The Formation of Wind Curl in the Marine Atmosphere Boundary Layer over the East China Sea Kuroshio in Spring

ZHANG Suping, and KONG Yang

Physical Oceanography Laboratory, and Ocean-Atmosphere Interaction and Climate Laboratory, Ocean University of China, Qingdao 266100, P. R. China

The Formation of Wind Curl in the Marine Atmosphere Boundary Layer over the East China Sea Kuroshio in Spring

ZHANG Suping*, and KONG Yang

Physical Oceanography Laboratory, and Ocean-Atmosphere Interaction and Climate Laboratory, Ocean University of China, Qingdao 266100, P. R. China

Various data are used to investigate the characteristics of the surface wind field and rainfall on the East China Sea Kuroshio (ESK) in March and April, 2011. In March, the wind speed maximum shows over the ESK front (ESKF) in the 10 meter wind field, which agrees with the thermal wind effect. A wind curl center is generated on the warm flank of the ESKF. The winds are much weaker in April, so is the wind curl. A rainband exists over the ESKF in both the months. The Weather Research and Forecasting (WRF) model is used for further researches. The winds on the top of the marine atmosphere boundary layer (MABL) indicate that in March, a positive wind curl is generated in the whole MABL over the warm flank of the ESKF. The thermal wind effect forced by the strong SST gradient overlying the background wind leads to strong surface northeasterly winds on the ESKF, and a positive shearing vorticity is created over the warm flank of the ESKF to generate wind curl. In the smoothed sea surface temperature experiment, the presence of the ESKF is responsible for the strong northeast winds in the ESKF, and essential for the distribution of the rainfall centers in March, which confirms the mechanism above. The same simulation is made for April, 2011, and the responses from the MABL become weak. The low background wind speed weakens the effect of the thermal wind, thus no strong Ekman pumping is helpful for precipitation. There is no big difference in rainfall between the control run and the smooth SST run. Decomposition of the wind vector shows that local wind acceleration induced by the thermal wind effect along with the variations in wind direction is responsible for the pronounced wind curl/divergence over the ESKF.

the East China Sea Kuroshio front; Ekman pumping; thermal wind effect; precipitation

1 Introduction

In recent years, due to the use of high-resolution satellite observations, we have learned much more about how the ocean and atmosphere interact. In regions of western boundary currents such as the Gulf Stream and the Kuroshio, the big sea surface temperature (SST) gradient can induce curl and divergence in the Marine Atmosphere Boundary Layer (MABL), thus making further influences on the free atmosphere.

Some literatures explained the in situ effect of SST front. The downward momentum mixing mechanism (Sweet et al., 1981; Wallace et al., 1989) over warmer SSTs accounts for the intensified downward momentum transport acting from aloft to accelerate the surface wind, and constitutes a main mechanism over the Yellow Sea in winter (Xie et al., 2002). Higher wind speed is in agreement with the warm tongue due to this mechanism. The pressure adjustment mechanism (Small et al., 2003; Cronin et al., 2003; Song et al., 2006; Minobe et al., 2008, 2010; Tokinaga et al., 2009; Koseki and Watanabe, 2010) can also explain wind field changes. This mechanism suggests that SST modifies the boundary layer air temperature, and the resulting relatively low (high) pressure anomalies produce pressure a gradient in the SST front and thus lead to higher wind speed over SST front. Zhang et al. (2010) suggested that the intense SST gradient can affect the surface wind field by thermal wind effect, another form of the pressure adjustment mechanism, and contributes to a strong surface wind over the East China Sea Kuroshio (ESK) front in spring. Takatama et al. (2012) used a regional atmosphere model to find out that the pressure adjustment mechanism plays a primary role over the Gulf Stream.

Different wind speed generates wind curl and divergence in SST fronts. When winds blow over the SST front, wind curl and divergence in the MABL exhibit coherent structures, which are mainly determined by the spatial variations of the SST field under different wind directions (Xie, 2004; Chelton et al., 2004; Small et al., 2008; Chelton and Xie, 2010). Divergence of surface wind and stress are both found to be linear functions of the downwind component of the SST gradient (front); the curl of surface wind and stress are likewise both found to be linear func-tions of the crosswind component of the SST gradient (Chelton et al., 2001, 2004; O’Neill et al., 2003, 2005; Larry O’Neill, Naval Research Laboratory, and colleagues, personal communication, 2010). These influences from the ocean on the overlying atmosphere can be found in monthly or even longer timescales.

