Abstract

The period from April to June is the time of transition from spring to summer over the north Indian Ocean and the South China Sea. Analysis shows that precipitation anomaly changes from April to June may indicate summer (June–August) mean precipitation anomalies over the South China Sea and the Arabian Sea. This study documents and compares the evolution of precipitation, surface wind, and sea surface temperature (SST) anomalies during the spring to summer transition corresponding to April–June precipitation anomaly changes and April–June mean precipitation anomalies over the South China Sea and the Arabian Sea. Over the South China Sea, a clear signal of local air–sea interaction is identified corresponding to the precipitation anomaly change, as indicated by a sequence of less precipitation, higher SST, more precipitation, and lower SST. In contrast, the mean precipitation anomaly features a response to remote SST forcing and a local forcing of atmosphere on the ocean. The evolution of surface heat flux anomalies supports the air–sea interaction over the South China Sea during the transition season. Over the Arabian Sea, local SST forcing contributes to both precipitation anomaly changes and mean precipitation anomalies through modulating atmospheric stability. A local negative feedback of atmosphere on SST is observed in the Arabian Sea as in the South China Sea. The surface heat fluxes make a large contribution to local SST change before May in the South China Sea but a small one in the Arabian Sea. Surface heat fluxes are important for local SST change after May in both the South China Sea and the Arabian Sea.

1. Introduction

Over the north Indian Ocean and the South China Sea, climatological summer monsoon onset occurs around May (e.g., Wu and Wang 2000). As such, the period from April to June signifies the transition from spring to summer. Climate variability during this transition season may have features different from that in boreal winter and summer, which have been the focus of most previous studies. In particular, during the spring to summer transition, the simultaneous impact of remote El Niño–Southern Oscillation (ENSO) forcing is not as strong as in winter since ENSO events are often in the developing and decaying stages during the transition season. Yet, there have been few studies of climate variability during the transition season. This motivates the present study, which will focus on factors and processes for climate variability over the north Indian Ocean and the South China Sea during the spring to summer transition season.

The climate variability over the north Indian Ocean through the western North Pacific is not only influenced by remote forcing, such as ENSO (e.g., Lau and Nath 2000; Wang et al. 2001; Chou et al. 2003; He and Wu 2014; Hu et al. 2014), but also involves regional air–sea interaction (e.g., Wang et al. 2003; Wu and Kirtman 2005, 2007; He and Wu 2013). The evolution of sea surface temperature (SST) anomalies in this region is associated with cloud–radiation and wind–evaporation processes (Wang et al. 2003; Wu and Kirtman 2007; Wu et al. 2008; He and Wu 2013). During the onset of the South China Sea summer monsoon, a strong coupled variation is observed between precipitation, wind, and SST changes over the tropical western Pacific and the South China Sea region (Wu 2009; Roxy and Tanimoto 2012). Wu et al. (2008) demonstrated the importance of air–sea coupled processes in the evolution of SST anomalies in the tropical Indian Ocean from spring to summer. Because of their influence on the evolution of SST anomalies, the air–sea interaction processes in the presummer season may contribute to summer rainfall anomalies. As such, climate anomalies and the associated processes during the spring to summer transition period may have consequences for summer climate. Thus, unraveling the factors and processes for the climate variability during the spring to summer transition can help improve the prediction of rainfall not only in the transition season, but also in summer.

Operational forecast of summer rainfall is made several months before summer, say in March. At that time, the states of ocean during the spring to summer transition are unknown. Yet, the regional ocean states during the transition season may interfere with the impacts of remote ocean states on summer rainfall. This may be a factor for the skill of summer rainfall prediction. Thus, it is important to understand how the oceanic states evolve and what the main factors are for their temporal evolutions during the transition season. This may help to understand why the skill of summer rainfall prediction varies from year to year. This is another motivation for the present study.

The organization of the rest of the text is as follows. Section 2 describes the datasets and methods used in the present study. Evidence will be presented in section 3 for the implication of anomaly changes during the spring to summer transition season for summer rainfall anomalies. Evolution of precipitation, SST, and low-level wind anomalies will be analyzed with respect to anomaly changes and seasonal mean anomalies in the South China Sea and the Arabian Sea in section 4. The air–sea interaction processes will be investigated in section 5. A summary and discussion are given in section 6.

2. Datasets and methods

The precipitation dataset used in the present study is the Global Precipitation Climatology Project (GPCP) version 2.2 monthly precipitation dataset (Adler et al. 2003; Huffman et al. 2009). The GPCP precipitation has a resolution of 2.5° × 2.5° and is available from January 1979 to the present. The GPCP dataset is provided by the NOAA/OAR/ESRL Physical Science Department (PSD) (http://www.esrl.noaa.gov/psd/).

The present study uses monthly mean SST from the NOAA optimum interpolation (OI) dataset version 2 (Reynolds et al. 2002). This dataset is available on a 1° × 1° grid for the period of December 1981 to the present. The OI SST is obtained from http://www.esrl.noaa.gov/psd/.

Monthly mean horizontal winds at 10 m, wind, air temperature, and humidity at pressure levels used in the present study are from the National Centers for Environmental Prediction (NCEP)–Department of Energy (DOE) reanalysis 2 (Kanamitsu et al. 2002), which are also provided by NOAA/OAR/ESRL PSD (http://www.esrl.noaa.gov/psd/). The reanalysis dataset is available from 1979 to the present. The pressure level wind, air temperature, and humidity are on a 2.5° × 2.5° grid and the 10-m winds are on a T62 Gaussian grid. Following previous studies (Roxy and Tanimoto 2007; Wu 2010), we define an instability index using the difference of equivalent potential temperature between 1000 and 700 hPa. A larger value of the index indicates a more unstable lower atmosphere.

Monthly mean surface heat fluxes and associated meteorological variables used in this study are from the National Oceanography Centre in Southampton flux dataset version 2.0 (NOCS V2.0; Berry and Kent 2009; http://badc.nerc.ac.uk). This is a gridded dataset of marine surface meteorology and fluxes constructed using optimal interpolation based on ship data only from the International Comprehensive Ocean–Atmosphere Dataset (ICOADS). The NOCS dataset has a 1° × 1° spatial resolution for the period 1973–2009. In this study, the convention for shortwave radiation (latent heat flux) is positive for downward (upward) flux.

