• Chang, E. K. N., 1993: Downstream development of baroclinic waves as inferred from regression analysis. J. Atmos. Sci., 50 , 20382053.

    • Search Google Scholar
    • Export Citation
  • Chang, E. K. N., 2001: GCM and observational diagnoses of the seasonal and interannual variations of the Pacific storm track during the cool season. J. Atmos. Sci., 58 , 17841800.

    • Search Google Scholar
    • Export Citation
  • Chang, E. K. N., , and I. Orlanski, 1993: On the dynamics of a storm track. J. Atmos. Sci., 50 , 9991015.

  • Chang, E. K. N., , and D. B. Yu, 1999: Characteristics of wave packets in the upper troposphere. Part I: Northern Hemisphere winter. J. Atmos. Sci., 56 , 17081728.

    • Search Google Scholar
    • Export Citation
  • Deng, Y., , and M. Mak, 2005: An idealized model study relevant to the dynamics of the midwinter minimum of the Pacific storm track. J. Atmos. Sci., 62 , 12091225.

    • Search Google Scholar
    • Export Citation
  • Hakim, G. J., 2003: Developing wave packets in the North Pacific storm track. Mon. Wea. Rev., 131 , 28242837.

  • Harnik, N., , and E. K. M. Chang, 2004: The effects of variations in jet width on the growth of baroclinic waves: Implications for midwinter Pacific storm track variability. J. Atmos. Sci., 61 , 2340.

    • Search Google Scholar
    • Export Citation
  • Held, I. M., , M. Ting, , and H. Wang, 2002: Northern winter stationary waves: Theory and modeling. J. Climate, 15 , 21252144.

  • James, I. N., 1987: Suppression of baroclinic instability in horizontally sheared flows. J. Atmos. Sci., 44 , 37103720.

  • Kitoh, A., 2007: ENSO modulation by mountain uplift. Climate Dyn., 28 , 781796.

  • Lee, S., , and I. M. Held, 1993: Baroclinic wave packets in models and observations. J. Atmos. Sci., 50 , 14131428.

  • Lee, W-J., , and M. Mak, 1996: The role of orography in the dynamics of storm tracks. J. Atmos. Sci., 53 , 17371750.

  • Manabe, S., , and T. B. Terpstra, 1974: The effects of mountains on the general circulation of the atmosphere as identified by numerical experiments. J. Atmos. Sci., 31 , 342.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., 1992: Midwinter suppression of baroclinic wave activity in the Pacific. J. Atmos. Sci., 49 , 16291642.

  • Nakamura, H., , and T. Sampe, 2002: Trapping of synoptic-scale disturbances into the North Pacific subtropical jet core in midwinter. Geophys. Res. Lett., 29 , 1761. doi:10.1029/2002GL015535.

    • Search Google Scholar
    • Export Citation
  • Orlanski, I., , and J. Katzfey, 1991: The life cycle of a cyclone wave in the Southern Hemisphere. Part I: Eddy energy budget. J. Atmos. Sci., 48 , 19721998.

    • Search Google Scholar
    • Export Citation
  • Orlanski, I., , and E. K. M. Chang, 1993: Ageostrophic geopotential fluxes in downstream and upstream development of baroclinic waves. J. Atmos. Sci., 50 , 212225.

    • Search Google Scholar
    • Export Citation
  • Peixoto, J., , and A. Oort, 1992: Physics of Climate. American Institute of Physics, 520 pp.

  • Penny, S., , G. H. Roe, , and D. S. Battisti, 2009: The source of the midwinter suppression in storminess over the North Pacific. J. Climate, 23 , 634648.

    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., , N. A. Rayner, , T. M. Smith, , D. C. Stokes, , and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15 , 16091625.

    • Search Google Scholar
    • Export Citation
  • Robinson, D. P., , R. B. Black, , and B. A. McDaniel, 2006: A Siberian precursor to midwinter intraseasonal variability in the North Pacific storm track. Geophys. Res. Lett., 33 , L15811. doi:10.1029/2006GL026458.

    • Search Google Scholar
    • Export Citation
  • Shaw, T. A., , M. Sigmond, , T. G. Shepherd, , and J. F. Scinocca, 2009: Sensitivity of simulated climate to conservation of momentum in gravity wave drag parameterization. J. Climate, 22 , 27262742.

    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., , and B. J. Hoskins, 1979: The downstream and upstream development of unstable baroclinic waves. J. Atmos. Sci., 36 , 12391254.

    • Search Google Scholar
    • Export Citation
  • Son, S-W., , M. Ting, , and L. M. Polvani, 2009: The effect of topography on storm-track intensity in a relatively simple general circulation model. J. Atmos. Sci., 66 , 393411.

    • Search Google Scholar
    • Export Citation
  • Vallis, G. K., , and E. P. Gerber, 2008: Local and hemispheric dynamics of the North Atlantic Oscillation, annular patterns and the zonal index. Dyn. Atmos. Oceans, 44 , 184212.

    • Search Google Scholar
    • Export Citation
  • Yu, J-Y., , and D. L. Hartmann, 1995: Orographic influences on the distribution and generation of atmospheric variability in a GCM. J. Atmos. Sci., 52 , 24282443.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. Q., , and I. M. Held, 1999: A linear stochastic model of a GCM’s midlatitude storm tracks. J. Atmos. Sci., 56 , 34163435.

  • Zurita-Gotor, P., , and E. K. M. Chang, 2005: The impact of zonal propagation and seeding on the eddy–mean flow equilibrium of a zonally varying two-layer model. J. Atmos. Sci., 62 , 22612273.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (left) Boundary conditions for the CCM3 simulations. Mountains higher than 500 m are shaded. (right) Seasonal march of the upper-level (300 hPa) North Pacific storminess (m) averaged between 140°E and 140°W, for the (a) M100, (b) M75, (c) M50, and (d) M20 experiments.

  • View in gallery

    Zonal wind averaged over the longitudinal interval of 60°–100°E for (a) January and (b) October simulated by CCM3.

  • View in gallery

    (a) Seasonal march of the low-level baroclinicity (day−1) and (b) baroclinic conversion (W m−2) over the North Pacific, averaged between 140°E and 140°W, for (left) M100 and (right) M50.

  • View in gallery

    Midwinter (16 Dec–15 Feb mean) anomalous (a) 300-hPa storminess (shading, m) and (b) baroclinic conversion (shading, W m−2) calculated from the differences between M100 and M50 (M100 − M50). The contour lines indicate climatological mean storminess in (a) and baroclinic conversion for M50 in (b).

  • View in gallery

    Same as Fig. 4 but for anomalous (a) mean-amplitude (shading, meters) and (b) frequency (shading, day) of 300-hPa storminess calculated from the differences between M100 and M50 (M100 − M50). See the text for the details on defining the amplitude and frequency. The contour lines indicate climatological mean amplitude in (a) and frequency for M50 in (b).

  • View in gallery

    Sensitivities of midlatitude (32°–67°N), zonal-mean fractional changes of (a) high-pass filtered BCC and (b) unfiltered total BCC (both shown as squares with solid lines) to the varying central Asian mountains. The zonal-mean fractional changes of 300-hPa storminess are plotted as circles with dotted lines for comparison.

