• Anderson, J. R., , and R. D. Rosen, 1983: The latitude-height structure of 40–50 day variations in atmospheric angular momentum. J. Atmos. Sci., 40 , 15841591.

    • Search Google Scholar
    • Export Citation
  • Hartmann, D. L., , and F. Lo, 1998: Wave-driven zonal flow vacillation in the Southern Hemisphere. J. Atmos. Sci., 55 , 13031315.

  • Higgins, R. W., , and K. C. Mo, 1997: Persistent North Pacific circulation anomalies and the tropical intraseasonal oscillation. J. Climate, 10 , 223244.

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., , and W. Shi, 2001: Intercomparison of the principal modes of interannual and intraseasonal variability of the North American monsoon system. J. Climate, 14 , 403417.

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., , A. Leetmaa, , Y. Xue, , and A. Barnston, 2000: Dominant factors influencing the seasonal predictability of U.S. precipitation and surface air temperature. J. Climate, 13 , 39944017.

    • Search Google Scholar
    • Export Citation
  • Hoerling, M. P., , J. W. Hurrell, , T. Xu, , G. T. Bates, , and A. S. Phillips, 2004: Twentieth century North Atlantic climate change. Part II: Understanding the effect of Indian Ocean warming. Climate Dyn., 23 , 391405.

    • Search Google Scholar
    • Export Citation
  • Hsu, H-H., 1996: Global view of the intraseasonal oscillation during northern winter. J. Climate, 9 , 23862406.

  • Hsu, H-H., , B. J. Hoskins, , and F-F. Jin, 1990: The 1985/86 intraseasonal oscillation and the role of the extratropics. J. Atmos. Sci., 47 , 823839.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , M. P. Hoerling, , A. S. Phillips, , and T. Xu, 2004: Twentieth century North Atlantic climate change. Part I: Assessing determinism. Climate Dyn., 23 , 371389.

    • Search Google Scholar
    • Export Citation
  • Janowiak, J. E., , G. D. Bell, , and M. Chelliah, 1999: A Gridded Data Base of Daily Temperature Maxima and Minima for the Conterminous United States: 1948–1993. NCEP/Climate Prediction Center Atlas 6, 50 pp.

    • Search Google Scholar
    • Export Citation
  • Jeong, J-H., , C-H. Ho, , B-M. Kim, , and W-T. Kwon, 2005: Influence of the Madden–Julian Oscillation on wintertime surface air temperature and cold surges in east Asia. Geophys. Res. Lett., 110 .D11104, doi:10.1029/2004JD005408.

    • Search Google Scholar
    • Export Citation
  • Jones, C., 2000: Occurrence of extreme precipitation events in California and relationships with the Madden–Julian oscillation. J. Climate, 13 , 35763587.

    • Search Google Scholar
    • Export Citation
  • Kayano, M. T., , and V. E. Kousky, 1999: Intraseasonal (30–60 day) variability in the global tropics: Principal modes and their evolution. Tellus, 51A , 373386.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , and K. M. Weickmann, 1987: 30-60 day atmospheric oscillations: Composite life cycles of convection and circulation anomalies. Mon. Wea. Rev., 115 , 14071436.

    • Search Google Scholar
    • Export Citation
  • Lee, S., , and S. B. Feldstein, 1996: Mechanism of zonal index evolution in a two-layer model. J. Atmos. Sci., 53 , 22322246.

  • Li, S., , M. P. Hoerling, , and S. Peng, 2006: Coupled ocean–atmosphere response to Indian Ocean warmth. Geophys. Res. Lett., 33 .L07713, doi:10.1029/2005GL025558.

    • Search Google Scholar
    • Export Citation
  • Lorenz, D. J., , and D. L. Hartmann, 2003: Eddy-zonal flow feedback in the Northern Hemisphere winter. J. Climate, 16 , 12121227.

  • Madden, R. A., , and P. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28 , 702708.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., , and P. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29 , 11091123.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., , and P. Julian, 1994: Observations of the 40–50-day tropical oscillation—A review. Mon. Wea. Rev., 122 , 814837.

  • Maloney, E. D., , and D. L. Hartmann, 2000: Modulation of eastern North Pacific hurricanes by the Madden–Julian oscillation. J. Climate, 13 , 14511460.

    • Search Google Scholar
    • Export Citation
  • Matthews, A. J., , and M. P. Meredith, 2004: Variability of Antarctic circumpolar transport and the Southern Annular Mode associated with the Madden–Julian Oscillation. Geophys. Res. Lett., 31 .L24312, doi:10.1029/2004GL021666.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , S. Zhou, , and S-K. Yang, 2003: Relationship of the Arctic and Antarctic oscillations to outgoing longwave radiation. J. Climate, 16 , 15831592.

    • Search Google Scholar
    • Export Citation
  • Quadrelli, R., , and J. M. Wallace, 2002: Dependence of the structure of the Northern Hemisphere annular mode on the polarity of ENSO. Geophys. Res. Lett., 29 .2132, doi:10.1029/2002GL015807.

    • Search Google Scholar
    • Export Citation
  • Robinson, W., 1991: The dynamics of the zonal index in a simple model of the atmosphere. Tellus, 43A , 295305.

  • Rui, H., , and B. Wang, 1990: Development characteristics and dynamic structure of tropical intraseasonal convection anomalies. J. Atmos. Sci., 47 , 357379.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2006: The NCEP Climate Forecast System. J. Climate, 19 , 34833517.

  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 10001016.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2001: Regional climate impacts of the Northern Hemisphere annular mode. Science, 293 , 8589.