The region of this study is the tropical north Indian Ocean and tropical western North Pacific. The relevant period is April to June (AMJ), which is the transition time from spring to summer over most of the above regions as the climatological summer monsoon onset happens in May (e.g., Wu and Wang 2000). In the present study, we are concerned with both the change of anomaly from April to June and the mean anomaly during AMJ. The change of anomaly from April to June signifies an accelerated or a decelerated transition from spring to summer and the mean anomaly during AMJ represents a persistent state anomaly of the transition season. In the following analysis, we compare the features corresponding to AMJ mean anomaly and those corresponding to anomaly change from April to June (the anomaly in June minus the anomaly in April, denoted as the JmA anomaly). Such comparison provides useful information for the contribution of different processes to the changing and persisting anomalies during the transition season.

3. Implications of anomaly changes during the transition season for summer anomalies

To motivate the readers, we provide evidence in the following that the anomaly change from April to June may indicate summer mean anomalies over the South China Sea and the Arabian Sea. Before that, we first show the climatological mean change of rainfall and low-level winds to provide a background for the readers. Figure 1 displays change of climatological mean precipitation and 850-hPa zonal winds from April to June (the difference of June minus April).

Fig. 1.

Difference of (a) climatological mean precipitation (mm day−1) and (b) 850-hPa zonal wind (m s−1) between April and June for the period 1979–2010.

Fig. 1.

Difference of (a) climatological mean precipitation (mm day−1) and (b) 850-hPa zonal wind (m s−1) between April and June for the period 1979–2010.

A large and pronounced precipitation increase from April to June is observed over the Arabian Sea, the Bay of Bengal, the South China Sea, and southwest Philippine Sea (Fig. 1a). The largest increase is along the west coast of the Indian Peninsula, the Indochina Peninsula, and the Philippine Islands, indicating the effect of topography. Another region of noticeable precipitation increase is from the lower reaches of the Yangtze River to southwestern Japan, signifying the start of the mei-yu–baiu rainy season. In contrast, a precipitation decrease is seen over Indonesia and northern Australia.

A positive change in zonal winds from April to June is observed over a zonal band extending from the Arabian Sea to the Philippine Sea along 5°–15°N (Fig. 1b). This signifies an enhancement in westerlies. The magnitude of zonal wind change decreases eastward, which is dynamically consistent with the precipitation increase. The zonal wind increase is accompanied by an enhanced cyclonic vorticity and convergence to the north side, leading to the increase in precipitation. A positive change is also seen over the subtropics, extending from southern China to south of Japan, which is consistent with the precipitation increase along the mei-yu–baiu zone. South of the equator, the zonal wind change is negative, indicating an enhancement in easterly winds.

The change of rainfall anomaly from April to June seems to be a good indicator for summer rainfall anomalies over the north Indian Ocean and the South China Sea. This is demonstrated in Fig. 2a, which displays the pointwise correlation of the JmA rainfall anomaly with JJA mean rainfall anomalies. High correlation coefficients are observed in two regions. One is the Arabian Sea where the correlation coefficient exceeds 0.8 and the other is the South China Sea where the correlation coefficient reaches 0.6. The correlation coefficient between the area-mean JmA rainfall anomaly and JJA mean rainfall anomaly is as high as 0.88 for the region 5°–20°N, 52.5°–72.5°E and reaches 0.68 for the region 5°–20°N, 110°–120°E (Table 1). In other regions of the analysis domain, the correlation coefficient is less than 0.6 except for some small areas. For example, the correlation coefficient is about 0.29 in the region 5°–20°N, 80°–100°E.

Fig. 2.

Pointwise correlation of June minus April (JmA) and June–August (JJA) (a) mean anomaly of precipitation and (b) 850-hPa zonal wind for the period 1979–2010. The two boxes denote the domain of the South China Sea (5°–20°N, 110°–120°E) and the Arabian Sea (5°–20°N, 52.5°–72.5°E).

Fig. 2.

Pointwise correlation of June minus April (JmA) and June–August (JJA) (a) mean anomaly of precipitation and (b) 850-hPa zonal wind for the period 1979–2010. The two boxes denote the domain of the South China Sea (5°–20°N, 110°–120°E) and the Arabian Sea (5°–20°N, 52.5°–72.5°E).

Table 1.

Correlation coefficient of area-mean precipitation and 850-hPa zonal wind anomaly between JJA and JmA or between JJA and AMJ in the South China Sea (SCS; 5°–20°N, 110°–120°E), the Arabian Sea (AraS; 5°–20°N, 52.5°–72.5°E), and the Bay of Bengal (BoB; 5°–20°N, 80°–100°E) for the period 1979–2010.

Correlation coefficient of area-mean precipitation and 850-hPa zonal wind anomaly between JJA and JmA or between JJA and AMJ in the South China Sea (SCS; 5°–20°N, 110°–120°E), the Arabian Sea (AraS; 5°–20°N, 52.5°–72.5°E), and the Bay of Bengal (BoB; 5°–20°N, 80°–100°E) for the period 1979–2010.
Correlation coefficient of area-mean precipitation and 850-hPa zonal wind anomaly between JJA and JmA or between JJA and AMJ in the South China Sea (SCS; 5°–20°N, 110°–120°E), the Arabian Sea (AraS; 5°–20°N, 52.5°–72.5°E), and the Bay of Bengal (BoB; 5°–20°N, 80°–100°E) for the period 1979–2010.

A relatively high correlation coefficient between the JmA anomaly and summer mean anomaly is also seen in 850-hPa zonal wind field over the Arabian Sea and the South China Sea (Fig. 2b). The correlation coefficient between the JmA and JJA zonal wind anomalies exceeds 0.6 to the north of 5°N over the South China Sea and reaches 0.6 over part of the Arabian Sea. The correlation coefficient between the area-mean JmA zonal wind anomaly and JJA mean zonal wind anomaly is 0.69 for the South China Sea and 0.64 for the Arabian Sea (Table 1). For comparison, the correlation coefficient is 0.39 for the Bay of Bengal.

The above results suggest that the April-to-June anomaly change is indicative of summer mean anomaly in the South China Sea and the Arabian Sea. Although statistically the above correlation may be partly due to the inclusion of June in the calculation of both the JmA and JJA mean anomaly, we argue that the anomaly change has physical implication. A positive anomaly change suggests an accelerated seasonal transition from spring to summer, whereas a negative anomaly change denotes a slow spring to summer transition. For example, given a negative atmospheric state in April and a normal atmospheric state in June, which corresponds to a positive anomaly change, an accelerated pace of seasonal transition from April to June is needed to capture the normal state in June. Thus, the correlation implies the dependence of summer mean rainfall anomaly upon the pace of seasonal transition from spring to summer.

