1. Introduction
Sea ice is a critical component of the climate system, influenced by both the atmosphere and the ocean. Many studies have found that fluctuations in sea ice primarily result from a combination of the variations in wind stress (dynamic) and the perturbations in the surface energy balance induced by the temperature anomalies (thermodynamic) (e.g., Agnew 1993; Fang and Wallace 1994; Deser et al. 2000). In addition to a large seasonal cycle, sea ice in the North Pacific exhibits large variability on interannual and decadal time scales as well (e.g., Fang and Wallace 1994; Parkinson et al. 1999; Liu et al. 2004). Parkinson (1990) demonstrated that changes in sea ice cover in the Sea of Okhotsk are associated with the anomalous positions and/or intensities of the Siberia high and the Aleutian low. Overland and Pease (1982) showed that shifts of storm track influence the sea ice edge in the Bering Sea. Wind and heat flux anomalies associated with a wave train over the Pacific, which bears some resemblance to the North Pacific Oscillation (Rogers 1981), lead to anomalies of opposite sign in the Sea of Okhotsk and the Bering Sea (Cavalieri and Parkinson 1987; Fang and Wallace 1994).
Variations in sea ice in turn modulate climate by influencing surface albedo, and exchanges of surface turbulent fluxes between the atmosphere and the ocean. Observational analyses and atmospheric general circulation model simulations have indicated that sea ice anomalies can affect the overlying atmosphere, which may cause feedback to large-scale atmospheric circulation. Herman and Johnson (1978) found a significant response in the simulated sea level pressure, 700-mb temperature, and 300-mb geopotential height to the ideal ice edge difference (maximum − minimum) in the Arctic. They also noted that the full atmospheric response could not be explained by local thermodynamics, suggesting that dynamical processes are important for the far-field anomalies. Murray and Simmonds (1995) examined the simulated atmospheric response to idealized specifications of sea ice concentrations in the Arctic. They found that reduced ice cover leads to a local monotonic but nonlinear increase in surface air temperature and a weakening of the midlatitude westerlies. Less ice also results in a significant decrease in the speeds and intensities of storms poleward of 45°N. In the North Pacific, Honda et al. (1999) examined the atmospheric response to maximum and minimum sea ice extent in the Sea of Okhotsk, where the difference between the two ice states was specified to be approximately twice as large as what has been observed. The model produced significant responses both locally and downstream over the Bering Sea, Alaska, and North America in the form of a stationary wave train in the troposphere, which was excited thermally by the anomalous surface turbulent heat fluxes.
In this study, we attempt to address some issues that are related to sea ice variability in the North Pacific and its associations with the east Asia–North Pacific climate with a focus on winter season and interannual time scales: What are the temporal and spatial patterns of sea ice variability in the North Pacific? Are sea ice anomalies of opposite sign between the Sea of Okhotsk and the Bering Sea found in Cavalieri and Parkinson (1987) and Fang and Wallace (1994) a robust and dominant feature? Are other processes or patterns of similar importance? What is the nature of the associations between 1) anomalies in local and large-scale atmospheric circulation and 2) anomalies in the east Asia–North Pacific climate, and the most dominant patterns of sea ice variability in the North Pacific? What are possible linkages between sea ice variability in the North Pacific and some of the most prominent large-scale modes of climate variability?
2. Data and analysis procedure
Datasets used in this study include 1) the Arctic sea ice concentrations (SICs) retrieved from the Scanning Multichannel Microwave Radiometer on the Nimbus 7 satellite and the Special Sensor Microwave/Imager on several defense meteorological satellites (based on a bootstrap algorithm; see Comiso et al. 1997 for details), which provide the longest quality-controlled record for studying interannual and decadal sea ice variability; 2) the surface turbulent heat flux (latent heat + sensible heat), air temperature at 17 pressure levels (from 1000- to 10-mb), 200- and 850-mb winds, 500-mb geopotential height, and sea level pressure from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis (Kalnay et al. 1996); 3) the surface air temperature from the National Aeronautics and Space Administration (NASA) Goddard Institute for Space Studies (GISS; the input data for this analysis are collected by many national meteorological services around the world; see Hansen et al. 1999 for details); and 4) the precipitation from the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP; based on estimates from gauge measurements and satellite retrievals; see Xie and Arkin 1997 for details). The common time period for all datasets is from October 1978 to September 2002, except for the precipitation data, which is from October 1979 to September 2002.