  • Thompson, D. W. J., , S. Lee, , and M. P. Baldwin, 2002: Atmospheric processes governing the Northern Hemisphere annular mode/North Atlantic Oscillation. The North Atlantic Oscillation: Climatic Significance an Environmental Impact, Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 1–35.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., , and S. J. S. Khalsa, 1990: The shift of convection from the Indian Ocean to the western Pacific Ocean during a 30–60 day oscillation. Mon. Wea. Rev., 118 , 964978.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., , G. R. Lussky, , and J. E. Kutzbach, 1985: Intraseasonal (30–60 day) fluctuations of outgoing longwave radiation and 250 mb streamfunction during northern winter. Mon. Wea. Rev., 113 , 941961.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., , and H. H. Hendon, 2004: An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132 , 19171932.

    • Search Google Scholar
    • Export Citation
  • Zhang, C., 2005: Madden–Julian oscillation. Rev. Geophys., 43 .RG2003, doi:10.1029/2004RG000158.

  • Zhou, S., , and A. J. Miller, 2005: The interaction of the Madden–Julian oscillation and the Arctic Oscillation. J. Climate, 18 , 143159.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    WH real-time multivariate MJO index shown in phase space (phases 1–8). Values outside of the unit circle project onto the leading PCs, RMM1 and RMM2. The location of the convectively active signal of the MJO is labeled on the perimeter (figure courtesy of Matthew Wheeler, Bureau of Meteorology Research Centre).

  • View in gallery

    (top) Percentage of MJO days (categorized by phase based on the WH index) when the AO index is in the positive polarity (gray shading) vs the negative polarity (black shading). The percentage difference between the positive and negative AO for MJO phase 4 exceeds the 95% significance level based on a Monte Carlo simulation of 1000 random samples. (bottom) As in (top) but showing the percentage of days when the tendency of the AO index is positive (gray) vs negative (black). MJO phase 7 exceeds the 95% significance level, and MJO phases 1, 2, and 6 exceed the 80% level. A 10-day running mean filter is applied to the AO index.

  • View in gallery

    (a) Composite difference of 300-hPa velocity potential anomalies during November–March where the average of positive AO days greater than one standard deviation is subtracted from negative AO days (>1σ). (b) As in (a) but for the difference between days when the MJO WH index is in phases 2, 3, and 4 subtracted from MJO phases 7 and 8. Velocity potential anomalies are divided by 103. (remaining panels) As in (top) but showing (c), (d) zonal-mean zonal wind, (e), (f) 1000-hPa geopotential height, and (g), (h) 2-m surface temperature. All data are from the NCEP–NCAR reanalysis.

  • View in gallery

    (a) Composite of pentad-averaged 300-hPa velocity potential anomalies with the 120-day running mean removed (NCEP–NCAR reanalysis) for November–March days prior to and following peaks in the positive polarity of the AO index (>1σ). Time step 0 represents the average of −2 days to +2 days, time step +1 is the average of +3 to +7 days, and so on. The pentad-averaged values of the AO index are also displayed to the side of the panels. The thick line represents the average of the individual events. (b) As in (a) but for troughs in the negative polarity of the AO index. (c), (d) As in (a), (b) but using model output from a CMIP free run of the CFS. Velocity potential anomalies are divided by 103. The anomalies displayed exceed the 95% significance level based on a two-tailed test of the t statistic.

  • View in gallery

    (top) Leading EOF shown as a regression map of monthly November–March 1000-hPa geopotential height anomalies (20.6% of variance explained). (middle) Regression map of the leading EOF for November–March days when the MJO WH index is in phases 2, 3, and 4 [18.4% of variance explained relative to the composite mean (or 3.6% relative to all November–March days)]; 1000-hPa geopotential height anomalies have been smoothed with a 10-day running mean filter. (bottom) As in (middle) but for MJO WH index days in phases 7 and 8 [20.8% of variance explained relative to the composite mean (or 3.5% relative to all November– March days)]; 7% of the leading EOFs created from 10 000 random subsets of the data exceeded the spread found between the first EOF (shown above) and second EOF (not shown).

  • View in gallery

    (top) Composite difference of November–March gridded co-op surface temperature (°C) for MJO days in phases 7 and 8 minus MJO days in phases 2, 3, and 4. Compare to Fig. 3h. (bottom) Using the area average of the grid box displayed in (top) (42°–45°N, 75°–70°W), the percent of November–March MJO days when the temperature is warmer (gray shading) and colder than average.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 253 253 30
PDF Downloads 106 106 8

Boreal Winter Links between the Madden–Julian Oscillation and the Arctic Oscillation

View More View Less
  • 1 Climate Prediction Center, Camp Springs, Maryland
© Get Permissions
Full access

Abstract

There is increasing evidence that the Madden–Julian oscillation (MJO) modifies the mid- to high-latitude circulation and, in particular, appears to have a relationship to the leading mode of extratropical variability, the Arctic Oscillation (AO). In this study, new insights into the observed similarities between the MJO and the AO are explored. It is shown that the eastward progression of the convectively active phase of the MJO is associated with a corresponding shift in the tendency and sign of the AO index. Moreover, the AO and the MJO share several analogous features not only in the global circulation, but also in surface temperature fields. Also, the AO is linked to a pattern of eastward-propagating MJO-like variability in the tropics that is partially reproduced in free runs of the NCEP Climate Forecast System (CFS) model. Finally, it is shown that the structure of the AO, as defined by the leading mode in the 1000-hPa geopotential height field, is significantly altered based on the phase of the MJO.

