Embed Size (px)
Citation preview
Reply
HYO-SEOK PARK
Department of Environmental Science and Engineering, California Institute of Technology, Pasadena, California
JOHN C. H. CHIANG
Department of Geography, and Center for Atmospheric Sciences, University of California, Berkeley, Berkeley, California
SEOK-WOO SON
Department of Atmospheric and Ocean Sciences, McGill University, Montreal, Quebec, Canada
(Manuscript received 11 April 2011, in final form 9 August 2011)
We thank Chang and Lin for their thoughtful and
constructive comments on our study (Park et al. 2010).
In Park et al. (2010), we did not explicitly state that the
topography-forced stationary waves are the direct cause
for the reduced downstream transient eddy kinetic energy
(EKE). The response of stationary waves to topography
may saturate even with a relatively small mountain (Cook
and Held 1992); furthermore, their magnitudes are much
smaller than thermally forced stationary waves (Chang
2009; Held et al. 2002). Instead, we suggest that quasi-
stationary waves generated by the central Asian moun-
tains may strongly affect North Pacific storminess by
changing the year-to-year variability of westerly winds
over the eastern Eurasian continent. Observational anal-
yses indicate that the midwinter suppression of North
Pacific storminess does not occur every year. Some years
experience stronger and more meridionally confined
zonal winds over the western North Pacific, leading to
stronger midwinter suppression (Harnik and Chang
2004; Nakamura and Sampe 2002).
In our atmospheric general circulation model (AGCM)
analyses, the interannual variability of westerly winds and
storminess over the North Pacific decrease substantially
in the absence of the central Asian mountains; a year
with strong midwinter suppression over the North Pacific
occurs rarely in this simulation. Moreover, it is still un-
clear why the presence of the central Asian mountains
strengthens the interannual variability of westerly jets and
storminess over the western North Pacific. We believe
understanding the cause of this strong interannual var-
iability is key to understanding the mechanisms for the
midwinter suppression. Indeed, fundamental questions
still remain with regard to the dynamics of quasi-stationary
waves, such as how mountains affect the diabatic heating
field (Held et al. 2002) and the convergence of eddy mo-
mentum fluxes (Chang 2009).
We share the concern of Chang and Lin (2011) that the
AGCM integration time (18 yr in our study) may be in-
sufficient to accurately capture the quantitative response
of downstream storminess. Also, we agree with Chang
and Lin (2011) that a multimodel ensemble approach will
be required to better quantify the impact of the moun-
tains on downstream storminess. Along these lines, we
tested the robustness of our results using the global at-
mospheric model, version 2.1 (AM2.1), developed at
the Geophysical Fluid Dynamics Laboratory (GFDL;
Anderson et al. 2004). This version of AM2.1 uses a
finite-volume dynamical core (Lin 2004) with 2.58 3 2.08
horizontal resolution (M45) and 24 vertical levels (L24).
Seasonally varying insolation and climatological sea surface
temperatures (SSTs) are prescribed in the model. The SSTs
are from 50 yr of monthly mean Reynolds reconstructed
historical SST analysis, spanning from 1950 to 2000 (Smith
et al. 1996). We ran the model for 60 yr, and the last 54 yr
are used for the analysis, tripling the integration period of
our previous work. Unlike the previous paper, for which an
8-day high-pass filter was used, we used a 10-day high-pass
filter to define transient eddies. This method slightly re-
duces uncertainty by increasing the spectral band by 2 days
and is widely used for defining synoptic-scale transients.
Overall results are virtually identical with those from the
8-day high-pass filtering method used in Park et al. (2010).
Corresponding author address: Hyo-Seok Park, California In-
stitute of Technology, MC 100-23, Pasadena, CA 91125.
E-mail: [email protected]
2804 J O U R N A L O F T H E A T M O S P H E R I C S C I E N C E S VOLUME 68
DOI: 10.1175/JAS-D-11-096.1
� 2011 American Meteorological Society
We found that the sensitivity of downstream stormi-
ness to the presence of the central Asian mountains in
AM2.1 is slightly weaker than what we found in Commu-
nity Climate Model 3.10 (CCM3; Kiehl et al. 1998). In
CCM3, the removal of the Altai-Sayan Mountains and the
northern part of the Tibetan Plateau [i.e., the M50 exper-
iment in Park et al. (2010)] increased downstream stormi-
ness by 20%–30%. On the other hand, we had to remove
the entire central Asian mountains (hereafter referred to as
the MN05 experiment) to get a comparable response of
downstream storminess in AM2.1. However, the response
of downstream storminess in AM2.1 is still substantially
stronger than what Chang and Lin (2011) suggest.