The Kuroshio is a well known warm oceanic western boundary current. When the middle part of this current, known as the ESK (Feng et al., 1999), flows along the continental slope, a sharp SST front – the ESK front (ESKF) – forms between the warm Kuroshio and the cold shelf water. Unlike the Gulf Stream, the ESK is located in the region of East Asian monsoon, where the background wind field has significant seasonal variations. The intensity of the ESKF is most robust from January to April in its annual circle (Zhang et al., 2010), but the background wind field is quite different between March and April because of the typical winter monsoon in March and the pronounced weakened winds in April. Such differences in wind field may lead to different impacts of the ESKF on the atmosphere. In this paper, we will discuss the effect of the ESKF on the local wind field and the precipitation in March and April.

2 Data and Model

2.1 TRMM, NGSST and Blended Wind

In recent years, TRMM data are used widely in regional precipitation analysis, the study of seasonal-scale rainfall diurnal variation and validation of the simulation of water condensate content (Yang and Smith, 2008). We used the Tropical Rainfall Measuring Mission satellite (TRMM) rainfall product 3B43 data, version 6 for the analysis of spatial characteristics of the rainband over the ESK and for the comparison with model simulations. The 3B43 rain rates are monthly averages gridded over 0.25× 0.25 degree lat/lon boxes (http://rain.atmos.colostate.edu/ CRDC/datasets/ TRMM_3B43.html).

We also used the New Generation SST (NGSST) product (Guan and Kawamura, 2004). NGSST combines infrared radiometer (the moderate-resolution imaging spectroradiometer MODIS and the advanced very high resolution radiometer AVHRR) and microwave radiometer (the advanced microwave scanning radiometer-earth observing system AMSR-E) measurements, which are quality-controlled with high spatial resolution on a 0.05 degree grid, and available since 2003.

Sea surface wind (10 m) is from the QSCAT and NCEP blended wind data provided by NOAA. The wind speeds were blended from multiple satellites (up to six, including Scatterameters (QuikSCAT), SSMIs, TMI and AMSR-E) observations, on a global 0.25-degree grid. The blended wind speeds were then decomposed into (u, v) components using the NCEP Re-analysis 2 (NRA-2) wind directions interpolated onto the blended speed grid (http://www. ncdc.noaa.gov/oa/rsad/blendedseawinds.html).

2.2 WRF

The Weather Research and Forecasting (WRF) model version 3.3.1 (WRF v3.3.1) is used to simulate the ESKF and helps make further discussions on the MABL because of the very high vertical resolution of this model. WRF is a new generation mesoscale model and widely used for regional climate research (http://www.wrf-model.org/ index.php). Table 1 lists the main physics options of the model.

Table 1 The main physical options of the model WRF v3.3.1

The National Centers for Environmental Prediction (NCEP) Final (FNL) Global Tropospheric Analyses is used for WRF as initial and boundary conditions. FNL is available at a time interval of 6 hours and a spatial resolution of 1?×1?, with 26 levels from 1000 to 10 hPa. Daily OI (Optimum Interpolation) SST with spatial resolution of 0.25? is used for sea-air boundary condition analysis. The model has a horizontal grid spacing of 21 km with 115×105 grid points and centers at 126?E, 30?N with Lambert projection. It has 40 uneven vertically spaced sigma levels with enhanced resolution in the boundary layer (19 levels below 0.8 sigma or about 2 km). To examine the effect of the ESKF on the atmosphere above, the simulation period should be in spring when the ESKF is most robust. As mentioned above, the background wind field may also have in situ effect. Therefore, March and April in 2011 are selected randomly because the wind field in the free atmosphere has great differences between the two months. We performed two experiments for each month, one with observed SST from OISST (referred to as CTL run), and the other with smoothed SST field (referred to as SmSST run) where the ESKF is removed (Fig.5(b)). The simulation periods are from March 1 to 31 and from April 1 to 30, 2011, respectively. In this paper we discuss the March and April means.