  • View in gallery

    The seasonal march of (a) baroclinic conversion (W m−2) and (b) 300-hPa storminess (m) over the eastern Eurasian continent (60°–140°E mean), for (left) M100 and (right) M50.

  • View in gallery

    (a) The 300-hPa eddy streamfunction for M100 in January. (b)–(d) Anomalous eddy streamfunction calculated from the differences between M100 and others: (b) M100 − M75, (c) M100 − M50, and (d) M100 − M20 in January. The contour interval is 3 × 106 m2 s−1.

  • View in gallery

    Midwinter (16 Dec–15 Feb mean) anomalous (a) 8-day high-pass filtered transient EKE (shading, m2 s−2) and (b) 100-day low-pass filtered stationary EKE (shading, m2 s−2) calculated from the differences between M100 and M50 (M100 − M50). The contour lines indicate the climatological mean, midwinter transient EKE in (a) and stationary EKE for M50 in (b).

  • View in gallery

    One-point 2-day lag correlation of the 8-day high-pass filtered eddy streamfunction at 300 hPa for (left) M100 and (right) M50. The contours are correlation coefficients at lag 2 days, which corresponds to 2 days after the position of (a) 55°N, 140°E, (b) 55°N, 120°E, and (c) 55°N, 80°E (these positions are indicated by ×). Values higher than 0.2 or lower than −0.2 are contoured with interval of 0.05.

  • View in gallery

    Midwinter (16 Dec–15 Feb mean), anomalous longitudinal traveling distance of wave packets (shadings: 1000 km) calculated from differences between M100 and M50 (M100 − M50). The contour lines indicate climatological mean longitudinal traveling distance of wave packets for M50. Contours start from 3000 km with intervals of 500 km.

  • View in gallery

    Sensitivities of midlatitude (32°–67°N) zonal-mean fractional changes of (a) 8-day high-pass filtered transient EKE and the (b) zonal extent of wave packets (both shown as squares with solid lines) to the varying central Asian mountains. The zonal-mean fractional changes of 300-hPa storminess are plotted as circles with dotted lines for comparison.

  • View in gallery

    Midwinter (16 Dec–15 Feb mean) anomalous (a) 700-hPa zonal wind speed (shading, m s−1) and (b) 780-hPa baroclinicity (shading, day−1) calculated from differences between M100 and M50 (M100 − M50). The contour lines indicate the climatological mean (a) 700-hPa zonal wind speed and (b) 780-hPa baroclinicity for M50.

  • View in gallery

    January zonal wind anomalies (shading) averaged over (a) central Asia (60°–100°E) and (b) the western Pacific (110°–150°E) calculated from the difference between M100 and M50 (M100 − M50). Contours are climatological mean, January zonal winds for M50. Contours start from 10 m s−1 with intervals of 10 m s−1.

  • View in gallery

    Sensitivities of the westerly index over the western Pacific (110°–150°E) in the (a) upper and (b) lower troposphere to the varying central Asian mountains. A larger (smaller) westerly index implies meridionally wide (broad) zonal wind structure. The North Pacific, area-averaged (45°–65°N, 140°–220°E) storminess is plotted as circles for comparison.

  • View in gallery

    Sensitivities of midlatitude (32°–67°N mean) zonal-mean fractional changes of 300-hPa storminess to the varying subgrid variability of mountains (G-series experiments; circles with solid lines) and to the varying height plus subgrid variability of mountains (M-series experiments; circles with dotted lines).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 34 34 4
PDF Downloads 23 23 2

The Role of the Central Asian Mountains on the Midwinter Suppression of North Pacific Storminess

View More View Less
  • 1 Department of Geography, and Center for Atmospheric Sciences, University of California, Berkeley, Berkeley, California
  • | 2 Department of Atmospheric and Ocean Sciences, McGill University, Montreal, Quebec, Canada
© Get Permissions
Full access

Abstract

The role of the central Asian mountains on North Pacific storminess is examined using an atmospheric general circulation model by varying the height and the areas of the mountains. A series of model integrations show that the presence of the central Asian mountains suppresses the North Pacific storminess by 20%–30% during boreal winter. Their impact on storminess is found to be small during other seasons. The mountains amplify stationary waves and effectively weaken the high-frequency transient eddy kinetic energy in boreal winter. Two main causes of the reduced storminess are diagnosed. First, the decrease in storminess appears to be associated with a weakening of downstream eddy development. The mountains disorganize the zonal coherency of wave packets and refract them more equatorward. As the zonal traveling distance of wave packets gets substantially shorter, downstream eddy development gets weaker. Second, the central Asian mountains suppress the global baroclinic energy conversion. The decreased baroclinic energy conversion, particularly over the eastern Eurasian continent, decreases the number of eddy disturbances entering into the western North Pacific. The “barotropic governor” does not appear to explain the reduced storminess in our model simulations.

Corresponding author address: Hyo-Seok Park, University of California, Berkeley, 531 McCone, Berkeley, CA 94720–4740. Email: hspark@berkeley.edu

A comment/reply has been published regarding this article and can be found at http://journals.ametsoc.org/doi/abs/10.1175/JAS-D-11-021.1 and http://journals.ametsoc.org/doi/abs/10.1175/JAS-D-11-096.1

Abstract

The role of the central Asian mountains on North Pacific storminess is examined using an atmospheric general circulation model by varying the height and the areas of the mountains. A series of model integrations show that the presence of the central Asian mountains suppresses the North Pacific storminess by 20%–30% during boreal winter. Their impact on storminess is found to be small during other seasons. The mountains amplify stationary waves and effectively weaken the high-frequency transient eddy kinetic energy in boreal winter. Two main causes of the reduced storminess are diagnosed. First, the decrease in storminess appears to be associated with a weakening of downstream eddy development. The mountains disorganize the zonal coherency of wave packets and refract them more equatorward. As the zonal traveling distance of wave packets gets substantially shorter, downstream eddy development gets weaker. Second, the central Asian mountains suppress the global baroclinic energy conversion. The decreased baroclinic energy conversion, particularly over the eastern Eurasian continent, decreases the number of eddy disturbances entering into the western North Pacific. The “barotropic governor” does not appear to explain the reduced storminess in our model simulations.

Corresponding author address: Hyo-Seok Park, University of California, Berkeley, 531 McCone, Berkeley, CA 94720–4740. Email: hspark@berkeley.edu

A comment/reply has been published regarding this article and can be found at http://journals.ametsoc.org/doi/abs/10.1175/JAS-D-11-021.1 and http://journals.ametsoc.org/doi/abs/10.1175/JAS-D-11-096.1

1. Introduction

Extratropical storminess in the North Pacific exhibits peculiar characteristics; it is generally weaker in winter than in fall and spring (e.g., right panel of Fig. 1a) although low-level baroclinicity peaks in winter (e.g., Fig. 3a). This so-called midwinter suppression of the North Pacific storminess or storm track (hereafter midwinter suppression) was first noted by Nakamura (1992). A number of studies have proposed hypotheses based on both observational data and climate models. These include weakened eddy seeding-feeding processes associated with downstream development (Chang 2001; Zurita-Gotor and Chang 2005), advection by strong westerly jets (Nakamura 1992; Harnik and Chang 2004), the barotropic wind shear effect (Deng and Mak 2005), trapping of the baroclinic waves in higher altitudes by stronger subtropical jet (Nakamura and Sampe 2002), and diabatic heating (Chang 2001). The definitive mechanisms, however, still remain to be determined.

