Download Topographic Effects on a Wintertime Cold Front in Taiwan

NOVEMBER 2006
CHIEN AND KUO
3297
Topographic Effects on a Wintertime Cold Front in Taiwan
FANG-CHING CHIEN
Department of Earth Sciences, National Taiwan Normal University, Taipei, Taiwan
YING-HWA KUO
Mesoscale and Microscale Meteorology Division, National Center for Atmospheric Research, Boulder, Colorado
(Manuscript received 19 August 2005, in final form 27 January 2006)
ABSTRACT
This paper describes an observational and numerical study of an intense wintertime cold front that
occurred in Taiwan on 8 January 1996. The front was associated with rope clouds at the leading edge, and
a broad area of stratiform clouds behind. The front was blocked by the Central Mountain Range of Taiwan
and divided into two sections on each side of the mountain range. As the cold air moved southward along
the east coast, the increasing westward Coriolis force induced a landward acceleration. After the cold air
piled up against the mountains, a coastal pressure ridge developed. The cold air damming yielded a
geostrophic balance between the westward Coriolis force and the eastward component of the pressure
gradient force in the x direction, and a southward acceleration in the y direction mainly caused by the
southward pressure gradient force component. Over the Taiwan Strait, southward pressure gradient forces
increased when the low-level stable cold air was confined over the Taiwan Strait, leading to a southward
acceleration of the cold air. The formation of a windward ridge off the northwest coast of Taiwan contributed to a large southward acceleration, resulting in the development of a coastal jet. Over the Taiwan Strait,
the cold air moved southward the fastest due to the channeling effect. The air parcels along the east coast
of Taiwan experienced a downgradient acceleration from the cold air damming and advanced at a slower
speed. Those traveling over the western plains and the nearshore coast advanced at the slowest speed. Two
sensitivity runs, one without Taiwan’s topography (flat land only) and the other without Taiwan’s landmass,
demonstrated the influences of Taiwan’s terrain and water–land contrast on the airflow. The run with no
surface fluxes showed that the ocean modified the low-level cold air by supplying surface heat and moisture
fluxes. This weakened the front, reduced low-level stability, and increased forced shallow convection
(formation of rope clouds) at the leading edge.
1. Introduction
The influence of mountains on fronts has been intensively addressed by researchers around the world since
the early twentieth century. From a theoretical point of
view, Egger and Hoinka (1992) provided an extensive
review. Bannon (1983), Zehnder and Bannon (1988),
and Williams et al. (1992) showed the weakening (or
retardation) and strengthening (or acceleration) of the
front on the upslope and downslope of a mountain,
respectively. These studies indicated that mountaininduced divergence and convergence modify the inten-
Corresponding author address: Fang-Ching Chien, Dept. of
Earth Sciences, National Taiwan Normal University, 88, Section
4, Ting-Chou Rd., Taipei 116, Taiwan.
E-mail: [email protected]
© 2006 American Meteorological Society
MWR3255
sity of the front when it passes a mountain. Frontal
retardation and deformation by topography were examined in idealized modeling studies (e.g., Davies 1984;
Schumann 1987; Gross 1994). Braun et al. (1999a,b)
documented the deceleration on the upstream crosscoast flow, which reduced the frontal speed. They also
showed an along-coast jet that was intensified by the
Coriolis force and weakened by friction. Blumen and
Gross (1987a,b) used a simple three-dimensional model
to show that a front, represented by a passive scalar,
can experience horizontal distortion due to anticyclonic
flow around the mountain. Li et al. (1996), in a threedimensional numerical study, showed that large frontal
distortion can occur for smaller circular barriers (width:
200–500 km) and the front can regain its intensity after
passing a mountain.
A large number of observational studies addressed
3298
MONTHLY WEATHER REVIEW
FIG. 1. Terrain height of Taiwan. Shading starts at 200 m, with
an interval of 200 m. The frontal positions were drawn chronologically from 2200 UTC 7 Jan to 0800 UTC 8 Jan 1996.
the effect of topography on a front. Smith (1986) provided a comprehensive review on many phenomena associated with fronts approaching the Alps, including
prefrontal shallow foehn winds, frontal distortion, flow
splitting, and postfrontal bora winds. Observational
studies of Freytag (1990), Kurz (1990), and Hoinka and
Heimann (1988) showed the retardation and modification effects of the Alps on synoptic-scale fronts. There
has also been a significant number of studies dealing
with mountain effects on fronts in North America. For
example, Colle and Mass (1999), Chien et al. (2001),
and Doyle (1997) examined mesoscale phenomena resulting from frontal modification by coastal mountains
of the western United States. Hartjenstein and Bleck
(1991) and Colle and Mass (1995) studied orographically influenced cold fronts by the Rockies. Schumacher et al. (1996) and O’Handley and Bosart (1996)
examined the frontal retardation effects of the Appalachians. There have also been many studies examining
the processes associated with cold fronts that were
modified by the terrain in other parts of the world, such
as New Zealand (Smith et al. 1991), Australia (Colquhoun et al. 1985; Howells and Kuo 1988), and southern
Africa (Anh and Gill 1981).
Like the aforementioned regions around the world,
Taiwan is also ideally suited for studying the influence
of terrain on a front. As shown in Fig. 1, Taiwan is
surrounded by ocean, with complex terrain consisting
of steep mountains (exceeding 3000 m) such as the Cen-
VOLUME 134
tral Mountain Range (CMR). The unique terrain of the
CMR often plays an important role in modifying a front
and produces mesoscale phenomena that could be different than those in other regions. Since the Taiwan
Area Mesoscale Experiment (TAMEX; Kuo and Chen
1990) held in 1987, many studies of terrain effects have
focused on the mei-yu fronts that usually occurred in
early summer in east Asia. Chen et al. (1989) and Chen
and Hui (1990, 1992) show that the front is blocked and
distorted as the shallow cold air moves around the
mountains. This modification can lead to the formation
of low-level jets northwest of Taiwan (Wang and Chen
2002; Yeh and Chen 2003) and influence local rainfall
characteristics (Yeh and Chen 2002; Yu et al. 2001;
Chien et al. 2002). Trier et al. (1990) show that the
CMR acts as a barrier to both pre- and postfrontal
flows, modifying the frontal intensity. The mei-yu front
is often split by the CMR into two segments as it approaches northern Taiwan, with the portion along the
east coast of Taiwan moving faster than that along the
west coast (Trier et al. 1990; Chen 1992; Chen and Li
1995).