Corresponding author address: M. L. L’Heureux, Development Branch, Climate Prediction Center, NOAA/NWS/NCEP, 5200 Auth Road, Camp Springs, MD 20746. Email: michelle.lheureux@noaa.gov

Abstract

There is increasing evidence that the Madden–Julian oscillation (MJO) modifies the mid- to high-latitude circulation and, in particular, appears to have a relationship to the leading mode of extratropical variability, the Arctic Oscillation (AO). In this study, new insights into the observed similarities between the MJO and the AO are explored. It is shown that the eastward progression of the convectively active phase of the MJO is associated with a corresponding shift in the tendency and sign of the AO index. Moreover, the AO and the MJO share several analogous features not only in the global circulation, but also in surface temperature fields. Also, the AO is linked to a pattern of eastward-propagating MJO-like variability in the tropics that is partially reproduced in free runs of the NCEP Climate Forecast System (CFS) model. Finally, it is shown that the structure of the AO, as defined by the leading mode in the 1000-hPa geopotential height field, is significantly altered based on the phase of the MJO.

Corresponding author address: M. L. L’Heureux, Development Branch, Climate Prediction Center, NOAA/NWS/NCEP, 5200 Auth Road, Camp Springs, MD 20746. Email: michelle.lheureux@noaa.gov

1. Introduction

The Madden–Julian oscillation (MJO) is the dominant pattern of subseasonal (∼30–60 day) climate variability in the global tropics. It is characterized by a large-scale, eastward-propagating, zonal circulation cell that is observed within equatorial latitudes (Madden and Julian 1971, 1972, 1994). In the eastern hemisphere, a distinctive convective signal slowly shifts eastward (∼5 m s−1) over the warm waters of the Indian Ocean and western Pacific (Zhang 2005). As the MJO progresses over the cooler waters of the western hemisphere, the deep convection dissipates, but coherent and faster eastward propagation (∼15 m s−1) remains apparent in fields of upper- and lower-level winds and velocity potential (Knutson and Weickmann 1987; Weickmann and Khalsa 1990). Although the dynamical forcing of the MJO is unknown, it is well established that the MJO significantly influences tropical weather and climate (e.g., Rui and Wang 1990; Kayano and Kousky 1999; Maloney and Hartmann 2000; Higgins and Shi 2001).

Even though the primary MJO features are confined to the equatorial plane, many studies have noted that its associated circulations also impact extratropical latitudes. For example, Anderson and Rosen (1983) show that the midlatitudes receive anomalous atmospheric angular momentum from the tropics with a 40–50-day time scale. Weickmann et al. (1985) identified significant 30–60-day planetary-scale oscillations in outgoing longwave radiation (OLR) and 250-hPa streamfunction. Moreover, the MJO is explicitly identified as a global, not just tropical, phenomenon in studies by Knutson and Weickmann 1987, Hsu et al. 1990, and Hsu 1996, who note the modification of the midlatitude circulation as the tropical convective anomalies evolve.

While the MJO clearly appears in filtered global fields, recent work shows that the MJO has a distinctive signature within the unfiltered day-to-day and week-to-week variations of extratropical weather. When the MJO is active, a Rossby wave train may originate from the western Pacific, often contributing to anomalous precipitation along the western coast of North America (Higgins and Mo 1997; Higgins et al. 2000; Jones 2000). Also, Jeong et al. (2005) demonstrate a significant relationship between the MJO and the frequency of cold surges over eastern Asia.

The most pervasive MJO impacts in the NH occur during the boreal winter, which coincides with the seasonal amplification of the Arctic Oscillation (AO). The AO [also known as the Northern Annular Mode (NAM)] is the leading mode of mid- to high-latitude climate variability in the NH, and is distinguished by a high degree of zonal symmetry that corresponds to the north–south vacillation of the midlatitude jet (Thompson and Wallace 2000). The positive polarity of the AO denotes anomalous zonal-mean westerlies at ∼55°N and anomalous zonal-mean easterlies at ∼35°N (contraction of the midlatitude jet), whereas the negative polarity is defined by anomalies of the opposite sign (expansion of the midlatitude jet). The equivalent barotropic structure of the zonal-mean zonal wind anomalies is evidence of forcing by high-frequency anomalies of the eddy momentum flux in the midlatitudes (Hartmann and Lo 1998; Lorenz and Hartmann 2003). A positive feedback between the eddy momentum fluxes and the zonal-mean flow may explain the persistence of the AO on weekly, monthly, and seasonal time scales (Robinson 1991; Lee and Feldstein 1996; Thompson et al. 2002). The AO also reflects the transfer of mass between mid- and high latitudes. Contraction (expansion) of the midlatitude jet leads to less (more) frequent cold air outbreaks over the midlatitudes (Thompson and Wallace 2001; Higgins et al. 2000).

Recently, there has been increasing interest in the possible wintertime relationship between the MJO and the extratropical AO. For example, Hsu (1996) shows that the MJO is linked to a standing oscillation in the tropics and a related equivalent barotropic, zonally symmetric structure that propagates poleward. Miller et al. (2003) and Zhou and Miller (2005) demonstrate that when MJO-related convection is enhanced (diminished) over the Indian Ocean, the AO index tends to favor its high (low) polarity. They argue that Rossby wave trains associated with the MJO affect the North Pacific sector, thereby influencing the polarity of the AO index. Also, Matthews and Meredith (2004) note that the Southern Hemisphere parallel to the AO, the Southern Annular Mode (SAM), is related to MJO modified convection in the Indian Ocean during the austral winter.