For each year, we generated winter (December–February) averages for the aforementioned datasets. Anomalies were then computed by subtracting the climatology from each individual winter. The empirical orthogonal function (EOF) analysis (including statistical significance) was performed on the winter sea ice concentration anomalies to identify the most dominant patterns of sea ice variability in the North Pacific. The indices were defined using the standardized principal components (PCs) of the most dominant modes of North Pacific sea ice variability. Regression and correlation analyses (including statistical significance) were then carried out to determine the anomalies in the surface turbulent heat flux, atmospheric circulation, surface air temperature, and precipitation during winter associated with the sea ice indices.
3. Results
a. North Pacific sea ice variability
Figure 1 shows the spatial pattern of the standard deviations of the winter SIC anomalies in the North Pacific. Large sea ice variability is not confined to sea ice edges, but is in the northwest portion of the Sea of Okhotsk and the northeast portion of the Bering Sea. Similar results are also shown in Deser et al. (2000) based on a different sea ice dataset.
The time series of the total sea ice extent (the cumulative area of grid boxes covering at least 5% SIC) anomalies in the Sea of Okhotsk and Bering Sea are shown in Fig. 2a. Here the Sea of Okhotsk (the Bering Sea) is defined as the area of 42°–63°N and 131°–161°E (53°–66°N and 161°E–158°W). The magnitude of the total ice extent anomalies in the Sea of Okhotsk is comparable to that of the Bering Sea. The correlation between the two time series is −0.26, which partially supports the idea that the ice anomalies in the Sea of Okhotsk tend to be out of phase with those of the Bering Sea (Cavalieri and Parkinson 1987; Fang and Wallace 1994). However, the negative correlation is not statistically significant, only accounting for ∼9% shared variance. In fact, the total ice extent anomalies of the two seas were in phase in 1979, 1983, 1985, 1992, 1993, 1994, 1995, 1996, 1998, and 2001 (10 of the total 24 yr), suggesting that another process may exert strong influence on sea ice across the two seas (Fig. 2a). Given that the negative correlation is low, and another process seems to produce an in-phase relationship between the two seas at times, it is instructive to quantify sea ice variability in the North Pacific in terms of different processes (using the EOF approach), rather than to generate one variability index for sea ice in the North Pacific using the total sea ice extent in the two seas individually or collectively.
To test the idea that two processes, one causing opposing effects in the two seas and the other leading to similar effects in the two seas, are at play, we performed the EOF analysis on the winter SIC anomalies in the North Pacific. The explained variance of the first two EOFs (40% for EOF1 and 22% for EOF2) is much larger than other EOFs (i.e., 8% for EOF3 and 6% for EOF4). The first two EOFs are statistically significant, and distinct from each other and the other EOFs, based on the method proposed by North et al. (1982). The first EOF mode features a dipole pattern between the Sea of Okhotsk (with the largest variability in the central-north Sea of Okhotsk) and the Bering Sea (with the largest variability in the eastern Bering Sea; Fig. 3a). The second EOF mode is characterized by less or more ice over in the entire North Pacific, with the largest variability in the northwest Sea of Okhotsk and Bering Sea (Fig. 3b). As shown in Fig. 3c and discussed later, the first EOF mode has a strong association with the Arctic Oscillation (AO; Thompson and Wallace 2000). To further insure that the second EOF mode (more uniform ice changes in the two seas) is not an artifact of the EOF approach, we conducted another EOF analysis on the residual winter SIC anomalies after removing the regressed impacts of the AO as shown in Fig. 3c. The resulting new first EOF mode (not shown) shows similar and more uniform ice changes to that of Fig. 3b.