In contrast to the many studies of mei-yu fronts and
the accompanying mountain effects, little study has
been done on the wintertime cold fronts in Taiwan. Lau
and Chang (1987) reviewed the cold surges in east
Asia based on the Winter Monsoon Experiment
(WMONEX; Greenfield and Krishnamurti 1979). Chen
and Lin (1999) conducted a diagnostic case study of a
low-level front over southern China. Chen et al. (1999)
performed an observational study of an east Asian cold
front associated with roll clouds as the front passed
northern Taiwan. Chen et al. (2002) examined an east
Asian cold surge event. All these studies dealt with
wintertime cold fronts from a large-scale point of view.
The cold front described in Chen et al. (2002) was associated with a clear cloud pattern that included rope
clouds near the leading edge, stratiform clouds behind,
and cellular clouds farther behind the front. The front
was significantly modified by the CMR when it moved
over the island, which made it an excellent case for
studying Taiwan’s terrain effects on a cold front. Although Chen et al. (2002) partially addressed this subject by comparing the weather at different locations,
they did not examine the details of the cold front’s
mesoscale structures and how these structures are
modified by the mountain barrier. Using the fifthgeneration Pennsylvania State University–National
Center for Atmospheric Research (PSU–NCAR) Mesoscale Model (MM5), we carried out a modeling study
on this same event. We primarily focused on the evolution of the three-dimensional frontal structure, fron-
NOVEMBER 2006
CHIEN AND KUO
tal speed, and force balances when the front was influenced by Taiwan topography. We also examined the
effects of ocean surfaces and land–water contrasts
on the cold front. Wintertime cold fronts are different from mei-yu fronts generally in terms of having
larger baroclinity, higher static stability, lower moisture, and stronger winds. Thus, the terrain effects of
the two could be very different. We will compare our
results with the previous mei-yu frontal studies in
Taiwan.
2. Synoptic overview of the event
Chen et al. (2002) discussed the large-scale analyses
of the January 1996 frontal case in great detail. Since we
focus primarily on the mesoscale features, only necessary synoptic analyses are presented here. At 1200
UTC 7 January 1996, a continental high of 1056 hPa
was located in Mongolia. A midlatitude cyclone of 1015
hPa formed over the Sea of Japan (Fig. 2a). A cold
front stretched from the center of the low southwestward to southeast China. Surface winds behind the cold
front were northwesterly. Prefrontal winds were very
weak. Twelve hours later, the low had deepened to
1000 hPa and moved to the west of Japan. The
cold front moved southeastward to arrive at northern
Taiwan (Fig. 2b). Cold air behind the front rushed off
the continent, resulting in strong northerly or northeasterly winds over the ocean. Upper-level analyses revealed a short-wave trough moving eastward from
northeastern China to Japan during this time period
(not shown).
The visible satellite picture at 0033 UTC 8 January
1996 (Fig. 3a) showed rope clouds over the western
Pacific at the leading edge of the front. Behind the
front, there was a broad area of stratiform clouds 400–
500 km wide. Over the East China Sea, cellular clouds
were clearly visible as the continental cold air moved
over the ocean. The size of the cellular clouds was progressively larger as the air flowed southward. At
around 30°N, the cloud fields changed into a very different structure, with some deep convective cells embedded within the generally stratiform clouds. Because
of the steep topography of Taiwan, the cold front and
its clouds were blocked by the CMR. They were split
into two parts on the two sides of the CMR. The eastern
section had a clear leading edge and a small cloud-free
area immediately off the coast near Hualien. The
clouds associated with the western portion of the cold
front were relatively detached and the leading edge was
fuzzy.
Along the east coast of Taiwan, surface winds behind
the cold front were very strong, generally 30–40 kt. The
3299
rope clouds and the front moved southward rapidly
with a speed of about 50 km h⫺1. They reached the
southeast coast of Taiwan by 0233 UTC 8 January (Fig.
3b). Cold air damming along the east coast of Taiwan
played an important role in determining the fast frontal
speed. The cloud-free area appeared at nearly the same
location as 2 h earlier, although the front had moved
southward. Its stationary nature indicated a link with
topography. The central mountain areas, southern low
plains, and nearshore western ocean were all free of
clouds at this time.
As the front crossed Taiwan, hourly frontal positions
were plotted using all available data (see Fig. 1). The
front was deformed when making landfall in northern
Taiwan at about 2200 UTC 7 January 1996. After 0000
UTC 8 January, the front was blocked by the CMR,
separating into eastern and western portions. The eastern part of the front quickly moved southward along
the coast. The western part was delayed over the western plains. After 0400 UTC 8 January, the western part
of the front weakened. Its passage was barely detectable.
Figure 4 presents time series plots for sea level pressure, surface temperature, and wind at five selected surface stations in Taiwan: three on the east coast and two
on the west coast. Frontal passage at Keelung was
clearly evident at about 2200 UTC 7 January 1996 by
the sharp temperature drop, pressure rise, and rapid
wind shift from southwesterly to northerly. Six hours
later, the temperature at Keelung had dropped to 12°C.
Farther south, the cold front passed Hualien at about
0100 UTC 8 January, as indicated by a gentle temperature drop and clear wind shift. In southeastern Taiwan,
Dawu experienced a sharp temperature drop at 0400
UTC (noon local time) on 8 January as the heated air in
the morning increased the temperature difference before and after the frontal passage. At this station, surface winds shifted to northeasterly 2 h before the arrival
of the front, but wind speeds did not increase until the
front passed. Pressure changes within these three eastern stations overall show that pressure rose after the
frontal passage and the amount decreased as the front
moved progressively to the south. Hsinchu, a northwest
coastal station, observed a temperature drop and an
increasing northerly wind shortly after 2300 UTC 7
January. The front, indicated by a weak temperature
drop, passed Wuchi before 0100 UTC 8 January. Much
earlier, winds had shifted to northeasterly. On the
southwest low plains, the frontal passage was hardly
identified at most stations (not shown). During frontal
passage, there was scattered light rainfall at the north
coast of the island.
3300
MONTHLY WEATHER REVIEW
FIG. 2. Surface analyses at (a) 1200 UTC 7 Jan and (b) 0000 UTC 8 Jan 1996. The
contour interval for sea level pressure (solid lines) is 4 hPa. Frontal symbols and station
models follow standard plotting conventions. For winds, full barbs are 10 kt and half barbs
are 5 kt.
VOLUME 134
NOVEMBER 2006
CHIEN AND KUO
3301
in the vertical. The initial and lateral boundary conditions were obtained from the Central Weather Bureau
(CWB) global spectral model (Liou et al. 1997) with 1°
horizontal resolution. An objective-analysis procedure
based on successive correction (i.e., the Cressman
scheme) was performed in D1 to incorporate traditional upper-air and surface observations. The MM5
was initialized at 0000 UTC 7 January 1996, about 24 h
before the cold front made landfall at Taiwan. The
model was run for 48 h.