To further demonstrate the implications of JmA anomalies, we compare the above correlation with that between AMJ and June–August (JJA) mean anomalies of precipitation and 850-hPa zonal wind. Note that both the AMJ and JJA anomalies include the anomaly in June. The correlation of precipitation in the South China Sea is much lower and the high correlation in the Arabian Sea covers a smaller area compared to the correlation of the April-to-June precipitation change (cf. Fig. 3a and 2a). The correlation of 850-hPa zonal wind is smaller in both the South China Sea and the Arabian Sea compared to the correlation of the April-to-June zonal wind change (cf. Figs. 3b and 2b). The correlation coefficients between area-mean AMJ and JJA mean precipitation anomalies are 0.69 for the region 5°–20°N, 52.5°–72.5°E and 0.44 for the region 5°–20°N, 110°–120°E, both of which are smaller than the correlation corresponding to the respective JmA precipitation anomaly (Table 1). The correlation coefficients between the area-mean AMJ and JJA mean zonal wind anomalies are 0.68 for the South China Sea and 0.57 for the Arabian Sea, respectively.

Fig. 3.

Pointwise correlation of April–June (AMJ) and JJA (a) mean anomaly of precipitation and (b) 850 hPa zonal wind for the period 1979–2010. The two boxes denote the domain of the South China Sea (5°–20°N, 110°–120°E) and the Arabian Sea (5°–20°N, 52.5°–72.5°E).

Fig. 3.

Pointwise correlation of April–June (AMJ) and JJA (a) mean anomaly of precipitation and (b) 850 hPa zonal wind for the period 1979–2010. The two boxes denote the domain of the South China Sea (5°–20°N, 110°–120°E) and the Arabian Sea (5°–20°N, 52.5°–72.5°E).

To examine the correspondence, we have inspected the year-to-year variations of the JmA, AMJ, and JJA rainfall anomalies in the South China Sea and the Arabian Sea. We define an anomalous year when the area-mean anomaly exceeds the 0.5 standard deviation. Table 2 shows the number of cases with different correspondence of JJA mean precipitation anomaly to the JmA precipitation anomaly and AMJ mean precipitation anomaly. In the South China Sea region, there are 8 cases when the JJA anomaly is of the same sign as both the JmA and AMJ anomalies, 6 cases when the JJA anomaly is of the same sign as the JmA anomaly only, and 3 cases when the JJA anomaly is of the same sign as AMJ anomaly only. There are 4 cases when the JJA anomaly is of the opposite sign to AMJ anomaly, and 2 cases when the JJA anomaly is of the opposite sign to JmA anomaly. Thus, the JmA anomaly is a better indicator than the AMJ anomaly for the JJA anomaly in the South China Sea region. In the Arabian Sea region, there are 11 cases when the JJA anomaly is of the same sign as both the JmA and AMJ anomalies, 3 cases when the JJA anomaly is of the same sign as the JmA anomaly, and 1 case when the JJA anomaly is of the opposite sign to the AMJ anomaly. Thus, the JmA anomaly is a better indicator of the JJA mean anomaly than the AMJ mean anomaly in the Arabian Sea region as well.

Table 2.

Number of different cases during the period 1979–2010 when area-mean JJA precipitation anomaly is of the same sign as or opposite sign to JmA and AMJ precipitation anomaly in the South China Sea and the Arabian Sea based on the 0.5 standard deviation as a criterion for an anomalous year.

Number of different cases during the period 1979–2010 when area-mean JJA precipitation anomaly is of the same sign as or opposite sign to JmA and AMJ precipitation anomaly in the South China Sea and the Arabian Sea based on the 0.5 standard deviation as a criterion for an anomalous year.
Number of different cases during the period 1979–2010 when area-mean JJA precipitation anomaly is of the same sign as or opposite sign to JmA and AMJ precipitation anomaly in the South China Sea and the Arabian Sea based on the 0.5 standard deviation as a criterion for an anomalous year.

As pointed out above, the anomaly change from April to June indicates the pace of the transition from spring to summer. On the other hand, the mean anomaly during April through June represents a persistent state. Thus, their variations may have different reasons. In the following, we analyze the spatial and temporal evolution of precipitation, SST, and wind anomalies corresponding to the anomaly change and mean anomaly of precipitation, respectively. Our purpose is to find out whether there are differences in the factors and processes for the anomaly change and mean anomaly. We are also concerned with differences between the South China Sea and the Arabian Sea.

4. Evolution of precipitation, SST, and wind anomalies

In this section, we examine and compare the evolution of precipitation, SST, and wind anomalies corresponding to the JmA precipitation anomaly and AMJ mean precipitation anomaly, respectively, in both the South China Sea and the Arabian Sea regions. The purpose is to unravel how the anomaly change and mean anomaly are related to remote forcing and regional air–sea interaction and to identify common features and differences between the South China Sea and the Arabian Sea regions and between the anomaly change and mean anomaly. We examine the evolution of anomalies associated with the anomaly change first and then those associated with the mean anomaly. In the following, we present results first for the South China Sea region and then for the Arabian Sea region. Comparisons of results between the two regions are made where appropriate.

a. The South China Sea

The evolution of anomalies associated with the precipitation anomaly change over the South China Sea displays a pronounced local coupled variation. Figure 4 displays anomalies obtained by regression upon area-mean JmA precipitation anomaly in the South China Sea. The precipitation anomaly over the South China Sea is negative in April and positive in June (Fig. 4a). Consistently, an anomalous anticyclone and cyclone appear over the South China Sea in April and June, respectively. The magnitudes of the precipitation and wind anomalies are larger in June than in April. The SST anomaly in the South China Sea is positive in May and negative in July (Fig. 4b). This leads to a sequence of less precipitation, higher SST, more precipitation, and lower SST in the South China Sea region. Another feature to note on Fig. 4 is that the precipitation anomaly over the Arabian Sea tends to be opposite to that over the South China Sea with large anomalous westerlies over the tropical Indian Ocean in June (Fig. 4a). This feature may be related to the spatial variability in the atmosphere–ocean interaction (Roxy et al. 2013; Roxy 2014). It may also be associated with the time lag for northeastward propagating intraseasonal precipitation anomalies to reach the two regions (Xie et al. 2007).

Fig. 4.