The principal components of the first two EOF modes seem to exhibit low-frequency behavior, though the relatively short period of record (24 yr) limits our ability to analyze this behavior. A significant increasing trend is found for PC1 (0.59 per decade), while no obvious trend exists for PC2 (Fig. 2b). To explore how much variance is explained by the trend in EOF1, we repeated the EOF analysis on the residual winter ice concentration anomalies after removing the linear trends. The first two EOF modes are extremely similar to those that are not detrended (not shown), though their explained variance is changed slightly.
b. Local responses associated with EOF1 and EOF2
As mentioned earlier, ice changes are primarily a reflection of atmospheric circulation and temperature anomalies. However, changes in sea ice can influence the overlying atmosphere by changing the surface albedo and air–sea turbulent fluxes. In winter, the albedo effect is suppressed due to the low insolation. The exchanges of surface turbulent heat fluxes by contrast are more important due to strong winds and temperature gradients. The regression patterns of the surface turbulent heat flux (latent + sensible) anomalies upon the standardized PC1 and PC2 are shown in Fig. 4. Positive values indicate above-normal heat transfer from the ocean to the atmosphere. As expected, where the ice is reduced (enhanced), there are above-normal (below-normal) heat flux anomalies. The largest heat flux change can be ∼20 W m−2 (∼−40 W m−2) in the Sea of Okhotsk (Bering Sea) for PC1, and ∼50 W m−2 (∼30 W m−2) in the Sea of Okhotsk (Bering Sea) for PC2.
The large surface heat flux perturbations associated with the ice changes suggest that the overlying atmosphere might be sensitive to the identified sea ice variations in the North Pacific. Figure 5 shows the longitude–height cross section of the regression patterns of the NCEP air temperature on the standardized PC1 and PC2. For PC1 (Fig. 5a), the reduced ice cover in the Sea of Okhotsk is associated with a positive tropospheric temperature anomaly that decays from the surface (∼1.6°C) to the tropopause, with an anomaly of opposite sign in the stratosphere, indicating the possible presence of deep vertical modes. The positive temperature anomaly in the troposphere tilts westward with height, and seems to be connected to a strong warm anomaly in Siberia (as shown later). By contrast, the enhanced ice cover in the Bering Sea is associated with a negative temperature anomaly that extends almost uninterrupted from the surface (∼−1.8°C) to the top of the atmosphere. The exception is a narrow band of weak warm anomaly in the tropopause. For PC2 (Fig. 5b), the widespread reduced ice cover in the North Pacific is associated with a positive temperature anomaly, which dominates the entire atmospheric column, with two maximum centers: one is at the surface over the Sea of Okhotsk (∼1°C), and the other, in the stratosphere, is centered over the western Sea of Okhotsk (∼1.4°C).
For both PCs, there is little longitudinal displacement of the temperature anomalies from the surface to ∼700 mb; the temperature anomalies are located near the source of the ice anomalies at those heights. Above 700-mb, the longitudinal displacement emerges and increases with additional ascent. Also, a good correspondence between the warm (cold) anomalies in Figs. 5a,b and high (low) 500-mb geopotential height anomalies is found in Figs. 6c,d, indicating that there are vertically extended hydrostatic anomalies. It is noted that the sea ice, surface heat flux, and air temperature anomalies are nearly collocated for both the PC1 and PC2 cases, suggesting that the surface heat flux perturbations that result from the ice changes are very important for the local thermal response. The model study of Alexander et al. (2004) also indicated that the surface heat flux plays a more important role in determining the local thermal response than the advective process.
c. Large-scale atmospheric circulation associated with EOF1 and EOF2
To further investigate changes in large-scale atmospheric circulation associated with the two dominant patterns of sea ice variability in the North Pacific, we examined the regression patterns of the NCEP 200-mb zonal wind, 500-mb geopotential height, 850-mb zonal and meridional winds, and sea level pressure with respect to the standardized PC1 and PC2 (Fig. 6).
Corresponding to one positive standard deviation change in PC1, the reduced ice cover in the Sea of Okhotsk and the enhanced ice cover in the Bering Sea are associated with a weakening of the east Asian jet stream (upper-level westerly flow) near 30°–50°N, and a strengthening of the westerly flow to the north (Fig. 6a). As the east Asian jet stream weakens, the east Asian trough, which is climatologically anchored along the coast of east Asia, extending from northeast China, through Japan to the Sea of Okhotsk (not shown), becomes shallower, as shown by an increase in 500-mb geopotential height in the upstream portion of the trough (Fig. 6c). At low levels, a zonally oriented anomalous anticyclonic circulation occupies the regions north of 40°N, which is also consistent with above-normal sea level pressure there (not shown). The anomalous easterlies in the southern flank of the anomalous anticyclone impinge on north China across Japan and Korea. The anomalous easterlies seem also to be coupled to an anomalous cyclonic circulation over the subtropical central North Pacific (Fig. 6e).
