Arctic sea ice and the Madden–Julian Oscillation (MJO)
- First Online:
- 2.2k Downloads
Arctic sea ice responds to atmospheric forcing in primarily a top-down manner, whereby near-surface air circulation and temperature govern motion, formation, melting, and accretion. As a result, concentrations of sea ice vary with phases of many of the major modes of atmospheric variability, including the North Atlantic Oscillation, the Arctic Oscillation, and the El Niño-Southern Oscillation. However, until this present study, variability of sea ice by phase of the leading mode of atmospheric intraseasonal variability, the Madden–Julian Oscillation (MJO), which has been found to modify Arctic circulation and temperature, remained largely unstudied. Anomalies in daily change in sea ice concentration were isolated for all phases of the real-time multivariate MJO index during both summer (May–July) and winter (November–January) months. The three principal findings of the current study were as follows. (1) The MJO projects onto the Arctic atmosphere, as evidenced by statistically significant wavy patterns and consistent anomaly sign changes in composites of surface and mid-tropospheric atmospheric fields. (2) The MJO modulates Arctic sea ice in both summer and winter seasons, with the region of greatest variability shifting with the migration of the ice margin poleward (equatorward) during the summer (winter) period. Active regions of coherent ice concentration variability were identified in the Atlantic sector on days when the MJO was in phases 4 and 7 and the Pacific sector on days when the MJO was in phases 2 and 6, all supported by corresponding anomalies in surface wind and temperature. During July, similar variability in sea ice concentration was found in the North Atlantic sector during MJO phases 2 and 6 and Siberian sector during MJO phases 1 and 5, also supported by corresponding anomalies in surface wind. (3) The MJO modulates Arctic sea ice regionally, often resulting in dipole-shaped patterns of variability between anomaly centers. These results provide an important first look at intraseasonal variability of sea ice in the Arctic.
KeywordsMadden–Julian Oscillation Arctic sea ice Intraseasonal variability
Arctic sea ice is a complex component of the Earth climate system. Part of its complexity comes from its sensitivity to the atmosphere on a range of spatial and temporal scales. For example, decades of observational and modeling studies of sea ice have confirmed that its variability is primarily a top-down process (Liu et al. 2004; Deser and Teng 2008), where the atmosphere provides the primary forcing mechanisms (Hopsch et al. 2012). In response, sea ice tends to organize—via motion, formation, melting, and accretion—in accordance with large-scale patterns of atmospheric circulation (Walsh and Johnson 1979; Overland and Pease 1982; Fang and Wallace 1994; Slonosky et al. 1997; Prinsenberg et al. 1997; Overland and Wang 2010). Because of these responses to the atmosphere, concentrations of sea ice have been found to be correlated with several of the major modes of atmospheric variability, including the North Atlantic Oscillation (NAO) (Deser et al. 2000; Kwok 2000; Parkinson 2000; Partington et al. 2003), the Arctic Oscillation (AO) (Wang and Ikeda 2000; Rigor et al. 2002; Belchansky et al. 2004) (the NAO and AO are often referred to as part of the Northern Hemisphere annular mode; Wallace 2000), the El Niño-Southern Oscillation (ENSO) (Liu et al. 2004), and longer-period oscillations (Polyakov et al. 2003). Furthermore, the leading mode of atmospheric intraseasonal variability, the Madden–Julian Oscillation (MJO; Madden and Julian 1972), has been shown to modulate the high-latitude (Zhou and Miller 2005; Cassou 2008) and Arctic (L’Heureux and Higgins 2008; Yoo et al. 2011) atmosphere. However, connections between sea ice and the MJO remain largely unexplored. Therefore, the purpose of this paper is to examine variability of Arctic sea ice concentration by phase of the MJO.
Tropical convection, which is the primary driver of the MJO, has been found to affect atmospheric circulation in high latitudes (e.g., Ferranti et al. 1990; Higgins and Mo 1997; Matthews et al. 2004). Vecchi and Bond (2004) found that geopotential height, specific humidity, and surface air temperature in the Arctic varied by phase of the MJO, and the response of surface air temperature in Canada to the MJO was confirmed by Lin and Brunet (2009). Lee et al. (2011) noted that the “polar amplification” in surface temperatures was in response to poleward-propagating Rossby waves excited by MJO-related tropical convection. Yoo et al. (2012) further confirmed that the MJO-driven, poleward propagating wave train drove changes in the Arctic overturning circulation, heat flux, and downward infrared radiation, and Flatau and Kim (2013) noted that that the MJO forces the annular modes (the AO and NAO) on intraseasonal time scales. All of these Arctic parameters affected by the MJO, from atmospheric circulation to temperature to radiation, have potentially significant impacts on sea ice concentration. However, the specific effects of MJO-driven atmospheric variability on sea ice concentration are not yet known. Therefore, the purpose of this paper is to explore variability in sea ice concentration and atmospheric parameters for two periods, one in the winter freeze-up season, November–January, NDJ, and another in the summer melt season, May–July, MJJ, and then to connect the observed variability to specific phases of the MJO. Both seasonal and monthly variability will be examined on timescales of the MJO. The rest of this paper is organized as follows: datasets and methodology are described in Sect. 2, results are presented in Sect. 3, and discussion and conclusions are presented in Sect. 4.