Anomalies of (a) monthly mean precipitation (mm day−1; shading) and 10-m wind (m s−1, vector, scale at top) and (b) SST (°C) from March to July obtained by regression with respect to June minus April precipitation anomaly averaged over 5°–20°N, 110°–120°E for 1983–2010. Thick contours denote regions where the precipitation and SST anomalies are significant at the 90% confidence level according to the Student’s t test. Only wind vectors that are significant at the 90% confidence level are plotted.

Fig. 4.

Anomalies of (a) monthly mean precipitation (mm day−1; shading) and 10-m wind (m s−1, vector, scale at top) and (b) SST (°C) from March to July obtained by regression with respect to June minus April precipitation anomaly averaged over 5°–20°N, 110°–120°E for 1983–2010. Thick contours denote regions where the precipitation and SST anomalies are significant at the 90% confidence level according to the Student’s t test. Only wind vectors that are significant at the 90% confidence level are plotted.

The anomalies display a systematic evolution and a large-scale feature corresponding to the mean precipitation anomaly over the South China Sea. Figure 5 show anomalies obtained by regression upon area-mean AMJ mean precipitation anomaly in the South China Sea. Negative precipitation anomaly is observed over the equatorial western Pacific during March–May (Fig. 5a). This is accompanied by anomalous lower-level easterlies along the equator and anomalous anticyclones to the north and south of the equator. These precipitation and wind anomalies appear as responses to negative SST anomalies in the equatorial central Pacific (Fig. 5b). With the weakening of SST anomalies after May, the precipitation and lower-level wind anomalies weaken accordingly.

Fig. 5.

As in Fig. 4, but for regression with respect to AMJ mean precipitation anomaly.

Fig. 5.

As in Fig. 4, but for regression with respect to AMJ mean precipitation anomaly.

Positive precipitation and convergent lower-level wind anomalies are observed over the Maritime Continent in March (Fig. 5a). These anomalies expand and move northward with an intensification of an anomalous lower-level cyclone over the South China Sea and western North Pacific in April. The areal extent of these precipitation and wind anomalies reduces in May and the anomalies move eastward after June. A similar eastward movement is seen in positive SST anomalies in the western North Pacific along with a weakening in the magnitude of anomalies (Fig. 5b). The coherent evolution is typical of that in the ENSO decaying years (Wang et al. 2000; Wu et al. 2003), which is associated with a coupled process between SST and atmospheric wind changes (Wang et al. 2000, 2003).

The tropical Indian Ocean displays an asymmetric precipitation and wind anomaly distribution about the equator during March–May (Fig. 5a) and a meridional SST anomaly gradient during March–April (Fig. 5b). Negative SST anomalies appear in the north Indian Ocean in May, which is followed by a reversal of precipitation anomalies in June. These are features of a spring asymmetric mode (Wu et al. 2008; Wu 2009) that are observed during the ENSO decaying years (Du et al. 2009) but can also occur in the absence of ENSO (Wu et al. 2008; Wu and Yeh 2010). As demonstrated by Wu et al. (2008), the evolution of these anomalies in the tropical Indian Ocean involves a series of atmosphere–ocean coupled processes. The precipitation anomalies are opposite in the South China Sea and the Arabian Sea in June with anomalous lower-level westerlies over the north Indian Ocean and the South China Sea (Fig. 5a), a feature that has been pointed out above.

The precipitation anomalies in the South China Sea during the spring to summer transition are contributed by SST anomalies in both the tropical Pacific and tropical south Indian Ocean (Hu et al. 2014). Thus, remote forcing is a major factor for the mean precipitation anomalies during the transition season. This differs from the anomaly change that is mainly related to local air–sea interaction. The precipitation and wind anomalies, in turn, induce SST cooling in the South China Sea (Fig. 5b) through wind–evaporation, cloud–radiation, and wind-induced oceanic processes (He and Wu 2013). Thus, strong atmospheric feedback on local SST is present for the seasonal mean anomaly, similar to the case of anomaly change.

The temporal relationship between local precipitation and SST changes is summarized in Figs. 6a and 6b, which display the evolution of area-mean monthly mean SST and precipitation anomalies in the South China Sea region. Corresponding to the anomaly change, the precipitation anomaly changes from significantly negative in April to large and significantly positive in June, with a significant positive SST anomaly intervening in May; there is also a significant negative SST anomaly in July (Fig. 6a). The negative precipitation anomaly in April corresponds to a positive SST tendency and the positive precipitation anomaly in June corresponds to a negative SST tendency, signifying an atmospheric forcing of the SST change (Wu et al. 2006). Corresponding to the mean anomaly, positive precipitation anomalies are maintained until June (Fig. 6b). These positive precipitation anomalies cannot be explained by local SST anomalies that are small and insignificant before May. This differs from the case of the anomaly change (Fig. 6a). SST anomalies in other regions may play a role in the formation of the South China Sea precipitation anomalies. For example, significant positive SST anomalies in the western North Pacific before June (Fig. 6b) may be one of the factors contributing to the South China Sea precipitation anomalies through a Rossby-type response (Wu et al. 2014; Hu et al. 2014). The large precipitation anomaly during AMJ is accompanied by a decrease in the SST anomaly and the appearance of significant negative SST anomalies in and after May. This feature is similar to the case of the anomaly change (Fig. 6a), signifying a feedback of the atmosphere on the ocean.

Fig. 6.

(a) Monthly mean anomalies of area-mean precipitation (mm day−1, black line) and SST (°C; red line) averaged over the region 5°–20°N, 110°–120°E obtained by regression upon the JmA precipitation anomaly over 5°–20°N, 110°–120°E for 1983–2010. (b) As in (a), but for regression upon AMJ mean precipitation anomaly. Included in (b) is area-mean SST (°C; red dashed line) averaged over 5°–20°N, 120°–150°E. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to the Student’s t test.

Fig. 6.