2.3 Ekman Pumping Calculation

Wind curl in the MABL can produce vertical velocity through the turbulent friction effect on the top of the Ekman layer (MABL), which leads to the air mass exchangebetween the boundary layer and the free atmosphere, including water vapor, aerosol and other trace substances. This effect is called the Ekman pumping. We consider the spring MABL to be neutral, thus the Ekman pumping velocity can be estimated as

where h is the MABL height, ζg the geostrophic wind curl and π the ratio of the circumference (Sheng et al., 2003). H can be calculated by WRF; the wind at the top of the MABL is used instead of the geostrophic wind.

3 Results

3.1 Observations

The blended wind shows that in March there is a wind speed peak on the ESKF, rather than the warm tongue (Fig.1(a)). This phenomenon can be explained by the thermal wind effect, in which the directions of wind anomalies forced by the SST gradient over the ESKF are consistent with the winter monsoon, thus leading to strong northerly/northeasterly local winds over the ESKF (Zhang et al., 2010).

In April (Fig.1(b)), the ESKF is as strong as it is in March. But due to the reversal of the monsoon in the free atmosphere, the 10 m wind speed is much lower than that in March, and the wide wind band on the ESKF can hardly be recognized. The wind speed on the SST front is still higher than it is on the same latitude, indicating that the thermal wind effect still exists in the surface wind field.

The surface wind curl fields are quite different between March and April (shown in Figs.2(a) and (b)). There is a positive wind curl center over the warm flank of the ESKF, near the warm tongue in March. In April, positive curl can hardly be seen in the surface wind field. As shown above, strong wind curl in the MABL generates Ekman pumping, and possibly enhances the local precipitation. We will focus on the formation of the wind curl center during the following discussion.

Previous studies showed that there is a rainband over the warm flank of the ESKF, with enhanced precipitation over the warm tongue from March to May (Xu et al., 2010). In Fig.2(c), the rainband is obvious in March 2011, and there is a large rainfall center over the warm flank of the ESKF rather than on the warm tongue. This center is located north of the 22℃ warm tongue, consistent with the wind curl center in Fig.2(a). It is likely forced by the strong wind curl at the same position. The precipitation area in April is wider than that in March, and the rainfall centers move to the warm tongue.

3.2 WRF CTL Run

3.2.1 Surface wind field

In CTL run, WRF reproduces the characteristic of the surface wind filed in both March and April. In March, the wind band (Fig.3(a)) lies over the ESKF rather than over the warm tongue, though the band is a little bit narrower than it is in Fig.1(a). A positive curl center exists over the warm flank of the ESKF. In April, the ESKF is still the same as it is in March. But the 10 m wind speed is much lower because of the reversal of the monsoon in the free atmosphere. Even if the large wind band in March is hard to recognize here, the wind speed is still higher on the SST front than it is on the same latitude (Fig.3(b)). Accompanied with the low wind speed, the surface wind curl (Fig.3(d)) becomes weak in April.

3.2.2 Wind field on the top of the MABL

As a boundary between the MABL and free atmosphere, the top of the MABL is an important layer for our study. We extracted the physical quantity of each grid point on the top of the MABL (TBL) through vertical interpolation as a whole layer.

As shown in Fig.4(a), the curl center still exists on the TBL in March, which indicates that the positive wind curl fills in the whole MABL over the warm flank of the ESKF. Then Ekman pumping can be estimated by wind curl and the height of MABL. In April (Fig.4(b), the wind speed on the TBL is much lower than that in March, too. There is no strong wind curl in Fig.4(b), thus no significant Ekman pumping exists in the whole MABL.

Fig.1 10 meter wind vector and wind speed (shaded, m s-1) from blended wind in March (a) and April (b), 2011. Monthly mean NGSST (℃) is plotted in contours.

Fig.2 10 meter wind curl ((a) and (b), shaded, 10-6) from blended wind and precipitation ((c) and (d), shaded, mm) observed by TRMM. (a) and (c) are for March; (b) and (d) are for April, 2011. Monthly mean NGSST (℃) is plotted in contours.

Fig.3 10 meter wind speed (a and b, shaded, m s-1) and wind curl (c and d, shaded, 10-6) from WRF CTL run. (a) and (c) are for March and (b) and (d) for April, 2011. Monthly mean OISST (℃) is plotted in contour.

