<<

Orographic Effects on South China Sea Summer Climate

Haiming Xu∗

Department of Atmospheric Sciences,

Nanjing University of Information Science and Technology, Nanjing, China

Shang-Ping Xie and Yuqing Wang

International Pacific Research Center and Department of Meteorology,

University of Hawaii, Hawaii, USA

Wei Zhuang, and Dongxiao Wang

LED,South China Sea Institute of Oceanology,

Chinese Academy of Science, Guangzhou, China

Submitted to Meteorology and Atmospheric Physics

April, 2007

∗ Corresponding author address: Dr. Haiming Xu, Department of Atmospheric Sciences, Nanjing University of Information Science and Technology, 114 New Street, Pancheng, Pukou District, Nanjing 210044, China. E-mail: [email protected] Abstract

New satellite observations reveal several distinct features of the South China Sea

(SCS) summer climate: an intense low-level southwesterly wind jet off the coast of

South Vietnam, a band on the western flank of the north-south running

Annam range, and a rainfall shadow to the east in the western SCS off the east coast of Vietnam. A high-resolution full-physics regional atmospheric model is used to investigate the mechanism for the formation of SCS summer climate. A comparison of the control model simulation with a sensitivity experiment with the artificially removed demonstrates that the aforementioned features form due to orographic effects of the Annam . Under the prevailing southwesterly monsoon, the mountain range forces the ascending motion on the windward and subsidence on the lee side, giving rise to bands of active and suppressed convection, respectively. On the south edge of the mountain range, the southwesterlies are accelerated to form an offshore low-level wind jet. The mid-summer cooling in the SCS induced by this wind jet further helps reduce precipitation over the central SCS. A reduced-gravity ocean model is used to investigate the ocean response to the orographically induced wind forcing, which is found to be important for the formation of the double-gyre circulation observed in the summer in SCS, in particular for the northern cyclonic circulation. Thus, is a key to shaping the SCS summer climate both in the atmosphere and in the ocean. 1. Introduction

The South China Sea (SCS) is a large semi-enclosed marginal ocean basin with a

total area of 3.5 million km2. It is connected to the East China Sea to the northeast, the

Pacific Ocean to the east, and the Indian Ocean to the southwest. The SCS climate is

part of the East Asian monsoon system [Lau et al., 1998]. In winter the SCS is

dominated by the strong northeasterly monsoon while in summer the winds reverse

the direction to southwesterly [e.g., Liu and Xie, 1999].

It is well known that in summer the southwesterly winds reach a maximum east

of Ho Chi Minh City [e.g., Xie et al., 2002]. Figure 1a shows the summer surface

wind climatology along with land topography over the Indochina peninsula. A narrow

mountain range, called the Annam cordillera, runs in a north-south direction on the

east coast of Indochina peninsula on the Vietnam-Laos borders and ends just north of

Ho Chi Minh City. Noting that the SCS wind jet is located just offshore of the south edge of the Annam cordillera, Xie et al. [2003] suggest that the orographic blockage

of the southwesterly monsoon and the wind acceleration at the south corner of the

mountain range are the cause of the wind jet. They go on to show that the strong curls

of this wind jet are the major drive of the summer SCS circulation, giving rise to a

number of important climatic features of the region such as a cold upwelling filament

[Huang et al., 1994; Kuo et al., 2000] that disrupts the summer warming. The wind jet

varies with the El Nino/Southern Oscillation (ENSO), driving interannual variability

of the SCS in both ocean circulation and sea surface temperature (SST). While

plausible, the orographic hypothesis of Xie et al. [2003] has never been rigorously

1 tested in numerical models.

The Annam cordillera also leaves a clear signature on precipitation. Figure 1b

shows the summer precipitation climatology in the region. As the southwesterly

monsoon impinges on the mountain range, the rising motion creates a windward rain

band and the subsequent subsidence produces a on the lee. Such

orographic rain bands are also observed on the west coasts of Myanmar, Thailand,

Cambodia, and the Philippines. With a numerical experiment, Xie et al. [2006]

suggest that such orographic rain bands are not simply a local phenomenon but exert important remote influences on the continental monsoon because of strong interaction between circulation and convection in the region during summer. The Annam

cordillera is more than 500 m high on average but only 200 km or less in width, posing a serious problem for numerical simulations. With typical horizontal resolution of 2-3o, the current global general circulation models represent the Annam mountain

range poorly and fail to simulate either the SCS wind jet or the orographic rain

band/shadow (not shown).

The summer circulation of the central SCS is dominated by a double-gyre

circulation, with an eastward inter-gyre jet in between that advects the cold coastal

water to form the cold filament east of south Vietnam. This pair of anticyclonic and cyclonic gyres south and north of roughly 12oN are observed from in-situ current

measurements [Fang et al., 2002] and satellite altimetry [Shaw et al. 1999; Ho, et al.,

2000]. With its strong wind curls, the orographic-induced southwest wind jet is

considered to be the major cause of this double-gyre circulation pattern [Xie et al.,

2 2003; Gan et al., 2006; Wang et al., 2006].

