Seaice Variability During 20032008

The Russian edition of this monograph published in 2007 presents analyses of sea ice data dating to 2003. For the English edition of this monograph, we have extended our studies to analyze significant and sometimes extreme (2007) changes observed in the Arctic since 2003, and we decided to extend our studies and explain these changes based on previously discussed theories and hypotheses. Some of the published works were not reflected in the Russian edition of the monograph and the authors have tried, as far as possible, to fill this gap. In this section we aim to investigate the cyclic variability of multiyear changes in the various Arctic climate system hydrometeoro-logical parameters, in order to assess their reliability as well as their role in the total dispersion of ice extent. Cyclic variability data on scales from decades to centuries (in the Holocene) and tens of millennia (late Pleistocene) are now available for a number of Earth's climate indexes (Monin and Sonechkin, 2005).

As discussed in the second chapter of this monograph, changes in the ice extent of the Eurasian Arctic seas in the twentieth century exhibited a long-term negative trend accompanied by cyclic variations with periods of 50-60, about 20, 9-12, 7-8, and fewer years. However, the variability of sea ice extent in the western region

(Greenland to Kara Seas) was mainly influenced by a long-term trend and low-frequency cycles (about 60 and 20 years), whereas in seas of the eastern region (Laptev, East Siberian, and Chukchi), the influence of higher-frequency cycles prevailed (10 years and less).

Studies by many scientists confirm the above cycles in interdecadal variability of ice cover area and other characteristics of Earth's climate system. These studies include the following.

Alekseev et al. (2003) reveal differences in the occurrence of warming and cooling epochs throughout the twentieth century as well as their association with atmospheric circulation. Monin and Sonechkin (2005) analyze large-scale climatic cycles related to alternation of glacial and interglacial epochs in Earth's history with special attention to 60- and 180-year cycles. The 180-year cycle, which we conventionally refer to as 200-year cycle, may be responsible for intra-century trends in climate change.

Klyashtorin and Lyubushin (2006) obtained 60-year oscillations in global surface air temperature based on instrumental measurements for the last 140 years. Using mean annual air temperature reconstructed from the oxygen isotope O18 concentration in ice cores from the Greenland glaciers for the last 1500 years, these authors carried out a spectral analysis that supported the existence of 60-year and 200-year cycles in climate changes. The results of these studies suggest that natural cycles, rather than greenhouse gases, may be the dominant factor in variability of the Earth's climate.

Proshutinsky and Johnson (1997) and Polyakov and Johnson (2000) present studies of changes in the Arctic Ocean regime associated with the Arctic Oscillation (AO), including cycles lasting 10-20 years. Minobe (1997) provides the results of studies of air and water temperature changes for different regions of the Pacific Ocean, North America, and the part of the Indian Ocean that adjoins the Asian continent from the south. This author concludes that in all the above regions, a characteristic feature of climate change is cyclic oscillations that last for 50-70 years and that are associated with alternating warm and cold epochs. The oscillation phases in most of these regions virtually coincide with similar Arctic cycles. The periods of 1870-1889, 1925-1947, and 1977-1990 qualify as warm epochs in these regions, and 1890-1924 and 1948-1976 as cold epochs. The oscillation phase is opposite only in the western part of the Pacific Ocean (Japan), which is due to the influence of the rear part of the Aleutian Low, where atmospheric pressure varies in accordance with alternating warm and cold epochs. The Aleutian Low deepens in the warm epochs and is partly filled during the cold ones. As will be shown below, climate changes in the northern part of the Atlantic Ocean have similar features.