Fig.4 Wind curl (shaded, 10-6) on the TBL from WRF CTL run in (a) March and (b) April, 2011.

Zhang et al. (2010) suggested that the thermal wind forces strong surface northerly/northeasterly winds on the ESKF in spring. Different wind speed generates shearing vorticity, so the strong curl centers in Fig.3(c) and Fig.4(a) are likely to be related to the thermal wind effect forced by the ESKF. We designed an SST smoothed experiment, the SmSST run, to verify this hypothesis.

3.3 SmSST Run

3.3.1 Surface wind

The SmSST run is very helpful to investigate the impact on the MABL from the ESKF; we smoothed the original SST field with 9-points smooth method for 200 times to make sure the ESKF is removed well. As shown in Fig.5(a), the wind band in March becomes narrow when the SST gradient decreases. The curl center (Fig.5(c)) on the ESKF is significantly reduced at the same time. In April, no wind acceleration is found in the former position of the ESKF, and the thermal wind effect can hardly be recognized in the surface wind field. The curl center in CTL run (Fig.3(d)) disappears in the SmSST run. The facts associaterd with Fig.5 reflect the direct relationship between the in situ ESKF effect and the curl center.

Fig.5 10 meter wind speed (a and b, shaded, m s-1) and wind curl (c and d, shaded, 10-6) from WRF SmSST run. (a) and (c) are for March and (b) and (d) for April, 2011. Monthly mean OISST (℃) is plotted in contours.

3.3.2 Wind field on the top of the MABL

As expected, the wind curl center on the TBL gets weaker as soon as the ESKF is removed in March (Fig.6(a)), proving that it is formed by the ESKF. In the SmSST run (Fig.6(b)) there is little change in the wind curl field on the TBL in April. The very weak curl center in the CTL run almost disappeares.

We subtracted the wind field in SmSST run from CTL run, and discuss the thermal wind effect on the TBL (Fig.7) in March. The wind speed difference on the ESKF in Fig.7 is around 1 m s-1over the ESKF, much bigger than it is over the warm/cold flank of the ESKF, and the wind direction is in line with the SST front, which obviously shows the acceleration effect of the ESKF on the local wind. Fig.7 indicates that the thermal wind effect in March is not only trapped in the surface, but also reaches the top of the MABL. The same method was also used for April, but the wind speed is lower (not shown).

As shown above, the effect of thermal wind overlying the northeasterly winds in March leads to a big wind speed difference over the warm flank of the ESKF. Thus a shearing vorticity is possible to generate strong curl in the whole MABL.

Fig.6 Wind curl (shaded, 10-6) on the TBL from WRF SmSST run in (a) March and (b) April, 2011.

Fig.7 Wind (vector, m s-1) in CTL run minus that in SmSST run and SST (contour, ℃) on the TBL in March 2011.

3.4 Wind Components

In order to further investigate the relation between wind direction and wind curl and divergence, we decomposed the wind vector on the TBL into downwind component (northeasterly wind) and crosswind component (northwesterly wind) based on the main trend of the ESKF axial. According to the thermal wind effect, the surface wind speed is higher over the ESKF than over the surroundings. As suggested by previous studies (Chelton et al., 2001, 2004; O’Neill et al., 2003, 2005; Larry O’Neill, Naval Research Laboratory, and colleagues, personal communication, 2010), shearing vorticity arises when the wind blows parallel to the SST front, leading to wind curl; acceleration (deceleration) happens when the wind blows across the SST front, leading to wind divergence (convergence). The results for March (Figs.8(a), 8(c)) show that, positive (negative) wind curl happens over the warm (cold) flank of the ESKF, as above. The wind ite convergence/divergence is also quite well consistent with what we inferred above. Positive wind curl in Fig.8(a) is calculated without curvature vorticity, so we can consider it as a result of the thermal wind shearing.

The SST gradient in the ESKF is still intense in April, so the TBL wind field can show the same characteristics in Figs.8(b) and 8(d), but much weaker than they are in March. Compared with the two months, the background wind in March has a stronger downwind component, and leads to strong northeasterly winds with thermal wind effect on the ESKF. The resulting positive wind curl center over the warm flank of the ESKF is strong enough to fill the whole MABL. In April, on the contrary, significant wind shearing can not appear because of the very weak background wind field, so the wind curl can’t reach the top of the MABL.