Associated with one positive standard deviation changes in PC2, the reduced ice cover throughout the North Pacific is accompanied by a strengthening of the east Asia jet stream (∼25°–45°N) and the westerly flow in the subtropical North Pacific (∼20°–40°N), and a weakening of the westerly flow in the North Pacific (∼40°–60°N; Fig. 6b). This deepens the east Asian trough, as shown by a decrease in 500-mb geopotential height in the central and downstream portion of the trough, and an increase in 500-mb geopotential height at the ridge location (the Sea of Okhotsk; Fig. 6d). Concomitant with the changes in the upper-level circulation, the most pronounced feature at low levels is a broad anomalous cyclonic circulation over the entire North Pacific, which is also consistent with below-normal sea level pressure there (Fig. 10), linking east Asia at one end and western North America at the other end. The anomalous northerlies in the western flank of the anomalous cyclone control the coast of east Asia. The anomalous cyclonic circulation seems also related to the weakened trade winds over the subtropical central North Pacific (Fig. 6f).
d. East Asia–North Pacific climate associated with EOF1 and EOF2
What are the associations between the aforementioned changes in large-scale atmospheric circulation and the east Asia–North Pacific winter climate? The east Asian winter monsoon (EAWM) is one of the most important systems that controls the east Asia–North Pacific winter climate. The associated dramatic shift of the northwesterlies and the outbreak of cold surges dominate the winter weather and climate in the east Asian region (e.g., Zhang et al. 1997; Jhun and Lee, 2004). The EAWM can also exert strong impacts through the midlatitudes and on tropical planetary-scale circulation, and even influence sea surface temperature in the tropical western Pacific (e.g., Chang et al. 1979; Chang and Lau 1982; Lau and Li 1984; Compo et al. 1999; Trenberth et al. 2000). Many studies have shown that the intensity of the EAWM is closely tied to low-level meridional winds (e.g., Zhang et al. 1997). In this study, we defined an index, that is, the average of the NCEP 850-mb meridional wind over the region of 25°–55°N, 110°–150°E (see the boxed area in Figs. 6e,f), representing the intensity of EAWM over the coast of east Asia. As shown in Table 1, the EAWM is significantly correlated with PC2 (−0.48), suggesting that the reduced ice cover throughout the North Pacific is associated with the intensified EAWM, and vice versa. By contrast, the relationship between EAWM and PC1 is much weaker (−0.17).
Figure 7 further shows the correlation patterns between the GISS surface air temperature and CMAP precipitation, and PC1 and PC2. During the positive phase of PC1, as expected, the local thermal responses are a warming over the Sea of Okhotsk and a cooling over the Bering Sea. In addition, the warming extends westward from the Sea of Okhotsk/Japan, through northeast China, and into Siberia (Fig. 7a). The positive phase of PC1 is also associated with above-normal precipitation over much of China and Siberia, with the exception of a narrow band of below-normal precipitation centered roughly at 50°N. The warm and wet anomalies in northeast China and central Siberia are associated with the anomalous low-level southeasterlies/easterlies (Fig. 6e). The anomalous southeasterly/easterly flows tend to block off the climatological northwesterly/westerly flows, which deprive cold and dry air from northeast China and central Siberia, leading to warming and above-normal precipitation there. The above-normal precipitation to the east of the boundary of the Tibetan Plateau might be due to the anomalous advection and ascent of warm and moist air from the east, extending from the North Pacific across Japan and Korea to northern China.
During the positive phase of PC2, a strong local warming occurs over the North Pacific north of 40°N (including the two seas). By contrast, a widespread cooling dominates much of continental east Asia and extends eastward to the central North Pacific (Fig. 7b). In addition, an anomalous warming is observed in the subtropical central North Pacific. The positive phase of PC2 is also associated with below-normal precipitation over much of continental east Asia but extends northeastward to the North Pacific (Fig. 7d). The cold and dry conditions over east Asia are coincident with the anomalous low-level northerlies (Fig. 6f), extending from eastern Siberia to the subtropical western North Pacific. The anomalous northerlies enhance the climatological low-level northwesterly flow, thereby intensifying the EAWM, which brings cold and dry air from the high latitudes and suppresses precipitation over east Asia. Furthermore, the intensified EAWM strengthens the local Hadley cell, which in turn strengthens the east Asian jet stream. As shown in Fig. 8, associated with the positive phase of PC2, there is a strong anomalous rising between 0° and 15°N and a pronounced anomalous sinking between 20° and 40°N, indicating the intensification of the local Hadley cell. The anomalous strong subsidence between 20° and 40°N, in addition to the aforementioned cold and dry air advection, dries the troposphere and clears clouds, causing a precipitation deficit over subtropical east Asia.