In the model simulations, the National Centers for
Environmental Prediction’s (NCEP’s) Medium-Range
Forecast model (MRF) PBL parameterization scheme
was used to represent planetary boundary layer processes (Hong and Pan 1996). The hydrological cycle
included the Kain–Fritsch subgrid-scale convective parameterization (Kain and Fritsch 1993), and a gridresolvable explicit moisture scheme in which cloud water, rainwater, and ice were predicted (Grell et al. 1994;
Dudhia 1993). Other schemes used in the MM5 simulation included the shallow convection, the cloud radiation, and the two-way interactive nested-grid schemes.
The control experiment (CT) used all of the physics
options described above. Simulation data of this run
were used to examine the three-dimensional structure
of the front during its passage over Taiwan. To determine the effect of Taiwan’s topography, land–water
contrast, and surface fluxes on the front and its movement, we performed three other experiments: the noTaiwan terrain (NT), the no-land (NL), and the nosurface fluxes (NF) runs. The terrain height of Taiwan
island was set to zero in the NT run. In the NL run, the
land of Taiwan was completely removed (land was replaced by ocean). In the NF run, surface heat and moisture fluxes were excluded during the simulation.
4. Results of the control run
In this section, the output of the CT run is used to
describe and diagnose the three-dimensional structural
evolution of the front and the influence of terrain.
a. Surface evolution
FIG. 3. Visible satellite imaginary at (a) 0033 UTC 8 Jan and
(b) 0233 UTC 8 Jan 1996.
3. Model and experimental design
The model used in this study was the MM5. The
domain settings included three nested domains: D1 (45
km), D2 (15 km), and D3 (5 km), with 32 sigma levels
Figure 5a shows that the simulated synoptic pressure
pattern at 12 h compared very well with the observation
at 1200 UTC 7 January 1996 (Fig. 2a), including locations and intensities of both the continental high and
the low over the Japan Sea. The cold front extending
from the low center to southeastern China was also well
reproduced in the simulation. The northerly or northeasterly winds behind the cold front were stronger than
those observed at the surface stations over land. This is
3302
MONTHLY WEATHER REVIEW
FIG. 4. Time series of sea level pressure (solid lines, hPa), surface temperature (dashed
lines, °C), and wind (kt) from 0000 UTC 7 Jan to 1200 UTC 8 Jan 1996 at five selected
stations: Keelung, Hualien, Dawu, Hsinchu, Wuchi. See Fig. 1 for locations of the stations.
Times of frontal passage denoted by rhombuses are defined by a sharp temperature drop,
pressure rise, and wind shift.
VOLUME 134
NOVEMBER 2006
CHIEN AND KUO
FIG. 5. Simulated sea level pressure (solid lines in 4-hPa intervals), temperature at 100 m
MSL (dashed lines with shading at 5°C intervals), and 10-m winds (kt) in domain 1 at (a) 12
and (b) 24 h into the simulation.
3303
3304
MONTHLY WEATHER REVIEW
because they were 10-m winds obtained by extrapolating model winds to the height of 10 m MSL. Observed
winds were generally taken at a height of 1–2 m. Twelve
hours later, the high had moved southward, resulting in
a cold air surge and the southeastward movement of the
cold front (Fig. 5b). On the synoptic scale, this agreed
very well with the observations at 0000 UTC 8 January
(Fig. 2b).
In addition to sea level pressure and winds, vertically
integrated cloud water and ice quantitatively representing the simulated cloud is presented in Fig. 6 for domain
D2. The frontal positions shown here and after were
defined by examining the simulated temperature distribution. By 24 h into the simulation, the front had
reached northern Taiwan (Fig. 6a). It had strong northerly winds (⬃30 kt) and was followed by cold air rushing over the ocean. Compared with Fig. 1 at 0000 UTC
8 January, the simulated front was slightly delayed (for
less than 1 h). The cloud pattern was similar to that in
the satellite picture: a broad band of stratiform clouds
(⬃450 km) behind the front and a narrow belt of denser
clouds1 near the leading edge of the front. A close comparison, however, showed noticeable differences between the simulation and observations. First, there
were no ribbonlike rope clouds (⬍5 km width) near the
simulated front. Instead, the simulation showed a wider
belt of dense clouds at the leading edge of the front.
Second, the cloud belt was broken near shore northeast
of Taiwan. The wide cloud belt was to be expected
because it is not possible for a 15-km grid to reproduce
the rope clouds of ⬍5 km width. The latter was a failure
of the model probably resulting from deficiencies in the
simulation. Compared with the front farther to the
northeast where wind shifts and clouds were better
simulated, the front near Taiwan had very weak wind
shifts in the simulation. This might be the reason why
convection did not extend toward the northeast coast of
Taiwan.
Four hours later, the front was split into two parts by
the CMR (Fig. 6b), similar to the observations. Compared with Figs. 1 and 3b, the simulated front had a
phase delay of about 1 h. We will take into account this
difference in making comparisons between the simulation and observations. Similar to that at 24 h, clouds
near the front, at this time, became thinner near the
east coast of Taiwan. Northerly winds in the cold air
were strong (about 12–15 m s⫺1), with large wind shifts
appearing only approximately east of 125°E. West of
125°E, winds showed only speed differences across the
1
Dense clouds are represented by high integrated hydrometeor
content.
VOLUME 134
front. At this time, the wind direction turned northerly
or northeasterly ahead of the front. This was because of
a continuous north-to-south pressure gradient across
the front. Off the east coast of Hualien, there was a
cloud-free region that matched the satellite picture
(e.g., Fig. 3b). The CMR blocked the low-level cold air
in northern Taiwan, resulting in dense clouds on the
windward side and clear sky over the southwest plains
on the leeward side. The western part of the front
moved southward along the Taiwan Strait. Its position
was difficult to identify by wind or pressure patterns.
The three-dimensional frontal structures were further examined using the simulation of D3 (the 5-km
grid). At 24 h, a narrow band of dense clouds extended
behind the front from the northeast coast of Taiwan
toward the Taiwan Strait (Fig. 7a). The leading edge of
the cold air was not clear because clouds also formed
far ahead of the front. Twenty-eight hours into the
simulation, the eastern portion of the front had approached the southeast coast of the island (Fig. 7b).