Our follow-up studies were aimed at revealing spatial-temporal features of approximately 60-year cycles in the Northern Hemisphere during the first and second warming epochs in the twentieth century. The differences in these epochs may be associated with a longer cycle that is evident in the linear trends of variations in different climate-system indicators. Figure 6.4a, b (see color section) shows the distribution of surface air temperature differences averaged for the winter and summer periods of 1980-2000 compared to 1930-1950 (Frolov et al., 2009). From the first to the second warming epoch, the temperature increased over Greenland, the mid-latitudes of Eurasia and North America, the near-Pacific Arctic, and the northern Pacific Ocean in winter, and over the Arctic basin, the Siberian shelf seas, the northern Pacific Ocean, Central Asia, and southern Siberia in summer. However, in the same period, winter air temperature dropped significantly in the North Atlantic and over a major part of the Arctic Ocean, including areas adjoining the Baffin Sea and the Arctic seas of the Eurasian shelf from the Barents Sea to the East Siberian Sea. As shown in Section 2.2, a decrease in air temperature was accompanied by growing ice extent in the Barents Sea in winter, which amounted (on average for October-February) to about 130,000 km2 in 50 years. Decreased air temperature in this period also occurred in the summer half of the year in the region west of Greenland, over the North Atlantic, Western Europe, eastern Siberia and the southern Asian continent.

Considering that cooling was recorded at the same time in the Antarctic (Gudkovich et al., 2008), it should be acknowledged that although "global warming" occurred as a global average, this was not uniform spatially or temporally, and cooling was recorded over large areas of our planet in both winter and summer during the last decades of the twentieth century. Note that the value of anomalies characterizing an epoch depends on the period for which the climatic "norm" is determined.

It was noted in Sections 4.1 and 4.2 that air temperature at mid and high latitudes primarily depends on dynamic processes in the atmosphere (Alekseev, 2000; Alekseev et al., 2003; Vorobiev and Smirnov, 2003). They influence air temperature due to both advective processes and the impact of cloudiness, which depend on the type of baric system in play. In winter, this influence is particularly high in areas where anticyclones are common. Weakening of anticyclones results in increasing temperature and cloudiness. Variation in cloudiness is one of the main causes of climate change is indicated by Sherstyukov (2008).

Maps showing differences in mean sea level pressure for the winter and summer periods (not given here) for 1980-2000 compared to 1930-1950 and characterizing changes occurring from the first to the second epoch of warming confirm the pattern mentioned above (Frolov et al., 2009). In the winter half of the year, atmospheric pressure over the Arctic during the last warming epoch was much lower than it was during the first warming. Pressure also dropped in the Siberian, Canadian, Greenland, and Arctic Highs. Due to the deepening and southward displacement of the Icelandic Depression (the most important atmospheric center of action) and a significant pressure increase over the North Atlantic, intense zonal transport in the atmosphere shifted from the high- to the mid-latitudes. This was a direct cause of a significant mid-latitude air temperature increase over the Eurasian and North American continents, where seasonal anticyclones are commonly located at this time of the year (Klimenko, 2007; Wallace et al., 1995). Temperatures dropped where thermal conditions were influenced by the rear areas of baric depressions (Baffin Sea, North Atlantic, Barents Sea) (Alekseev, 2000, 2004; Klimenko, 2007; Wallace et al., 1995).

A similar map characterizing the summer half of the year points to a major atmospheric pressure drop over the Arctic basin in the last warming epoch along with a slight increase over Eurasia and northern regions of the Atlantic and Pacific oceans. This created a favorable temperature background (for ice decrease) over the Eurasian Arctic seas and an unfavorable one over the northwest Atlantic Ocean, including the Baffin Sea.

Walsh et al. (1995) describe a major reduction in atmospheric pressure at sea level over the Arctic near the end of the 20th century. Thus, air temperature variations during the time period between the warming epochs can be accounted for by corresponding variations in the average pressure fields that characterize atmospheric dynamics. These changes correspond to climatic variations in the condition of polar (circumpolar) vortices (Dmitriev and Belyazo, 2006; Gudkovich et al., 2008). It is known that cyclonic rotation of the troposphere and the lower atmosphere from west to east around the poles is associated with polar vortices. In the lower layers of the atmosphere, over the Arctic Basin in winter, cyclonic vorticity changes its sign: an Arctic High forms here. In summer, a cyclonic field is commonly found near the surface (Dolgin, 1968; Anon. (E)). A somewhat similar pattern of general atmospheric circulation is observed in the Antarctic, through its different distribution of land and sea, as well as the presence of a thick Antarctic glacier, result in certain differences (Anon. (D)).