e. Teleconnections
As a first step to investigate the possible linkages between sea ice variability in the North Pacific and some of the most prominent large-scale modes of climate variability, we calculated the correlations between PC1 and PC2, and various climate indices (Table 1), including the AO, Pacific decadal oscillation (PDO), Pacific–North America pattern (PNA), and El Niño–Southern Oscillation (ENSO: Niño-3; http://www.cdc.noaa.gov/ClimateIndices/Analysis).
PC1 has significant positive correlation with the AO (0.69) and negative correlation with the PDO (−0.42), but has little correlation with other indices (PNA and ENSO). Fig. 9 shows the regression patterns of the GISS surface air temperature on the standardized PC1 and AO, respectively. Here we multiplied the regression coefficients by 0.59 and 0.31, which, respectively, are the values of the 10-yr linear trend in PC1 and the AO. Thus, the resulting regression patterns are representative of the surface air temperature associated with the 10-yr trend in PC1 and AO. The similarity of the temperature patterns associated with PC1 and the AO is rather striking, indicating that PC1 has a strong association with the AO. However, the magnitude of the temperature anomalies associated with the 10-yr trend in PC1 is 2 times larger than that of the AO. By contrast, PC2 has no significant relationship with any indices (except EAWM).
4. Discussion and summary
In this study, we investigated sea ice variability in the North Pacific and its associations with the east Asia–North Pacific climate with a focus on wintertime and interannual time scales.
By conducting the EOF analysis on the satellite-based sea ice concentrations, we identified that two processes, one causing opposing ice changes in the Sea of Okhotsk and the Bering Sea, and the other leading to more uniform ice changes in the aforementioned two seas, are at play and together explain ∼62% of the interannual variance. Additionally, a significant increasing trend is found for the principal component of the first process (PC1; 0.59 per decade), while no obvious trend exists for the principal component of the second process (PC2).
The changes in the surface heat flux associated with the two dominant patterns of sea ice variability in the North Pacific are very important for the local atmospheric thermal response, since the ice, surface heat flux, and air temperature anomalies are nearly collocated. That is, sea ice reduction (enhancement) is associated with above-normal (below-normal) heat transfer from the ocean to the atmosphere, and warming (cooling) for the overlying atmosphere. The local atmospheric thermal response identified here is also supported by the direct linear response to sea ice anomalies as demonstrated in previous model studies (e.g., Honda et al. 1999; Alexander et al. 2004).
Results from regression and correlation analyses show that corresponding to the positive polarity of PC1, at 200-mb, the east Asian jet stream weakens significantly near 30°–50°N, and the westerly flow strengthens significantly to the north, leading to a significant weakening of the east Asian trough at 500 mb. At 850 mb, the zonally oriented anomalous anticyclonic circulation extends from the North Pacific to Siberia. The associated anomalous southeasterlies/easterlies deprive northeast China and central Siberia of cold and dry air, leading to warming and above-normal precipitation there. The advection and ascent of warm and moist air associated with the anomalous easterlies in the southern flank of the anomalous cyclone might be responsible for the above-normal precipitation to the east of boundary of the Tibetan Plateau, but further study is required to understand this relationship. It is interesting that changes in the surface air temperature associated with the positive polarity of PC1 resembles the Cold Ocean–Warm Land pattern (COWL; Wallace et al. 1995) that is also associated with the positive phases of the AO.
