not been extensively explored. A possible explanation for why few studies exist that model future snow cover is that both air temperature and precipitation need to be accurately simulated to correctly reproduce snow cover at relatively small regional scales. This is probably still beyond the capabilities of current GCMs given all the inherent uncertainties. However, as GCMs are improved, more reliable experiments will be carried out and a better understanding of future snow cover should develop. Still, some limited literature exists and it will be discussed below.

Some early GCM studies tried to determine changes in soil moisture in a warmer climate. One soil moisture study was by Kellogg and Zhao (1988), who compared soil moisture over North America for winter and summer from five different GCMs. They found that during the winter, soil moisture increased over higher latitudes. It can be inferred from this that more precipitation fell in liquid rather than solid form and that more existing snow cover melted in a doubled CO2 climate compared to the present climate. From their results the most probable conclusion is that snow cover would be less even at higher latitudes during winter in the warmer climate.

Another similar study that explored changes in soil moisture in an increased CO2 climate was by Manabe and Wetherald (1987). In the paper, they compare the results from two GCM experiments; the GCM used is the Geophysical Fluid Dynamics Laboratory (GFDL) GCM. They conducted two types of experiments: a control run (atmospheric CO2 concentration at 300 parts per million (ppm)) and a perturbation experiment with CO2 increased to up to 600 and 1200 ppm. Snow cover is discussed in so far as it is relevant to soil moisture. As might be expected in a warmer climate, precipitation remains liquid longer during the fall and winter season and snow melt commences earlier in the winter and spring (resulting in an earlier date at which snow cover has completely disappeared) when compared to the control climate. Despite greater precipitation in the winter (due to a warmer and more moist atmosphere), total snowfall in the winter is less, so the duration and quantity of snow on the ground is less in the increased CO2 modeling experiments when compared with the control climate.

In contrast, a study of a future doubled CO2 climate by Ye and Mather (1997) found that global snow mass will increase in the polar regions (poleward of 60°N). They compared temperature and precipitation differences among three GCMs using current and doubled CO2 values. The three GCMs compared are: the Goddard

Table 4.2 Total changes in water equivalent in snow (1012). Taken from Ye and Mather (1997).

GFDL

GISS

UKMO

North polar region

-38.14

-3.27

7.55

South polar region

500.08

925.20

1783.08

Total north and south

461.94

921.93

1790.63

Institute for Space Studies (GISS) GCM, the GFDL GCM and United Kingdom Meteorological Office (UKMO) GCM. In the UKMO GCM the liquid water equivalent (LWE) of snowfall increased in both the northern and southern polar regions. While in the GISS and GFDL GCMs, the LWE of snowfall only increased over Antarctica and decreased in the Northern Hemisphere's polar regions. The only regions in the Northern Hemisphere to show a net gain in snowfall were central and northern Greenland. However the total LWE of snowfall for both hemispheres increased in the doubled CO2 simulations for all three GCMs (see Table 4.2). They estimate that the increase in total annual flux of water out of the oceans onto the polar ice caps for doubled CO2 to be about one-tenth the magnitude of the current exchange of water from the oceans to the land from autumn (maximum sea level) to spring (minimum sea level). The increase of snow mass in the high latitudes agrees with an observational study by Miller and de Vernal (1992). They studied climate proxy data for the past 130000 years before the present day and concluded that the conditions most favorable for glacial inception are climate conditions similar to the present climate. They conclude that the elevated winter temperatures predicted at high latitudes for doubled CO2 will increase snowfall rates in the Canadian and Russian land areas adjacent to the Arctic Ocean. With increased snow mass and assuming negligible increases in summer temperatures, conditions are more favorable for ice sheet growth in a warmer climate.

A paper by Cohen (1994) briefly discusses different scenarios for snow cover in a doubled CO2 climate. Using a control simulation and a doubled CO2 simulation of the GISS GCM, Cohen found snow cover and snow mass to decrease uniformly across the Northern Hemisphere for all 12 months of the year. Snow cover decreases by nearly 30% and snow mass even more, by close to 40%.

More in-depth GCM studies include those by Boer et al. (1992) (see Fig. 4.18) and Essery (1997a) (see Fig. 4.19). Boer et al. (1992) compare simulations for the current climate and a doubled CO2 climate using the Canadian Climate Center (CCC) GCM. They separately compare permanent and seasonal snow cover between the two climate simulations. Their findings for the CCC GCM is consistent with that of Ye and Mather (1997) for three different GCMs. Total snow mass

Figure 4.18. The left panel shows Northern Hemisphere winter (DJF) distribution of simulated seasonal snow mass for doubled CO2. The right panel shows the change in the position of the simulated snow line between the control (fine shading) and the doubled C02 cases (coarse shading). The snow line is denoted by values of snow mass greater than 10 kg m~2. Taken from Boer et al. (1992).