(a) Monthly mean anomalies of area-mean precipitation (mm day−1, black line) and SST (°C; red line) averaged over the region 5°–20°N, 110°–120°E obtained by regression upon the JmA precipitation anomaly over 5°–20°N, 110°–120°E for 1983–2010. (b) As in (a), but for regression upon AMJ mean precipitation anomaly. Included in (b) is area-mean SST (°C; red dashed line) averaged over 5°–20°N, 120°–150°E. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to the Student’s t test.

b. The Arabian Sea

The evolution of anomalies corresponding to the precipitation anomaly change over the Arabian Sea is shown in Fig. 7. A large positive precipitation anomaly is observed over the Arabian Sea in June, which is accompanied by lower-level wind convergence (Fig. 7a). The precipitation anomaly is negative, but small, in April along with weak anticyclonic lower-level wind anomalies. The SST anomaly is positive in the northern Arabian Sea in March and April and increases in magnitude and expands in area in May (Fig. 7b). After June, the SST anomaly decreases. This indicates a sequence of weakly less precipitation, an increase of positive SST anomaly, more precipitation, and a decrease of positive SST anomaly. An opposite precipitation anomaly along with large anticyclonic lower-level winds appears over the South China Sea with large anomalous easterlies over southern South China Sea, the Bay of Bengal, and equatorial central Indian Ocean (Fig. 7a). This is followed by a positive SST anomaly in the South China Sea in June and July (Fig. 7b). The evolution of precipitation, wind, and SST anomalies in the Indian Ocean displays features similar to those corresponding to the Indian Ocean basin mode. This indicates a role of ENSO in the development of anomalies in the Indian Ocean via processes discussed in previous studies (e.g., Du et al. 2009, 2013). It suggests that the JmA anomaly in the Arabian Sea may be indirectly related to ENSO. Here, we are mainly concerned with how the local SST anomalies, after they are induced, affect the atmosphere.

Fig. 7.

Anomalies of (a) monthly mean precipitation (mm day−1; shading) and 10-m wind (m s−1; vector, scale at top) and (b) SST (°C) from March to July obtained by regression with respect to the JmA precipitation anomaly averaged over 5°–20°N, 52.5°–72.5°E for 1983–2010. Thick contours denote regions where the precipitation and SST anomalies are significant at the 90% confidence level according to a Student’s t test. Only wind vectors that are significant at the 90% confidence level are plotted.

Fig. 7.

Anomalies of (a) monthly mean precipitation (mm day−1; shading) and 10-m wind (m s−1; vector, scale at top) and (b) SST (°C) from March to July obtained by regression with respect to the JmA precipitation anomaly averaged over 5°–20°N, 52.5°–72.5°E for 1983–2010. Thick contours denote regions where the precipitation and SST anomalies are significant at the 90% confidence level according to a Student’s t test. Only wind vectors that are significant at the 90% confidence level are plotted.

A feature common to the South China Sea and the Arabian Sea is that the precipitation anomaly change involves a coupled process between SST and atmospheric precipitation and wind changes. Above-normal precipitation is preceded by a positive SST anomaly and followed by a decrease in the SST anomaly. This signifies a local air–sea coupled process, which will be discussed further in the next section. One difference is that before May the SST anomaly experiences a transition from negative to positive in the South China Sea but an increase in magnitude in the Arabian Sea. This difference is associated with the difference in the negative precipitation anomaly in April, which is large in the South China Sea but small in the Arabian Sea. Another difference is that after May the SST anomaly experiences a reversal of sign in the South China Sea but only a decrease in magnitude in the Arabian Sea. Thus, the SST anomaly change is larger in the South China Sea than in the Arabian Sea both before and after May.

The anomalies corresponding to mean precipitation anomaly over the Arabian Sea appear more regional and local SST anomalies appear more important compared to the South China Sea. Figure 8 displays anomalies obtained by regression upon the area-mean AMJ mean precipitation anomaly in the Arabian Sea. Positive SST anomalies are present in the Arabian Sea in March, increase in April, and weaken after that (Fig. 8b). Positive precipitation anomalies develop over the Arabian Sea in May and are enhanced in June, accompanied by anomalous lower-level convergent winds (Fig. 8a). In June, opposite precipitation anomalies are present in the Arabian Sea and the South China Sea with anomalous lower-level easterlies over the north Indian Ocean and the South China Sea.

Fig. 8.

As in Fig. 7, but for regression with respect to AMJ mean precipitation anomaly.

Fig. 8.

As in Fig. 7, but for regression with respect to AMJ mean precipitation anomaly.

Positive SST anomalies are maintained in southwest tropical Pacific Ocean with some displacement in the location (Fig. 8b). Positive SST anomalies increase in the northwest tropical Pacific Ocean from March to June. As a response, an anomalous lower-level cyclone is seen to the northwest of the SST anomalies in May and June (Fig. 8a). Positive SST anomalies appear in the South China Sea in July (Fig. 8b) following negative precipitation anomalies in June (Fig. 8a).

The evolution of precipitation and SST anomalies and their interrelation corresponding to the mean anomaly appear similar to those corresponding to the anomaly change in the Arabian Sea. In both cases, local SST impact appears important for precipitation anomalies. Another common feature is that the precipitation change feeds back on the ocean, leading to a decrease in SST anomalies. A difference is that a negative (though small) precipitation anomaly is present in April in the case of anomaly change, but not in the case of mean anomalies.

The temporal relationship between local precipitation and SST changes in the Arabian Sea region is summarized in Figs. 9a and 9b. Corresponding to the anomaly change, the precipitation anomaly changes from weakly negative in April to significantly positive in June, with a significant positive SST anomaly in between in May, and a negative SST tendency is seen in June accompanying positive precipitation anomaly (Fig. 9a). These features are similar to those in the South China Sea. However, different from the South China Sea, precipitation anomaly in April is small, positive SST anomaly persists for at least 5 months before May, and the SST anomaly does not fall to negative after the peak precipitation anomaly. Thus, the SST anomaly change appears smaller in the Arabian Sea than in the South China Sea. We note that positive SST anomalies during April cannot induce positive precipitation anomalies, likely because the background SST is still unfavorable in the Arabian Sea and the SST anomalies are not large enough. Corresponding to the mean anomaly, the precipitation anomaly is small in March and April (Fig. 9b). A persistent positive SST anomaly is seen before May, leading to a positive precipitation anomaly. This may indicate a local SST impact on precipitation. The SST anomaly decreases in May–June when the precipitation anomaly is positive (Fig. 9b), indicative of a negative feedback of the atmosphere on the SST. These features are similar to the case of the anomaly change (Fig. 9a).

Fig. 9.

(a) Monthly mean anomalies of area-mean precipitation (mm day−1, black line) and SST (°C, red line) averaged over 5°–20°N, 52.5°–72.5°E obtained by regression upon the JmA precipitation anomaly over 5°–20°N, 52.5°–72.5°E for 1983–2010. (b) As in (a), but for regression upon AMJ mean precipitation anomaly. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to a Student’s t test.