The positive polarity of PC2 shows that, at 200 mb, the east Asia jet stream intensifies significantly (∼25°–45°N) and the westerly flow weakens significantly in the North Pacific (∼40°–60°N), leading to a significant deepening of the east Asian trough at 500 mb. At 850 mb, the strong anomalous cyclonic circulation forms over the North Pacific, which intensifies the EAWM, thereby bringing cold and dry air from the high latitudes to the entire east Asian coast. In fact, the cold and dry conditions extend from the eastern half of the Asian continent to the central North Pacific. The intensified EAWM also strengthens the local Hadley cell, which in turn strengthens the east Asian jet steam and leads to the precipitation deficit over subtropical east Asia through the enhanced subsidence. A tripole of temperature anomalies over the ocean is somewhat reminiscent of the positive polarity of the Pacific decadal oscillation.
Additionally, the wind anomalies (i.e., due to the variations of atmospheric circulation) associated with the two dominant modes help reinforce the ice changes (Figs. 6e,f). For example, the Sea of Okhotsk experiences anomalous easterly flow under the positive phase of both PC1 and PC2. These anomalous easterly flows lessen impacts of cold continental air from Siberia and help explain the ice reduction in the Sea of Okhotsk. In the Bering Sea, which experiences more ice with the positive phases of PC1 and less ice with the positive phases of PC2, an anomalous northwesterly flow helps explain the former, while the aforementioned anomalous easterly flow helps explain the latter.
Correlations between PC1 and PC2, and various climate indices support the idea that PC1 has a strong association with the AO. As demonstrated in Fig. 3c, the AO can also leads to similar spatial anomalies in sea ice to those associated with PC1 in the North Pacific through an advective process (ice advection and temperature advection; Liu et al. 2004). However, our further analysis suggests that the response of sea ice to the AO is 2 times less than those associated with PC1 (not shown). This raises the possibility that a positive feedback exists between the AO and sea ice in the North Pacific. Moreover, the magnitude of the COWL pattern (the warm anomalies in Siberia and cold anomaly in the North Pacific north of 45°N) associated with the 10-yr trend in PC1 is much larger than that of the AO. Trend analysis of the surface air temperature obtained from the International Arctic Buoy Program (1979–97) shows a significant warming of up to 2°C per decade over central Siberia, extending north over the Laptev Sea in winter (Fig. 9 of Rigor et al. 2000). Thus, as shown in Fig. 9 PC1 better explains recent warming in central Siberia than the AO, though PC1 is locally defined, while AO is hemispherically defined. The implication is that the atmospheric circulation change associated with the AO variations might set up the ice anomalies in the Sea of Okhotsk and the Bering Sea through ice advection and temperature advection. Thereafter, the ice anomalies in the two seas tend to sustain and amplify the signals, and have a positive feedback on the atmospheric circulation via the surface turbulent heat flux.
PC2, which is significantly correlated with the EAWM, has no significant relationship with other climate indices, though its signature in the surface air temperature over the ocean resembles the PDO. Recent studies have suggested that the interannual variability of sea ice in the North Pacific, particularly in the Bering Sea, is strongly tied to large-scale atmospheric circulation associated with the Aleutian low and relatively local atmospheric fluctuations related with pressure anomalies over Alaska (e.g., Sasaki and Minobe 2005). Figure 10 shows the composite difference of the sea level pressure between PC2 greater and less than one standard deviation. The positive phases of PC2 are associated with a broad below-normal sea level pressure over the entire North Pacific, indicating the intensification/expansion of the Aleutian low. It is well known that EAWM depends strongly on the development of the Aleutian low and Siberian high (e.g., Jhun and Lee 2004). The strengthening of the Aleutian low as shown in Fig. 10 leads to an increase in the pressure gradients between the Aleutian low and Siberian High, which results in the strengthening of EAWM.
The present study mainly focuses on the relationship between sea ice variability in the North Pacific and winter climate over east Asia–North Pacific. However, one possible mechanism of the causal relationship that can be inferred from this study is that changes in atmospheric circulation (i.e., due to large-scale modes and storm tracks) would affect sea ice in the Sea of Okhotsk and Bering Sea dynamically (wind stress) and thermodynamically (surface heat flux anomalies). The ice changes can 1) sustain and amplify the signals, and have the positive feedback on the large-scale atmospheric circulation via the surface turbulent heat flux, and 2) alter the meridional temperature gradient at the northern edge of the Aleutian low and relative intensities/positions between the Aleutian low and the Siberia High, which would influence the baroclinicity of the atmosphere, leading to the variations of EAWM.