The purpose of this work is to highlight the remarkable similarities between the signature of the MJO and the AO during an extended boreal winter (November–March). We build upon previous work by showing 1) that the eastward progression of the convectively active phase of the MJO is associated with a corresponding shift in the tendency of the AO index; 2) that the AO and the MJO share several analogous features not only in the extratropical NH circulation but also in surface temperature fields; 3) that the AO is linked to a pattern of eastward-propagating, MJO-like variability in the tropics that is partially reproduced in the National Centers for Environmental Prediction (NCEP) Climate Forecast System (CFS) model; and 4) that the structure of the AO in the geopotential height field significantly varies based on the phase of the MJO. We will then comment on the potential to exploit the correspondence between the AO and MJO to improve subseasonal climate forecasts for the contiguous United States.

The data and methodology are given in section 2. Similarities between the MJO and the AO are described in section 3. A summary and discussion is presented in section 4.

2. Data and methods

The daily MJO index used in this study is described in Wheeler and Hendon (2004) (hereafter WH MJO index) and is obtained from the Bureau of Meteorology Web site (http://www.bom.gov.au/bmrc/clfor/cfstaff/matw/maproom/RMM/). The amplitude and phase, or location, of the MJO is determined by calculating a combined empirical orthogonal function (EOF) of equatorial zonal-mean 850-hPa zonal wind, 200-hPa zonal wind, and OLR. Before the combined EOF is computed, variability linked to the El Niño–Southern Oscillation (SST1 index) is linearly removed, as well as the 120-day mean of the most recent 120 days at each point. Because the index uses satellite-derived OLR, the analysis is restricted to the period 1979–2006. The two leading principal components (PCs), known as Real-Time Multivariate MJO series 1 and 2 (RMM1 and RMM2), respectively, are a pair in approximate quadrature, representing the eastward propagation of the MJO. While the index shows spectral peaks at 30–60 days, the spectrum does provide evidence that the WH index projects onto slightly faster-moving wave-1 Kelvin waves, potentially slightly inflating the number of canonical MJO days (M. Wheeler 2006, personal communication). The amplitude and phase information is obtained from RMM1 and RMM2, which indicate the location of the convectively active component of the wave through eight phases, 1–8 (Fig. 1).

In this study, MJO days and events are identified using a pentad-averaged version of the WH index subject to three major requirements: (i) the index must have an amplitude [(RMM12 + RMM22)1/2] greater than one for consecutive pentads; (ii) the index must include phases that are in numerical order (i.e., phases 6, 7, 8, 1, 2); and (iii) events must continue for at least six consecutive pentads (30 days) and cannot remain in a particular phase for more than four pentads (20 days). A select few cases in which the amplitude of one pentad dips slightly below one are included as they were clearly part of a larger MJO event.

We also make use of the daily AO index, available from (National Oceanic and Atmospheric Administration) NOAA/Climate Prediction Center (http://www.cpc.ncep.noaa.gov). The daily index is obtained by projecting daily data onto the first EOF of monthly 1000-hPa geopotential height anomalies from the NCEP–National Center for Atmospheric Research (NCAR) reanalysis. Daily values of 1000-hPa geopotential height, 300-mb velocity potential, zonal-mean zonal wind, and 2-m temperature are also obtained from the NCEP–NCAR reanalysis. Surface air temperature data are obtained from the gridded daily cooperative dataset of Janowiak et al. (1999). Finally, a 100-yr free run of the NCEP Climate Forecast System (CFS), initialized 1 January 2002, is obtained from the T126 Coupled Model Intercomparison Project (CMIP-1) (http://cfs.ncep.noaa.gov/). Documentation of the CFS is found in Saha et al. (2006). While the objective of this study is not to assess model biases in the CFS, we make use of CFS output in order to evaluate the reproducibility of the results and identify potential model errors.

Significance is assessed by using Monte Carlo simulations and by subdividing composites to ensure that the anomalies remain stable in both halves of the dataset. Because the MJO daily index stays in a particular phase for several consecutive days, the actual degrees of freedom must be reduced. For example, during the period November–March 1979–2006 (27 seasons), the WH index is in phases 2, 3, and 4 for 833 days, but there are only 58 unique MJO events. The Monte Carlo test used to determine significance in Fig. 2 reflects the random selection of N-days for each phase based on the number of nonconsecutive, independent MJO events (see Table 1). Based on 1000 random samplings, statistical significance is evaluated by computing how often the spread exceeds the percentage difference between the positive and negative values of the AO.

Filtering is not used on the atmospheric and surface variables, except for the regression maps of geopotential height (Fig. 5) and the velocity potential anomalies (Fig. 4). The 1000-hPa geopotential height field is smoothed using a 10-day running-mean filter to isolate the low-frequency (greater than 10 day) component of the AO. This smoothing has the effect of increasing the spread between the first and second EOFs (see Fig. 5 for the fraction of explained variance). A Monte Carlo test was performed using randomly drawn data and shows that the spread between the first and second EOFs was exceeded only 7% of the time.

3. Results

The sign and tendency of the AO index is evaluated as a function of the phase of the MJO using the WH index in Fig. 2; a 10-day running-mean filter has been applied to the AO index. The AO tendency, or slope, is computed as the difference of the AO on day t + 1 and the AO on day t – 1, divided by 2. Significance is assessed using a Monte Carlo test described in the previous section and is noted in the figure caption. The results shown in the top panel of Fig. 2 reveal that the AO is more likely to be positive on days when the convectively active phase of the MJO is in the eastern hemisphere (phases 2, 3, and 4). By contrast, the negative polarity of the AO is preferred when the convectively active phase of the MJO is located over the western hemisphere (phases 7, 8). These results are consistent with Zhou and Miller (2005), who found that the high (low) AO phase was more likely when convection was enhanced (suppressed) over the Indian Ocean. The bottom panel of Fig. 2 reveals a clear evolution from a positive AO tendency when the MJO is in phases 1, 2, and 3 to a negative AO tendency when the MJO is in phases 6 and 7 (the AO tendency in phases 1, 2, 5, 6, and 7 is evident in both the first and second halves of the dataset).