2 Data and methods
The analyses in this study were based on three publicly available datasets. First, to gain an understanding of the state of the Arctic atmosphere under different phases of the MJO at both surface and mid-tropospheric levels, daily data from the National Centers for Environmental Prediction (NCEP)–Department of Energy (DOE) reanalysis 2 (Kanamitsu et al. 2002) were examined. Variables included in the atmospheric analysis were 500-hPa geopotential height, mean sea level pressure, 2-m surface temperature, and 10-m winds, for the period 1979 to 2011. Daily composite anomalies of pressure, height, temperature and wind were created for both winter and summer months by phase of the MJO using the methodology described below.
Third, the MJO itself was defined using the daily real-time multivariate MJO (RMM) index (Wheeler and Hendon 2004). The RMM phases were used to divide the reanalysis and daily change in sea ice concentration datasets. The daily RMM index oscillates between eight phases, each corresponding to the broad location of an MJO-enhanced equatorial convective signal (Wheeler and Hendon 2004). The index is created such that the MJO generally progresses eastward, from phase 1 to 8 and back to phase 1 again. Days during which the magnitude of the MJO vector was less than one standard deviation from zero were not considered, following the compositing methodology of other recent studies (e.g., Zhou et al. 2012; Virts et al. 2013; Zhang 2013; Barrett and Gensini 2013). Anomalies in daily ΔSIC, 500-hPa geopotential height, sea level pressure, 2-m air temperature, and 10-m wind were then found by averaging the dayn means for each MJO phase and subtracting them from the overall monthly mean.
To isolate relationships between the MJO and Arctic sea ice, and to remove some of the effects of the long-period decline in overall sea ice cover (e.g., Serreze et al. 2007), only daily ΔSIC beyond one standard deviation (either positive or negative) from the normal daily change for that month was used to calculate monthly anomalies; all other daily ΔSIC were not considered for the analysis. In addition to focusing on extreme values of anomalous ΔSIC, a minimum number of days threshold was imposed at each grid box, such that only those boxes in which anomalous ΔSIC values were above (or below) one standard deviation for at least 5 days, for a particular MJO phase and month, were considered. Significance testing was performed using the Student’s t test, and both atmospheric and sea ice anomalies were examined for significance at the 95 % confidence level.
3.1 Seasonal atmosphere variability
3.2 Monthly atmosphere and ice variability
Similar agreement between the atmosphere and anomalous daily change in sea ice concentration was found in the Pacific sector (Fig. 7). For example, on days when the MJO was in phase 2 (Fig. 7, top row), negative surface pressure anomalies were located over the Bering Strait and Chukchi Sea, leading to northerly surface wind anomalies over the Bering Sea and below-normal surface temperatures (as much as 6 K below January normal) in the Sea of Okhotsk, concurrent with an increase in sea ice concentration in both locations. On days when the MJO was in phase 6 (Fig. 7, bottom row), sea level pressure anomalies in the Bering Strait and Chukchi Sea were positive, leading to strong southerly surface wind anomalies (up to 5 m s−1) and positive surface temperature anomalies over the Sea of Okhotsk, concurrent with a decrease in sea ice concentration. For the Pacific sector, surface temperature anomalies seemed to be most strongly related to anomalous change in sea ice concentration. However, in the Atlantic sector surface wind anomalies seemed to be most strongly related to anomalous change in sea ice concentration, in good agreement with Prinsenberg et al. (1997), Deser et al. (2000), and DeWeaver and Bitz (2006), who all noted important effects of surface wind anomalies on Atlantic sea ice concentration in winter.
4 Discussion and conclusions
The goal of this paper was to explore variability in the Arctic atmosphere and sea ice concentration, and to connect such variability with phases of the MJO. Recent studies have heralded significant association between phase of the MJO and high-latitude terrestrial surface air temperature, atmospheric circulation, geopotential height, specific humidity (Ferranti et al. 1990; Higgins and Mo 1997; Matthews et al. 2004; Vecchi and Bond 2004; Zhou and Miller 2005; Cassou 2008; Lin and Brunet 2009). In addition, modulation of the Arctic atmosphere specifically by phase of MJO has also been documented (L’Heureux and Higgins 2008; Yoo et al. 2011). However, none of the previous works cited have considered associations between sea ice concentration and phase of MJO.