Behind the front, a pressure ridge formed along the east
coast as a result of cold air damming. The cold air was
blocked against the steep terrain on the right-hand side
as it surged southward. Offshore of Hualien near 24°N,
the cloud-free region was clearly present within the
broad cloud band behind the front. This area was located in the lee of the gap between two major peaks of
the CMR. Westerly flow above the cold air in the upstream passed through the gap and resulted in leeside
subsidence that suppressed cloud development in this
small region close to the land (not shown).
b. Cross sections
Figure 8 shows potential temperature, cloud (water
plus ice) mixing ratio, and winds for D2 at 28 and 30 h
for cross section A (see Fig. 6b for position of the cross
section). It is evident from Fig. 8a that while moving
southward, cold air behind the front was blocked
against the CMR within ⬃120 km of the east coast. The
isentropes tilted upward along the steep mountain
slope. The strongest northerly wind (⬃18.8 m s⫺1) was
found at a low level within the core of the cold air, and
it increased to ⬃19 m s⫺1 at 30 h (Fig. 8b). These findings suggest that cold air damming produced the pressure ridge along the east coast shown in Fig. 7b, resulting in a southward acceleration of the cold air. Low
clouds were found above the well-mixed layer. To the
west of the CMR over the strait, cold air was well mixed
below ⬃800 m MSL and was confined between the
CMR and the continent of China. The isentropes
bulged upward toward the slope of the wedge of the
colder air over land. In addition to the cold air surging
NOVEMBER 2006
CHIEN AND KUO
FIG. 6. Simulated sea level pressure (contour lines at 1-hPa intervals), integrated cloud
water (shading), and 10-m winds (kt) in domain 2 at (a) 24 and (b) 28 h into the simulation. Line A in (b) represents the position of the cross section presented in Fig. 8.
3305
3306
MONTHLY WEATHER REVIEW
VOLUME 134
FIG. 8. Potential temperature, cloud mixing ratio, normalsection wind, and along-section circulation of D2 at (a) 28 and (b)
30 h for cross section A shown in Fig. 6b. Shading index of cloud
mixing ratio (g kg⫺1) is denoted in bottom. Contour intervals for
potential temperature (thick lines) and normal-section wind
(dashed lines, minus sign represents out-of-section direction) are
1 K and 2 m s⫺1, respectively. Wind vectors denote along-section
wind and vertical wind components, with scales shown in bottom.
FIG. 7. Simulated sea level pressure (contour lines at 1-hPa
intervals), integrated cloud water (shading), and 10-m winds (kt)
in domain 3 at (a) 24 and (b) 28 h into the simulation. Lines B and
C in (b) represent the positions of cross sections presented in
Fig. 9.
southward down the strait, there was also cold air rushing off the coast in southeast China, enhancing the intensity of the cold air over the ocean. By 30 h (Fig. 8b),
cold air over the strait had reached a minimum potential temperature of ⬃285 K, and the maximum lowlevel northerly winds had increased from ⬃17.5 to ⬃18
NOVEMBER 2006
3307
CHIEN AND KUO
m s⫺1. The increase in northerly winds was due to a
channeling effect that resulted from the low-level stable
cold air traveling over the ocean with the Wu-yi Mountains of China on the right and the CMR on the left. A
similar type of gap flow was documented for southwesterly monsoonal flow during the mei-yu season (e.g.,
Chen and Li 1995) and for northeasterly flow associated with a typhoon (Lin et al. 2002). With low-level
cold air being more stable, the channeling effect in winter could be more prominent than in the other seasons.
Two other cross sections (along lines B and C in Fig.
7b) of D3 are shown to examine the vertical structure of
the front. The western cross section B, along the strait,
shows that convection was weak and the associated
cloud pattern was detached near the leading edge of the
front at approximately the 100-km mark (Fig. 9a).
Capped by a very stable layer about 1 km MSL, the cold
air was heated below by the ocean surface and became
well mixed below 600 m. As a result, stratiform clouds
developed above the mixed layer. The eastern cross
section C shows that the surface front was located near
the 140-km mark (Fig. 9b). About 30 km behind, a
narrow region of forced convection associated with a
large cloud mixing ratio occurred in the low troposphere. Between the 200- and 280-km marks, evident
from the undulated potential temperature pattern at
1–2 km MSL, clouds were suppressed by leeside subsidence from over-mountain flow through the gap of the
CMR. North of the 280-km mark, stratiform clouds
with gentle upward motion were evident from 700 to
2100 m MSL. A shallow layer (⬃100 m) of unstable air
was observed near the surface in the cold air because of
heat fluxes from the warm ocean beneath. Above, air
was well mixed with a depth of about 500 m.
c. Force balances
Figure 7b shows a pressure ridge forming behind the
front along the east coast of Taiwan. This was similar to
that described for cases over the western United States
(Chien et al. 1997), southeastern Australia (Colquhoun
et al. 1985), and southern Africa (Anh and Gill 1981).
These studies documented that the coastal ridge, with
mountains on the right (for the Northern Hemisphere),
created a downgradient acceleration along the coast.
We, therefore, computed the pressure gradient force
(PGF), the Coriolis force (CF), and the u component of
the ageostrophic winds in order to examine the force
balances along the east and west coasts. Figure 10a
shows that to the northeast of Taiwan there was a
northeast–southwest-oriented region (shading) with a
large westerly ageostrophic wind component behind
the front at 24 h. From
FIG. 9. Potential temperature, cloud mixing ratio, relative humidity, and winds of D3 at 28 h along cross sections (a) B and (b)
C shown in Fig. 7b. Shading index of the cloud mixing ratio
(g kg⫺1) is denoted at the bottom. Contour intervals for potential
temperature (thick lines) and relative humidity (dashed lines) are
1 K and 10%, respectively. Wind vectors denote along-section
wind and vertical wind components, with scales shown at the bottom.
d
V ⫽ ⫺g⵱z ⫺ f k ⫻ V ⫹ residual
dt
⫽ f k ⫻ Va ⫹ residual,
共1兲
where V is the wind vector, g is gravity, z is geopotential height, f is the Coriolis parameter, and Va is the
3308
MONTHLY WEATHER REVIEW
VOLUME 134
FIG. 10. The 1000-hPa pressure gradient force (thick arrow),
Coriolis force (thin arrow), and u component of ageostrophic
wind (contour and shading) of D3 at (a) 24, (b) 26, and (c) 28 h.
Vector scales (10⫺4 m s⫺2) are shown at the bottom. The contour
interval is 10 m s⫺1, with dark shading representing positive (⬎10
m s⫺1) values and light shading used for negative values (⬍⫺10
m s⫺1). Those values below ground level and the zero line are
omitted.
ageostropic wind vector, the westerly ageostropic wind
indicated a large southward acceleration of cold air, if
residual (or friction) terms are neglected. The southeastward PGF and the westward CF contributed to the
acceleration. Along the northeast coast of Taiwan, the
cold air was advancing farther southward. Cold air was
blocked over northern Taiwan where the ageostrophic
winds were weak westerly, or even easterly, as a result
NOVEMBER 2006
CHIEN AND KUO
of weak PGFs. Over the strait, the southward PGF
component contributed to an along-strait acceleration,
which was especially large near the front. The acceleration was toward the southwest behind the cold front
near the coast of China.