Evidently there is a significant temporal relationship between the AO and the MJO, but to what extent are the leading modes spatially analogous? Motivated by the findings in Fig. 2, the MJO is separated into two categories: 1) phases 2/3/4, when enhanced convection is located over the eastern hemisphere, and 2) phases 7/8, when convection is located over the western hemisphere. Note that while phase 7/8 is referred to as “convectively active” over the western hemisphere, upper-level divergence is not characterized by the deep convection that is present in the eastern hemisphere. Category 1) has a sample size of 833 days and category 2) includes 612 days. To examine the relationships between the MJO and the circulation, composite difference maps of several fields are constructed as the difference of category (1) and (2) composites (Fig. 3). In a similar manner, composite difference maps are constructed by subtracting AO positive polarity days greater than 1σ from AO negative polarity days greater than 1σ. Significance is assessed by ensuring that the anomalies remain in subdivisions of the dataset (not shown).

AO anomalies and MJO anomalies of 300-hPa velocity potential (Figs. 3a,b), zonal-mean zonal wind (Figs. 3c,d), 1000-hPa geopotential height (Figs. 3e,f), and 2-m surface temperature (Figs. 3g,h) from the NCEP–NCAR reanalysis are shown in Fig. 3. Figures 3a,c,g,e emphasize the negative polarity of the AO while Figs. 3b,d,f,h show the anomalies associated with the MJO when convection is located in the western hemisphere. As expected, the velocity potential anomalies related to the MJO have a planetary, wave-one structure centered on the equator, with upper-level divergence and convergence located over the western and eastern hemispheres, respectively (Fig. 3b). The AO composite shows upper-level convergence at polar latitudes, which is consistent with the higher-than-average surface pressure linked to a low-index AO (Fig. 3a). Less expected are the extensive planetary-scale, wave-one velocity potential anomalies that extend into the tropics and that are collocated with the MJO anomalies in the phases shown.

There are also pronounced similarities between the MJO and anomalies in the mid- to high latitudes, which are typically associated with the AO. For example, the negative polarity of the AO is linked to an equivalent barotropic zonal-mean zonal wind dipole of anomalous easterlies at roughly 55°N and anomalous westerlies at roughly 35°N (Thompson and Wallace 2000). Similarly, when MJO enhanced convection is in the western hemisphere, an extratropical zonal wind dipole aligns with the node of the AO (Figs. 3c,d). In addition the AO is characterized by a zonally symmetric signal in the 1000-hPa geopotential height field from 20° to 90°N (Fig. 3e). Correspondingly the MJO-related circulation is associated with a considerable degree of hemispheric symmetry, particularly in the Atlantic and Pacific sectors—only the polar cap anomaly is not prominent (Fig. 3f). Finally, it is well known that the low-index phase of the AO is tied to lower-than-average temperature anomalies in the midlatitudes (Fig. 3g) as discussed in Thompson and Wallace (2001). Remarkably, the MJO influences extratropical temperature anomalies in nearly the same way, albeit weaker in amplitude, but still impacting vast areas of Eurasia, Canada, and the north-central and northeast United States. Collectively, the results of Fig. 3 indicate that MJO-related enhanced convection in the western hemisphere tends to be associated with the low-index polarity of the AO.

The global wave-one signatures of the 300-hPa velocity potential anomalies associated with the AO for days prior to and following the maxima/minima in the AO are shown in Fig. 4. Distinctive peaks/troughs in the AO index greater than 1σ are isolated from the November–March record for a total of 30–32 samples. The velocity potential anomalies and AO index are averaged into pentads, so that time step zero represents the average of days −2 to +2, time step one represents days +3 to +7, and so on. Only velocity potential anomalies significant at the 95% confidence level, based on a two-tailed t test with a null hypothesis that the sample means are significantly different from the November–March mean state, are shown. The velocity potential anomalies have the 120-day running mean removed to isolate the subseasonal signal. Figures 4a,b show pentad anomalies based on the NCEP–NCAR reanalysis while Figs. 4c,d show anomalies based on the CFS CMIP-1 free run.

The negative polarity of the AO (Fig. 4b) shows anomalous upper-level convergence centered over the Maritime Continent with an adjacent region of anomalous divergence over the western hemisphere (analogous to MJO phase 7/8). The positive polarity of the AO (Fig. 4a) is characterized by anomalies of the opposite sign (analogous to MJO phase 2/3/4). Both the NCEP–NCAR reanalysis (Figs. 4a,b) and CFS free run (Figs. 4c,d) replicate the sign of the wave-one anomalies. However, most striking is the apparent eastward movement that is captured using data from the NCEP–NCAR reanalysis (Figs. 4a,b). While the anomalies are not as robust by the third time step, it is still apparent from the days leading up to the peak/trough in the AO to the days that follow (across 34 days) that there is an eastward shift in the location of the velocity potential anomalies. There are also hints of an eastward shift in the CFS free run, but they are not as clear and may be more indicative of the model lacking the relevant and unknown physics. Thus, by keying in on the maxima/minima in the AO index, it is possible to isolate wave-one, eastward progression that is reminiscent of the MJO.