Figure 1a shows the difference in the 10-day high-pass
filtered EKE between the MN100 and MN05 experi-
ments (MN100 minus MN05) during midwinter (from 15
December to 14 February). Transient EKE is reduced over
a wide range of midlatitudes in the presence of the central
Asian mountains. In general, the magnitude of the EKE
reduction over the North Pacific is around 20%, which is
a little bit smaller than what Park et al. (2010) found, but it
reaches up to 30% in some areas sporadically. Consistent
with a transient EKE response, the standard deviation of
the 10-day high-pass filtered geopotential heightffiffiffiffiffiffiffiffiZ92
p
decreases over a wide range of midlatitudes (Fig. 1b). The
magnitude of the response is around 15%–25%, which can
substantially deepen the midwinter suppression signal.
Figure 2a shows the wintertime stationary waves, de-
fined by the 300-hPa eddy streamfunction, simulated by
AM2.1. The climatological mean amplitude of stationary
waves simulated by AM2.1 is about 10% weaker than
what CCM3 simulates. The anomalously strong station-
ary waves over the North Atlantic Ocean, appearing in
CCM3 (Park et al. 2010) and in an old version of the
GFDL model (Held et al. 2002), appear muted in AM2.1.
In particular, strong positive and negative dipoles near
508N over North America, which appear in CCM3 (Park
et al. 2010) and in the previous version of the GFDL
model (Held et al. 2002), are substantially weakened.
Figure 2b shows stationary waves forced by the central
Asian mountains, calculated by the difference between
MN100 and MN05. Overall, the magnitude of the re-
sponse is smaller than what CCM3 simulated in Park
et al. (2010), but larger than what Chang and Lin (2011)
found.
As we mentioned earlier, midwinter suppression does
not occur every year. Thus, the climatologically aver-
aged impact of the central Asian mountains on down-
stream storminess can substantially vary depending on
the model used and integration period chosen. We plan
to further analyze our AM2.1 simulations to better un-
derstand why the central Asian mountains enhance the
interannual variability of North Pacific storminess in the
context of quasi-stationary waves.
Acknowledgments. We thank Yohai Kaspi for read-
ing this manuscript and providing constructive com-
ments.
FIG. 1. Anomalous (a) 10-day high-pass filtered transient EKE
(shading, m2 s22) at 300 hPa, calculated from the differences be-
tween MN100 and MN05 (MN100 2 MN05). The contour lines
indicate the climatological mean transient EKE. (b) As in (a), but
for geopotential height.
FIG. 2. (a) The 300-hPa eddy streamfunction for MN100.
(b) Anomalous eddy streamfunction calculated from the differences
between MN100 and MN05. The contour interval is 3 3 106 m2 s21.
NOVEMBER 2011 C O R R E S P O N D E N C E 2805
REFERENCES
Anderson, J. L., and Coauthors, 2004: The new GFDL global at-
mosphere and land model AM2-LM2: Evaluation with pre-
scribed SST simulations. J. Climate, 17, 4641–4673.
Chang, E. K. M., 2009: Diabatic and orographic forcing of northern
winter stationary waves and storm tracks. J. Climate, 22, 670–
688.
Chang, X., and X. Lin, 2011: Comments on ‘‘The role of the central
Asian mountains on the midwinter suppression of North Pa-
cific storminess.’’ J. Atmos. Sci., 68, 2800–2803.
Cook, K. H., and I. M. Held, 1992: The stationary response to large-
scale orography in a general circulation model and a linear
model. J. Atmos. Sci., 49, 525–539.
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,
23–40.
Held, I. M., M. Ting, and H. Wang, 2002: Northern winter sta-
tionary waves: Theory and modeling. J. Climate, 15, 2125–
2144.
Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson,
and P. J. Rasch, 1998: The National Center for Atmospheric
Research Community Climate Model: CCM3. J. Climate, 11,
1131–1149.