The present study test the above orographic hypothesis for the formation of the

wind jet and the couplet of the rain band and shadow in the summer SCS by using a

high-resolution regional atmospheric model. The 0.2º grid size of the model is

equivalent to T520 resolution for a global spectral model. Our results show that indeed, the Annam cordillera exerts a great influence on the wind and precipitation distributions over the Indochina peninsula and SCS. We further apply the atmospheric model results to assess the orographic effect on ocean circulation using a reduced-gravity ocean model. We find a strong effect on the northern cyclonic gyre north of the wind jet.

The rest of the paper is organized as follows. Section 2 describes the regional

atmospheric model, experimental design, and observational data sets used for

verification. Section 3 presents the atmospheric simulation results and investigates the

orographic effects of the Annam mountain range. Section 4 describes the ocean model

and presents the experiment results. Section 5 presents a summary and discussion.

2. Regional Atmospheric Model and Experimental Design

2.1. Atmospheric Model

The regional atmospheric model (RAM) developed at the International Pacific

Research Center (IPRC), University of Hawaii, is used in this study. It is a primitive

equation model with sigma as the vertical coordinate, solved on a longitude-latitude

grid system. The model domain is 5ºS-25ºN, 90º-135ºW, including the SCS,

Indochina peninsula, east part of the Bay of Bengal, and part of the western Pacific

3 (Fig. 2). The model uses a grid spacing of 0.2º in both longitude and latitude, and has

28 levels in the vertical. A detailed description of the model and its performance in

simulating regional climate of East Asia can be found in Wang et al. [2003]. The

model has also been used to simulate the regional climate over the eastern Pacific,

including the atmospheric response to tropical instability ocean waves [Small et al.,

2003], boundary layer over the southeast Pacific [Wang et al., 2004], the effects of the Andean and Central American mountains [Xu et al., 2004; Xu et al.,

2005], and more recent the diurnal cycle of precipitation over the Maritime continent region (Zhou and Wang 2006; Wang et al. 2007).

The model includes a detailed microphysics scheme for grid-scale moist processes [Wang, 2001]. The mixing ratios of cloud water, rainwater, cloud ice, , and graupel are all prognostic variables in the model. Condensation (evaporation) of

cloud water takes place instantaneously when the air is supersaturated (subsaturated).

Subgrid-scale convective processes, such as shallow convection, midlevel convection,

and penetrative deep convection, are considered based on the mass flux cumulus

parameterization scheme originally developed by Tiedtke [1989] and later modified by

Nordeng [1995].

The subgrid-scale vertical mixing is accomplished by the so-called E⎯ε closure

scheme, in which both the turbulence kinetic energy (TKE) and its dissipation rate are

prognostic variables [Detering and Etling, 1985]. Turbulent fluxes at the ocean

surface are calculated using the TOGA COARE algorithm [Fairall et al., 1996; Wang

2002]. Over the land, the bulk aerodynamic method is used in the land surface model,

4 which uses the Biosphere-Atmosphere Transfer Scheme [Dickinson et al., 1993]. Soil

moisture is initialized using a method described by Giorgi and Bates [1989] such that

the initial soil moisture depends on the vegetation and soil type defined for each grid

cell.

The radiation package originally developed by Edwards and Slingo [1996] and

later modified by Sun and Rikus [1999] is used, which includes seven/four bands for

longwave/shortwave radiation. Seasonal-varying climatological ozone and a constant

mixing ratio of carbon dioxide for the present climate are used.

2.2. Experimental Design

The initial and lateral boundary conditions are obtained from the National Centers

for the Environmental Prediction/National Center for Atmospheric Research

(NCEP/NCAR) global reanalysis [Kalnay et al., 1996], available on a 2.5º×2.5º grid

with 17 vertical pressure levels. They are interpolated onto the model grid by cubic

spline interpolation in the horizontal and linear interpolation in both the vertical and time based on the four times daily reanalysis. Over the ocean, the NOAA optimal interpolation weekly SST dataset on a 1º×1º grid is used as the lower boundary condition [Reynolds et al. 2002].

The following three sets of experiments are carried out to examine the effects of the Annam cordillera on the summer climate of SCS. Each consists of an ensemble of three simulations that are initialized at 0000UTC on 26, 27, and 28 June 2001, respectively, and integrated to the end of July 2001. July 2001 is selected during which the SCS wind jet is well developed (Fig. 3a). The rest of the paper discusses the

5 July means, averaged for three ensemble simulations based on their hourly output.

·Control (CTL) run. The model topography is based on the U. S. Geophysical Survey

1/12º topographic dataset and smoothed with an envelope topographic algorithm

(Wang et al. 2003). At our model resolution, the main Annam cordillera is represented reasonably well (Fig. 2a), at about 0.5 km high and 200 km in width.

·No-topography (NoTop) run. The Annam range south of 17.5ºN is flattened by setting land elevation at 0.5 m (Fig. 2b). Strictly speaking, the design of this no-topography run may be physically inconsistent with the imposed lateral boundary conditions that are influenced by the presence of these Annam mountains in the first place. Nevertheless, a comparison of the CTL and NoTop runs can help identify the regional effects of the mountain range within the context of this model.