Figure 4.18. The left panel shows Northern Hemisphere winter (DJF) distribution of simulated seasonal snow mass for doubled CO2. The right panel shows the change in the position of the simulated snow line between the control (fine shading) and the doubled C02 cases (coarse shading). The snow line is denoted by values of snow mass greater than 10 kg m~2. Taken from Boer et al. (1992).

90N 60N 30N

Snow mass change (kg/m2)

Snow mass change (kg/m2)

10 100

10 100

Temperature change (K)

Temperature change (K)

90N 60N 30N

Snowfall change (kg/m2/month)

Snowfall change (kg/m2/month)

90N 60N 30N

Figure 4.19. Differences between 2030-2050 averages from the climate-change simulation and 130-year averages from the control of the Hadley GCM for (a) snow mass (b) temperature and (c) snowfall. Taken from Essery (1997a). (Plate 4.19.)

Figure 4.19. Differences between 2030-2050 averages from the climate-change simulation and 130-year averages from the control of the Hadley GCM for (a) snow mass (b) temperature and (c) snowfall. Taken from Essery (1997a). (Plate 4.19.)

for permanent snow cover for the globe increases for doubled CO2. However the results differ when the two hemispheres are compared separately. In the Northern Hemisphere over Greenland there is a dramatic decrease in the snow accumulation rate, which would probably result in a depleted ice cap. In Antarctica, snow-cover accumulation increases and therefore would probably produce a mass increase in

the southern ice cap. Taken together the increase in snow mass over Antarctica more than compensates for the decrease in Greenland.

Boer et al. (1992) also explored changes in the snow-cover extent in addition to snow mass. They found that seasonal snow cover consistently decreases in the doubled CO2 climate for both hemispheres. In the Northern Hemisphere, winter snow is shallower and the snow line retreats poleward. The total change in winter snow cover is a reduction of 8.3 x 1012 km2 or about 20% (see Fig. 4.19). Currently in winter about 50% of the Northern Hemisphere's land surface is snow covered. Therefore in a doubled CO2 climate only 40% of land surface will be snow covered in winter. In summer the total decrease in snow cover is 6.1 x 1012km2 orabout50%.

Essery (1997a) (see Fig. 4.19) used the Hadley Centre GCM to look at the impact of climate change on seasonal snow cover. In a GCM simulation where anthropogenic aerosols and CO2 are gradually increased, surface temperatures over the Northern Hemisphere land masses warmed by about 2 °C, with high-latitude amplification. Total precipitation increases but a larger percentage of it falls in the liquid form. Snow mass generally decreases over North America, Europe, and Western Asia with the largest decreases occurring over the northern Rockies, Alaska, and Scandinavia. In contrast, snowfall does increase in parts of the Canadian Arctic and Eastern Asia. The greatest retreat of snow cover occurs over North America and Europe and less so over Asia. Essery attributes this difference to a larger percentage of shallow snow cover (defined as snow cover with a depth between 1 and 10 cm) in North America and Europe compared to Asia. He concludes that based on the GCM simulation, global warming will reduce a greater amount of North American snow cover than Eurasian snow cover.

Frei and Gong (2005) (see Fig. 4.20) studied decadal trends in North American snow extent for the twentieth and twenty-first centuries among coupled atmosphere-ocean GCMs participating in the IPCC fourth assessment report. They found that compared with the observed snow extent most GCMs underestimate the North American snow-cover extent. They also found that there is no temporal correlation in decadal scale variability among the individual models or the observations and the models, individually or combined as a mean. In the twenty-first century they found a robust decreasing trend in the North American snow-cover extent that is statistically significant above the 99% confidence level. They conclude that the coupled models predict a significant decrease in snow extent, which can be used as an indicator of anthropogenically forced climate change. Table 4.3 summarizes the main findings for snow cover in a doubled CO2 climate among the GCM studies discussed above.

Snow melt constitutes a large percentage of water used for consumption and irrigation. Runoff makes up 33% of the world's irrigation waters and can be as high as 100% (Rango, 1997). Given the importance of snow regionally, in addition

Table 4.3 GCM studies evaluating changes in snow cover in a doubled CO2 climate. The table lists the direction of change in snow-cover extent and/or mass for both hemispheres. If a region is not discussed, the column is left blank. Final column for the entire globe is for snow mass only.