The intensity of circumpolar vortices varies within a year, driven by seasonal air temperature gradient variations between low and high latitudes; in the winter period in both hemispheres, atmospheric circulation intensifies, and in the 6-month summer period it weakens. The association between the state of polar vortices and air temperature is quite different in the climatic variability of warming and cooling epochs. In warming epochs, atmospheric pressure and geopotential values within the troposphere and the lower stratosphere decrease in the zone of polar vortices. This results in intensification of zonal flows in the atmosphere of mid-latitudes, which are apparent in indices of general atmospheric circulation, such as the North Atlantic Oscillation, the Arctic Oscillation, high-latitude zonation, and others in the Northern Hemisphere, and the South Polar Oscillation in the Antarctic. In cooling epochs, zonal flows become weaker (Gudkovich et al., 2008). It should be noted that the Arctic High weakens with an intensifying northern polar vortex; with weakening of the vortex the Arctic High strengthens (Dmitriev and Belyazo, 2006).

The patterns of variation in the intensity of zonal flows in the atmosphere of mid-latitudes described above are confirmed in Section 4.2. Particularly, variation in the mean annual zonality index is shown to express the difference in atmospheric pressure at sea level between 40°N and 65°N during the 20th century (Figure 4.9). In addition to a characteristic increase in the index from cold to warm epochs, the pattern reveals an intensification of zonal transport from the first to the second warming epoch due to the fact that the belt of intensified zonal transport in the atmosphere displaces from high to mid-latitudes as a result of the extension and the deepening of the polar vortex.

What causes these variations in atmospheric circulation?

Variations in circumpolar vortices may be caused by both external and internal factors. Among the internal factors, until recently, most climatologists placed major emphasis on the effect of accumulating anthropogenically generated greenhouse gases

(mainly C02) in the Earth's atmosphere. Section 5.1 provides reasons why the "greenhouse theory'' has weak foundations.

A series of papers by G. V. Alekseev and his co-workers examines the low-frequency cyclic oscillations of climate with a period of 60-80 years. In these papers, it is presumed that the first twentieth-century warming period was characterized by higher surface air temperatures in the near-Atlantic Arctic, and the second warming period exhibited higher surface air temperatures in the near-Pacific region and other latitudinal zones (Alekseev, 2003; Alekseev et al., 2003; Alekseev and Ivanov, 2003). Because anthropogenic emission of greenhouse gases to the atmosphere in the first warming epoch was far less, and became apparent only by the time of the second warming, the authors made a presumption: warming in the first half of twentieth century was caused by natural oscillations of the climate system, and the last warming "cannot be accounted for without regard for the anthropogenic factors.'' Greenhouse gases resulting from burning fuel and, partly, from emissions by volcanic activity (Katsov, 2003; Johannessen et al., 2004; Vinnikov et al., 1999) were recognized as such factors by those who carry out coupled models of the atmosphere and the ocean. Note that, based on temperature diagrams provided in IPCC reports (2001, 2007), the strongest recent volcanic eruptions only impacted the Earth's climate for a maximum of 3 years, and thus they cannot be the cause of climate changes on the scale of decades.

The Report of the Nongovernmental International Panel on Climate Change (NIPCC) (Singer, 2008) criticized the main IPCC conclusions regarding the intensification of anthropogenic global warming in recent years. The NIPCC report argues that the magnitude of global warming is essentially overestimated due to the influence of urban heat islands on measured surface air temperature. Rapp (2008) discusses the inadequacies of the surface temperature measurement system in some detail. Nevertheless, it cannot be argued that the Earth has not warmed significantly in the 20th century. A major problem for climate models is how to deal with putative increases in humidity resulting from increases in global temperature due solely to increased C02. Most models treat humidity as a global average, and since water vapor is a powerful greenhouse gas, this greatly amplifies the temperature increase due to increased C02. However, Lindzen (1997) emphasized that the degree of water vapor feedback as a heating force in any region depends on the absolute humidity. In desert regions with very low absolute humidity, an increase in humidity provides a significant heating force. However, in regions with high absolute humidity, an increase in humidity provides a very modest heating force. Tropical regions that already have high humidity, do not gain much additional heating from an increase in humidity.