·No-topography and smoothed SST (NoTopSmSST) run. While the comparison of the

CTL and NoTop runs can identify the direct effect of the Annam mountain range, this mountain range leaves marked signatures in the SST field in the form of a cold filament through the action of the wind jet [Fig. 2a; Xie et al., 2003]. This cold filament, in turn, affects the atmospheric circulation and convection, and may be considered as an indirect effect of Annam cordillera. In the NoTopSmSST run, we remove this orographically induced anomaly in SST as well as the mountain range.

The smoothed SST field is obtained by first setting SST to 29.0℃ if it is less than this value off the southeast coast of Vietnam, then applying a 5-point smoother 5 times in the coastal region. Fig. 2b shows the resultant SST field, which is nearly uniform near the southeast coast of Vietnam.

6 2.3. Observational Data

To evaluate the model simulations, we use 3-hourly Tropical Rainfall Measuring

Mission (TRMM) Real-Time (RT) multi-satellite precipitation analysis (3B42RT)

available since January 2002 on a 0.25º grid covering 50ºN to 50ºS [Huffman et al.,

2005]. The 3B42RT product combines measurements by the TRMM satellite's

Microwave Imager (TMI) and TMI-calibrated infrared estimates of precipitation from

geosynchronous satellites. We also use TRMM Precipitation Radar (PR) surface rainfall product 3A25G2 [Kummerow et al., 2000] from December 1997 to August

2005 on a 0.5º grid. The PR makes the best estimation of precipitation but its narrow

swath introduces larger sampling errors. We limit the use of the PR-only 3A25G2

product to construct a multi-year climatology in Fig. 1b while using the 3B42RT

product for the July 2001 analysis.

The microwave scatterometer on the National Aeronautics and Space

Administration’s (NASA) QuikSCAT satellite measures daily surface wind velocity

over the world ocean [Liu et al., 2000]. QuikSCAT observations have revealed rich

wind structures on short spatial scales around the world [Chelton et al., 2004]. Here

we use a monthly product for wind velocity available from August 1999 to August

2005 on a 0.25ºgrid.

3. Effect on Wind and Precipitation

3.1. Control Run

Figure 3 compares the simulated 10-m surface wind velocity averaged for July

2001 with QuikSCAT observations over the ocean for the same period. The model

7 simulates well the surface wind field over the SCS, with southwesterly winds north of

about 5ºN and southerly winds to the south. In particular, the model captures the

strong wind jet off the southeast coast of Vietnam with a maximum wind speed above

9 m s-1, in good agreement with QuikSCAT observations. The model also reproduces

the reduced wind speed lee of the Annam range. In the lee weak-wind zone, the model

exaggerates the alongshore variations in wind speed due to the smoothed orography

with two artificial saddles. For reasons unclear at this time, the winds tend to be

weaker than in observations south of the wind jet.

Figure 4 compares the monthly mean precipitation in the model with TRMM

3B42RT observations for July 2001. The model captures the orographic effects of

Indo-China mountains, including the rain bands on the Cambodia coast at the foothills

of the Cardamom Hills and on the west slope of the Annam cordillera as well as the

broad rain shadow lee of the Annams. We note that the rain band windward of the

Annams is not reproduced in a lower-resolution (0.5o) version of the IPRC RAM [Xie

et al., 2006], suggesting the importance of sufficient resolution. The model

underestimates precipitation west of the Philippines, a problem also noted in Xie et al.

[2006], illustrating the difficulty in simulating the convection-circulation interaction.

The southwesterly winds blowing west of Luzon Island are weak by 2 m/s compared

to QuikSCAT observations (Fig. 3), presumably reducing orographic effects of the island. Despite all these deficiencies, the model captures the salient features induced

by Indochina mountains, encouraging us to study the orographic effect of the Annams

with an experimental approach.

8 3.2. Topographic Effects

With the Annam mountain range removed in the NoTop run, the surface wind field is markedly changed. Without the mountain barrier, air flows freely across the

Indochina peninsula, and the southwesterly winds become much smoother in space over the northern SCS (Fig. 5a), especially off the east coast of Vietnam. In particular, wind speed markedly increases off the east coast of Vietnam, and the wind jet core disappears. The effect of the Annam cordillera on the SCS wind field can be better seen in the surface wind difference field between the CTL and NoTop runs (Fig. 5b).

The presence of the Annam range reduces the southwesterlies in the lee by as much as

5 m/s while increases the wind speed at its southern corner by up to 2 m/s. Thus the

SCS wind jet is indeed a result of the orographic blockage of the Annams, which forces the air to rush through the southern flank of the mountain range.

Without the Annam mountain range, the precipitation distribution is also markedly changed (Fig. 6a). The rain band on the windward side of the Annam cordillera disappears and the associated rainfall shadow weakens in the NoTop run, resulting in a much smoother precipitation distribution over the Indochina peninsula and western

SCS. The rainfall minimum on the coast of North Vietnam appears due to the low-level wind divergence as wind accelerates offshore due to reduced surface roughness over the ocean. The orographic effect on precipitation can be better seen in the difference field between the CTL and NoTop runs (Fig. 6b). A nearly zonally oriented dipole of rainfall anomalies, positive and negative roughly west and east of coastal line of Vietnam, respectively, indicates that the Annam mountain range acts to

9 force convection on the windward side while suppressing it on the leeside.