Lin, S.-J., 2004: A ‘‘vertically Lagrangian’’ finite-volume dynamical
core for global models. Mon. Wea. Rev., 132, 2293–2307.
Nakamura, H., and T. Sampe, 2002: Trapping of synoptic-scale dis-
turbances into the North-Pacific subtropical jet core in mid-
winter. Geophys. Res. Lett., 29, 1761, doi:10.1029/2002GL015535.
Park, H.-S., J. C. H. Chiang, and S.-W. Son, 2010: The role of the
central Asian mountains on the midwinter suppression of
North Pacific storminess. J. Atmos. Sci., 67, 3706–3720.
Smith, T. M., R. W. Reynolds, R. E. Livezey, and D. C. Stokes,
1996: Reconstruction of historical sea surface temperatures
using empirical orthogonal functions. J. Climate, 9, 1403–1420.
2806 J O U R N A L O F T H E A T M O S P H E R I C S C I E N C E S VOLUME 68
CORRESPONDENCE
Comments on ‘‘The Role of the Central Asian Mountains on theMidwinter Suppression of North Pacific Storminess’’
EDMUND K. M. CHANG
School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York
WUYIN LIN
Atmospheric Sciences Division, Brookhaven National Laboratory, Upton, New York
(Manuscript received 28 December 2010, in final form 9 May 2011)
In a recent study, Park et al. (2010) conducted experi-
ments using an atmospheric general circulation model
(AGCM) to study the impact of the central Asian moun-
tains on Pacific storm-track activity. Their results suggested
that ‘‘the presence of the central Asian mountains sup-
presses the North Pacific storminess by 20%–30% during
boreal winter’’ and can ‘‘amplify stationary waves and ef-
fectively weaken the high-frequency transient eddy kinetic
energy in boreal winter.’’ We are intrigued by such strong
sensitivity of the winter storm track and stationary waves
to the central Asian mountains, since in a previous study
(Chang 2009) we have conducted numerical experiments
to examine the impact of mountains on Northern Hemi-
sphere winter storm tracks and stationary waves by remov-
ing all mountains and found apparently weaker impacts
than those found by Park et al. (2010). Since the configu-
ration of the experiments examined by Chang (2009) is
different from those presented in Park et al., we have per-
formed some experiments similar to those discussed in Park
et al. to more directly compare our results to theirs.
The AGCM used in this study is the Community At-
mospheric Model version 3.1 (CAM3.1; see Collins et al.
2006), run at a resolution of T42 in the horizontal, with
26 hybrid sigma levels in the vertical. Compared to the
model used by Park et al. [version 3 of the National Center
for Atmospheric Research (NCAR) Community Climate
Model (CCM3) run at T42 and 18 levels; see Kiehl et al.
1998], the model used in this study has updated physics
and higher vertical resolution. Major changes in model
physics from CCM3 to CAM3 are described in Collins
et al. (2004). All experiments are run with climatological
SST as a lower boundary condition.
Park et al. (2010) conducted a series of experiments to
examine the impacts of the Altai-Sayan Mountains on the
Pacific storm track. Their control experiment (M100) is
run with full orography. They then conducted sensitivity
experiments by systematically reducing the orography
over central Asia (see their Table 1 and Fig. 1). The
experiments that they called M75 and M50 have the
Altai-Sayan mountains largely removed, with part of
the Tibetan Plateau also removed in the latter case, and
their M20 experiment has most of the Tibetan Plateau
reduced. We have conducted a similar series of experi-
ments, with the orography used in our four experiments
shown in our Fig. 1. Comparing our orography to theirs
(see their Fig. 1), our reductions are slightly more ag-
gressive (with more of the mountains removed); hence
we named our experiments MN100 (control with full
orography), MN70, MN40, and MN05, respectively
(see Fig. 1). As in Park et al. (2010), over regions where the
orography is reduced, the subgrid-scale variability (stan-
dard deviation) in orography (used in gravity wave drag
parameterization) is also reduced by the same ratio. Note
that outside of central Asia the orography is unchanged.
Since the results of Park et al. (2010) suggested that
the central Asian mountains have the largest impacts on
the Northern Hemisphere (NH) storm tracks during
midwinter, we will focus on this season here. To obtain
a longer time series for midwinter without having to
Corresponding author address: Edmund K. M. Chang, School of
Marine and Atmospheric Sciences, Stony Brook University, Stony
Brook, NY 11794-5000.