Fig.8 Wind curl/convergence (shaded, 10-6) and wind components (vector, m s-1) from CTL run. (a) and (c) for curl and downwind component; (b) and (d) for convergence and crosswind component. (a) and (b) are for March and (c) and (d) for April, 2011. Monthly mean OISST (℃) is plotted in contours.

3.5 Wind Curl and Precipitation

In March, the positive wind curl generates vertical movement – the Ekman pumping – in the whole MABL. We shall make a brief study on the impact of the Ekman pumping on the local rainfall in this section.

In the CTL run for March, the location of the rainband (Fig.9(a)) is consistent with the observed data (Fig.1(b)). The rainfall center is north of the 22℃ warm tongue,coinciding with the location of the wind curl center (Fig.3(c) or Fig.4(a)). After the ESKF has been smoothed, the precipitation changes a lot. The rainband is still there, but the precipitation centers change. There is a distinct precipitation center near the warm tongue in the CTL run; it disappears along with the weakening of the ESKF. It is reasonable to consider these changes in the precipitation as the results of the in situ ESKF effect.

Considering that the ESKF is basically unchanged from March to April, and the wind field changes a lot, the simulation results of the precipitation in April can be used as a control. Compared with Figs.9(b) and 9(d), the rainband and the location of the large rainfall centers do not change significantly in April after we have removed the ESKF. The same thing happens in the wind curl field in Fig.6(b). The effect of the SST front on the precipitation seems decreased in a lower speed wind field.

The Ekman pumping caused by the wind curl (Fig.4(a)) in March is calculated based on Eq. (1). Its magnitude is 10-3m s-1, the same as the average vertical velocity in the MABL. The positive extremum area of the Ekman pumping is consistent with the biggest precipitation center strength (Fig.10) and indicates the important role of Ekman pumping in the development of the precipitation center in March.

The vertical section shows the existence of the Ekman pumping directly. As shown in Fig.11(a), ascending motion is formed south from the ESKF below 900 hPa, and two centers appear. One is on 28.5?N, the warm flank of the ESKF. It corresponds with the large Ekman pumping area in Fig.10 and is accompanied by large amount of rain water. The other center of the ascending motion next to the Ekman pumping is on 27?N, just upon the warm tongue. In the SmSST run, the upward motion and the rain band disappears at the same time as in the MABL. The rainfall center in March can be generated as follows: the water vapor is transported from the bottom to the upper atmosphere through Ekman pumping, then condensation occurs while the vapor is rising, heats the atmosphere and increases the MABL precipitation.

In addition, in the CTL run there is another ascending motion center in the free atmosphere on 770 hPa. The two centers connect each other. Vapor may continually riseinto the free atmosphere and get to the upper troposphere after they reaches the top of the MABL, where the Ekman pumping plays a role of agency. Such an interaction between free atmosphere and boundary layer is essential in the progress of precipitation, we will not discuss it in this paper, but it needs to be further studied.

Fig.9 Precipitation (shaded, mm) from WRF CTL run (a and b) and SmSST run (c and d). (a) and (c) are for March and (b) and (d) for April, 2011. Monthly mean OISST (℃) is plotted in contours.

Fig.10 Ekman pumping speed (shaded, 10-3m s-1) on the TBL and precipitation (contours, mm) in CTL run, March 2011.

4 Conclusion and Discussions

We use various data and a regional atmospheric model to investigate the in situ effect of the ESKF in the MABL in March and April, 2011. The Blended wind data shows a significant thermal wind effect feature of the surface wind field, in which the wind speed is the maximum over the ESKF rather than over the warm tongue. According to the TRMM data, there is a rainband in both March and April, and a precipitation center over the warm flank of the ESKF in March. These features can be captured by WRF model.

The numerical experiments, the CTL runs and the SmSST runs have revealed the effect of the ESKF in detail and help make further discussions. Thermal wind effect plays a dominant role in the MABL in March. By contrasting CTL run with SmSST run, we find that with a strong background wind filed in March, the thermal wind effect can reach the top of the MABL. Shearing vorticity induces a strong wind curl over the warm flank of the ESKF, thus leads to a maximum precipitation center in the same location. The background winds in April, on the other hand, are not as strong as it is in March because of the monsoon shifting. The wind curl still exists at the surface, but is not obvious in the whole MABL. The shallow wind curl cannot generate strong Ekman pumping, thus the rainband in CTL run (Fig.9(b)) and SmSSTrun (Fig.9(d)) are nearly the same.