Now we examine further the mechanism by which the Annam mountain range affects the precipitation pattern. Figure 7 presents the cross sections of specific humidity and cloud liquid water content in both the CTL and NoTop runs, along with vertical velocity in the CTL run. The mountains force an updraft on the windward side

(Fig. 7c) with elevated cloud liquid water content (Fig. 7a) and precipitation (Fig. 4b).

On the lee side, on the other hand, the mountains cause a strong downdraft (Fig. 7c), which along with the windward precipitation depresses specific humidity by advecting dry air downward. About 100 km away from the coast, boundary layer moisture recovers to values high enough for convection to become active east of 112º

E.

In the NoTop run, the specific humidity (Figs. 7b, d) and vertical velocity (not shown) do not change much across the Indochina peninsula. Without mountains, moisture depletion by orographic rainfall is greatly reduced (Fig. 7b), and so is the subsidence off the east coast of Vietnam (not shown). Both effects act to increase moisture in the lower atmosphere over the western SCS near the Indochina peninsula in the NoTop run compared to the CTL run. The warming and drying induced by orographic subsidence on the SCS reach as far as 100-200 km offshore, amounting to

2℃ and -1.5 g kg-1 in magnitude, respectively (Fig. 7d).

3.3. SST Effects

Surface winds become uniform off the southeast coast of Vietnam in the NoTop run (Fig. 5a), indicating that it might be reasonable to assume that the SST field

10 would be uniform offshore without the Annam cordillera. Thus, we use a smoothed

SST field (Fig. 2b) to remove the indirect orographic effect of the Annams on summer

SCS climate.

The simulated precipitation pattern over the SCS in the NoTopSmSST run is

similar to that in the NoTop run, i.e., with a precipitation distribution quite uniform

over the Indochina peninsula and western SCS (not shown). The removal of the cold

filament in the western SCS enhances precipitation there by about 2 mm day-1, as seen

clearly in the precipitation difference field between the NoTopSmSST and NoTop run

(Fig. 8). Noticeable negative rainfall anomalies also exist on the south edge of the

Indochina Peninsula, possibly a remote response to increased convection over the mid-SCS, which forces subsidence Rossby waves to the west. The indirect effect of the cold filament on rainfall is about 30-50% of the direct orographic effect over the mid-SCS and is smaller on the Indochina Peninsula.

4. Effect on Ocean Circulation

This section assesses the orographic effect of the Annam cordillera on the SCS

summer circulation by using a 1.5-layer reduced-gravity ocean model. The model

consists of two layers: the upper active layer represents the upper ocean while the

lower layer represents the deep motionless ocean. At the interface, the entrainment or

detrainment is parameterized following McCreary and Yu [1992].

The model domain covers 0º-24ºN, 100ºE-124.5ºE, and the coastline is set by the

50 m isobath. The model grid size is 5´in both longitude and latitude. The water exchange between the SCS and the adjacent seas is mainly through the Luzon Strait,

11 with all other straits closed. The inflow velocity at the southern open boundary

(Luzon Strait) is calculated from the National Centers for Environmental Predication

Ocean Data Assimilation System (NCEP-ODAS) tropical Pacific Ocean analysis [Ji et

al., 1995]. The outflow velocity is determined by balancing volume transport. Zhuang

et al. [2006] gives a detailed description of this ocean model and its performance in

the SCS.

The ocean model is forced by QuikSCAT monthly climatological winds, spun up from the resting state and integrated for 5 years. The annual-mean thermocline depth

(102 m) and SST (22.6º℃) averaged in the SCS basin (deeper than 200 m) are used as the initial thickness and temperature of the upper layer, respectively. Sea surface

heat flux is parameterized as a Haney [1971] type relaxation toward observed SST

plus the Southampton constrained version of net heat flux [Grist et al., 2003]. The

annual cycle shows no appreciable drift in the last three years, indicating that the

model has reached a quasi-steady state. Therefore, results of the fifth year are saved

for analyses. Besides this ocean control (OcnCTL) simulation, we conduct an

experiment by removing the orographic effect from the QuikSCAT wind forcing

(OcnNoTop run). For simplicity, we define the orographic effect as the weighted July

wind velocity difference between the atmospheric CTL and NoTop runs. The weights

are (0.0, 0.5, 1.0, 1.0, 0.5, 0.0) for (May, June, July, August, September, October) to

mimic observations that the southwest wind jet begins to develop in June, is in full

strength in July and August, and weakens in September.

Figure 9 presents the three-month mean stream function fields averaged for July,

12 August, and September (JAS). In the OcnCTL run, the simulated upper layer circulation displays a strong anticyclonic gyre in the southern SCS, a weak cyclonic gyre to the north, and an offshore jet between the two gyres, which is in broad agreement with observations of Fang et al. [2002]. In addition, two weak cyclonic centers are embedded in the northern gyre with one center just off central Vietnam, and the other over the northeast SCS around 20ºN, 116ºE (Fig. 9a).