E-mail: [email protected]
2800 J O U R N A L O F T H E A T M O S P H E R I C S C I E N C E S VOLUME 68
DOI: 10.1175/JAS-D-11-021.1
� 2011 American Meteorological Society
run the experiments for many decades, we have con-
ducted persistent January forcing experiments. Previous
studies (e.g., Zhang and Held 1999; Chang 2009) have
shown that persistent forcing experiments using AGCMs
can reproduce NH winter climate very well, and Zhang
and Held (1999) also showed that the midwinter sup-
pression in Pacific storm-track activity can be repro-
duced in persistent forcing experiments run with forcing
from the different months. Each of the four experiments
is run for 15 yr under persistent 15 January insolation
and SST forcings, with data from the final 10 yr (or
120 months) analyzed to examine the sensitivity of
the Northern Hemisphere winter climate to differences
in orography.
The response of the winter stationary waves to changes
in orography is shown in Fig. 2 (this figure should be
compared to Fig. 8 of Park et al). The stationary wave
response is represented by the zonal asymmetrical (or
eddy) part of the 300-hPa streamfunction. The stationary
waves simulated in the control experiment are shown
in Fig. 2a. This pattern is very similar to that shown in
Fig. 8a of Park et al. (2010), but the amplitude is a bit
weaker here. However, our stationary wave amplitude
agrees better with observed January stationary waves
derived from National Centers for Environmental
prediction (NCEP)–NCAR reanalysis (e.g., see Fig. 1a of
Held et al. 2002).
Examining the responses of the stationary waves to
changes in orography, our results are consistent with
those of Park et al. (2010) in that the central Asian moun-
tains clearly enhance the stationary waves, especially the
part of the wave spreading from Asia across the Pacific into
western North America. However, it is clear that the am-
plitudes of the responses in our experiments are signifi-
cantly smaller than those shown in Fig. 8 of Park et al. For
example, in their M50 experiment, the negative center
over East Asia is reduced by over 18 3 106 m2 s21
(compared to a climatological amplitude of about 224 3
106 m2 s21), while in all of our experiments, even the one
with most of Tibet removed (MN05), the reduction never
exceeds half that value.
The storm-track response to changes in orography is
shown in Fig. 3 (this should be compared to Fig. 4a of
Park et al.). Similar to Park et al., to represent ‘‘storminess,’’
we use 8-day high-pass filtered standard deviation of
300-hPa geopotential height. In Fig. 3a, the differences
between the control and MN40 experiments are shown
by shades (with the 25-m difference contour shown by
dotted contours), while the storm tracks simulated by
the MN40 experiment are shown by the solid contours.
FIG. 1. Orography over central Asia used in the four different experiments: (a) MN100, (b) MN70, (c) MN40, and (d)
MN05. Mountains higher than 500 m are shaded.
NOVEMBER 2011 C O R R E S P O N D E N C E 2801
We have used the same shade interval (10 m) and con-
tour interval as those used in Park et al. (2010) so that our
results can be directly compared to theirs. Comparing our
Fig. 3a to their Fig. 4a, it is clear that in both experiments,
the central Asian mountains strongly suppress storm-
track activity locally. However, downstream over east-
ern Asia, the Pacific, North America, and the Atlantic,
changes in orography generate much weaker responses in
our experiments than in those conducted by Park et al.
(2010). In Fig. 3b, the storm-track response to the MN05
experiment is shown. Even after removing nearly the
entire Tibetan Plateau, the increase in storminess never
exceeds 15 m in amplitude over the Pacific east of 1708E,
much less than the over 30-m increase in amplitude found
by Park et al. (2010) for their M50 experiment.
As in Park et al. (2010), we have also computed the
storm-track response in terms of 300-hPa filtered eddy
kinetic energy (EKE). The results for the MN05 experi-
ment are shown in Fig. 3c. Consistent with the results for
storminess (Fig. 3b), the EKE is strongly suppressed near
the mountains, but the reduction is significantly less away
from them. Compared to Fig. 9a in Park et al. (2010),1
over the central and eastern Pacific, the response in our
experiments is clearly much weaker even with much
more of Tibet removed.