So far, the focus has been on changes to the mean state during different stages of the MJO and AO. However, another way to examine the relationship between the AO and the MJO is to examine how the structure of the AO, as the leading pattern of Northern Hemisphere variability, fluctuates depending on the phase of the MJO. Quadrelli and Wallace (2002) demonstrate that the signature of the AO is significantly different based on the state of ENSO. The top panel of Fig. 5 shows the AO as it is typically defined—the leading EOF in the monthly mean, 1000-hPa geopotential height field (November–March). After applying a 10-point running-mean filter to daily data, the leading EOF in 1000-hPa geopotential height is computed for days when the MJO is in phase 2/3/4 (middle panel) and when the MJO is in phase 7/8 (bottom panel). During MJO phase 2/3/4 (convection mostly over the eastern hemisphere), the structure of the AO changes so that the greatest amplitudes are found in the polar cap anomaly, which becomes more pronounced over the Siberian sector (Fig. 5, middle panel). Moreover, the midlatitude anomalies become weaker and the Pacific center of action vanishes. This is in contrast to the AO during MJO phase 7/8 (convection primarily over the western hemisphere), where the structure better resembles the canonical AO: the Pacific center of action reemerges and the polar center of action favors the Atlantic sector (Fig. 5, top and bottom panels).

4. Summary and implications

The results presented in this study confirm the striking similarities between the anomalies of the AO and MJO during boreal winter, particularly in mid- and high latitudes. The analyses capture several unique aspects of the relationship between the AO and MJO:

  1. The progression of the MJO, as measured by the WH index, through phases 1–8 corresponds to a nearly linear response in the AO index. Specifically, the tendency of the AO is gradually more positive through phases 1, 2, and 3, resulting in a preference for positive polarity of the AO during MJO phase 2/3/4. An increasingly negative AO tendency is apparent in phases 6 and 7 and culminates in the enhanced likelihood for the negative polarity of the AO in MJO phases 7 and 8.
  2. Several extratropical atmospheric and surface anomalies associated with the AO project onto the anomalies of the MJO. When the MJO shifts into the western hemisphere (phases 7/8), anomalies of geopotential height, zonal-mean zonal wind, and surface temperature resemble the negative polarity of the AO. Even though the anomalies of the MJO confined to the tropics generally vary across longitude, the related anomalies in the extratropics exhibit a high degree of hemispheric symmetry.
  3. The AO exhibits an anomalous planetary wave-one signature in velocity potential that is commonly associated with the MJO. Composites keyed onto the days prior to and following maxima and minima of the AO index reveal an eastward shift in the velocity potential anomalies. CFS CMIP free runs reproduce the sign of the velocity potential anomalies, but do not capture the eastward movement sufficiently.
  4. The structure of the AO is significantly altered depending on the location of the convectively active phase of the MJO. During phases 2/3/4, the North Pacific center of the AO is nonexistent, and the polar cap anomaly has larger amplitude over the Siberian sector. In contrast, during phases 7/8, the canonical structure of the AO reemerges with two distinct centers over the Atlantic and Pacific. The return of the Pacific center may be the result of the eastward extension of the jet stream that occurs as the MJO shifts into the western hemisphere during phases 7/8 (Higgins and Mo 1997). Relying on modeling studies, Quadrelli and Wallace (2002) hypothesize that the Pacific center might be more amplified because of the existence of a more pronounced jet exit region.

These results are also interesting in the context of several studies that suggest a linkage on decadal time scales between coupled ocean–atmosphere interactions in the Indian Ocean, where the MJO has its greatest signal, and variability in the North Atlantic, where the AO has its strongest teleconnection (Hoerling et al. 2004; Hurrell et al. 2004). Li et al. (2006) use models that show that warming in the Indian Ocean can force the positive polarity of the North Atlantic Oscillation (NAO). While the mechanism, or even the direction, of the MJO and AO relationship remains unclear, further diagnostic studies might be able to elucidate the processes that result in a causal relationship on intraseasonal time scales.

Given that the MJO evolves on a time scale of 30–60 days, there is the enticing possibility that changes in the AO may be predicted beyond the skill of current numerical weather prediction models, perhaps out to weeks 3–4. For example, the negative polarity of the AO is commonly associated with an enhanced likelihood of cold air outbreaks over the United States (Thompson and Wallace 2001). A composite of temperature anomalies for the United States (Fig. 6), which shows the sign of the MJO in phases 7/8 (cf. Fig. 3h), indicates that anomalously cold (warm) air is prevalent over the upper Midwest and New England (western United States) during November–March. These anomalies are also evident in both the first and second halves of the data record. The grid points in a selected region (42°–45°N, 75°–70°W) in New England (Fig. 6a), when averaged, show an increased chance of below average temperatures during MJO phase 8, and an increased likelihood of above average temperatures during MJO phases 3–6 (Fig. 6b). Potentially, composites based on the predicted phase of the MJO could be integrated into the suite of real-time MJO monitoring tools to forecast the probability of temperature anomalies over the United States.

While the dynamics behind the onset and demise of the MJO are not well understood, these results suggest that once an MJO emerges, there are systematic changes to the extratropical circulation as it evolves. Future work will focus on elucidating the MJO–AO relationship to determine the extent of subseasonal predictability that can be exploited.

Acknowledgments

We are grateful for the insight imparted by the reviewers of this manuscript: Huug van den Dool, Mike Halpert, David W. J. Thompson, Klaus Weickmann, and one anonymous reviewer. Also thanks to Jon Gottschalck and Vernon Kousky for providing helpful comments throughout the course of this study.