With the orographic effect removed in the OcnNoTop run, the northern cyclonic gyre retreats northeastward (Fig. 9b). The weak cyclonic recirculation off central

Vietnam in the OcnCTL is replaced by a weak anticyclone in the OcnNoTop run. The offshore jet shows a meander pattern, which previous numerical studies show is due to the north-south asymmetry in the wind stress forcing with a northeast tilted zero-curl line [Yasuda et al., 1996]. The orographic effect on the ocean circulation can be seen more clearly in the OcnCTL minus OcnNoTop stream function difference map (Fig. 9c), which is dominated by negative stream function anomalies east of

Central Vietnam lee of the Annams, with weak positive anomalies south of the wind jet. This anomalous circulation pattern is consistent with the anomalous wind curl that is much larger to the north of the wind jet lee of the Annam cordillera than to the south (Fig. 9d). Thus, the cyclonic (re-)circulation north of the wind jet owes its existence much to the orographic blockage by the Annams, which forces strong positive wind curls there. The anticyclonic gyre and recirculation, on the other hand, are only weakly affected by the Annam orographic effect. To the first order, the negative wind curls south of the SCS wind jet are part of lager-scale atmospheric

13 circulation rather than due to local orography, a result not obvious without numerical

experiments. Indeed, the southwesterly winds begin to transit to a southerly

cross-equatorial wind regime around 6-7oN in both observations and the CTL simulation (Fig. 3).

5. Summary and Discussion

A high-resolution regional atmospheric model is used to study the effects of the

Annam cordillera on the summer climate of the South China Sea. The model

reproduces the salient features of SCS summer climate as compared to QuikSCAT and

TRMM observations, including an intense wind jet off the southeast coast of Vietnam,

a precipitation band on the windward side of the Annam mountain range, and a

rainfall shadow on the lee side. Our experiments with and without the Annam

cordillera demonstrate that these features are indeed due to the orographic effects of

the mountain range. As the southwesterly winds impinge on the Annam range, the

induced ascending motion promotes deep convection on the windward side while the

subsidence suppresses convection on the lee side. At the southern tip of the mountain

range, the southwesterly winds rush through to form a strong wind jet offshore. Our

results also show that the cold filament under the coast wind jet, which itself results from orographic effect, is an additional factor helping suppress atmospheric

convection east of Vietnam.

A 1.5-layer reduced-gravity ocean model is also used to study the effect of

orographically-induced wind changes on the SCS ocean circulation. As the ocean

model is forced by monthly mean climatological QuikSCAT winds, the simulated

14 summer ocean circulation is characterized by a strong anticyclonic gyre in the

southern SCS, a weaker cyclonic gyre in the northern SCS, and a strong offshore

ocean jet in between off south Vietnam, in agreement with observations. With the orographic effects removed from the QuikSCAT winds, the simulated ocean circulation changes markedly, with the northern cyclonic gyre shifting northward. As a result, the eastward ocean jet weakens east of south Vietnam. Thus, the Annam cordillera exerts significant influences on the double-gyre circulation in the summer

SCS, especially the northern gyre and the inter-gyre eastward offshore jet.

This study joins a growing body of literature on coastal orography-induced air-sea

interaction phenomena [e.g., Xie et al., 2001; Chelton et al., 2004; Xie et al., 2005].

Owing to the increasing computing power, some of these phenomena have been successfully simulated in numerical models, for example, west of Hawaii [Sakamoto et al., 2004; Sasaki and Nonaka, 2006] and Central America [Sasai et al., 2007]. Over

the Asian summer monsoon region, narrow mountains such as the Annam play an

important role in the spatial distribution of convection [Chang et al., 2005; Xie et al.,

2006]. On a smaller regional scale over the SCS, the orographic effect of the Annams

leaves clear signatures in ocean circulation and SST. A regional coupled

ocean-atmosphere model [Xie et al., 2007] is being developed to study the

ocean-atmosphere interaction over the SCS and its effect on local and broad-scale

monsoon.

15 Acknowledgments. We wish to thank Jan Hafner for archiving the TRMM-PR and

QuikSCAT data from Remote Sensing Systems’ Web site and Y. Zhang for archiving

the NCEP reanalysis dataset. This work is supported by NSFC (40575045), 973

Program (2006CB403607; 2004CB418304), Chinese Academy of Sciences, NASA,

and the Japan Agency for Marine-Earth Science and Technology (JAMSTEC) through

its sponsorship of International Pacific Research center at University of Hawaii. IPRC

publication #xxx and SOEST publication #yyyy.

16 References

Chang, C.-P., Z. Wang, J. McBride and C. H. Liu (2005), Annual cycle of Southeast

Asia - Maritime Continent rainfall and the asymmetric monsoon transition, J.

Climate, 18, 287-301.