Dynamics of downstream development of baroclinic wave packets have been intensively investigated (Simmons and Hoskins 1979; Lee and Held 1993; Orlanski and Katzfey 1991; Orlanski and Chang 1993; Chang and Orlanski 1993; Chang 1993; Hakim 2003) and applied to the midwinter suppression (Chang 2001; Zurita-Gotor and Chang 2005). By investigating the eddy kinetic energy (EKE) budget of the eddy life cycle, Orlanski and Chang (1993) found that wave packets primarily decay by transferring energy downstream, specifically in the form of ageostrophic geopotential flux. Zurita-Gotor and Chang (2005) performed a number of experiments using a two-layer quasigeostrophic model by placing a strong eddy damping (or wave packet barrier) in a limited region. It is found that localized eddy damping tends to reduce eddy amplitude over vast downstream areas and its recovery takes place over a long distance, specifically in the Northern Hemisphere (NH) winter condition when the basic flow has fast eddy group velocity.

Previous studies have suggested that the topography is important for the presence of the Pacific storm tracks (Lee and Mak 1996), but its quantitative impact has only been recently investigated (Son et al. 2009, hereafter S09). Using a dry atmospheric dynamical core model with an imposed basic state, S09 found that a strong single jet (qualitatively similar to the jet over the Eurasian continent) has well-organized wave packets traversing the globe in few weeks. Eddies grew not only through local baroclinicity, but also by eddy seeding/feeding processes. This organized wave packet, having uniquely long zonal extent and temporal persistence, got substantially shorter when idealized topography was introduced in midlatitudes. It resulted in a localized storm track downstream of the topography whose intensity was much weaker than the one in the absence of the topography. They further showed that the topographic effect on storm-track intensity is highly sensitive to background flow. When the background flow has weak double-jet pattern (qualitatively similar to the Atlantic jet), storm-track intensity rather slightly increases. These results suggest that the topographical impact on the two different storm tracks in the Northern Hemisphere, the Pacific and Atlantic storm tracks, might be very different.

This study is also motivated by the observational studies of Penny et al. (2009). Using reanalysis data, Penny et al. (2009) found that central Asia (including Siberia) is the area where the suppression of eddy seeding occurs. Their Lagrangian tracking methods revealed that the amplitude and frequency of cyclogenesis in the lee of the central Asian mountains reduce in winter, whereas the local amplitude and frequency of cyclogenesis over the western Pacific do not decrease, suggesting that the reduced eddy seeding from central Asia and Siberia is a key factor in midwinter suppression of the North Pacific storminess.

In light of the above studies, we investigate the role of the central Asian mountains on North Pacific storminess using a comprehensive atmospheric general circulation model (AGCM). It has merit over the simple models used by Zurita-Gotor and Chang (2005) and S09 in that the AGCM incorporates realistic surface boundary conditions and latent heating, simulating much more realistic atmospheric stationary waves. This will allow us to identify relevant mechanisms that cannot be captured in idealized models.

2. Model and analysis method

We used the Community Climate Model 3.10 (CCM3) at T42 horizontal resolution and standard 18 levels in the vertical. In this study, monthly-varying climatological surface boundary conditions such as land surface type, sea surface temperatures (SSTs), and sea ice are prescribed in the model. The climatological monthly mean SSTs are calculated from the 40 years of monthly mean Reynolds optimum interpolation data (Reynolds et al. 2002), spanning 1960 to 2000.

In all model integrations, the same boundary conditions are used except for topography over central Asia, which is gradually reduced from high latitudes to midlatitudes as summarized in Table 1. The control experiment, M100, maintains all the mountains at present-day heights (Fig. 1a). In the M75 experiment, we reduced the height of the Altai-Sayan mountains (covering 42°–55°N, 55°–115°E) by 95% from their original heights (Fig. 1b), while mountain heights from 42° to 37°N are reduced gradually so that they merge smoothly with the unaltered topography south of 37°N. In the M50 experiment, we further removed the northern part of the Tibetan Plateau. Latitudes north of 37°N are reduced by 95%, while the transitional regions from 32° to 37°N are similarly reduced to merge smoothly with the unaltered topography south of 32°N (Fig. 1c). The entire Tibetan Plateau area is further reduced in the M20 experiment, retaining only the southern part of the Tibetan Plateau, the Himalayas (Fig. 1d). It should be noted that the imposed subgrid variability of the mountain height, which is used for the gravity wave drag parameterization, has also been reduced with mountain height (so by 95% if the mountain height is reduced by 95%).

To separate the role of the gravity wave drag associated with the subgrid variability of the central Asian mountains from the effect of mountain height, we conducted an additional series of experiments. While we fix the height of the central Asian mountains to present-day values, we gradually reduced the subgrid variability of the mountains from high latitudes to midlatitudes. The areas of reduction are consistent with the previous experiments, which are summarized in Table 1.

Each experiment ran for 22 years, and the last 18 years are used for the analysis. As climatological monthly-mean SSTs are prescribed, four years of integration is sufficient for spinup. We used daily mean output to calculate time-mean flow and various eddy statistics, such as storminess, eddy heat flux, and baroclinic energy conversion (BCC). The low-level baroclinicity is measured by Eady parameter, ( f/N)(dU/dz), where f and N represent the Coriolis parameter and vertical static stability, respectively. It is evaluated at 780 hPa, using a finite difference method between two adjacent vertical levels (i.e., 700 and 850 hPa) to evaluate the vertical wind shear. The storminess, or storm-track intensity, is defined as 8-day high-pass filtered upper-level (300 hPa) geopotential height variations, similar to the method used by Nakamura (1992). The resulting high-frequency signal has been smoothed by a 20-day running mean, and the smoothed data averaged to produce the annual mean. The BCC is calculated using the equation in Peixoto and Oort (1992),
i1520-0469-67-11-3706-eq1
where ω′ and α′ denote transient pressure velocity and specific volume, respectively; MT and TOA respectively indicate the mountain top (or surface) and the top of atmosphere. Both ω′ and α′ are filtered by using an 8-day high-pass filter1 to exclude low-frequency eddies. Since the resulting BCC might be dependent on the filtering method, we also calculated the unfiltered BCC, , where the square brackets denote zonal-mean climatology. We found that the BCC calculated from the 8-day high-pass filtered eddies explains about 65% of the total unfiltered BCC. In the case when the 20-day high-pass filtered eddies are used, the filtered transient BCC explains more than 80% of the total unfiltered BCC.

The eddy kinetic energy, u2 + υ2, is computed using daily zonal and meridional wind fields. The seasonal cycle, derived from 18-yr daily climatology and smoothed with a 20-day running average, is first removed from the data. The EKE spectrum is then partitioned into transient and stationary (including quasi-stationary) EKEs. The former is calculated by using 8-day high-pass filtered eddy fields, whereas the latter is defined with low-pass filtered (longer than 100 days) eddies. Strictly speaking, the stationary EKE we defined includes the quasi-stationary or some low-frequency transient EKE.