The three principal findings of the current study are as follows. (1) The MJO projects onto the Arctic atmosphere in both winter (NDJ) and summer (MJJ) seasons. This projection was evident from the distinct wavy pattern in 500-hPa geopotential height anomalies (Figs. 3, 4), and it confirms the earlier work of Vecchi and Bond (2004) and Yoo et al. (2012). Both location and sign of height anomalies displayed a tendency to flip every 3–5 phases of the MJO. This MJO-mid tropospheric connection also proved robust, being visible in 3-month seasonal plots with statistically significant anomalies at the 95 % level, in both winter and summer seasons. Furthermore, in NDJ, height anomalies in phase 2 resembled positive AO polarity while height anomalies in phases 6 and 7 resembled negative AO polarity, in good agreement with Flatau and Kim (2013). (2) Variability in sea ice concentration by phase of MJO was found in both summer and winter seasons, and this variability was supported by corresponding anomalies in the state of the atmosphere. The magnitude of variability tended to shift largely with the migration of the ice margin poleward (equatorward) during the summer (winter) period. By computing anomalous ΔSIC per month, and binning by phase of MJO, active regions of coherent ice concentration variability were identified in both Atlantic and Pacific sectors for specific phases during January (Figs. 6, 9) and for North Atlantic and Siberian sectors during July. The signs of anomalies (positive or negative) for specific MJO phases changed with season. In January, areas of positive (negative) ΔSIC in the Atlantic sector were collocated with southerly (northerly) wind anomalies, with southerly (northerly) winds pushing ice toward (away from) land resulting in anomalously positive (negative) change in concentration. In the Pacific sector in January, areas of positive (negative) ΔSIC were collocated with negative (positive) surface temperature anomalies, with colder (warmer) surface temperatures promoting local increases (decreases) in ice concentration. In July, areas of positive (negative) ΔSIC in both the North Atlantic and Siberian sectors were collocated with northerly (southerly) surface wind anomalies, as unlike in January, northerly (southerly) winds acted to push ice away from (toward) the primary ice source region (the central Arctic), leading to positive (negative) changes in ice concentration. Sea level pressure anomalies were found to support the observed variability in surface wind. (3) The MJO modulates Arctic sea ice regionally, often resulting in dipole-shaped variability between anomaly centers. The most commonly observed dipoles occurred between the Barents and Greenland seas in January, in agreement with Ivanova et al. (2012). All four sectors (Atlantic and Pacific in winter, and North Atlantic and Siberian in summer) demonstrated instances of ice anomalies that changed sign approximately every 3–4 phases of the MJO, as evidenced from the January and July examples presented in this study. These changes in sign of anomalous ΔSIC corresponded with similar changes in surface pressure, surface wind, and mid-tropospheric geopotential height, and suggest a physical robustness to the MJO-sea ice relationship.
It is important to note that accelerating decline in extent of multi-year sea ice over the last several decades has cast some doubt on earlier findings of ice-climate relationships, particularly between sea ice and phase of the NAO. For example, during winter, cyclonic surface air flow promotes ice export through the Fram Strait (Jung and Hilmer 2001), particularly export of multi-year ice (Deser and Teng 2008), leaving the newer, thinner pack more vulnerable to forcings including enhanced downward longwave radiation (Francis and Hunter 2006) and circulation (Comiso 2006; Maslanik et al. 2007; Francis and Hunter 2007). This process has accelerated with the changing character of sea ice, and perhaps also expanded the ice margins that are susceptible to changes in atmospheric circulation and temperature that vary by phase of the MJO. To mitigate potential effects of the long-term decline in overall sea ice extent, in this study, we imposed several restrictions on the sea ice concentration data. First, we only examined daily ΔSIC that was more than one standard deviation above (or below) normal. Second, we excluded grid points from the analysis with fewer than 6 days of non-zero daily change in sea ice concentration (thus ensuring focus on the ice margins). Third, only sea ice anomalies that were statistically significant at the 95 % confidence interval were plotted in Figs. 6, 7, 8, 9 (column four of each figure). These three restrictions served to amplify the MJO-ice signal by removing regions of insignificant variability, particularly toward the center of the Arctic.
The results presented in this paper show statistically significant variability in Arctic sea ice by phase of the MJO that is well supported by corresponding tendencies in surface wind and surface air temperature. While the specific phase relationships may well change, the MJO will continue to project onto the Arctic and modify sea ice at the ice margins, and may become even more prominent due, in particular, to the decline in thicker multi-year ice. With the tendency for thinner and more vulnerable first-year ice to occupy a greater fraction of the Arctic, the MJO-sea ice relationship shown here may become even more prominent under our changing Arctic climate. A follow-on study is underway to explore these future relationships.
The authors thank American Society for Engineering Education (ASEE) Science and Engineering Apprenticeship Program (SEAP) interns Anna Haschert and Cassandra Marino for assistance in manuscript preparation. Funding for this research was provided by the National Science Foundation under Grant ARC-1203843.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.