Chelton, D. B., M.G. Schlax, M.H. Freilich, and R.F. Milliff (2004), Satellite

measurements reveal persistent small-scale features in ocean winds, Science,

303, 978-983.

Detering, H.W., and D. Etling (1985), Application of the E-ε turbulence model to the

atmospheric boundary layer, Bound. Layer Meteor., 33, 113-133.

Dickinson, R.E., A. Henderson-Sellers, and P.J. Kennedy (1993),

Biosphere-Atmosphere Transfer Scheme (BATS), Version 1e as coupled to the

NCAR Community Climate Model, NCAR Tech. Note NCAR/TN-387+STR, 72 pp.

Edwards, J.M., and A. Slingo (1996), Studies with a flexible new radiation code. I:

Choosing a configuration for a large-scale model, Quart. J. Roy. Meteor. Soc., 122,

689-719.

Fang W., G. Fang, P. Shi, Q. Huang, and Q. Xie (2002), Seasonal structures of upper

layer circulation in the southern South China Sea from in situ observations, J.

Geophys. Res., 107 ,(C11), 3202, doi: 10.1029/2002 JC001343.

Fairall, C.W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young (1996), Bulk

parameterization of air-sea fluxes for Tropical Ocean-Global Atmosphere Coupled

Ocean Atmosphere Research Experiment, J. Geophys. Res., 101, 3747-3764.

Gan, J.P., H. Li, E.N. Curchitser, and D.B. Haidvogel (2006), Modeling South China

sea circulation: Response to seasonal forcing regimes, J. Geophys. Res., 111,

17 C06034, doi:10.1029/2005JC003298.

Giorgi, F., and G. T. Bates (1989), The climatological skill of a regional model over

complex terrain, Mon. Weather Rev., 117, 2325-2347.

Grist, J. P., and S. A. Josey (2003), Inverse analysis adjustment of the SOC air-sea

flux climatology using ocean heat transport constraints, J. Climate, 16,

3274-3295.

Haney, R. L. (1971), Surface thermal boundary conditions for ocean circulation

models, J. Phys. Oceanogr., 1, 241-248.

Huffman, G.J., R.F. Adler, S. Curtis, D.T. Bolvin, and E.J. Nelkin (2005), Global

rainfall analyses at monthly and 3-hr time scales, Measuring Precipitation from

Space: EURAINSAT and the Future, V. Levizzani, P. Bauer, and J. Turk, Ed.,

Springer Verlag (Kluwer Academic Pub. B.V.), Dordrecht, The Netherlands, in

press.

Huang, Q.-Z., W.-Z. Wang, Y. S. Li, and C. W. Li (1994), Current characteristics of

the South China Sea, in Oceanology of China Sea, edited by D. Zhou et al., pp

39-47, Kluwer Acad., Norwell, Mass.

Ho, C., Q. Zheng, Y. Song, N. Kou, and J. Hu (2000), Seasonal variability of sea

surface height in the South China Sea observed with TOPEX/Poseidon Altimeter

data, J. Geophys. Res., 105, 13981-13990.

Ji, M., A. Leetmaa, and J. Derber (1995), An ocean analysis system for seasonal to

interannual climate studies, Mon. Wea. Rev., 123, 460-481.

Kalnay, E., and Coauthors (1996), The NCEP/NCAR 40-Year Reanalysis Project, Bull.

Am. Meteor. Soc., 77, 437-472.

18 Kou, N.-J., Q. Zheng, and C. R. Ho ( 2000), Satellite observation of upwelling along

the western coast of the South China Sea, Remote Sens. Environ., 74, 463-470.

Kummerow, C., and Coauthers (2000), The status of the Tropical Rainfall Measuring

Mission (TRMM) after two years in orbit, J. Appl. Meteor., 39, 1965-1982.

Lau, K. M., H. T. Wu, and S. Yang (1998), Hydrologic processes associated with the

first transition of the Asian summer monsoon: A pilot satellite study, Bull. Am.

Meteor. Soc., 79, 1871-1882.

Liu, W. T., and X. Xie (1999), Space-based observations of the seasonal changes of

South Asian monsoons and oceanic response, Geophy. Res. Lett., 26, 1473-176.

Liu, W.T., X. Xie, P.S. Polito, S.-P. Xie and H. Hashizume (2000), Atmospheric

manifestation of tropical instability waves observed by QuikSCAT and Tropical

Rain Measuring Mission, Geophys. Res. Lett., 27, 2545-2548.

McCreary, J. P., and Z. Yu (1992), Equatorial dynamics in a 2 1/2-layer model, Prog.

Oceanogr., 29,61-132.

Nordeng, T. E. (1995), Extended versions of the convective parameterisation scheme

at ECMWF and their impact upon the mean climate and transient activity of the

model in the Tropics, ECMWF Research Department Tech. Memo., 206, 41 pp.

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. Clim., 15, 1609-1625.

Sakamoto T. T., A. Sumi, S. Emori, T. Nishimura, H. Hasumi, T. Suzuki, M. Kimoto

(2004), Far-reaching effects of the in the CCSR/NIES/FRCGC

high-resolution climate model, Geophys. Res. Lett., 31, L17212,

doi:10.1029/2004GL020907.