Climate models assume that the main factor affecting the atmosphere is the greenhouse effect of carbon dioxide, but they do not account for the Earth's inter-decadal climate changes that affect the evolution of circumpolar vortexes discussed above. Moreover, Gudkovich et al. (2008) averaged the calculations for fields of atmospheric pressure from five models (HAD, CNRN, EHAM, GFDL, INM) and found that they significantly overestimate atmospheric pressure over the Arctic basin during climate warming periods. This contradicts the finding by Vize (1944b) and confirmed in subsequent years that air temperature and ice extent in the Arctic seas are fundamentally dependent on the degree of development of the Arctic High. The model calculations also contradict the observation that, as shown above, air temperature anomalies are primarily dependent on dynamic processes in the atmosphere. Errors in the calculated temperature would undoubtedly lead to major inaccuracies in model predictions of ice cover conditions and other climate characteristics.

In our opinion, a reliable argument contesting the decisive role of the anthropogenic factor in climate change is the decrease in winter air temperature over large regions of the Arctic. It is known that, in the winter season, long-wave radiation plays a decisive role in the heat balance of polar seas. Long-wave outgoing radiation could theoretically be affected by the concentration of greenhouse gases in the atmosphere, which strongly increased by the end of the twentieth century (IPCC, 2007). Nevertheless, there did not seem to be any effect on the atmosphere at high latitudes; instead of warming, cooling was recorded over vast spaces. Even accounting for the positive temperature anomalies recorded in the first decade of the twenty-first century, average air temperature in the second warming epoch was not higher than in the first one.

Water vapor played a role in the air temperature increase at the end of the twentieth century in regions where seasonal anticyclones occur in winter; note that water vapor has a greater influence on effective radiation of the atmosphere than greenhouse gases of anthropogenic origin. An increase in cloudiness and water vapor content in the atmosphere over continents (but not over desert continental regions) in this period was due to a decrease in atmospheric pressure, which caused more intense cyclonic activity. This is confirmed by a corresponding growth in river runoff (see Section 4.7). These factors have a major influence on global air temperature as does the larger area occupied by the mid-latitudes compared to the high-latitudes. As a result, unlike the pattern for air temperature in the Arctic, the first warming epoch is less prominent in global temperature records than the second one.

Short-wave solar radiation is the most significant summer-season forcing, or, more precisely, the part of it that depends on albedo and absorption by the ice cover and the sea. Due to changes in albedo not related to greenhouse gases of anthropogenic origin, this heat balance constituent can vary by several dozen W/m2 in polar regions, or one order of magnitude greater than the most optimistic assessments of the influence of greenhouse gases.

As an alternative to the "greenhouse theory'' as a main cause of climate change in the late twentieth and the beginning of the twenty-first centuries, the effect of solar activity (SA) on atmospheric processes attracts considerable attention. Section 5.2 provides a review of papers (mainly by Russian scientists) on the relationship in time between changes in ice extent and other characteristics of the climate system with SA parameters (mainly Wolf numbers). Luk'yanova (2007) offers interesting facts related to these issues that have largely been discovered by non-Russian scientists.

Satellite measurements have brought new understanding of the influence of SA on the total solar irradiance (TSI) to the Earth (Frolich and Lean, 1998) and its climate (Douglass and Clader, 2002). Gudkovich et al. (2005) suggest a positive linear trend in SA (Wolf numbers) in the twentieth century as a possible cause of corre sponding climate changes in the Arctic. In Figure 6.5, borrowed from Soon (2005), changes in annual air temperature anomalies north of 62oN (Polyakov et al., 2003) are compared with TSI values estimated by Hoyt and Schatten (1993), as well as with CO2 content in the atmosphere from 1875-2000. The variation of temperature matches the TSI curve far better than it matches the CO2 curve. However, the Hoyt and Schatten model for TSI is just one of many, and other models lead to very different patterns for TSI vs. year. Furthermore, climate modelers would argue that the temperature curve in the second warming epoch represents the continuation of the first warming epoch, interrupted by a period from about 1940 to about 1980 when increasing aerosol concentrations outweighed the effect of increasing greenhouse gases. Therefore, Figure 6.5 is just one representation of many that could be derived. Nevertheless, if Figure 6.5 were taken at face value, the temperature and TSI varia-

Figure 6.5.