3. Sensitivity of the North Pacific storminess to the central Asian mountains

The response of the low-level eddies to the mountains would be different from the upper-level eddies because the upper troposphere has faster eddy group speed, conveying the upstream disturbances more efficiently. Reanalysis data shows that the midwinter suppression of North Pacific storminess is much weaker in the lower troposphere than in the upper troposphere (Nakamura 1992). In light of this, we first examine how the upper- and low-level eddies respond to the mountains. Seasonal progression of upper-level storminess is shown in section 3a and that of low-level eddy activities is examined in section 3b. We will particularly focus on the differences between the M100 and M50 experiments, since the midwinter suppression is largely gone (although slight suppression still exists) by the M50 experiment.

a. Seasonal march of upper-level storminess

The M100 (control) experiment captures midwinter suppression of North Pacific (from 140°E to 140°W) storminess (Fig. 1a) and the wintertime baroclinicity maximum (Fig. 3a) reasonably well. The storminess peaks in October and rapidly decays from November until it is greatly suppressed in midwinter. On the other hand, baroclinicity has a single maximum strength in winter, consistent with previous results.

The wintertime storminess gradually increases as the area and height of the central Asian mountains decrease (right panels of Fig. 1). The distinct double maximum signal almost disappears when the entire Altai-Sayan mountains are removed and the Tibetan Plateau is substantially reduced (M20 in Fig. 1d). Interestingly, the North Pacific storminess during fall or spring rarely responds to the mountains. Using a series of idealized model integrations, S09 showed that storminess is sharply decreased by a mountain when the background westerly is a strong single-jet state. However, only minimal sensitivity is found if the background westerly is a weak double-jet state. Figure 2 shows the longitudinally averaged zonal wind over central Asia (from 60° to 100°E) in January and October. While the January jet profile is reminiscent of a single jet, the October jet is qualitatively similar to weak double-jet state in S09. Thus, the results of S09 appear applicable to interpreting our situation.

To further investigate storm-track sensitivity to the background flow, we examined storminess over the North Atlantic after reducing the area and height of the Rocky Mountains. Unlike the Eurasian continent, the wintertime westerly jet over the eastern Pacific and North America is broad, although not as broad as the double-jet state in S09. The wintertime storminess over the North Atlantic increases only by 5% when the Rockies are removed (not shown). This is again consistent with the findings of S09.

b. Seasonal march of low-level baroclinic eddy activities

As addressed above, the seasonal evolution of the baroclinicity has a single wintertime maximum (left panel of Fig. 3a) and the structure remains almost unchanged even after the entire Altai-Sayan mountains are removed (right panel of Fig. 3a). Rather, the wintertime baroclinicity slightly “strengthens” over 30°–45°N in the presence of the Altai-Sayan mountains even though the storms become less frequent. Note that baroclinicity is based on the linear theory and does not necessarily indicate the behavior of actual baroclinic eddy activities, which can be highly nonlinear.

To better understand the puzzling relationship between upper-level storminess and lower-level baroclinicity, we define the baroclinic eddy activities by the baroclinic energy conversion. In the M100 control experiment, the BCC exhibits a slight suppression in midwinter relative to its shoulder seasons (left panel of Fig. 3b). Similarly, the 700-hPa eddy heat flux exhibits a similar mild midwinter suppression (not shown). However, the magnitude of suppression is much weaker than that of the upper level, which is consistent with reanalysis data (Nakamura 1992). When the Altai-Sayan mountains and the northern Tibetan Plateau are removed, the BCC is somewhat enhanced in boreal fall and winter, especially during November and December (right panel of Fig. 3b). While a hint of midwinter minimum still exists, midwinter BCC increases up to 20%. Note that the M50 experiment exhibits enhanced eddy activities over the high-latitude regions (50°–60°N) relative to the ones in the M100 experiment, indicating that the mountains tend to meridionally confine baroclinic eddy activities southward.

4. Baroclinic energy conversion and the upper-troposphere storminess

In this section, we further elucidate the possible linkage between the lower- and upper-level eddy activities. We first examine the global anomalous patterns of storminess and BCC caused by the mountains in section 4a. Their possible relationships are further investigated and discussed in section 4b.

a. Altered midwinter storminess and BCC

Figure 4a shows the response of the NH midwinter storminess to the Altai-Sayan mountains and the northern Tibetan Plateau (M100 − M50). The “midwinter” refers to the 60-day annual mean from 16 December to 15 February. It can be seen that the mountains affect storminess not only near downstream of the mountains but also far downstream, indicating that the mountains modify eddy activities through the whole hemisphere. Over the eastern Eurasian continent, storminess decreases more than 30 m (shadings of Fig. 4a), which is about 40% of the background storminess of M50 (contours of Fig. 4a). The North Pacific also experiences pronounced weakening of storminess, up to about 30%.

The changes in the upper-level eddy activities are further examined by partitioning the reduced storminess into the changes in amplitude and the frequency of storminess. Note that reduced storminess could result from weaker amplitude of individual storms and/or less frequent development of storms. The standard deviation of the geopotential height at 300 hPa in the NH midlatitudes during midwinter is about 120 m in the control experiment (M100). We arbitrary set the deviation of 140 m, slightly larger than climatological standard deviation, as the threshold value of a storm event.2 The storm events passing this threshold are used to measure storm amplitude and frequency. The amplitude is quantified by averaging anomalous geopotential height for those events. The frequency is estimated by counting the number of events. To avoid overcounting associated with consecutive perturbations, consecutive perturbations occurring within 5 days are counted as a single event.

Figure 5 shows differences of storm amplitude and frequency between the M100 and M50 experiments. A simple visual comparison of the anomalous amplitude and frequency indicates that decrease in frequency is a main factor for weakening the storms in the presence of full topography. The storm frequency reduces up to 40% over the vast midlatitude areas (Fig. 5b) and the spatial pattern is consistent with the anomalous storminess (Fig. 4a). On the other hand, the amplitude change is less than 10% from the mean (Fig. 5a).

While the decrease in storminess or storm frequency persistently occurs over the whole NH extratropics, the reduction of BCC occurs somewhat sporadically (Fig. 4b). It does not necessarily mean that upper-level storminess is not influenced by lower-level eddy activities. Reanalysis data consistently exhibit that the maximum EKE in the upper troposphere is located downstream of the maximum baroclinicity in the midtroposphere (Vallis and Gerber 2008). Considering that BCC is tied with baroclinicity (i.e., linear eddy growth rate), the maximum storminess may occur somewhere downstream of the maximum BCC. Therefore, the decreased BCC over the eastern Eurasian continent (Fig. 4b) might contribute to decreasing the North Pacific storminess (Fig. 4a).