Fig. 9.

(a) Monthly mean anomalies of area-mean precipitation (mm day−1, black line) and SST (°C, red line) averaged over 5°–20°N, 52.5°–72.5°E obtained by regression upon the JmA precipitation anomaly over 5°–20°N, 52.5°–72.5°E for 1983–2010. (b) As in (a), but for regression upon AMJ mean precipitation anomaly. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to a Student’s t test.

5. Local air–sea interaction processes

As indicated by the temporal relationship between SST and precipitation anomaly changes in the previous section, the anomaly change during the spring to summer transition in the South China Sea and the Arabian Sea involves atmosphere–ocean interactions. This section further investigates the local air–sea relationship and diagnoses roles of different processes in the SST anomaly change during the atmosphere–ocean interaction.

The local air–sea relationship displays a notable spatial variation during the transition season. Figure 10 shows local correlations between precipitation and SST, precipitation and SST tendency, and SST and precipitation tendency for the period April–June. The former two correlations have been used in previous studies to detect the nature of local atmosphere–ocean interaction (Wu et al. 2006; Wu and Kirtman 2007). Here, we add the third one to characterize the delayed impact of SST on precipitation change. The tendency of SST and precipitation at a specific month is calculated as the difference of SST and precipitation anomaly in the succeeding month minus that in the preceding month. Based on the sign and magnitude of the above three correlations, we may infer where the atmosphere and ocean are coupled. When the precipitation–SST correlation is large positive and the precipitation–SST tendency and SST–precipitation tendency correlation is small, it indicates that the atmospheric effect on the SST change is small and the SST forcing of the atmosphere is a dominant one. In such case, we may infer a simultaneous impact of SST anomalies on precipitation. When the precipitation–SST tendency correlation is large negative and the SST–precipitation tendency correlation is positive, it indicates a sequence of positive SST leading to positive precipitation leading to SST cooling. In such cases, we may infer a coupled nature of precipitation and SST change.

Fig. 10.

Pointwise correlation of (a) precipitation and SST, (b) precipitation and SST tendency, and (c) SST and precipitation tendency during AMJ for 1983–2010.

Fig. 10.

Pointwise correlation of (a) precipitation and SST, (b) precipitation and SST tendency, and (c) SST and precipitation tendency during AMJ for 1983–2010.

The Arabian Sea, the Bay of Bengal, and the South China Sea display features distinct from most other regions. In these regions, a positive SST–precipitation tendency correlation (Fig. 10c) is accompanied by a negative precipitation–SST tendency correlation (Fig. 10b), whereas the simultaneous precipitation–SST correlation is weak (Fig. 10a). In contrast, there is a positive precipitation–SST correlation over the equatorial central-eastern Pacific Ocean, western equatorial Indian Ocean, and southwest tropical south Indian Ocean (Fig. 10a), which indicates a quick response of atmosphere to local SST anomaly (Wu et al. 2006; Wu and Kirtman 2007). In these regions, the precipitation–SST tendency correlation and SST–precipitation tendency correlation are small (Figs. 10b,c). The above correlation suggests a nearly simultaneous impact of SST anomalies on precipitation over the equatorial central-eastern Pacific and southwest tropical Indian Ocean, but a coupled feature in precipitation and SST changes over the north Indian Ocean and the South China Sea. The western equatorial Pacific appears as a transition region with an eastward extension of negative precipitation–SST tendency correlation (Fig. 10b) and a westward extension of positive precipitation–SST correlation (Fig. 10a).

One factor for SST change is surface heat flux. To examine roles of surface heat fluxes in SST changes, we show in Figs. 11a and 11b anomalies of area-mean surface net shortwave radiation, surface latent heat flux, cloud, and surface wind speed obtained by regression with respect to the JmA precipitation anomaly and AMJ mean precipitation anomaly, respectively, in the South China Sea. Also included in Fig. 11 is the SST tendency converted to the equivalent flux assuming a mixed layer depth of 25 m based on de Boyer Montégut et al. (2004) for a quantitative comparison with surface heat flux terms. The surface net longwave radiation anomaly is smaller than surface net shortwave radiation but with an opposite sign, thus cancelling part of shortwave radiation contribution. The surface sensible heat flux anomaly is of the same sign as the surface latent heat flux anomaly but with a smaller magnitude.

Fig. 11.

(a) Monthly mean anomalies of area-mean surface net shortwave radiation (W m−2; solid red line), surface latent heat flux (W m−2; solid blue line), cloud (%; dashed red line), surface wind speed (m s−1; dashed blue line), and SST tendency (W m−2; solid black line) averaged over 5°–20°N, 110°–120°E obtained by regression upon the JmA precipitation anomaly over 5°–20°N, 110°–120°E for 1983–2010. (b) As in (a), but for regression upon the AMJ mean precipitation anomaly. The SST tendency has been converted to the unit of heat flux assuming a mixed layer depth of 25 m. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to a Student’s t test. The convention for shortwave radiation (latent heat flux) is positive for downward (upward) flux.

Fig. 11.

(a) Monthly mean anomalies of area-mean surface net shortwave radiation (W m−2; solid red line), surface latent heat flux (W m−2; solid blue line), cloud (%; dashed red line), surface wind speed (m s−1; dashed blue line), and SST tendency (W m−2; solid black line) averaged over 5°–20°N, 110°–120°E obtained by regression upon the JmA precipitation anomaly over 5°–20°N, 110°–120°E for 1983–2010. (b) As in (a), but for regression upon the AMJ mean precipitation anomaly. The SST tendency has been converted to the unit of heat flux assuming a mixed layer depth of 25 m. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to a Student’s t test. The convention for shortwave radiation (latent heat flux) is positive for downward (upward) flux.

Corresponding to the anomaly change, the increase in the SST anomaly before May is contributed to by both an increase in net shortwave radiation and a decrease in latent heat flux (Fig. 11a). These surface heat flux anomalies are consistent with reduced cloudiness associated with less precipitation and surface wind speed (Fig. 11a). The decrease in the SST anomaly after May is contributed mainly by a decrease in net shortwave radiation with a supplementary contribution from surface latent heat flux (Fig. 11a). Surface wind speed is enhanced in June and August (Fig. 11a), which is associated with more precipitation, leading to an increase in surface latent heat flux. The latent heat flux anomalies, however, are small likely because of a cancellation of the wind speed effect by the sea–air humidity difference effect. The latent heat flux anomaly in July (Fig. 11a) is negative, mainly due to the sea–air humidity difference following negative SST anomalies (Fig. 6a), which provides a damping effect. We note that the SST tendency and surface heat fluxes in February display a large discrepancy. This discrepancy may be attributed to the use of a constant mixed layer depth of 25 m, the contributions of oceanic processes, and the uncertainty of surface heat fluxes.