19 Sasai, Y., H. Sasaki, K. Sasaoka, A. Ishida, Y. Yamanaka (2007), Marine ecosystem

simulation in the eastern tropical Pacific with a global eddy resolving coupled

physical-biological model, Geophys. Res. Lett., submitted.

Sasaki H., M. Nonaka (2006), Far-reaching Hawaiian Lee Countercurrent driven by

wind-stress curl induced by warm SST band along the current, Geophys. Res. Lett.,

33, L13602, doi:10.1029/2006GL026540.

Shaw, P. T., S. Y. Chao, and L. L. Full (1999), Sea surface height variations in the

South China Sea from satellite altimetry, Oceanol. Acta, 22, 1-17.

Small, R. J., S.-P. Xie and Y. Wang (2003), Numerical simulation of atmospheric

response to Pacific tropical instability waves, J. Climate, 16, 3722-3737.

Sun, Z., and L. Rikus (1999), Improved application of exponential sum fitting

transmissions to inhomogeneous atmosphere, J. Geophys. Res., 102, 6291-6303.

Tiedtke, M. (1989), A comprehensive mass flux scheme for cumulus parameterization

in large-scale models, Mon. Wea. Rev., 117, 1779-1800.

Wang, G., D. Chen, and J.L. Su (2006), Generation and life cycle of the dipole in the

South China Sea summer circulation, J. Geophys. Res., 111, C06002, doi:

10.1029/2005JC003314.

Wang, Y. (2001), An explicit simulation of tropical cyclone with a triply nested

movable mesh primitive equation model: TCM3. Part I: Model description and

control experiment, Mon. Wea. Rev., 129, 1370-1394.

Wang, Y. (2002), An explicit simulation of tropical cyclones with a triply nested

movable mesh primitive equations model–TCM3. Part II: Model refinements and

sensitivity to cloud microphysics parameterization. Mon. Wea. Rev., 130,

20 3022-3036.

Wang, Y., O.L. Sen, and B. Wang (2003), A highly resolved regional climate model

(IPRC–RegCM) and its simulation of the 1998 severe precipitation event over

China. Part I: Model description and verification of simulation, J. Climate, 16,

1721-1738.

Wang, Y., S.-P. Xie, H. Xu, and B. Wang (2004), Regional model simulations of boundary layer clouds over the Southeast Pacific off South America. Part I: Control experiment, Mon. Wea. Rev., 132, 274-296.

Wang, Y., L. Zhou, and K.P. Hamilton (2007), Effect of convective

entrainment/detrainment on simulation of tropical precipitation diurnal cycle. Mon.

Wea. Rev., 135, 367-385.

Xie, Q., W. Q. Wang, and Q. Mao (2002), Comparison among four kinds of data of

sea surface wind stress in the South China Sea, Acta Oceanol. Sin., 21, 263-273.

Xie, S.-P., H. Xu, N. H. Saji, and Y. Wang (2006), Role of narrow mountains in

large-scale organization of Asian monsoon convection, J. Climate, 19, 3420-3429.

Xie, S.-P., H. Xu, W.S. Kessler, and M. Nonaka (2005), Air-sea interaction over the

eastern Pacific warm pool: Gap winds, thermocline dome, and atmospheric

convection, J. Clim., 18, 5-25.

Xie, S.-P., Q. Xie, D. Wang, and W. T. Liu (2003), Summer upwelling in the South

China Sea and its role in regional climate variations, J. Geophy. Res., 108,

doi:10.1029/2003JC001867.

Xie, S.-P., T. Miyama, Y. Wang, H. Xu, S. P. de Szoeke, R. J. Small, K.J. Richards, T.

Mochizuki, and T. Awaji (2007), A regional ocean-atmosphere model for eastern

Pacific climate: Towards reducing tropical biases, J. Climate, 20, 1504-1522.

21 Xie, S.-P., W.T. Liu, Q. Liu and M. Nonaka (2001), Far-reaching effects of the

Hawaiian Islands on the Pacific Ocean-atmosphere system, Science, 292,

2057-2060.

Xu, H., S.-P. Xie, Y. Wang, and R. J. Small (2005), Effects of Central American

mountains on the eastern Pacific winter ITCZ and moisture transport, J. Climate,

18, 3856-3873.

Xu, H., Y. Wang, and S.-P. Xie (2004), Effects of the Andes on eastern Pacific climate:

A regional atmospheric model study, J. Climate, 17, 589-602.

Yasuda, T., and K. Hanawa (1996), Influence of asymmetric wind stress curl on the

general ocean circulation using an eddy-resolving quasi-geostrophic model, J.

Oceanogr., 52,189-206.

Zhou, L., and Y. Wang (2006), TRMM observations and regional model study of

precipitation diurnal cycle in the New Guinean region. J. Geophys. Res., 111,

D17104, doi:10.1029/2006JD007243.

Zhuang, W., D. Wang, J. Hu, and W. Ni (2006), Response of the cold water mass in

the western South China Sea to the wind stress curl associated with the summer

monsoon, Acta Oceanol. Sin., 25, 1-13.