Snow

Snow

Snow

Snow

Study

extent NH

mass NH

extent SH

mass SH

Globe

Kellogg and Zhao (1988)

Decrease

Decrease

Manabe and Wetherald (1987)

Decrease

Decrease

Boer etal. (1992)

Decrease

Decrease

Decrease

Increase

Increase

Cohen (1994)

Decrease

Decrease

Ye and Mather (1997)

Decrease

Decreasea

Decrease

Increase

Increase

Essery (1997a)

Decrease

Decrease

Frei and Gong (2005)

Decreaseb

a one GCM study shows an increase. b North America only.

a one GCM study shows an increase. b North America only.

Figure 4.20. Annual time series (thin line), overlaid with nine-year running means (thick line), of ensemble-mean January North American snow-cover extent, including both twentieth and twenty-first century scenarios, for nine available coupled atmosphere-ocean GCMs. Taken from Frei and Gong (2005). (Plate 4.20.)

to the global-scale numerical simulations of the impacts of climate change on snow cover, more finer-scale experiments need to be done when the appropriate skill can be achieved. Some initial studies have been completed concentrating on local variations in snow cover in mountainous terrain. One study by Giorgi et al. (1997) looked at the elevation dependency of surface climate under doubled CO2 conditions using a regional model of the Alps nested in a GCM. In the control simulation of their model, snow accumulates starting from an elevation of 800 meters in winter and 1100 meters in spring. While in the doubled CO2 simulation, snow is completely depleted except at the highest elevations. This also results in the high-elevation amplification of warming in the doubled CO2 simulation.

Similar results were found by Rango (1997) using a snowmelt runoff model for the Rio Grande basin. In general, it was found that for a doubled CO2 climate there was an increase in rain events during winter, resulting in more immediate runoff and less snow accumulation. Also more of the snowpack melts during the winter with a net result of a reduced snowpack at the beginning of the traditional snow melt season (April 1). In particular, on April 1, the average LWE of snow cover was reduced from 58.1 to 20.4 cm. The spring runoff peaks would be two to four weeks earlier. Therefore there is increased runoff during winter and in the early part of the snow melt season and reduced runoff during later parts of the snow melt season (June-July).

4.5.4 Conclusions

Snow cover is an integral component of the climate system, the importance of which is just starting to be realized. Snow cover cools the atmosphere, not only due to its high albedo, but also due to its high thermal insulation. Therefore, because snow cover experiences large spatial and temporal variations it can significantly alter the surface energy balance. In addition to the local climate forcings, recent studies have shown that thermodynamic anomalies forced locally by snow cover can be transported remotely through teleconnections and result in dynamic anomalies downstream in space and time (Watanabe and Nitta, 1998; Cohen and Entekhabi, 1999; Clark and Serreze, 2000).

Snow-climate feedbacks are even more important when considered in the context of global warming. Future warming is not predicted to be uniform across the globe but less in the tropics and greater at high latitudes, especially the high latitudes of the Northern Hemisphere. Most of the earth's snow cover exists in the same regions where global warming is predicted to be greatest. Therefore changes in future climate will significantly impact snow cover and in turn changes in future snow cover will have important feedbacks for predicted climate change. However given how difficult it is for models to simulate correctly the variability of observed snow cover and its feedback on the climate, it will continue to be a challenge for current GCMs to accurately predict the role of snow cover in a rapidly changing climate system.

As discussed earlier, GCMs do not correctly simulate snow-cover variability as observed and even lack consistency among themselves. Similarly, snow-climate interactions as parameterized by GCMs exhibit a wide dispersion, while the observed snow-climate interactions are not well defined. Therefore it is not surprising that the early diagnosis of changes in snow cover and snow mass in the coming decades is conflicting. Future snow-cover extent has been universally modeled to recede, both towards the poles and the tops of mountains, as greenhouse gases continue to increase in the twenty-first century. However the impact of a more limited snow extent on future lower troposphere air temperatures shows large variability. The range varies from significantly amplifying warming at mid-high latitudes to slightly dampening warming at these same latitudes. And even though snow cover is expected to decrease, models conflict as to whether total snow mass will also decrease or may actually increase (even resulting in a drop in sea level). How future snow cover and snow mass responds and its role in amplifying or moderating a changing climate is a challenge for climate and snow modelers, which will require increased collaborative effort by the hydrologic and atmospheric community to resolve. Our confidence in future climate projections depends on correctly simulating snow-cover variability, its interactions with the climate system, and its response to increased greenhouse gases.

Renewable Energy 101

Renewable Energy 101

Renewable energy is energy that is generated from sunlight, rain, tides, geothermal heat and wind. These sources are naturally and constantly replenished, which is why they are deemed as renewable. The usage of renewable energy sources is very important when considering the sustainability of the existing energy usage of the world. While there is currently an abundance of non-renewable energy sources, such as nuclear fuels, these energy sources are depleting. In addition to being a non-renewable supply, the non-renewable energy sources release emissions into the air, which has an adverse effect on the environment.

Get My Free Ebook


Post a comment