Figure 6 shows how global BCC and storminess respond to decreasing central Asian mountains. Fractional changes of BCC generally keep pace with those of storminess, suggesting a strong connection between these two variables. Because the central Asian mountains decrease transient eddy heat flux (not shown), the vertical eddy heat flux (i.e., BCC) is likely to decrease as well. It is unclear if the topography-forced stationary waves are direct causes for decreasing transient eddy heat flux. It should be noted that the relationship between BCC and storminess presented in Fig. 6 is not completely linear. The high-frequency BCC of M50 is slightly stronger than that of M20 (Fig. 6a).

b. Possible impact of the upstream BCC on downstream storminess

The seasonal evolution of BCC over the eastern Eurasian continent exhibits much clearer wintertime suppression of wave activities than the North Pacific. The left panel of Fig. 7a indicates that the eastern Eurasian continent (60°–140°E) experiences dramatic suppression of BCC in winter when the full mountains are imposed in the model (M100). The BCC peaks in the late summer or early fall, up to 4.5 W m−2, and then rapidly decreases down to 2 W m−2 in November. In midwinter, it is only around 1 W m−2 or even less. Similarly, the seasonal march of upper-level storminess shows a distinct midwinter minimum (left panel of Fig. 7b). In the absence of the Altai-Sayan mountains and the northern Tibetan Plateau (M50), the BCC over the eastern Eurasian continent is substantially enhanced in winter (right panel of Fig. 7a). Although a double maximum signal (maximum BCC in late fall and early spring) still exists, the wintertime BCC of M50 is more than 100% stronger than that of M100. Likewise, the upper-level storminess of M50 in boreal winter is much stronger (right panel of Fig. 7b) than that of M100.

Since the eddy group velocity substantially increases and the traveling distance of wave packets gets much longer in winter (Zurita-Gotor and Chang 2005), this increase (decrease) in BCC over the eastern Eurasian continent might effectively strengthen (weaken) the North Pacific storminess. This argument is consistent with Penny et al. (2009) and with Zhang and Held (1999). Zhang and Held (1999) demonstrate, using a linear model framework with stochastically forced eddies, that the enhanced eddy (temperature or vorticity) forcing over central Asia can strengthen the North Pacific storminess.

What causes such a radical suppression of wintertime BCC over the eastern Eurasia continent? Penny et al. (2009) suggests that the strong near-surface static stability over Siberia could be a key factor for reducing the frequency and amplitude of cyclogenesis in boreal winter. Based on this idea, we calculated the near-surface static stability over the Eurasian continent for both the M100 and M50 experiments. While the near-surface static stability is substantially high in boreal winter over the eastern Eurasian continent, there are minimal changes to it after the Altai-Sayan mountains and the northern Tibetan Plateau are removed (not shown), suggesting that local BCC may not be always dependent on the near-surface static stability. In section 6, we will further discuss why the Altai-Sayan mountains and the northern Tibetan Plateau can effectively decrease BCC, especially over the eastern Eurasian continent.

5. Stationary waves and baroclinic wave packets

The most pronounced impact of the central Asian mountains on the NH winter circulations is the generation of stationary waves. The mountain-induced stationary waves affect the transient eddies by redistributing the EKE spectrum (Manabe and Terpstra 1974; Yu and Hartmann 1995). Since upper-level storminess has been defined by the high-pass filtered geopotential height, the decrease in transient EKE in the presence of the mountains may have a direct relationship with the decrease in storminess.

a. Stationary waves and the EKE spectrum

The linear wave response of the upper-tropospheric flow to the Tibetan Plateau is known to be limited in regional scale (Held et al. 2002), probably because of various damping effects such as Ekman damping. The nonlinear response, however, is known to occur over wide downstream areas as the Tibetan Plateau modifies a diabatic heating field (Held et al. 2002).

Figure 8a shows 300-hPa eddy streamfunction in January, simulated by the M100 experiment. Here, “eddy” denotes the deviation from the zonal mean. The simulated stationary waves have wavenumber-2 patterns, and their amplitudes are quite similar to observations [e.g., Fig. 1a of Held et al. (2002)]. The anomalous eddy streamfunctions in the other experiments—deviations from M100—are displayed in Figs. 8b–d (M100 − M75, M100 − M50, and M100 − M20). In all simulations, anomalies are found over the entire NH. It is worth noting that the magnitude of the anomalous stationary wave response is comparable in all three cases (Figs. 8b–d), although the M100 − M20 exhibits slightly wider stationary wave response than others (M100 − M75 and M100 − M50). As Held et al. (2002) noted, it is not surprising to see that the mountain height is not an essential factor for amplifying the stationary waves. The mountains in higher latitudes (i.e., the Altai-Sayan mountains) effectively amplify the wintertime stationary waves, even though their heights are significantly lower than the heights of the Tibetan Plateau. We are not sure if this simulated stationary wave response will consistently occur in other models as well. If it does, it will be interesting to further examine why the addition of the mountains in higher latitudes (i.e., the Altai-Sayan mountains) to the lower-latitude mountains (i.e., the Tibetan Plateau) is so efficient in exciting the wintertime stationary waves.

The magnitude of the stationary waves excited by the Altai-Sayan mountains is significant relative to the total stationary waves (cf. Figs. 8a and 8b). The EKE spectrum is hence likely to be modified by the mountains as well. Figure 9 shows the EKE difference between the M100 and M50 experiments during midwinter. As expected, transient EKE is reduced over a wide range of midlatitudes (Fig. 9a), whereas stationary and quasi-stationary EKE are substantially enhanced (Fig. 9b). The spatial pattern of reduced transient EKE is quite similar to that of storminess (Fig. 4a). The zonal-mean fractional changes of transient EKE in all experiments keep pace with those of upper-level storminess (Fig. 12a). The linear relationship between the transient EKE and storminess still holds if a 20-day high-pass filter is used to calculate the transient EKE (not shown). This result suggests that the decreased transient EKE change is strongly tied with the midwinter suppression. This result is somewhat different from S09, who found much weaker sensitivity of transient EKE to the mountain heights.

b. Wave packet structure

What causes the zonally extensive reduction of transient EKEs in Fig. 9a? To answer this, we consider the zonal propagation of baroclinic wave packets and the associated downstream development of transient eddies. The structure of wave packets can be inferred by calculating the one-point lag or lead correlations of the eddy streamfunction (Chang and Yu 1999; S09). The eddy streamfunction is calculated by 8-day high-pass filtered streamfunction to capture the transient eddy of interest. A reference gridpoint is chosen, and correlation coefficients are computed against all other grid points in the NH with a 2-day lag.

Figures 10a and 10b show the structure of wave packets over the western North Pacific (55°N, 140°E) and over East Asia (55°N, 120°E) individually. Wave packets in the M100 and M50 experiments are significantly different; while the wave packets in M100 are zonally restricted, those in M50 have wider zonal extent. Furthermore, the wave packets of M100 refract more equatorward compared to M50. This result resembles those of Penny et al. (2009) and Hakim (2003), who found that a large fraction of wave packets upwind of the Tibetan Plateau refract into the subtropics. Unlike the western Pacific or East Asia, wave packets over central Asia (55°N, 80°E; Fig. 10c) are only weakly sensitive to the mountains. This is in contrast to Robinson et al. (2006), who found distinct differences in the waveguide over central Asia between years when the midwinter suppression is strong and weak. It is likely that interannual variability in the waveguide depends on the variability of the background zonal flow. In this paper, since we are focusing on the climatological mean response of the waveguide to different topographic conditions, the interannual analogy does not apply.