At 26 h, the area of large southward acceleration (or
westerly ageostrophic wind) along the east coast had
moved with the front to the south of 24°N (Fig. 10b).
The large southward PGF dominated the force balances. Behind the front, cold air increased its southerly
speed. Consequently, the increasing westward CF induced a landward acceleration of cold air, and the cold
air piled up against the mountains and resulted in a
coastal pressure ridge. The cold air damming produced
a geostrophic balance between the westward CF and
the eastward component of PGF in the x direction, and
a southward acceleration in the y direction, mainly contributed by the southward PGF component. Over most
of the strait, the along-strait acceleration of cold air
increased from 2 h earlier as a result of the increasing
southward PGF component. The forces in the crossstrait direction were nearly in geostrophic balance between the westward CF and the eastward PGF component.
Two hours later at 28 h (Fig. 10c), the large southward acceleration region off the east coast had moved
farther south to the southeast coast. Behind the front,
there was an area of weak PGF as a result of a small
pressure variation as shown in Fig. 7b. The force balances show a westward acceleration that was mainly
contributed by the CF. There was neither cold air damming nor was there coastal pressure ridge development
in this region because cold air was still accelerating
toward the coast. This was another reason, besides
lee subsidence, to explain why there was no cloud formation around this area. Farther north, cold air damming and the pressure ridge were both prominent, resulting in significant southward acceleration of cold air.
To the north and northwest of Taiwan, blocking on
the windward slope reduced the southward acceleration. The formation of a windward ridge (Fig. 7b) off
the northwest coast of Taiwan contributed to large
positive ua near the region. This resulted in a large
southward acceleration and a coastal jet offshore of
the northwest coast. This was similar to the barrier jet,
but under the southwesterly flow condition during a
mei-yu season (Yeh and Chen 2003). Over the strait
was a wide area of westerly ageostrophic winds (southward acceleration) in the cold air. The large southward
pressure gradient, resulting from confined low-level
cold air, was responsible for the southward acceleration.
3309
FIG. 11. Trajectories of air parcels (from 24 to 34 h) that are
released behind the front at ␴ ⫽ 0.9 in domain 3. Arrows denote
the positions of air parcels at every hour, with black dots showing
those at 30 h except for air parcel number 8. The heights (␴
values) of air parcels are indicated by the sizes of arrows. Legend
is shown in the lower-right corner. Terrain height of D3 is contoured with a 500-m interval.
d. Frontal movement
Figure 11 presents trajectories of air parcels released
along a line about 60 km behind the cold front at a
height of ␴ ⫽ 0.9. The three-dimensional winds of D3
were used to integrate the paths of the air parcels from
24 to 34 h. Trajectories (T) 7 and 8 showed that lowlevel cold air was blocked and lifted by the terrain. The
path of T7 rose and circled northwest of the CMR and
traveled at a lower level over the western plains. The
path of T8 ascended to a high level, turned to the northeast coast, and moved over the ocean. Other trajectories generally moved around the island in two groups:
east and west.
The eastern group of air parcels traveled southward
with an anticyclonic curvature over the ocean along the
east coast of Taiwan. The T10 and T11 paths advanced
the fastest with an average speed of about 47 km h⫺1,
slightly slower than the observed frontal speed estimated in section 2. This is because T10 and T11 traveled a longer distance with a curved path, while the
observed frontal speed was estimated by a straight line.
Other similar trajectories, T9, T12, and T13, were
slightly slower than T10 and T11. It is noted that the
3310
MONTHLY WEATHER REVIEW
VOLUME 134
FIG. 12. Time–space distribution of ␪, sea level pressure, the ␷ component of winds at the lowest sigma level along cross sections
(a)–(c) B, and (d)–(f) C of D3. The abscissa denotes the time of simulation (h), and the ordinate indicates the distance from the southern
end (km). Contour intervals for ␪, SLP, and v are 0.5 K, 1 hPa, and 2 m s⫺1, respectively. Fronts are shown by gray lines. Transitions
of frontal speeds are denoted by thin horizontal lines.
distribution of these air parcels at 30 h had a shape
similar to that of the nearshore front at 28 h (Fig. 7b).
The western group of trajectories mostly proceeded
over the Taiwan Strait. Trajectories 1–4 rushed southward in nearly a straight line over the ocean with fast
traveling speeds of ⬃58 km h⫺1. They were faster than
T5 and T6, which, influenced by Taiwan’s topography,
advanced with a slower speed of ⬃40 km h⫺1. Of all the
air parcels examined, those over the strait (T1–4) traveled the fastest, only slightly influenced by the island.
Air parcels along the east coast of Taiwan proceeded at
a slower speed, which was still faster than those trajectories near the west coast (T5 and T6).
Figure 12 presents a time–space distribution of potential temperature ␪, sea level pressure, and the ␷ component of the winds at the lowest sigma level along
cross sections B and C. The western cross section (B)
shows a two-step frontal transition (Fig. 12a). North of
the 290-km mark, the front exhibited a tighter potential
temperature gradient and slower speed,2 indicating that
the front weakened and traveled faster (from ⬃40 to
⬃60 km h⫺1) as it moved southward along the strait.
With cold air surging behind the front, southward pressure gradients increased substantially along the strait
(Fig. 12b). As a result, northerly winds greatly intensified north of the 290-km mark (Fig. 12c). The strongest
northerly winds were found at 100–300 km of the cross
section, and they persisted for the entire period. The
2
This is judged from the slopes of lines that connect the leading
edge of the tight potential temperature gradient.
NOVEMBER 2006
CHIEN AND KUO
increased north–south pressure gradient over the strait
after the frontal passage suggested that the channeling
effect of the low-level cold air was responsible for the
increasing frontal speed and the intensifying northerly
winds.
Along the eastern cross section, the front experienced three stages with different frontal speeds (Fig.
12d). North of the 300-km mark, the front moved
southward with a speed of ⬃35 km h⫺1. Between 100
and 300 km of the cross section, the front moved faster
with a speed of ⬃50 km h⫺1. Farther south, it moved
slower with a speed of ⬃32 km h⫺1. The front became
weaker as it headed southward. Southward pressure
gradients were large between the 250- and 310-km
marks of the cross section (Fig. 12e). The large gradients were almost stationary throughout the entire period. This is a result of the cold air damming that was
significantly intensified around this region by the steep
mountains to the west. Therefore, the front increased
its speed and northerly winds intensified until the end
of the period (Fig. 12f).