Annual-mean Arctic-wide air temperature anomaly time series (dotted line) correlated with estimated total solar irradiance (solid line in the top panel) from the model by Hoyt and Schatten, and with the mixing ratio of atmospheric carbon dioxide (solid line in the bottom panel) (Soon, 2005).

tion charts would suggest the presence of both a positive "100-year" trend and quasi 60-year cyclic oscillations. This is also corroborated by the correlation coefficients between annual average air temperature in the Arctic, TSI, and SA anomalies given in Hoyt and Schatten (1993). With 10-year smoothing, the coefficients are 0.89 (with TSI) and 0.47 (with CO2). Hence, according to this particular model, the main cause of climate changes in the Arctic would be the dependence on TSI rather than buildup of greenhouse gases. However, there is no way to test the Hoyt and Schatten model, and other models for TSI exist with quite different results. While Figure 6.5 is suggestive, the fact remains that we really do not know how TSI varied prior to the advent of satellite measurements around 1980. Figure 6.5 demonstrates that the form of the variability of Arctic surface temperatures during the 20th century resembles the variability of the Hoyt and Schatten model for TSI. This is suggestive that variations in TSI may have been an important factor in 20th century climate change. Though the total variance of TSI from 1880 to 2000 according to Hoyt and Schatten was 3-4 W/m2, the simple spreading of this flow over the spherical area of the Earth is incorrect. As we show in this work, a significant part of TSI variance influences the high-latitude regions. Furthermore, as was noted in Section 5.4, Budyko (1969) concluded by calculations that solar constant variations of several tenths of % are sufficient to induce essential climate changes.

In seeking a relationship between solar variability and climate change, we may consider TSI and SA. The connection between TSI and climate is direct; TSI represents the fundamental heat input from the Sun that drives our climate. However, although SA represents fundamental aspects of the dynamics of the Sun, its connection to the total power emitted by the Sun is not quite clear. SA includes energetic particle emission, electromagnetic emission in the UV and higher frequency ranges and magnetic fields. It is manifested in the Earth's phenomena in the form of polar lights, magnetic storms, radio-communication blackouts, etc. A number of different indices are used to measure the level of SA, particularly sunspot indices (Wolf number, etc.), the intensity of solar wind, and various magnetic indices. Even though variations in TSI associated with changes in SA may be small, the impact on higher latitudes is significantly amplified by the interaction of charged solar wind particles with the Earth's magnetic field. As shown in our work, evidence exists that variability of SA is connected to Arctic climate variations. In addition, we have also shown in Section 5.4 that the interaction of the gravity fields of the Sun and the solar system planets ("dissymmetry of the solar system planets''), by which we mean a displacement of the Sun's center relative to the center of the system mass, can produce seasonal changes in the solar input to Earth, which would affect the climate of higher latitudes on a 60-year cycle.

The cause of a "100-year" trend in SA that may be associated with a 200-year cycle (Westbrook, 1998; Bashkirtsev and Mashnich, 2004; Raspopov, 2004) has not yet been reliably determined. However, the 60-year SA cycle (Fritz cycle) is probably due to the influence of "dissymmetry of the solar system'' (see Section 5.4), which changes the distance between the Earth and the Sun on a 60-year cycle (Gudkovich et al., 2005). Over the longer term, if we can confirm existence of a 60- year cycle in TSI, this might confirm the theory, and would also provide a basis for explaining the opposite signs found in the 60-year climate cycles of the Arctic and the Antarctic.