22 Figure Caption

Figure 1. Summer (JJA) climatology: (a) QuikSCAT 10-m wind velocity (m s-1), and (b) TRMM-PR precipitation (mm day-1). Land orography is plotted in (a) in gray shaded (at 0.5, 1.0, 1.5 km) and star marks the location of Ho Chi Minh City.

Figure 2. Model domain with topography (shaded at 0.25, 0.5, 1.0 km) and SST

(contours at 0.25℃ intervals) for (a) CTL and (b) NoTopSmSST runs.

Figure 3. Wind velocity (contours with 1 m s-1 intervals) at 10 m in (a) QuikSCAT observations, and (b) the CTL run, averaged for July 2001. Land orography (shaded at

0.25, 0.5, 1.0 km) is also plotted.

Figure 4. Monthly mean precipitation (mm day-1) for July 2001: (a) TRMM

3B42RT observations, and (b) the CTL run.

Figure 5. (a) 10-m wind velocity (m s-1) and land orography (shaded at 0.25, 0.5, and 1.0 km) in the NoTop run. (b) CTL-NoTop difference in 10-m wind velocity (0.5 m s-1 interval), with shade indicating values passing a two-sided t test at the 99% significant level.

Figure 6. (a) July 2001 precipitation (mm day-1) in the NoTop run. (b)

CTL-NoTop difference in precipitation (mm day-1), with shade indicating values passing a two-sided t test at the 99% significant level.

Figure 7. Vertical cross-sections for July 2001: specific humidity (contours in g kg-1) and cloud liquid water content (shaded in 10-2 g kg-1) along 15oN in (a) the CTL and (b) NoTop runs; (c) vertical velocity (in Pa S-1, with the zero contour omitted ) in the CTL run; (d) CTL - NoTop difference in temperature (contours at 0.5 oC intervals)

23 and specific humidity (shaded in g kg-1) between the CTL and NoTop runs

Topography is shaded black.

Figure 8. Precipitation difference of (contours in mm day-1 with the zero contour omitted) between the NoTopSmSST and NoTop runs, averaged for July 2001.

Shading denotes regions where the difference passes the two-sided t test at the 95% significant level.

Figure 9. July-September mean stream function (contours at 0.5 Sv intervals) in the ocean model for (a) the OcnCTL and OcnNoTop runs, with their difference shown in (c). (d) CTL-NoTop difference in surface wind stress (vectors in N/m2) and its curl

(contours at 1.0×10-7 N/m3 intervals).

24

Annam Cordillera

Figure 1. Summer (JJA) climatology: (a) QuikSCAT 10-m wind velocity (m s-1), and (b) TRMM-PR precipitation (mm day-1). Land orography is plotted in (a) in gray shaded (at 0.5, 1.0, 1.5 km) and star marks the location of Ho Chi Minh City.

25

Figure 2. Model domain with topography (shaded at 0.25, 0.5, 1.0 km) and

SST (contours at 0.25℃ intervals) for (a) CTL and (b) NoTopSmSST runs.

26

Figure 3. Wind velocity (contours with 1 m s-1 intervals) at 10 m in (a) QuikSCAT observations, and (b) the CTL run, averaged for July 2001. Land orography (shaded at 0.25, 0.5, 1.0 km) is also plotted.

27

Figure 4. Monthly mean precipitation (mm day-1) for July 2001: (a) TRMM 3B42RT observations, and (b) the CTL run.

28

Figure 5. (a) 10-m wind velocity (m s-1) and land orography (shaded at 0.25, 0.5, and 1.0 km) in the NoTop run. (b) CTL-NoTop difference in 10-m wind velocity (0.5 m s-1 interval), with shade indicating values passing a two-sided t test at the 99% significant level.

29

Figure 6. (a) July 2001 precipitation (mm day-1) in the NoTop run. (b) CTL-NoTop difference in precipitation (mm day-1), with shade indicating values passing a two-sided t test at the 99% significant level.

30

Figure 7. Vertical cross-sections for July 2001: specific humidity (contours in g kg-1) and cloud liquid water content (shaded in 10-2 g kg-1) along 15oN in (a) the CTL and (b)

NoTop runs; (c) vertical velocity (in Pa S-1, with the zero contour omitted ) in the CTL run; (d) CTL - NoTop difference in temperature (contours at 0.5 oC intervals) and specific humidity (shaded in g kg-1) between the CTL and NoTop runs Topography is shaded black.

31

Figure 8. Precipitation difference of (contours in mm day-1 with the zero contour omitted) between the NoTopSmSST and NoTop runs, averaged for July 2001. Shading denotes regions where the difference passes the two-sided t test at the 95% significant level.

32

a b

c d

Figure 9. July-September mean stream function (contours at 0.5 Sv intervals) in the ocean model for (a) the OcnCTL and (b) OcnNoTop runs, with their difference shown in (c). (d) CTL-NoTop difference in surface wind stress

2 -7 3 (vectors in N/m ) and its curl (contours at 1.0×10 N/m ).

33