5. Sensitivity studies
To further identify the effects of Taiwan’s topography, the land–water contrast, and surface fluxes on the
cold front, we performed three sensitivity experiments.
The NT run of D2 at 28 h (Fig. 13a) shows that the
simulated sea level pressure, clouds, and winds were
very similar to those of the CT run (Fig. 6b), except for
the region near Taiwan where the terrain was removed.
Comparisons of the two figures showed the orographical effects of Taiwan’s terrain on the front. These effects included flow splitting, frontal distortion, cold air
blocking, and cloud enhancement on the windward
side, and a coastal pressure ridge along the east coast of
Taiwan.
Differences of sea level pressure and 500-m winds in
D3 at 30 h between the CT and NT runs (Fig. 13b)
clearly show the effects of Taiwan’s topography on the
cold air.3 A pressure ridge developed on the windward
(or northern) side of the island because of blocking on
the cold air, resulting in an outward flow (from the high
center) of the wind difference fields. This indicates that
the 500-m airflow was blocked and split by the CMR in
northern Taiwan. The presence of terrain resulted in a
coastal pressure ridge along the east coast of Taiwan,
except offshore near 24°N. Since winds were generally
northerly in direction along the east coast, wind differences indicate that the airflow had a more westerly
3
The large sea level pressure over substantial terrain may contain errors and should be ignored.
3311
component near the east coast, a more northerly component near the coastal ridge, and a slower speed along
the southeast coast because of Taiwan’s terrain. The
slower speed at the southeast coast was not necessarily
in contradiction to the alongshore acceleration as discussed in section 4. In the NT run, cold air could rush
southward without much delay because of the absence
of terrain. In other words, the difference in the overall
synoptic-scale environment had overshadowed the terrain effects of the acceleration in a local area. To the
west, Taiwan’s topography caused low pressure to form
over the southwest lowlands and the strait. The northerly wind difference extended westward to the southeast coast of China. This implied that although Taiwan’s topography had the most impact on the flow near
the island, it still influenced the flow over the entire
strait. Without the CMR there would be no channeling
effect and the wind speed over the strait would be
smaller.
Figure 13c shows the same plot as Fig. 13b but for
differences between the NT and NL runs. It is evident
that the effect of the water–land contrast (friction, heat
capacity, etc.) produced higher pressure and weaker
northerly winds over land. These can be seen from Fig.
13a as well by comparing the pressure and winds over
land and over ocean. Because of the low heat capacity
over land, a lower temperature and thus a higher pressure formed over land behind the front. Over ocean,
slightly higher pressure was induced in the cold air and
winds had a component more outward from the island
because of the influences of the land. Off the southeast
coast near the location of the front, winds were more
northerly in the NT run than in the NL run, suggesting
that the small pressure change resulting from the water–land contrast could still contribute to the flow acceleration near the front along the east coast.
In the NF run, the surface heat and moisture fluxes
were ignored, but Taiwan’s terrain was retained. Figure
14a shows that without the surface fluxes from the
warm and moist ocean, the front preserved its continental structure and moved faster than in the CT run
(Fig. 6b). Behind the front were intense pressure and
temperature gradients and strong winds. These were
evident in Fig. 14b, which shows the differences in the
temperature and winds at 500 m MSL between the CT
and NF runs. Clouds behind the front in the NF run
showed a wider and more homogeneous pattern than in
the CT run (Fig. 14a). There was no narrow cloud band
near the leading edge of the front to the east of Taiwan.
This was because convection was suppressed in the NF
run when the low-level flow increased the stability in
the absence of a supply of surface heat and moisture
fluxes from the ocean. One exception to this pattern,
3312
MONTHLY WEATHER REVIEW
VOLUME 134
FIG. 13. (a) Same as in Fig. 6b but for the NT experiment. (b) Differences in sea level pressure and winds at 500 m MSL between
the CT and NT runs at 30 h. Contour interval is 0.3 hPa, with positive values shaded. A full wind barb denotes 5 m s⫺1. (c) Same as
in (b) but for differences between the NT and NL runs.
however, was noted along the southeast coast where
clouds developed as a result of strong cold air damming. The damming produced upward motion on the
mountain slopes because of the increasing westward
wind components as the air parcels moved south.
The above points are more clearly shown in the vertical frontal structure along cross section C (Fig. 14c).
The front was associated with an intense potential temperature gradient near the leading edge. The cold air,
capped by an extremely stable layer at 600–1000 m
NOVEMBER 2006
CHIEN AND KUO
3313
FIG. 14. (a) Same as in Fig. 6b except for the NF run. (b) Differences in temperature and winds at 500 m MSL between the CT
and NF runs at 28 h. Contour interval is 1 K, and a full wind barb denotes 5 m s⫺1. (c) Same as in Fig. 9b except for the NF run.
MSL, was nearly neutral in the low level probably because of mechanical mixing. The clouds near the leading edge of the front and those extending ⬃100 km
behind were a result of strong cold air damming. Between the 200- and 280-km marks, there was a cloudfree area with an undulated potential temperature pattern and weak sinking motion that were also present in
the CT run. The frontal speeds of the two experiments
were different, but the cloud-free area appeared at the
same location. This stationary phenomenon clearly indicated that it was related to orographical subsidence.
Since the cold front and the cold air behind in the NF
run were both much more intense than in the CT run,
the coastal ridge along the east coast and the channel-
3314
MONTHLY WEATHER REVIEW
ing effect over the strait in the NF run were also stronger. Consequently, the front of the NF run moved faster
than in the CT run as it passed Taiwan. For example,
along the east coast the frontal speed was ⬃56 km h⫺1
in the NF run, and ⬃50 km h⫺1 in the CT run.
6. Summary and conclusions
This paper describes an observational and numerical
study of an intense wintertime cold front and its associated cold air outbreak event that occurred in Taiwan
on 8 January 1996. The PSU–NCAR Mesoscale Model
(MM5) was used to examine the three-dimensional
flow structure of the front and how it was modified by
the mountains of Taiwan. Observations show that the
front was associated with rope clouds at the leading
edge and a broad area of stratiform clouds behind. After making landfall on Taiwan, the front and its clouds
were blocked by the CMR and were split into two parts.
The eastern section that moved quickly southward
along the coast had a clear shape at the leading edge
and a small cloud-free area behind the cold front off the
coast near Hualien. The clouds associated with the
western portion of the cold front were relatively detached.