Monin and Sonechkin (2005) provide some support for such an explanation of climate variations. They consider the cause of 60-year climate variability to be a triple cycle of solar magnetic activity (the Hale cycle) that is shown to last about 60 years by wavelet analysis of a number of hydrometeorological parameters. "In such a triple loop, the Sun follows a trajectory around the center of inertia in the form of a slightly unlatched trefoil, always behind the center of inertia at a distance of slightly more than the diameter of the Sun'' (Monin and Sonechkin, 2005, p.16). The same complex quasi-periodic motion also includes a cycle that averages 179 years in length. It is related to "... variations of the solar radiation incoming to the Earth, which change in many respects due to the gravitational interactions between the Sun and planets, especially Jupiter and Saturn (Monin and Sonechkin, 2005, p. 43).

Isotope analyses of ice cores drilled from glaciers in the Antarctic and Greenland using radionuclides of cosmic origin (14C and 10Be) allowed reconstruction of Wolf numbers as far back as the middle of the ninth century ad (Usoskin et al., 2003; Solanki et al., 2004). Figure 6.6a (see color section) presents the results of several versions of this reconstruction. Although there is considerable variance from model to model, the models all suggest that sunspot numbers appear to have been lower on average for a thousand years prior to the 20th century. While we do not have reliable models to connect sunspot activity to TSI and climate change, the fact that sunspot activity appears to have increased significantly in the 20th century suggests that climate change in the 20th century may, at least in part, be related to solar changes.

There have been many attempts to estimate the historical surface temperatures over the past few hundred years or in a few cases, as far back as two millennia. Studies based on temperature proxies (tree rings, ice cores, coral terraces, pollen counts, etc.) have estimated temperatures in various regions for various time periods. Mann et al. (1998, 1999,2003, 2004,2008) attempted to integrate the entire array of proxies into a single cohesive reconstructed estimate of the global average temperature over the past two millennia. While the results of such reconstructions by Mann and others vary considerably from study to study (see Figure 6.6b, see color section), they all lead to a common morphology of a temperature profile for two thousand years, followed by a sudden sharp rise in the 20th century. There is moderate evidence of the so-called "Medieval Warm Period'' or a "Little Ice Age'' in these results. It should be noted that the red curve at the far right of Figure 6.6b (see color section) (CRU instrumental record) is grossly exaggerated in vertical height. The figures indicate a global temperature rise of 1.3°C in the past century—about double the accepted value. This result has served the needs of global warming alarmists who view the sudden temperature rise in the 20th century after two millennia of small variations, as evidence of the anthropogenic impact on climate change. McIntyre and McKitrick (2003, 2005, 2006, 2007) and Wegman, Scott and Said (2006) found errors in the data reduction procedures used by Mann et al., and dubbed the temperature profile obtained by Mann et al. in derisive terms as the "hockey stick.'' Rapp (2008) describes this controversy in considerable detail. While Figure 6.6b (see color section) appears to underestimate the climate variations associated with the so-called "Medieval Warm

Period" or a "Little Ice Age" and it exaggerates the rise in the late 20th century, nevertheless, the comparison of Figures 6.6a and 6.6b (see color section) is suggestive that there may be a solar connection to long-term climate change over the past millennium.

The data in Appendix A show that the cover area of sea ice decreased during the period from 2000 to 2008. A small positive ice extent anomaly was recorded only in the eastern region during two of the nine years (2001 and 2004). The year 2007 appeared to be the warmest, when the maximum air temperature, the minimum ice extent of the Arctic seas, and other extremes of the observation series were recorded. The highest anomalies were reported from the East Siberian and Chukchi Seas (Frolov, 2007; NCDC, 2007), i.e., from the region where the role of short-term fluctuations is significant (see Section 2.3).

The reviews referenced above and a paper by Dmitriev (2007) consider the hydrometeorological features of 2007 in detail, and their findings include the following phenomena. Zonal circulation in the atmosphere of the Arctic was abnormally high in 2007. Negative anomalies of surface atmospheric pressure in the Eurasian sector and positive anomalies in the American sector resulted in advection of warm air masses from the Pacific Ocean to the Arctic, which remained stable throughout most of the year. A positive anomaly of average annual surface air temperature in the Eurasian sector of the Arctic and in the zone north of 70°N reached 2.5-2.8°C (relative to the period 1961-1990).