REFERENCES

  • Anderson, J. R., , and R. D. Rosen, 1983: The latitude-height structure of 40–50 day variations in atmospheric angular momentum. J. Atmos. Sci., 40 , 15841591.

    • Search Google Scholar
    • Export Citation
  • Hartmann, D. L., , and F. Lo, 1998: Wave-driven zonal flow vacillation in the Southern Hemisphere. J. Atmos. Sci., 55 , 13031315.

  • Higgins, R. W., , and K. C. Mo, 1997: Persistent North Pacific circulation anomalies and the tropical intraseasonal oscillation. J. Climate, 10 , 223244.

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., , and W. Shi, 2001: Intercomparison of the principal modes of interannual and intraseasonal variability of the North American monsoon system. J. Climate, 14 , 403417.

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., , A. Leetmaa, , Y. Xue, , and A. Barnston, 2000: Dominant factors influencing the seasonal predictability of U.S. precipitation and surface air temperature. J. Climate, 13 , 39944017.

    • Search Google Scholar
    • Export Citation
  • Hoerling, M. P., , J. W. Hurrell, , T. Xu, , G. T. Bates, , and A. S. Phillips, 2004: Twentieth century North Atlantic climate change. Part II: Understanding the effect of Indian Ocean warming. Climate Dyn., 23 , 391405.

    • Search Google Scholar
    • Export Citation
  • Hsu, H-H., 1996: Global view of the intraseasonal oscillation during northern winter. J. Climate, 9 , 23862406.

  • Hsu, H-H., , B. J. Hoskins, , and F-F. Jin, 1990: The 1985/86 intraseasonal oscillation and the role of the extratropics. J. Atmos. Sci., 47 , 823839.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , M. P. Hoerling, , A. S. Phillips, , and T. Xu, 2004: Twentieth century North Atlantic climate change. Part I: Assessing determinism. Climate Dyn., 23 , 371389.

    • Search Google Scholar
    • Export Citation
  • Janowiak, J. E., , G. D. Bell, , and M. Chelliah, 1999: A Gridded Data Base of Daily Temperature Maxima and Minima for the Conterminous United States: 1948–1993. NCEP/Climate Prediction Center Atlas 6, 50 pp.

    • Search Google Scholar
    • Export Citation
  • Jeong, J-H., , C-H. Ho, , B-M. Kim, , and W-T. Kwon, 2005: Influence of the Madden–Julian Oscillation on wintertime surface air temperature and cold surges in east Asia. Geophys. Res. Lett., 110 .D11104, doi:10.1029/2004JD005408.

    • Search Google Scholar
    • Export Citation
  • Jones, C., 2000: Occurrence of extreme precipitation events in California and relationships with the Madden–Julian oscillation. J. Climate, 13 , 35763587.

    • Search Google Scholar
    • Export Citation
  • Kayano, M. T., , and V. E. Kousky, 1999: Intraseasonal (30–60 day) variability in the global tropics: Principal modes and their evolution. Tellus, 51A , 373386.

    • Search Google Scholar
    • Export Citation
  • Knutson, T. R., , and K. M. Weickmann, 1987: 30-60 day atmospheric oscillations: Composite life cycles of convection and circulation anomalies. Mon. Wea. Rev., 115 , 14071436.

    • Search Google Scholar
    • Export Citation
  • Lee, S., , and S. B. Feldstein, 1996: Mechanism of zonal index evolution in a two-layer model. J. Atmos. Sci., 53 , 22322246.

  • Li, S., , M. P. Hoerling, , and S. Peng, 2006: Coupled ocean–atmosphere response to Indian Ocean warmth. Geophys. Res. Lett., 33 .L07713, doi:10.1029/2005GL025558.

    • Search Google Scholar
    • Export Citation
  • Lorenz, D. J., , and D. L. Hartmann, 2003: Eddy-zonal flow feedback in the Northern Hemisphere winter. J. Climate, 16 , 12121227.

  • Madden, R. A., , and P. Julian, 1971: Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28 , 702708.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., , and P. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40–50 day period. J. Atmos. Sci., 29 , 11091123.

    • Search Google Scholar
    • Export Citation
  • Madden, R. A., , and P. Julian, 1994: Observations of the 40–50-day tropical oscillation—A review. Mon. Wea. Rev., 122 , 814837.

  • Maloney, E. D., , and D. L. Hartmann, 2000: Modulation of eastern North Pacific hurricanes by the Madden–Julian oscillation. J. Climate, 13 , 14511460.

    • Search Google Scholar
    • Export Citation
  • Matthews, A. J., , and M. P. Meredith, 2004: Variability of Antarctic circumpolar transport and the Southern Annular Mode associated with the Madden–Julian Oscillation. Geophys. Res. Lett., 31 .L24312, doi:10.1029/2004GL021666.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , S. Zhou, , and S-K. Yang, 2003: Relationship of the Arctic and Antarctic oscillations to outgoing longwave radiation. J. Climate, 16 , 15831592.

    • Search Google Scholar
    • Export Citation
  • Quadrelli, R., , and J. M. Wallace, 2002: Dependence of the structure of the Northern Hemisphere annular mode on the polarity of ENSO. Geophys. Res. Lett., 29 .2132, doi:10.1029/2002GL015807.

    • Search Google Scholar
    • Export Citation
  • Robinson, W., 1991: The dynamics of the zonal index in a simple model of the atmosphere. Tellus, 43A , 295305.

  • Rui, H., , and B. Wang, 1990: Development characteristics and dynamic structure of tropical intraseasonal convection anomalies. J. Atmos. Sci., 47 , 357379.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and Coauthors, 2006: The NCEP Climate Forecast System. J. Climate, 19 , 34833517.