When we decompose the wind vector on the TBL into downwind component and crosswind component, the wind curl/convergence distribution based on the spatial variability of the wind direction emerges. The northeasterly winds lead to positive (negative) wind curl over the warm (cold) flank of the ESKF (Figs.8(a) and 8(c)), and the northwesterly winds correspond to wind convergence (divergence) over the ESKF warm (cold) flank (Figs.8(b) and 8(d)). The above effect for respectively March and April is different, and background wind direction is essential to the surface wind curl and divergence distribution.

Fig.11 The vertical-longitude section of vertical velocity (contours, m s-1) and rain water mixing ratio (shaded, kg kg-1) along 127.5?E in CTL run (a) and SmSST run (b), March 2011.

Acknowledgements

This work is supported by NSFC 41175006 and ‘973’2012CB955602.

Chelton, D. B., and Xie, S.-P., 2010. Coupled ocean-atmosphere interaction at oceanic mesoscales. Oceanography, 23: 52-69.

Chelton, D. B., Esbensen, S. K., Schlax, M. G., Thum, N., Freilich, M. H., Wentz, F. J., Gentemann, C. L., McPhaden, M. J., and Schopf, P. S., 2001. Observations of coupling between surface wind stress and sea surface temperature in the eastern tropical Pacific. Journal of Climate, 14: 1479-1498.

Chelton, D. B., Schlax, M. G., Freilich, M. H., and Milliff, R. F., 2004. Satellite radar measurements reveal short-scale features in the wind stress field over the world ocean. Science, 303: 978-983.

Cronin, M. F., Xie, S.-P., and Hashizume, H., 2003. Barometric pressure variations associated with Eastern Pacific tropical instability waves. Journal of Climate, 16: 3050-3057.

Dudhia, J., 1989. Numerical study of convection observed during the winter monsoon experiment using a mesoscale twodimensional model. Journal of the Atmospheric Sciences, 46: 2077-3107.

Dudhia, J., Hong, S. Y., and Lim, K. S., 2008. A new method for representing mixed-phase particle fall speeds in bulk microphysics parameterizations. Journal of the Atmospheric Sciences Japan, 86 (A): 33-44.

Feng, S. Z., Li, F. Q., and Li, S. J., 1999. Introduction to Marine Science. Higher Education Press, Beijing, 503pp (in Chinese).

Guan, L., and Kawamura, H., 2004. Merging satellite infrared and microwave SSTs: Methodology and evaluation of the new SST. Journal of Oceanography, 60: 905-912.

Hong, S. Y., Noh, Y., and Dudhia, J., 2006. A new vertical diffusion package with an explicit treatment of entrainment processes. Monthly Weather Review, 134: 2318-2341.

Kain, J. S., and Fritsch, J. M., 1990. A one-dimensional entraining/detraining plume model and its application in convective paramterization. Journal of the Atmospheric Sciences, 24 (1): 65-81.

Kain, J. S., and Fritsch, J. M., 1993. Convective parameterization for mesoscale models: The Kain-Fritcsh Scheme. In: The Representation of Cumulus Convection in Numerical Models. Emanuel, K. A., and Raymond, D. J., eds., the American Meteor Society, 246pp.

Koseki, S., and Watanabe, M., 2010. Atmospheric boundary layer response to mesoscale SST anomalies in the Kuroshio extension. Journal of Climate, 23: 2492-2507.

Minobe, S., Kuwano-Yohida, A., Komori, N., Xie, S.-P., and Small, R. J., 2008. Influence of the Gulf Stream on the troposphere. Nature, 452: 206-209, DOI: 10.1038/nature06690.

Minobe, S., Miyashita, M., Kuwano-Yoshida, A., Tokinaga, H., and Xie, S.-P., 2010. Atmospheric response to the Gulf Stream: Seasonal variations. Journal of Climate, 23: 3699-3719.