To better understand changes in the structure of wave packets, the zonal extent (coherence) of wave packets are quantified following S09. As in Fig. 10, a 2-day lag correlation is computed for each reference gridpoint; the zonal extent between this reference point and the farthest gridpoint with correlation coefficient higher than a threshold value3 (chosen to be 0.35) becomes the measure of the zonal extent for that reference point. The correlation field was slightly smoothed using Gaussian smoothing before measuring the zonal extent. This operation is done over all grid points, producing a hemispheric-wide map of zonal extent of wave packets. Figure 11 shows the differences of wave packet extent between the M100 and M50 experiments (M100 − M50). The decrease in wave packet extent is evident over the vast midlatitude storm track areas covering the North Pacific and the North America. More importantly, the overall pattern is qualitatively similar to that of storminess (Fig. 4a) and storm frequency (Fig. 5b). Figure 12b further shows that the zonal-mean fractional changes of wave packet extent keep pace with those of storminess in all experiments. This result suggests that the reduced efficiency of downstream eddy development is an important factor for the midwinter suppression in our model.

6. Other possible mechanisms

It is still unclear why global BCC decreases as the central Asian mountains increase. James (1987) showed that baroclinic instability can be limited by barotropic wind shear. The “barotropic governor” has been applied to the midwinter suppression problem with different methodologies (Nakamura 1992; Harnik and Chang 2004; Deng and Mak 2005). It is beyond the scope of this study to fully investigate the barotropic governor process. Instead, we will briefly diagnose the possible relationship between BCC and barotropic wind shear in our AGCM experiments. We also discuss the sensitivity of our results to the magnitude of gravity wave drag associated with the subgrid variability of mountains.

a. Barotropic governor effect

Figure 13a shows the altered zonal winds at 700 hPa (regarded as a steering level4 for transient eddies) resulting from the addition of the Altai-Sayan mountains and the northern Tibetan Plateau to the southern Tibetan Plateau (M100 − M50). Strengthening of subtropical westerlies (25°–40°N) and weakening of midlatitude westerlies (45°–65°N) are clearly seen in the western North Pacific (shadings of Fig. 13a), indicating that the westerlies are restricted to more equatorial latitudes when the mountains are present. The equatorward confinement of low-level westerlies is accompanied by the weakening (strengthening) of low-level meridional temperature gradient in midlatitudes (subtropics). Figure 13b shows that the low-level baroclinicity, which is directly related with the meridional temperature gradient, gets weaker (stronger) over the midlatitude (subtropical) North Pacific.5

The latitude–vertical height cross section of zonal wind over central Asia is shown in Fig. 14a. The central Asian mountains strengthen the westerlies at the equatorward flank of the jet (15°–25°N) but weaken them at the poleward flank of jet (30°–55°N), indicating an equatorward shift of the mean jet. A similar dipole pattern is also found over the western Pacific (110°–150°E), although over this region the anomalies act to strengthen the existing jet and restrict its poleward extent (Fig. 14b). It is noteworthy that there is little difference in the anomalous zonal wind structure among M100 − M75, M100 − M50, and M100 − M20 (now shown), indicating that the overall equatorward confinement is mostly explained by the Altai-Sayan mountains, not by the Tibetan Plateau.

Does the equatorward shift or confinement of zonal winds over central Asia suppress the downstream baroclinic instability? Previous studies showed that the meridionally confined narrow jets decrease storminess by horizontal shearing and stretching deformation of the flow in the Pacific (Deng and Mak 2005). The stronger zonal group velocity relative to the vertical wind shear is also known to weaken the downstream storminess (Harnik and Chang 2004). Using reanalysis data, Nakamura and Sampe (2002) further found that the years experiencing strong midwinter suppression have meridionally confined zonal wind structure over the western Pacific and more equatorward refraction of the waveguide.

To address the above questions, we examine the relationship between the North Pacific storminess and the meridional confinement of zonal winds driven by the central Asian mountains. The degree of zonal wind confinement is measured by calculating the difference of zonal wind speed between the latitude of jet core (∼28°–33°N) and the latitude of 45°N (hereafter the “westerly index”). The choice of 45°N is subjective, but our results using this index are insensitive to the choice of the latitude (between 40° and 55°N). Since the zonal wind structure over the central eastern North Pacific (180°–140°W) is not much affected by the mountains, we focus on the upstream winds averaged over 110°–150°E. A larger westerly index denotes stronger meridional confinement of jet, and stronger barotropic wind shear.

Figure 15 shows the sensitivity of the westerly index to the mountains in the upper troposphere (200–400-hPa mean) and in the lower troposphere (750–850-hPa mean) separately. The Altai-Sayan mountains effectively confine zonal winds equatorward (see the difference between M100 and M75), whereas decreasing the height of the Tibetan Plateau in addition (M50 and M20) hardly affected the zonal wind structure. On the other hand, North Pacific storminess increases as the mountain is progressively reduced. We infer from this that the barotropic governor effect does not appear to be a dominant factor in explaining our results.

b. Role of gravity wave drag

Topographically induced gravity waves play an important role for the upper-troposphere zonal momentum balance (e.g., Shaw et al. 2009). This raises the possibility that the substantial recovery of wintertime storminess in the M50 and M20 experiments (Figs. 1c and 1d) might be in part associated with the weakened gravity wave drag. To isolate the role of the gravity wave drag associated with the subgrid variability of the central Asian mountains, we conducted an additional series of experiments in which the subgrid-scale variation in the central Asian mountains was reduced in the same way as for the M75, M50, and M20 experiments, but the height of the mountains was kept to present-day values. Hereafter, we call these simulations G75, G50, and G20, respectively.

Figure 16 shows the sensitivity of storminess to different magnitudes of gravity wave drag (G75, G50, G20) over the central Asian mountains; by comparison, the corresponding M75, M50, and M20 results are also plotted. The difference between the G75 (G50, G20) and M75 (M50, M20) results thus indicates the non–gravity wave drag contribution to the change in storminess. The zonal-mean storminess increases only by 4.5% if the gravity wave drag is removed over the Altai-Sayan mountains (the G75 experiment) and the northern Tibetan Plateau regions (the G50 experiment); the contribution increases to around 9% if gravity wave drag is removed from the rest of the Tibetan Plateau (G20). By comparison, the increase in storminess in the M-series experiments is considerably larger than the G series; for example, the M50 experiment shows a 19% increase in storminess, compared to 4.5% for the G50 experiment. These results indicate that while the magnitude of gravity wave drag contributes to the change in storminess, the pure topographic effect of the height of the central Asian mountains plays a larger role. We note that the response of the troposphere circulation to gravity wave drag is dependent on the model resolution (Shaw et al. 2009). We have used a relatively coarse model with T42 resolution and 18 vertical levels; it is possible that with vertical resolution in the upper troposphere and the lower stratosphere, quantitatively different results could be found.

7. Summary and discussion

In this study, the role of the central Asian mountains in downstream storminess has been examined by varying the area and height of the central Asian mountains in the NCAR CCM3. We found that the midwinter suppression of the North Pacific storminess almost disappears when the central Asian mountains are removed. Interestingly, it is the Altai-Sayan mountains, located to the north of the central Asian mountains, that turn out to be most effective in reducing BCC and downstream storminess. The height of the Altai-Sayan mountains is significantly lower than the Tibetan Plateau, and the zonal winds impinging on the Altai-Sayan mountains are weaker than those impinging on the Tibetan Plateau. Nevertheless, the Altai-Sayan mountains effectively enhance the amplitude of North Pacific stationary waves and confine the zonal winds equatorward.