Causality and/or mechanisms are being verified with model experiments. For example, perpetual winter runs forced with the observed two dominant patterns of sea ice variability in the North Pacific are being conducted to shed light on the following questions: By what dynamic and thermodynamic processes does sea ice variability in the North Pacific influence the remote atmospheric circulation? What is the actual magnitude of the remote atmospheric responses to sea ice variability in the North Pacific? Experiments should also be designed to further our understanding of the positive feedback between the AO and PC1, and interactions among the Aleutian low and PC2. These will be addressed in future research. Given the identified relationships between sea ice variability in the North Pacific and temperature and precipitation in east Asia, to the extent that sea ice anomalies can be predicted, PC1 may be a valuable indicator of temperature variability in Siberia as climate warms, whereas PC2 may be a valuable predictor of the EAWM and associated temperature and precipitation.
Acknowledgments
This research is supported by the “Hundred Talent Program” of the Chinese Academy of Sciences, National Natural Science Foundation of China under Grant 40233032 as well as by NASA.
REFERENCES
Agnew, T., 1993: Simultaneous winter sea-ice and atmospheric circulation anomaly patterns. Atmos.–Ocean, 31 , 259–280.
Alexander, M. A., U. S. Bhatt, J. E. Walsh, M. S. Timlin, J. S. Miller, and J. D. Scott, 2004: The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter. J. Climate, 17 , 890–905.
Cavalieri, D. J., and C. L. Parkinson, 1987: On the relationship between atmospheric circulation and fluctuations in sea ice extents of the Bering and Okhotsk Seas. J. Geophys. Res., 92 , 7141–7162.
Chang, C-P., and K-M. Lau, 1982: Short-term planetary-scale interactions over the tropics and midlatitudes during northern winter. Part I: Contrasts between active and inactive periods. Mon. Wea. Rev., 110 , 933–946.
Chang, C-P., J. Erickson, and K-M. Lau, 1979: Northeasterly cold surges and near-equatorial disturbances over the winter MONEX area during 1974. Part I: Synoptic aspects. Mon. Wea. Rev., 107 , 812–829.
Comiso, J. C., D. J. Cavalieri, C. L. Parkinson, and P. Gloersen, 1997: Passive microwave algorithms for sea ice concentration: A comparison of two techniques. Remote Sens. Environ., 60 , 357–384.
Compo, G., G. Kiladis, and P. J. Webster, 1999: East Asian winter monsoon pressure surges and their relationship to tropical variability. Quart. J. Roy. Meteor. Soc., 125 , 29–54.
Deser, C., J. E. Walsh, and M. S. Timlin, 2000: Arctic sea ice variability in the context of recent atmospheric circulation trends. J. Climate, 13 , 617–633.
Fang, Z., and J. M. Wallace, 1994: Arctic sea ice variability on a timescale of weeks and its relation to atmospheric forcing. J. Climate, 7 , 1897–1914.
Hansen, J., R. Ruedy, J. Glascoe, and M. Sato, 1999: GISS analysis of surface temperature change. J. Geophys. Res., 104 , 30997–31022.
Herman, G. T., and W. T. Johnson, 1978: The sensitivity of the general circulation of Arctic sea ice boundaries: A numerical experiment. Mon. Wea. Rev., 106 , 1649–1664.
Honda, M., K. Yamazaki, H. Nakamura, and K. Takeuchi, 1999: Dynamic and thermodynamic characteristics of atmospheric response to anomalous sea ice extent in the Sea of Okhotsk. J. Climate, 12 , 3347–3358.
Jhun, J-G., and E-J. Lee, 2004: A new east Asian winter monsoon index and associated characteristics of the winter monsoon. J. Climate, 17 , 711–726.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437–471.
Lau, K-M., and M-T. Li, 1984: The monsoon of east Asia and its global associations—A survey. Bull. Amer. Meteor. Soc., 65 , 114–125.
Liu, J., J. A. Curry, and Y. Hu, 2004: Recent Arctic sea ice variability: Connections to the Arctic Oscillation and the ENSO. Geophys. Res. Lett., 31 .L09211, doi:10.1029/2004GL019858.
Murray, R. J., and I. Simmonds, 1995: Responses of climate and cyclones to reductions in Arctic sea ice. J. Geophys. Res., 100 , 4791–4806.