Mlawer, E. J., Taubman, S. J., Brown, P. D., Iacono, M. J., and Clough, S. A., 1997. Radiative transfer for inhomogeneous atmosphere: RRTM, a validate d correlated-k model for the longwave. Journal of Geophysical Research, 102 (D14): 16663-16682.

Monin, A. S., and Obukhov, A. M., 1954. Basic laws of turbulent mixing in the surface layer of the atmosphere. Tr. Akad. Nauk SSSR Geophiz. Inst., 24 (151): 163-187 (in Russian).

O’Neill, W. L., Chelton, D. B., Esbensen, S. K., and Wentz, F. J., 2005. High-resolution satellite measurements of the atmos-pheric boundary layer response to SST variations along the Agulhas Return Current. Journal of Climate,18: 2706-2723.

O’Neill, W. L., Chelton, D. B., and Esbensen, S. K., 2003. Observations of SST-induced perturbations of the wind stress field over the Southern Ocean on seasonal timescales. Journal of Climate,16: 2340-2354.

Sheng, P. X., Mao, J. T., Li, J. G., Zhang, A. C., Sang, J. G., and Pan, N. X., 2003. Atmospheric Physics. Peking University Press, Beijing, 522pp (in Chinese).

Small, R. J., deSzoekea, S. P., Xie, S.-P., O’Neill, L., Seo, H., Song, Q. T., Cornillon, P., Spall, M., and Minobe, S., 2008. Air-sea interaction over ocean fronts and eddies. Dynamics of Atmospheres and Oceans,45: 274-319.

Small, R. J., Xie, S.-P., and Wang, Y., 2003. Numerical simulation of atmospheric response to Pacific tropical instability waves. Journal of Climate,16: 3722-3737.

Song, Q., Cornillon, P., and Hara, T., 2006. Surface wind response to oceanic fronts. Journal of Geophysics Research,111, C12006, DOI: 10.1029/2006JC003680.

Sweet, W. R., Fett, R., Kerling, J., and La Violette, P., 1981. Air-sea interaction effects in the lower troposphere across the north wall of the Gulf Stream. Monthly Weather Review,109: 1042-1052.

Takatama, K., Minobe, S., Inatsu, M., and Small, R. J., 2012. Diagnostics for near-surface wind convergence/divergence response to the Gulf Stream in a regional atmospheric model. Atmospheric Science Letters,13: 16-21.

Tokinaga, H., Tanimoto, Y., Xie, S.-P., Sampe, T., Tomita, H., and Ichikawa, H., 2009. Ocean frontal effects on the vertical development of clouds over the western North Pacific: In situ and satellite observations. Journal of Climate,22: 4241-4260.

Wallace, J. M., Mitchell, T. P., and Deser, C., 1989. The influence of sea surface temperature on surface wind in the eastern equatorial Pacific: Seasonal and interannual variability. Journal of Climate,2: 1492-1499.

Xie, S.-P., 2004. Satellite observations of cool ocean-atmosphere interaction. Bulletin of the American Meteorological Society,85: 195-208.

Xie, S.-P., Hafner, J., Tanimoto, Y., Liu, W. T., Tokinaga, H., and Xu, H., 2002. Bathymetric effect on the winter sea surface temperature and climate of the Yellow and East China Seas. Geophysical Research Letters,29, 2228, DOI: 10.1029/2002 GL015884.

Xu, M. M., Xu, H. M., and Zhu, S. X., 2010. Ocean-to- atmosphere forcing in the vicinity of the sea surface temperature front in the East China Sea in spring. Chinese Journal of Atmosphere science,34: 1071-1087 (in Chinese).

Yang, S., and Smith, E. A, 2008. Convective-stratiform precipitation variability at seasonal scale from 8 yr of TRMM observations: Implications for multiple modes of diurnal variability. Journal of Climate,21: 4087-4114．

Zhang, S. P., Liu, J. W., and Meng, X. G., 2010. The effect of the East China Sea Kuroshio front on the marine atmospheric boundary layer. Journal of Ocean University of China,9: 210-218.

(Edited by Xie Jun)

(Received March 18, 2013; revised May 7, 2013; accepted July 1, 2014)

? Ocean University of China, Science Press and Springer-Verlag Berlin Heidelberg 2014

* Corresponding author. Tel: 0086-532-66781528

E-mail: zsping@ouc.edu.cn