We suggest two plausible mechanisms for how the central Asian mountains suppress storminess in midwinter: 1) by decreasing BCC, especially over the eastern Eurasian continent, and 2) by reducing downstream eddy development, through changing North Pacific stationary waves that alter the zonal progression of wave packets. The combined effects of the two mechanisms effectively reduce the wintertime North Pacific storminess, even though the low-level baroclinicity has its maximum in winter.

We further examined whether a barotropic governor effect could explain the decrease in BCC and storminess. The Altai-Sayan mountains confine the zonal winds equatorward, increasing the meridional shear of the zonal westerlies, and possibly contribute to suppressing the BCC through the barotropic governor. However, further removal of the Tibetan Plateau had little effect on the zonal wind structure in the northwestern Pacific; on the other hand, both the BCC and storminess increase further. This result suggests that the barotropic governor effect may not be a dominant factor in explaining our results.

A limitation of our present study is that surface boundary conditions—specifically SST, but also land surface type—are prescribed to the present day and do not interact with the topography-induced climate changes. SST, and thus atmospheric convection, are potentially sensitive to changes in large-scale mountains. Kitoh (2007) showed in a coupled model study that the global SST and precipitation field changed as the global mountains were uniformly reduced. Changes to the diabatic heating field significantly affect atmospheric circulation and atmospheric available potential energy, thus potentially changing the upper-level storminess. Our results are thus restricted to the direct effects of the central Asian mountains on North Pacific storminess; the role of interactive SST in modulating the response will be addressed in the future.

Acknowledgments

HSP would like to thank three anonymous reviewers and A. Friedman for constructive comments and suggestions. Comments from Dr. Inez Fung were particularly helpful. HSP and JC acknowledge support of the Berkeley Atmospheric Science Center Fellowship and the Gary Comer Science and Education Foundation. SWS acknowledges support of the NSERC Discovery Grant.

REFERENCES

  • Chang, E. K. N., 1993: Downstream development of baroclinic waves as inferred from regression analysis. J. Atmos. Sci., 50 , 20382053.

    • Search Google Scholar
    • Export Citation
  • Chang, E. K. N., 2001: GCM and observational diagnoses of the seasonal and interannual variations of the Pacific storm track during the cool season. J. Atmos. Sci., 58 , 17841800.

    • Search Google Scholar
    • Export Citation
  • Chang, E. K. N., , and I. Orlanski, 1993: On the dynamics of a storm track. J. Atmos. Sci., 50 , 9991015.

  • Chang, E. K. N., , and D. B. Yu, 1999: Characteristics of wave packets in the upper troposphere. Part I: Northern Hemisphere winter. J. Atmos. Sci., 56 , 17081728.

    • Search Google Scholar
    • Export Citation
  • Deng, Y., , and M. Mak, 2005: An idealized model study relevant to the dynamics of the midwinter minimum of the Pacific storm track. J. Atmos. Sci., 62 , 12091225.

    • Search Google Scholar
    • Export Citation
  • Hakim, G. J., 2003: Developing wave packets in the North Pacific storm track. Mon. Wea. Rev., 131 , 28242837.

  • Harnik, N., , and E. K. M. Chang, 2004: The effects of variations in jet width on the growth of baroclinic waves: Implications for midwinter Pacific storm track variability. J. Atmos. Sci., 61 , 2340.

    • Search Google Scholar
    • Export Citation
  • Held, I. M., , M. Ting, , and H. Wang, 2002: Northern winter stationary waves: Theory and modeling. J. Climate, 15 , 21252144.

  • James, I. N., 1987: Suppression of baroclinic instability in horizontally sheared flows. J. Atmos. Sci., 44 , 37103720.

  • Kitoh, A., 2007: ENSO modulation by mountain uplift. Climate Dyn., 28 , 781796.

  • Lee, S., , and I. M. Held, 1993: Baroclinic wave packets in models and observations. J. Atmos. Sci., 50 , 14131428.

  • Lee, W-J., , and M. Mak, 1996: The role of orography in the dynamics of storm tracks. J. Atmos. Sci., 53 , 17371750.

  • Manabe, S., , and T. B. Terpstra, 1974: The effects of mountains on the general circulation of the atmosphere as identified by numerical experiments. J. Atmos. Sci., 31 , 342.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., 1992: Midwinter suppression of baroclinic wave activity in the Pacific. J. Atmos. Sci., 49 , 16291642.

  • Nakamura, H., , and T. Sampe, 2002: Trapping of synoptic-scale disturbances into the North Pacific subtropical jet core in midwinter. Geophys. Res. Lett., 29 , 1761. doi:10.1029/2002GL015535.

    • Search Google Scholar
    • Export Citation
  • Orlanski, I., , and J. Katzfey, 1991: The life cycle of a cyclone wave in the Southern Hemisphere. Part I: Eddy energy budget. J. Atmos. Sci., 48 , 19721998.

    • Search Google Scholar
    • Export Citation
  • Orlanski, I., , and E. K. M. Chang, 1993: Ageostrophic geopotential fluxes in downstream and upstream development of baroclinic waves. J. Atmos. Sci., 50 , 212225.

    • Search Google Scholar
    • Export Citation
  • Peixoto, J., , and A. Oort, 1992: Physics of Climate. American Institute of Physics, 520 pp.

  • Penny, S., , G. H. Roe, , and D. S. Battisti, 2009: The source of the midwinter suppression in storminess over the North Pacific. J. Climate, 23 , 634648.

    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., , N. A. Rayner, , T. M. Smith, , D. C. Stokes, , and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15 , 16091625.

    • Search Google Scholar
    • Export Citation
  • Robinson, D. P., , R. B. Black, , and B. A. McDaniel, 2006: A Siberian precursor to midwinter intraseasonal variability in the North Pacific storm track. Geophys. Res. Lett., 33 , L15811. doi:10.1029/2006GL026458.

    • Search Google Scholar
    • Export Citation
  • Shaw, T. A., , M. Sigmond, , T. G. Shepherd, , and J. F. Scinocca, 2009: Sensitivity of simulated climate to conservation of momentum in gravity wave drag parameterization. J. Climate, 22 , 27262742.

    • Search Google Scholar
    • Export Citation
  • Simmons, A. J., , and B. J. Hoskins, 1979: The downstream and upstream development of unstable baroclinic waves. J. Atmos. Sci., 36 , 12391254.

    • Search Google Scholar
    • Export Citation
  • Son, S-W., , M. Ting, , and L. M. Polvani, 2009: The effect of topography on storm-track intensity in a relatively simple general circulation model. J. Atmos. Sci., 66 , 393411.

    • Search Google Scholar
    • Export Citation
  • Vallis, G. K., , and E. P. Gerber, 2008: Local and hemispheric dynamics of the North Atlantic Oscillation, annular patterns and the zonal index. Dyn. Atmos. Oceans, 44 , 184212.

    • Search Google Scholar
    • Export Citation
  • Yu, J-Y., , and D. L. Hartmann, 1995: Orographic influences on the distribution and generation of atmospheric variability in a GCM. J. Atmos. Sci., 52 , 24282443.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. Q., , and I. M. Held, 1999: A linear stochastic model of a GCM’s midlatitude storm tracks. J. Atmos. Sci., 56 , 34163435.