  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 10001016.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2001: Regional climate impacts of the Northern Hemisphere annular mode. Science, 293 , 8589.

  • Thompson, D. W. J., , S. Lee, , and M. P. Baldwin, 2002: Atmospheric processes governing the Northern Hemisphere annular mode/North Atlantic Oscillation. The North Atlantic Oscillation: Climatic Significance an Environmental Impact, Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 1–35.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., , and S. J. S. Khalsa, 1990: The shift of convection from the Indian Ocean to the western Pacific Ocean during a 30–60 day oscillation. Mon. Wea. Rev., 118 , 964978.

    • Search Google Scholar
    • Export Citation
  • Weickmann, K. M., , G. R. Lussky, , and J. E. Kutzbach, 1985: Intraseasonal (30–60 day) fluctuations of outgoing longwave radiation and 250 mb streamfunction during northern winter. Mon. Wea. Rev., 113 , 941961.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M. C., , and H. H. Hendon, 2004: An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132 , 19171932.

    • Search Google Scholar
    • Export Citation
  • Zhang, C., 2005: Madden–Julian oscillation. Rev. Geophys., 43 .RG2003, doi:10.1029/2004RG000158.

  • Zhou, S., , and A. J. Miller, 2005: The interaction of the Madden–Julian oscillation and the Arctic Oscillation. J. Climate, 18 , 143159.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

WH real-time multivariate MJO index shown in phase space (phases 1–8). Values outside of the unit circle project onto the leading PCs, RMM1 and RMM2. The location of the convectively active signal of the MJO is labeled on the perimeter (figure courtesy of Matthew Wheeler, Bureau of Meteorology Research Centre).

Citation: Journal of Climate 21, 12; 10.1175/2007JCLI1955.1

Fig. 2.
Fig. 2.

(top) Percentage of MJO days (categorized by phase based on the WH index) when the AO index is in the positive polarity (gray shading) vs the negative polarity (black shading). The percentage difference between the positive and negative AO for MJO phase 4 exceeds the 95% significance level based on a Monte Carlo simulation of 1000 random samples. (bottom) As in (top) but showing the percentage of days when the tendency of the AO index is positive (gray) vs negative (black). MJO phase 7 exceeds the 95% significance level, and MJO phases 1, 2, and 6 exceed the 80% level. A 10-day running mean filter is applied to the AO index.

Citation: Journal of Climate 21, 12; 10.1175/2007JCLI1955.1

Fig. 3.
Fig. 3.

(a) Composite difference of 300-hPa velocity potential anomalies during November–March where the average of positive AO days greater than one standard deviation is subtracted from negative AO days (>1σ). (b) As in (a) but for the difference between days when the MJO WH index is in phases 2, 3, and 4 subtracted from MJO phases 7 and 8. Velocity potential anomalies are divided by 103. (remaining panels) As in (top) but showing (c), (d) zonal-mean zonal wind, (e), (f) 1000-hPa geopotential height, and (g), (h) 2-m surface temperature. All data are from the NCEP–NCAR reanalysis.

Citation: Journal of Climate 21, 12; 10.1175/2007JCLI1955.1

Fig. 4.
Fig. 4.

(a) Composite of pentad-averaged 300-hPa velocity potential anomalies with the 120-day running mean removed (NCEP–NCAR reanalysis) for November–March days prior to and following peaks in the positive polarity of the AO index (>1σ). Time step 0 represents the average of −2 days to +2 days, time step +1 is the average of +3 to +7 days, and so on. The pentad-averaged values of the AO index are also displayed to the side of the panels. The thick line represents the average of the individual events. (b) As in (a) but for troughs in the negative polarity of the AO index. (c), (d) As in (a), (b) but using model output from a CMIP free run of the CFS. Velocity potential anomalies are divided by 103. The anomalies displayed exceed the 95% significance level based on a two-tailed test of the t statistic.

Citation: Journal of Climate 21, 12; 10.1175/2007JCLI1955.1

Fig. 5.
Fig. 5.

(top) Leading EOF shown as a regression map of monthly November–March 1000-hPa geopotential height anomalies (20.6% of variance explained). (middle) Regression map of the leading EOF for November–March days when the MJO WH index is in phases 2, 3, and 4 [18.4% of variance explained relative to the composite mean (or 3.6% relative to all November–March days)]; 1000-hPa geopotential height anomalies have been smoothed with a 10-day running mean filter. (bottom) As in (middle) but for MJO WH index days in phases 7 and 8 [20.8% of variance explained relative to the composite mean (or 3.5% relative to all November– March days)]; 7% of the leading EOFs created from 10 000 random subsets of the data exceeded the spread found between the first EOF (shown above) and second EOF (not shown).

Citation: Journal of Climate 21, 12; 10.1175/2007JCLI1955.1

Fig. 6.
Fig. 6.

(top) Composite difference of November–March gridded co-op surface temperature (°C) for MJO days in phases 7 and 8 minus MJO days in phases 2, 3, and 4. Compare to Fig. 3h. (bottom) Using the area average of the grid box displayed in (top) (42°–45°N, 75°–70°W), the percent of November–March MJO days when the temperature is warmer (gray shading) and colder than average.

Citation: Journal of Climate 21, 12; 10.1175/2007JCLI1955.1

Table 1.

The number of discrete MJO events found within each phase during the November–March 1979–2006 time period. The MJO events are extracted from Wheeler and Hendon’s real-time multivariate MJO index.

Table 1.
Save