These 2007 atmospheric processes created exceptionally favorable conditions for ice cover destruction in the Arctic seas studied herein, as well as in adjoining areas of the Arctic basin. Late onset of ice formation in the autumn of the preceding year, a low rate of ice cover formation, and intensified ice removal from the seas to the Arctic basin and further (to the Greenland Sea) in the winter of 2006/2007 resulted in the following: by the spring of 2007, the ice cover in these seas was mainly composed of first-year ice of decreased thickness but with inclusions of younger ice that formed in extensive flaw polynyas. Early onset of melting under these conditions resulted in rapid destruction of the ice cover, and stable drifting promoted ice removal beyond the boundaries of the seas. This was particularly evident in seas of the eastern region, which were in the zone of high baric gradients between the Arctic High and the Icelandic Depression extending far eastward.

As a result, as early as August 2007, high negative ice extent anomalies were recorded in these seas (Table 6.2). In September 2007, the ice edge in the East Siberian Sea sector approached 85°N, which had been never been recorded throughout the entire period of routine observations (i.e. since the 1930s). Intense melting and early disappearance of the ice cover from large areas of open water resulted in increased heating and freshening of surface water and late ice formation. All of this occurred

Table 6.2. Characteristics of ice extent anomalies in the Arctic seas in August 2007

Seas

Barents

Kara

Laptev

East Siberian

Chukchi

Beaufort

Anomalies (%)

-7

-3

-22

-76

-31

-35

against the background of the earlier positive anomaly of temperature (to +1.5°C) and greater thickness of the deep Atlantic water layer.

The major 2007 Arctic ice cover anomaly prompted many climatologists to revise their views on the intensity of Arctic ice area reduction connected with "global warming'' due to greenhouse gases. In press releases to the mass media, a number of climatologists said that 2007 Arctic ice conditions pointed to an acceleration of the global warming process. In some of the interviews, it was predicted that the Arctic ice would disappear in the next five years (!).

Such views on climate change can be accounted for by the fact that some scientists, unfortunately, are apparently unaware of a very important principle regarding patterns of time variability in hydrometeorological parameters: the average absolute value of anomalies decreases with an increase in the averaging time. Consequently, extrapolation of the changes observed during short intervals to long periods is not appropriate.

The SA models based on cosmic radionuclides, such as that of Solanki et al. (2004) indicate a quasi-periodic behavior for SA indices. The Solanki model suggests that the probability of the persistent elevated solar activity for the next five decades is only about 8%. To the extent that the Arctic climate is driven by variations in SA, it would seem unlikely that warming observed in the 20th century will persist far into the 21st century. The fact that there have recently been short-term contractions of sea ice extent in seas of the eastern region of the Arctic cannot be extended to long-term trends.

As we pointed out previously, 2007 was an anomalously warm year for the Arctic, but one year does not create a trend. As it turns out, 2008 was colder than 2007 and ice extent in all seas in the eastern region of the Russian Arctic in August increased by a value exceeding 0.6 • 106 km2 (Table 6.3), which corresponds to the reduction in ice cover area in all seas of the Eurasian shelf within the twentieth century (see Table 2.3). Significant increases in ice extent were also observed in the Arctic Basin and the whole Arctic Ocean. Assuming that by the end of September 2007 the area of the residual (first-year, second- and multiyear) ice in the Arctic Basin decreased to 2.92 million km2 (according to weekly ice analysis provided by the Arctic and Antarctic Research Institute), since total ice extent at September 2008 was 3.47 million km2, this represented an increase of 0.55 million km2 (Frolov, 2008, 2009). The same estimates for the whole Arctic Ocean, available on a basis of the hemispherical ice analysis provided by the US National/Naval Ice Center (IICWG, 2008), were 3.98 million km2 for the end of September 2007 and 4.66 million km2 for the end of September 2008, an increase of 0.66 million km2.1

1 According to other estimates based on daily passive microwave SSM/I ice products (NSIDC Notes, 2007, 2008). the minimum ice extent for the Arctic Ocean of 4.67 million km2 for 2008 was reached on 14 September 2008 and the minimum of 4.28 million km2 for 2007 was reached on 16 September 2007 (an increase of 0.39 million km2) . The difference between the ice charting analysis and the passive microwave estimates is mostly attributable to greater accuracy in ice analysis of the radar and visible satellite imagery used for the ice-charting purposes.