The control run reproduced the cloud pattern associated with the front, except that there were no rope
clouds. This is because the grids were not fine enough
to reproduce rope clouds of ⬍5 km width. The simulation also reproduced the cloud-free area behind the
front off the east coast near Hualien. Clouds were suppressed because westerly flow passed through the gap
between the two major peaks of the CMR and resulted
in subsidence near this region. As cold air surged southward along the east coast, the increasing westward Coriolis force induced a landward acceleration. The cold
air gradually piled up against the mountains, resulting
in a coastal pressure ridge. The cold air damming
yielded a geostrophic balance between the westward
Coriolis force and the eastward component of the pressure gradient force in the x direction, and a southward
acceleration in the y direction mainly caused by the
southward pressure gradient force component. The
damming did not happen near Hualien where air was
accelerating toward the coast. This was another reason
why there was no cloud formation around this region.
The western part of the front moved southward along
the Taiwan Strait. Cold air was confined over the strait
between the Wu-yi Mountains of China on the right
and the CMR on the left, resulting in an increasing
southward pressure gradient force component. Such a
channeling effect caused the along-strait acceleration of
cold air to increase over most of the strait. The forces in
VOLUME 134
the cross-strait direction were nearly in geostrophic balance. The formation of a windward ridge off the northwest coast of Taiwan contributed to significant southward acceleration, resulting in the development of a
coastal jet. This was similar to the barrier jet, but under
the southwesterly flow condition during a mei-yu season, documented by Yeh and Chen (2003), near this
region.
It is shown that the western part of the front was
intense before entering the Taiwan Strait. The front
weakened and moved faster as it moved southward
along the strait. Along the east coast, the front first
increased its speed off the coast where the cold air damming effect was most intense. Then, it slowed down
near the southeast coast after passing over the steepest
mountains of the east coast. The intensity of the front
decreased as the front headed southward. Trajectory
analyses further show that air parcels traveling along
the east coast experienced a downgradient acceleration
from cold air damming, advancing at a moderate speed
of approximately 47 km h⫺1. The western air parcels,
moving over the Taiwan Strait, traveled at the fastest
speed (⬃58 km h⫺1) because of the channeling effect.
Air parcels proceeding over the western plains and the
nearshore ocean were influenced by Taiwan’s topography, advancing at the slowest speed of ⬃40 km h⫺1.
These results were somewhat different from the previous mei-yu front studies that simply indicated that a
front along the east coast of Taiwan moved faster than
one along the west coast. An important factor that contributed to this difference appears to be the greater
channeling effect that occurs in winter because of more
stable low-level air.
The no-Taiwan terrain run clearly indicates the orographical effects of Taiwan’s terrain on the front. These
include flow splitting, frontal distortion, cold air blocking, and cloud enhancement on the windward side, and
a coastal pressure ridge along the east coast of the island. The effect of the water–land contrast produced
higher pressure and weaker northerly winds over land
behind the front. Off the southeast coast near the location of the front, the small pressure change resulting
from the water–land contrast could still contribute to
the flow acceleration near the front along the east coast.
Without the surface fluxes from the warm and moist
ocean, the front could reserve its continental structure
and move faster. Behind the cold front were intense
pressure and temperature gradients and strong winds.
The surface heat and moisture fluxes over the ocean
modified the low-level cold air behind the front, weakening the front, reducing low-level static stability, and
increasing the shallow convection (formation of rope
clouds) at the leading edge.
NOVEMBER 2006
CHIEN AND KUO
Acknowledgments. This research was supported by
the National Science Council of Taiwan (Grant NSC912111-M-003-004). We thank the Central Weather Bureau for numerical and observational data. The late Dr.
Shang-Wu Li contributed to the data collection. Mr.
Sheng-Feng Lin performed the numerical experiments
and Miss Chen-Wen Wu of the Central Weather Bureau Satellite Center provided satellite pictures. Their
assistance is greatly appreciated.
REFERENCES
Anh, N. N., and A. E. Gill, 1981: Generation of coastal lows by
synoptic-scale waves. Quart. J. Roy. Meteor. Soc., 107, 521–
530.
Bannon, P. J., 1983: Quasi-geostrophic frontogenesis over topography. J. Atmos. Sci., 40, 2266–2277.
Blumen, W., and B. D. Gross, 1987a: Advection of a passive scalar
over a finite-amplitude ridge in a stratified rotating atmosphere. J. Atmos. Sci., 44, 1696–1705.
——, and ——, 1987b: Semigeostrophic flow over topography in a
stratified rotating atmosphere. Part I: Steady threedimensional solutions over finite ridges. J. Atmos. Sci., 44,
3007–3019.
Braun, S., R. Rotunno, and J. B. Klemp, 1999a: Effects of coastal
orography on landfalling cold fronts. Part I: Dry, inviscid
dynamics. J. Atmos. Sci., 56, 517–533.
——, ——, and ——, 1999b: Effects of coastal orography on landfalling cold fronts. Part II: Effects of surface friction. J. Atmos. Sci., 56, 3366–3384.
Chen, G. T.-J., 1992: Mesoscale features observed in the Taiwan
mei-yu season. J. Meteor. Soc. Japan, 70, 497–516.
——, and K.-C. Lin, 1999: A diagnostic case study of a winter
low-level front over southern China. Mon. Wea. Rev., 127,
1096–1107.
Chen, T.-C., M.-C. Yen, S.-Y. Wang, and R. W. Arritt, 1999: Roll
clouds associated with an east Asian cold front. Bull. Amer.
Meteor. Soc., 80, 2199–2208.
——, ——, W.-R. Huang, and W. A. Gallus, 2002: An east Asian
cold surge: Case study. Mon. Wea. Rev., 130, 2271–2290.
Chen, Y.-L., and N. B.-F. Hui, 1990: Analysis of a shallow front
during the Taiwan Area Mesoscale Experiment. Mon. Wea.
Rev., 118, 2649–2667.
——, and ——, 1992: Analysis of a relative dry front during the
Taiwan Area Mesoscale Experiment. Mon. Wea. Rev., 120,
2442–2468.
——, and J. Li, 1995: Characteristics of surface airflow and pressure patterns over the island of Taiwan during TAMEX.
Mon. Wea. Rev., 123, 695–716.
——, Y.-X. Zhang, and N. B.-F. Hui, 1989: Analysis of a surface
front during the early summer rainy season over Taiwan.
Mon. Wea. Rev., 117, 909–931.
Chien, F.-C., C. F. Mass, and Y.-H. Kuo, 1997: Interaction of a
warm-season frontal system with the coastal mountains of the
western United States. Part I: Prefrontal onshore push,
coastal ridging, and alongshore southerlies. Mon. Wea. Rev.,
125, 1705–1729.