  • Zurita-Gotor, P., , and E. K. M. Chang, 2005: The impact of zonal propagation and seeding on the eddy–mean flow equilibrium of a zonally varying two-layer model. J. Atmos. Sci., 62 , 22612273.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(left) Boundary conditions for the CCM3 simulations. Mountains higher than 500 m are shaded. (right) Seasonal march of the upper-level (300 hPa) North Pacific storminess (m) averaged between 140°E and 140°W, for the (a) M100, (b) M75, (c) M50, and (d) M20 experiments.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 2.
Fig. 2.

Zonal wind averaged over the longitudinal interval of 60°–100°E for (a) January and (b) October simulated by CCM3.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 3.
Fig. 3.

(a) Seasonal march of the low-level baroclinicity (day−1) and (b) baroclinic conversion (W m−2) over the North Pacific, averaged between 140°E and 140°W, for (left) M100 and (right) M50.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 4.
Fig. 4.

Midwinter (16 Dec–15 Feb mean) anomalous (a) 300-hPa storminess (shading, m) and (b) baroclinic conversion (shading, W m−2) calculated from the differences between M100 and M50 (M100 − M50). The contour lines indicate climatological mean storminess in (a) and baroclinic conversion for M50 in (b).

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 5.
Fig. 5.

Same as Fig. 4 but for anomalous (a) mean-amplitude (shading, meters) and (b) frequency (shading, day) of 300-hPa storminess calculated from the differences between M100 and M50 (M100 − M50). See the text for the details on defining the amplitude and frequency. The contour lines indicate climatological mean amplitude in (a) and frequency for M50 in (b).

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 6.
Fig. 6.

Sensitivities of midlatitude (32°–67°N), zonal-mean fractional changes of (a) high-pass filtered BCC and (b) unfiltered total BCC (both shown as squares with solid lines) to the varying central Asian mountains. The zonal-mean fractional changes of 300-hPa storminess are plotted as circles with dotted lines for comparison.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 7.
Fig. 7.

The seasonal march of (a) baroclinic conversion (W m−2) and (b) 300-hPa storminess (m) over the eastern Eurasian continent (60°–140°E mean), for (left) M100 and (right) M50.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 8.
Fig. 8.

(a) The 300-hPa eddy streamfunction for M100 in January. (b)–(d) Anomalous eddy streamfunction calculated from the differences between M100 and others: (b) M100 − M75, (c) M100 − M50, and (d) M100 − M20 in January. The contour interval is 3 × 106 m2 s−1.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 9.
Fig. 9.

Midwinter (16 Dec–15 Feb mean) anomalous (a) 8-day high-pass filtered transient EKE (shading, m2 s−2) and (b) 100-day low-pass filtered stationary EKE (shading, m2 s−2) calculated from the differences between M100 and M50 (M100 − M50). The contour lines indicate the climatological mean, midwinter transient EKE in (a) and stationary EKE for M50 in (b).

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 10.
Fig. 10.

One-point 2-day lag correlation of the 8-day high-pass filtered eddy streamfunction at 300 hPa for (left) M100 and (right) M50. The contours are correlation coefficients at lag 2 days, which corresponds to 2 days after the position of (a) 55°N, 140°E, (b) 55°N, 120°E, and (c) 55°N, 80°E (these positions are indicated by ×). Values higher than 0.2 or lower than −0.2 are contoured with interval of 0.05.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 11.
Fig. 11.

Midwinter (16 Dec–15 Feb mean), anomalous longitudinal traveling distance of wave packets (shadings: 1000 km) calculated from differences between M100 and M50 (M100 − M50). The contour lines indicate climatological mean longitudinal traveling distance of wave packets for M50. Contours start from 3000 km with intervals of 500 km.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 12.
Fig. 12.

Sensitivities of midlatitude (32°–67°N) zonal-mean fractional changes of (a) 8-day high-pass filtered transient EKE and the (b) zonal extent of wave packets (both shown as squares with solid lines) to the varying central Asian mountains. The zonal-mean fractional changes of 300-hPa storminess are plotted as circles with dotted lines for comparison.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 13.
Fig. 13.

Midwinter (16 Dec–15 Feb mean) anomalous (a) 700-hPa zonal wind speed (shading, m s−1) and (b) 780-hPa baroclinicity (shading, day−1) calculated from differences between M100 and M50 (M100 − M50). The contour lines indicate the climatological mean (a) 700-hPa zonal wind speed and (b) 780-hPa baroclinicity for M50.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 14.
Fig. 14.

January zonal wind anomalies (shading) averaged over (a) central Asia (60°–100°E) and (b) the western Pacific (110°–150°E) calculated from the difference between M100 and M50 (M100 − M50). Contours are climatological mean, January zonal winds for M50. Contours start from 10 m s−1 with intervals of 10 m s−1.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 15.
Fig. 15.

Sensitivities of the westerly index over the western Pacific (110°–150°E) in the (a) upper and (b) lower troposphere to the varying central Asian mountains. A larger (smaller) westerly index implies meridionally wide (broad) zonal wind structure. The North Pacific, area-averaged (45°–65°N, 140°–220°E) storminess is plotted as circles for comparison.

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Fig. 16.
Fig. 16.

Sensitivities of midlatitude (32°–67°N mean) zonal-mean fractional changes of 300-hPa storminess to the varying subgrid variability of mountains (G-series experiments; circles with solid lines) and to the varying height plus subgrid variability of mountains (M-series experiments; circles with dotted lines).

Citation: Journal of the Atmospheric Sciences 67, 11; 10.1175/2010JAS3349.1

Table 1.

AGCM design of experiments (main experiments).

Table 1.

1

Since the storminess has been defined by 8-day high-pass filtered geopotential height, we consistently used 8-day high-pass filter for the calculations of transient BCC and transient EKE. However, a 20-day high-pass filter is more widely used for calculating the transient eddies. We also calculated transient BCC and transient EKE using 20-day high-pass filters to check the robustness of our interpretations. We found that both 8-day and 20-day high-pass filtered transient eddies provide qualitatively consistent results.

2

The fractional changes in the amplitude and frequency of storm activities are somewhat sensitive to the threshold value. However, for any given threshold we find that frequency has consistently much larger impact than amplitude.

3

The zonal extent of wave packets is somewhat sensitive to the choice of threshold value particularly when the threshold value is small. For example, if the threshold value is less than 0.2, the subtropics exhibit longer zonal extents of wave packet than do the midlatitudes. It is likely to be an artifact associated with the reflection of eddies from the tropics to the extratropics.

4

The state of the low-level westerlies, steering the low-level ridges and troughs, would be an important basic state modulating the downstream eddy propagation. The 700-hPa level is often referred to as a steering level, where westerly winds bear a direct relationship with the velocity of low-level disturbances. In a previous study, a simple mean of the 500-hPa and the 1000-hPa level (roughly representing the 700-hPa level) westerly is used to identify the position of storm tracks over the North Pacific (Nakamura 1992).

5

The equatorward confinement of baroclinicity is not well detected in the zonal-mean, annual-mean configuration (Figure 3a). As the annual-mean baroclinicity in the midlatitudes is much weaker than that of subtropics in boreal winter, the changes in the midlatitude baroclinicity are not well detected visually unless anomalies are explicitly calculated.

Save