The surface heat flux anomalies corresponding to the mean anomaly display both similarities to and differences from those corresponding to the anomaly change. Corresponding to the mean anomaly in the South China Sea, the positive SST tendency before April is accompanied by a large reduction in shortwave radiation associated with an increase in cloudiness (Fig. 11b). The latent heat flux anomalies are positive (though small) (Fig. 11b). As such, shortwave radiation and latent heat flux appear as a damping effect of the SST anomaly. This differs from the case of the anomaly change. The large negative SST tendency in May is contributed by the decrease in shortwave radiation (Fig. 11b). This feature is similar to the case of the anomaly change. The latent heat flux also contributes to the SST tendency, but its magnitude is small (Fig. 11b).

The roles of surface heat fluxes in SST changes in the Arabian Sea display some differences from those in the South China Sea. Figures 12a and 12b show anomalies of area-mean surface net shortwave radiation, surface latent heat flux, cloud, and surface wind speed obtained by regression with respect to the JmA precipitation anomaly and AMJ mean precipitation anomaly, respectively, in the Arabian Sea. Again, the SST tendency is included for comparison after it is converted to the equivalent flux assuming a mixed layer depth of 35 m based on de Boyer Montégut et al. (2004). Corresponding to the anomaly change, both net shortwave radiation and latent heat flux anomalies are small before May (Fig. 12a). This appears to be a reason for the small SST tendency in this region. After May, the decrease in the SST anomaly is contributed by both decrease in shortwave radiation and increase in latent heat flux (Fig. 12a), which is consistent with enhanced cloudiness (June and July) and surface wind speed (July). The increase in latent heat flux in June is likely due to an increase in sea–air humidity difference following positive SST anomaly (Fig. 9a). One important difference between the South China Sea and the Arabian Sea is the role of surface heat fluxes in the SST anomaly formation before May. The surface shortwave radiation and latent heat flux have large contributions to the formation of positive SST anomaly in the South China Sea, whereas their contributions are small in the Arabian Sea. In the latter region, the SST anomaly appears to have been formed about half year ago and then maintained through winter and spring.

Fig. 12.

As in Fig. 11, but for 5°–20°N, 52.5°–72.5°E and a mixed layer depth of 35 m.

Fig. 12.

As in Fig. 11, but for 5°–20°N, 52.5°–72.5°E and a mixed layer depth of 35 m.

Corresponding to the mean anomaly in the Arabian Sea region, shortwave radiation is reduced in association with increase in cloudiness in March (Fig. 12b). This appears to be a damping effect of the positive SST anomaly (Fig. 9b). This differs from the case of the anomaly change. The SST anomaly decrease during May–July is contributed by both shortwave radiation and latent heat flux anomalies associated with cloud and surface wind speed changes (Fig. 12b). This feature is similar to the case of the anomaly change.

The oceanic influence on atmosphere includes both dynamic and thermodynamic processes. The dynamic effect is related to remote forcing on a large spatial scale. According to the evolution of anomalies in previous section, the remote forcing is obvious for the mean anomaly over the South China Sea, but not so for the anomaly change over both the South China Sea and the Arabian Sea as well as the mean anomaly over the Arabian Sea. To examine the thermodynamic effect of local SST anomalies, we show in Fig. 13 anomalies of the area-mean instability index obtained by regression with respect to the JmA anomaly and AMJ mean anomaly over both the South China Sea and the Arabian Sea. In the Arabian Sea region, the instability index anomaly is large and significantly positive, corresponding to both the anomaly change and mean anomaly (Fig. 13b), indicative of a weakening of the atmospheric stability. This confirms the role of local SST anomalies with regard to precipitation through modifying the atmospheric stability over the Arabian Sea. In the South China Sea region, the instability index anomaly is positive in May (Fig. 13a) but insignificant corresponding to the anomaly change. The anomaly is large and significantly negative in May and June corresponding to the mean anomaly. The above results suggest that the thermodynamic effect is less important in the South China Sea than in the Arabian Sea.

Fig. 13.

(a) Monthly mean anomalies of stability index (K) averaged over 5°–20°N, 110°–120°E obtained by regression upon JmA (solid line) and AMJ mean (dashed) precipitation anomalies over 5°–20°N, 110°–120°E for 1983–2010. (b) As in (a), but for 5°–20°N, 52.5°–72.5°E. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to the Student’s t test.

Fig. 13.

(a) Monthly mean anomalies of stability index (K) averaged over 5°–20°N, 110°–120°E obtained by regression upon JmA (solid line) and AMJ mean (dashed) precipitation anomalies over 5°–20°N, 110°–120°E for 1983–2010. (b) As in (a), but for 5°–20°N, 52.5°–72.5°E. Marked points denote that the corresponding correlation coefficients are significant at the 90% confidence level according to the Student’s t test.

6. Summary and discussions

Analysis reveals that the change of precipitation anomaly from April to June is a good indicator, better than the April–June mean precipitation anomaly, for summer precipitation anomalies over the South China Sea and the Arabian Sea. This indicates that anomalies in summer may be related to the pace of the spring to summer transition. Motivated by this finding, the present study analyzes in detail the temporal evolution of precipitation, surface wind, and SST anomalies during the spring to summer transition season. We compare the evolution of anomalies corresponding to the precipitation anomaly change and the mean precipitation anomaly during the spring to summer transition and examine the differences between the South China Sea and the Arabian Sea.

In the South China Sea region, a local coupled variation is observed corresponding to the precipitation anomaly change during the spring to summer transition. There is a clear sequence of less precipitation, higher SST, more precipitation, and lower SST, indicative of a coupling between atmosphere and ocean. Differently, the mean precipitation anomaly during April–June is mainly induced by remote forcing. In both cases, the atmospheric change has a large feedback on local SST change. Thus, in the South China Sea region, the atmospheric influence on the SST appears robust, whereas local SST forcing of the atmosphere is mainly seen in the fast change component of the atmosphere.