Table 6.3. Ice extent values recorded in the Eurasian Arctic Seas in August 2007 and 2008, 103 km2

Sea

GS

BS

KS

LS

ESS

CS

Western seas

Eastern seas

Total

2007

304

42

236

144

0

0

582

144

726

2008

196

56

166

315

383

60

418

758

1176

Difference 2008-2007

-108

+14

-70

+171

+383

+60

-164

+611

+450

GS—Greenland Sea. BS—Barents Sea. KS—Kara Sea. LS—Laptev Sea. ESS—East Siberian Sea. CS— Chukchi Sea. The western seas encompass Greenland, Barents, and Kara, and the eastern seas Laptev, East Siberian, and Chukchi.

GS—Greenland Sea. BS—Barents Sea. KS—Kara Sea. LS—Laptev Sea. ESS—East Siberian Sea. CS— Chukchi Sea. The western seas encompass Greenland, Barents, and Kara, and the eastern seas Laptev, East Siberian, and Chukchi.

In light of the above it is interesting to consider the changes in the propagation of old (second- and multiyear) ice in the Arctic Basin in recent years that were not taken into account in Section 4.5. Figure 6.7 presents the changes of mean latitude of the old ice dominance boundaries (partial concentration 5 tenths and more) in late winter (March) for the three meridian sectors corresponding to the Laptev Sea, East Siberian, Chukchi and the adjacent areas of the Arctic Basin.

1990 1995 2000 2005 Years

Figure 6.7. Mean latitude of the old ice dominance boundary (thin lines) and its approximation by a polynomial to the power of 6 (thick lines) in March 1990-2008 for the three meridian sectors corresponding to the Laptev Sea (1), East Siberian (2) and Chukchi (3) Seas.

Figure 6.7. Mean latitude of the old ice dominance boundary (thin lines) and its approximation by a polynomial to the power of 6 (thick lines) in March 1990-2008 for the three meridian sectors corresponding to the Laptev Sea (1), East Siberian (2) and Chukchi (3) Seas.

This figure confirms that the gradual southward shift of the old ice boundary described in Section 4.5, which was observed in the second half of the twentieth century, continued at least until 2002. However, in the next 5-6 years, a substantial northward retreat of the old ice boundary occurred, which as noted above, was due to exceptionally favorable (for ice decrease) hydrometeorological conditions in the region associated with short-period cycles of atmospheric circulation. Calculations show that for period since 2002 the area of old ice in this sector of the Arctic Basin was reduced by approximately 2.106 km2. Such results do not contradict conclusions by Mahoney et al. (2008), although gaps in the data and 10-years running smoothing used in the analysis, distorts the temporal scale of fluctuations revealed by its authors. Similar fluctuations, although smaller in scale, were observed in the Laptev Sea and area of the Arctic Basin north of it in 1995-1997 (see Figure 3.3), as well as in the Beaufort Sea in 1998-2001.

Since 2007 there appears to have been a transition to a new phase of this oscillation, during which the boundary of the old ice started to move southward. This is confirmed by the changes that occurred from 2007 to 2008. In addition to the above changes in ice extent, the old ice boundary during the last year shifted southward which was revealed by the setting in September 2008 of a new North Pole drifting station (NP-36) onto a multiyear ice floe over 3 meters thick at 82°35'N, 172°07'E, where no ice was present in September 2007. In March 2009 the position of the MY ice boundary at the meridians of the East Siberian Sea was near 81 °N while during the previous year it retreated beyond the North Pole.

Arctic cooling is also corroborated by the fact that in the 1990s the sign of the trend of the high-latitude zonality index characterizing the mean difference in the elevation of the AT-500 surface between 60°N and 80°N changed from plus to minus, and the recurrence trend of the Arctic High became positive (Dmitriev, 2007). These variations point to a climate change turning point manifested as a start of filling of the Arctic circumpolar vortex. The consequences of this process will become clear in the coming decades.

Was this article helpful?

0 0

Post a comment