Overall, our results suggest that while the central
Asian mountains do enhance the stationary waves and
suppress storm-track activity during midwinter, the
sensitivity of both to the mountains may be much less
than what the results of Park et al. (2010) suggest. Our
results suggest that while the central Asian mountains
may indeed contribute to the midwinter suppression of
the Pacific storm track, merely removing these moun-
tains is unlikely to remove the suppression, and other
FIG. 2. (a) The 300-hPa eddy streamfunction for MN100. (b)–
(d) Anomalous eddy streamfunction calculated from the differ-
ences between MN100 and others: (b) MN100 2 MN70, (c)
MN100 2 MN40, and (d) MN100 2 MN05. Contour interval is
3 3 106 m2 s21.
FIG. 3. (a) Anomalous 300-hPa storminess (shading, interval
210 m; dotted line shows 25-m contour), calculated from the
differences between MN100 and MN40. The contour lines indicate
climatological mean storminess for MN40. (b) As in (a), but for the
MN05 experiment. (c) As in (b), but for anomalous 300-hPa fil-
tered EKE (shading interval 215 m2 s22).
1 The EKE values (contours) shown in Fig. 3c appear to be much
less than those shown in Fig. 9a of Park et al. (2010). However, note
that in our control experiment (MN100), the peak EKE values in
the eastern Pacific are only slightly smaller than those computed
based on 40-yr European Centre for Medium-Range Weather
Forecasts (ECMWF) Re-Analysis (ERA-40) data.
2802 J O U R N A L O F T H E A T M O S P H E R I C S C I E N C E S VOLUME 68
physical mechanisms are still needed to help explain
this phenomenon.
Currently, it is not clear why our results differ from
those of Park et al. (2010). Possibilities include differ-
ences in model physics as well as differences in vertical
resolution. Statistics may also play a small role, as Park
et al. (2010) only used data from 18 winter seasons, while
we analyzed 120 months of data. If we replot Fig. 2 using
only 18 months of data, the amplitude of the differences
between the control and sensitivity experiments would
have been larger in some of those plots. However, split-
ting our 120 months into six 18-month periods, we did not
find any period during which the change in stationary
waves is as large as those shown in Fig. 8 of Park et al.
With 120 months of data (see Fig. 2), the change in sta-
tionary waves (at least over Asia and the western Pacific)
is found to become systematically larger as more of the
mountains are removed, whereas with 18 months of data,
such systematic increase is not always observed, sug-
gesting that 18 seasons is probably not long enough to
quantitatively characterize the response.
Our results suggest that model-simulated storm-track
and stationary wave responses to changes in orographic
forcing appear to be very sensitive to the model used.
Further studies, perhaps using multimodel ensembles, as
well as efforts to understand what causes these large
model differences, will be needed to better quantify the
impacts of the central Asian mountains on Northern
Hemisphere winter climate.
Acknowledgments. One of us (EC) is supported by
NSF Grant ATM0757250. Most of the CAM runs are
conducted on the NCAR Bluefire supercomputer.
REFERENCES
Chang, E. K. M., 2009: Diabatic and orographic forcing of north-
ern winter stationary waves and storm tracks. J. Climate, 22,670–688.
Collins, W. D., and Coauthors, 2004: Description of the NCAR
Community Atmosphere Model (CAM 3.0). NCAR Tech.
Note NCAR/TN-4641STR, 210 pp. [Available online at http://
www.cesm.ucar.edu/models/atm-cam/docs/description/.]
——, and Coauthors, 2006: The formulation and atmospheric
simulation of the Community Atmospheric Model version 3
(CAM3). J. Climate, 19, 2144–2161.
Held, I. M., M. Ting, and H. Wang, 2002: Northern winter stationary
waves: Theory and modeling. J. Climate, 15, 2125–2144.
Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, D. L. Williamson,
and P. J. Rasch, 1998: The National Center for Atmospheric
Research Community Climate Model: CCM3. J. Climate, 11,
1131–1149.
Park, H.-S., J. C. H. Chiang, and S.-W. Son, 2010: The role of the
central Asian mountains on the midwinter suppression of
North Pacific storminess. J. Atmos. Sci., 67, 3706–3720.
Zhang, Y., and I. M. Held, 1999: A linear stochastic model of
a GCM’s midlatitude storm tracks. J. Atmos. Sci., 56, 3416–
3435.
NOVEMBER 2011 C O R R E S P O N D E N C E 2803