North, G. R., T. L. Bell, R. F. Cahalan, and F. J. Moeng, 1982: Sampling errors in the estimation of empirical orthogonal functions. Mon. Wea. Rev., 110 , 699–706.
Overland, J. E., and C. H. Pease, 1982: Cyclone climatology of the Bering Sea and its relation to cyclone extent. Mon. Wea. Rev., 110 , 5–13.
Parkinson, C. L., 1990: The impacts of the Siberian high and Aleutian low on the sea-ice cover of the Sea of Okhotsk. Ann. Glaciol, 14 , 226–229.
Parkinson, C. L., D. J. Cavalieri, P. Gloersen, H. J. Zwally, and J. C. Comiso, 1999: Arctic sea ice extents, areas, and trends, 1978–1996. J. Geophys. Res., 104 , 20837–20856.
Rigor, I. G., R. L. Colony, and S. Martin, 2000: Variations on surface air temperature in the Arctic, 1979–97. J. Climate, 13 , 896–914.
Rogers, J. C., 1981: Spatial variability of seasonal sea level pressure and 500 hpa height anomalies. Mon. Wea. Rev., 109 , 2093–2105.
Sasaki, Y. N., and S. Minobe, 2005: Seasonally dependent interannual variability of sea ice in the Bering Sea and its relation to atmospheric fluctuations. J. Geophys. Res., 110 .C05011, doi:10.1029/2004JC002486.
Thompson, D. W. J., and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 1000–1016.
Trenberth, K. E., D. P. Stepaniak, and J. M. Caron, 2000: The global monsoon as seen through the divergent atmospheric circulation. J. Climate, 13 , 3969–3993.
Wallace, J. M., Y. Zhang, and J. A. Renwick, 1995: Dynamic contribution to hemispheric mean temperature trends. Science, 270 , 780–783.
Xie, P., and P. A. Arkin, 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78 , 2539–2558.
Zhang, Y., K. R. Sperber, and J. S. Boyle, 1997: Climatology and interannual variation of the East Asian winter monsoon: Results from the 1979–95 NCEP/NCAR reanalysis. Mon. Wea. Rev., 125 , 2605–2619.
Std devs (%) of the winter sea ice concentration anomalies.
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
(a) Time series of the total winter sea ice extent anomalies (×105 km2) in the Sea of Okhotsk and Bering Sea. (b) Standardized principal components corresponding to the two dominant EOF modes in Fig. 3.
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
First two dominant EOF spatial patterns of the winter sea ice concentration anomalies: (a) EOF1, (b) EOF2, and (c) regression of the sea ice concentration anomalies on the standardized AO.
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Regression of the surface turbulent heat flux (W m−2) on the standardized (a) PC1 and (b) PC2. (The light shading indicates the 90% confidence level)
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Lon–height cross section of regression of the NCEP air temperature (°C) on the standardized (a) PC1 and (b) PC2. (For the grid points west of 162.5°E, the average of 50°–60°N is used for the Sea of Okhotsk, whereas for the grid points east of 162.5°E, the average of 57°–66°N is used for the Bering Sea; the light shading indicates the 95% confidence level)
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Regression of the NCEP (a), (b) 200-mb zonal wind (m s−1), (c), (d) 500-mb geopotential height (m), and (e), (f) 850-mb zonal and meridional winds (m s−1) on the standardized (a), (c), (e) PC1 and (b), (d), (f) PC2. (The topography above 2000 m is shaded dark, the area above 95% confidence level is shaded light, and the thick vectors denote that either the zonal or meridional component is above the 95% confidence level)
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Correlation between the (a), (b) GISS surface air temperature and (c), (d) CMAP precipitation, and (a), (c) PC1 and (b), (d) PC2. (The light shading indicates the 95% confidence level)
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Regression of the vertical velocity (Pa s−1, multiplied by 1000) on the standardized PC2 along 120°E. (The light shading indicates the 95% confidence level)
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Regression of the GISS surface air temperature (°C) on the standardized (a) PC1 and (b) AO (multiplied by the 10-yr linear trend in PC1 and AO, respectively).
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Composite difference of the NCEP sea level pressure (mb) between PC2 greater and less than one std dev.
Citation: Journal of Climate 20, 10; 10.1175/JCLI4105.1
Correlations between PC1 and PC2, and various climate indexes. (Correlations exceeding the 95% confidence level are in bold type)