——, ——, and P. J. Neiman, 2001: An observational and numerical study of an intense landfalling cold front along the northwest coast of the United States during COAST IOP 2. Mon.
Wea. Rev., 129, 934–955.
3315
——, Y.-H. Kuo, and M.-J. Yang, 2002: Precipitation forecasts of
the MM5 in Taiwan area during the 1998 mei-yu season. Wea.
Forecasting, 17, 739–754.
Colle, B. A., and C. F. Mass, 1995: The structure and evolution of
cold surges east of the Rocky Mountains. Mon. Wea. Rev.,
123, 2577–2610.
——, and ——, 1999: An observational and numerical study of a
cold front interacting with the Olympic Mountains during
COAST IOP5. Mon. Wea. Rev., 127, 1310–1334.
Colquhoun, J. R., D. J. Shepherd, C. E. Coulman, R. K. Smith,
and K. McInnes, 1985: The southerly buster of south eastern
Australia: An orographically forced cold front. Mon. Wea.
Rev., 113, 2090–2107.
Davies, H. C., 1984: On the orographic retardation of a cold front.
Beitr. Phys. Atmos., 57, 409–418.
Doyle, J. D., 1997: The influence of mesoscale orography on a
coastal jet and rainband. Mon. Wea. Rev., 125, 1465–1488.
Dudhia, J., 1993: A nonhydrostatic version of the Penn State/
NCAR Mesoscale Model: Validation tests and simulation of
an Atlantic cyclone and cold front. Mon. Wea. Rev., 121,
1493–1513.
Egger, J., and K. P. Hoinka, 1992: Fronts and orography. Meteor.
Atmos. Phys., 48, 3–36.
Freytag, C., 1990: Modifications of the structure of cold fronts
over the foreland and in a mountain valley. Meteor. Atmos.
Phys., 43, 69–76.
Greenfield, R. S., and T. N. Krishnamurti, 1979: The Winter Monsoon Experiment—Report of December 1978 field phase.
Bull. Amer. Meteor. Soc., 60, 439–444.
Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of
the fifth-generation Penn State/NCAR Mesoscale Model
(MM5). NCAR/TN-398⫹STR, 138 pp.
Gross, B. D., 1994: Frontal interaction with isolated orography. J.
Atmos. Sci., 51, 1480–1496.
Hartjenstein, G., and R. Bleck, 1991: Factors affecting cold-air
outbreaks east of the Rocky Mountains. Mon. Wea. Rev., 119,
2280–2292.
Hoinka, K. P., and D. Heimann, 1988: Orographic channeling of a
cold front by the Pyrenees. Mon. Wea. Rev., 116, 1817–1823.
Hong, S.-Y., and H.-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea.
Rev., 124, 2322–2339.
Howells, P. A. C., and Y.-H. Kuo, 1988: A numerical study of the
mesoscale environment of a southerly buster event. Mon.
Wea. Rev., 116, 1771–1788.
Kain, J. S., and J. M. Fritsch, 1993: Convective parameterization
for mesoscale models: The Kain–Fritsch scheme. The Representation of Cumulus Convection in Numerical Models, Meteor. Monogr., No. 46, Amer. Meteor. Soc., 165–177.
Kuo, Y.-H., and G. T.-J. Chen, 1990: The Taiwan Area Mesoscale
Experiment (TAMEX): An overview. Bull. Amer. Meteor.
Soc., 71, 488–503.
Kurz, M., 1990: The influence of the Alps on structure and behavior of cold fronts over southern Germany. Meteor. Atmos.
Phys., 43, 61–68.
Lau, K.-M., and C. P. Chang, 1987: Planetary scale aspects of the
winter monsoon and atmospheric teleconnection. Monsoon
Meteorology, C.-P. Chang and T. N. Krishnamurti, Eds., Oxford University Press, 161–202.
Li, S.-W., M. S. Peng, and R. T. Williams, 1996: A threedimensional study of the influence of mountains on a front. J.
Atmos. Sci., 53, 2757–2772.
Lin, Y.-L., D. B. Ensley, S. Chiao, and C.-Y. Huang, 2002: Oro-
3316
MONTHLY WEATHER REVIEW
graphic influences on rainfall and track deflection associated
with the passage of a tropical cyclone. Mon. Wea. Rev., 130,
2929–2950.
Liou, C. S., and Coauthors, 1997: The second-generation Global
Forecast System at the Central Weather Bureau in Taiwan.
Wea. Forecasting, 12, 653–663.
O’Handley, C., and L. F. Bosart, 1996: The impact of the Appalachian Mountains on cyclonic weather systems. Part I: A
climatology. Mon. Wea. Rev., 124, 1353–1373.
Schumacher, P. N., D. J. Knight, and L. F. Bosart, 1996: Frontal
interaction with the Appalachian Mountains. Part I: A climatology. Mon. Wea. Rev., 124, 2453–2468.
Schumann, U., 1987: Influence of mesoscale orography on idealized cold fronts. J. Atmos. Sci., 44, 3423–3441.
Smith, R. B., 1986: Mesoscale mountain meteorology in the Alps.
Scientific Results of the Alpine Experiment, GARP Publications Series, Vol. 2, World Meteorological Organization, 407–
423.
Smith, R. K., R. Ridley, M. Page, J. Steiner, and A. Sturman, 1991:
Southerly changes on the east coast of New Zealand. Mon.
Wea. Rev., 119, 1259–1282.
VOLUME 134
Trier, S. B., D. B. Parsons, and T. J. Matejka, 1990: Observations
of a subtropical cold front in a region of complex terrain.
Mon. Wea. Rev., 118, 2449–2470.
Wang, C.-C., and G. T.-J. Chen, 2002: Case study of the leeside
mesolow and mesocyclone in TAMEX. Mon. Wea. Rev., 130,
2572–2592.
Williams, R. T., M. S. Peng, and D. A. Zankofski, 1992: Effects of
topography on fronts. J. Atmos. Sci., 49, 287–305.
Yeh, H.-C., and Y.-L. Chen, 2002: The role of offshore convergence on coastal rainfall during TAMEX IOP 3. Mon. Wea.
Rev., 130, 2709–2730.
——, and ——, 2003: Numerical simulations of the barrier jet over
northwestern Taiwan during the mei-yu season. Mon. Wea.
Rev., 131, 1396–1407.
Yu, C.-K., B. J.-D. Jou, and D. P. Jorgensen, 2001: Retrieved
thermodynamic structure of a subtropic, orographically influenced, quasi-stationary convective line. Mon. Wea. Rev., 129,
1099–1116.
Zehnder, J. A., and P. R. Bannon, 1988: Frontogenesis over a
mountain ridge. J. Atmos. Sci., 45, 628–644.