In the Arabian Sea region, the local SST anomaly appears important for both the precipitation anomaly change and mean precipitation anomaly during the spring to summer transition. The local SST effect is clearly seen in the modulation of atmospheric stability. The atmosphere has a negative feedback on local SST change in both cases. Thus, there is a local coupled variability, which is similar to the South China Sea region. Different from the South China Sea region, the SST anomaly before the transition persists a long time and appears not to be generated by the surface heat flux in the spring.

Roles of shortwave radiation and latent heat flux anomalies in the SST changes during the spring to summer transition show differences between the South China Sea and the Arabian Sea and vary with time. Before May, both shortwave radiation and latent heat flux contribute to local SST anomaly change corresponding to the precipitation anomaly change in the South China Sea, whereas surface heat flux anomalies are small corresponding to the precipitation anomaly change in the Arabian Sea. This leads to different SST tendencies, large in the South China Sea but small in the Arabian Sea. Shortwave radiation has a negative effect on local SST anomaly change before April corresponding to mean precipitation anomaly in the South China Sea. Such a negative effect is also seen for local SST change in March corresponding to mean precipitation anomaly in the Arabian Sea. After May, the shortwave radiation is a main factor for the SST anomaly change in the South China Sea, whereas both shortwave radiation and latent heat flux contribute to the SST anomaly change in the Arabian Sea.

The local air–sea relationship during the spring to summer transition season displays a notable difference in the north Indian Ocean and the South China Sea from the equatorial central-eastern Pacific and southwest tropical Indian Ocean. This is likely linked to the particular location of the north Indian Ocean and the South China Sea. The spring to summer transition displays a pronounced change in both precipitation and winds over the north Indian Ocean and the South China Sea. This signifies the importance of studying air–sea interaction processes in the regions where seasonal transition occurs. Whether this applies to other regions during the transition season needs to be investigated in the future.

Why does the JmA anomaly have a higher correlation with the JJA mean anomaly than the AMJ anomaly? Based on the comparison of precipitation and SST anomaly evolution corresponding to JmA and AMJ precipitation in the South China Sea, we propose the following explanation. Corresponding to the AMJ anomaly, a remote forcing–induced positive precipitation anomaly persists from April to June (Fig. 6b). The precipitation anomaly, in turn, exerts a persistent impact on local SST change via surface heat flux changes (Fig. 11b), leading to large negative SST anomalies that persist for a relatively long time (Fig. 6b). These negative SST anomalies also appear earlier (starting in May). The effect of these negative SST anomalies cancels the remote forcing (Fig. 5), leading to weak precipitation anomalies in July and August (Figs. 5a and 6b). Corresponding to the JmA anomaly, local interaction is more prominent. A negative April precipitation anomaly induces a local positive SST anomaly in May (Fig. 6a) via surface heat flux changes (Fig. 11a). The positive SST anomaly in turn induces a positive June precipitation anomaly (Fig. 6a) that turns around the SST anomaly via surface heat flux changes (Fig. 11a). Because of a relatively short period of atmospheric effect, the resultant negative SST anomalies are not as significant in summer compared to the case corresponding to AMJ anomaly. In addition, the negative SST anomalies appear later (July) (Fig. 6a). As such, the negative effect of SST anomalies on summer precipitation is relatively smaller. Thus, the JJA anomaly is more consistent with the JmA anomaly than the AMJ anomaly.

The large-scale SST anomaly pattern and its influence on regional climate may be modulated by low-frequency changes. For example, the roles of the SST anomalies in tropical Indo-Pacific regions in southern China rainfall variability have experienced obvious interdecadal changes in the past (e.g., Wu et al. 2012; Chen et al. 2014). As such, the results obtained in the present study may change on the low-frequency time scales. Such change may be more prominent for the mean anomaly, which depends upon the large-scale SST anomaly pattern, than for the anomaly change, which is mainly dependent upon local air–sea interaction processes. Further studies are needed to address this issue.

Acknowledgments

The comments of three anonymous reviewers are appreciated. This study is supported by a National Key Basic Research Program of China grant (2014CB953902), National Natural Science Foundation of China grants (41275081 and 41475081), and a Hong Kong Research Grants Council grant (CUHK403612). The GPCP precipitation data were obtained from http://www.esrl.noaa.gov/psd/. The NOAA OI version 2 SST data were obtained from http://www.esrl.noaa.gov/psd/. The NCEP-DOE Reanalysis 2 data were obtained from http://www.esrl.noaa.gov/psd/. The NOCS V2.0 data were obtained from http://badc.nerc.ac.uk/.

REFERENCES

REFERENCES
Adler
,
R. F.
, and Coauthors
,
2003
:
The version 2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present)
.
J. Hydrometeor.
,
4
,
1147
1167
, doi:.
Berry
,
D. I.
, and
E. C.
Kent
,
2009
:
A new air–sea interaction gridded dataset from ICOADS with uncertainty estimates
.
Bull. Amer. Meteor. Soc.
,
90
,
645
656
, doi:.
Chen
,
J.-P.
,
Z.-P.
Wen
,
R.
Wu
,
Z.-S.
Chen
, and
P.
Zhao
,
2014
:
Interdecadal changes in the relationship between southern China winter-spring precipitation and ENSO
.
Climate Dyn.
,
43
,
1327
1338
, doi:.
Chou
,
C.
,
J.-Y.
Tu
, and
J.-Y.
Yu
,
2003
:
Interannual variability of the western North Pacific summer monsoon: Differences between ENSO and non-ENSO years
.
J. Climate
,
16
,
2275
2287
, doi:.
de Boyer Montégut
,
C.
,
G.
Madec
,
A. S.
Fischer
,
A.
Lazar
, and
D.
Iudicone
,
2004
:
Mixed layer depth over the global ocean: An examination of profile data and a profile-based climatology
.
J. Geophys. Res.
,
109
,
C12003
, doi:.
Du
,
Y.
,
S.-P.
Xie
,
G.
Huang
, and
K.
Hu
,
2009
:
Role of air–sea interaction in the long persistence of El Niño–induced north Indian Ocean warming
.
J. Climate
,
22
,
2023
2038
, doi:.
Du
,
Y.
,
S.-P.
Xie
,
Y.-L.
Yang
,
X.-T.
Zheng
,
L.
Liu
, and
G.
Huang
,
2013
:
Indian Ocean variability in the CMIP5 multimodel ensemble: The basin mode
.
J. Climate
,
26
,
7240
7266
, doi:.
He
,
Z.
, and
R.
Wu
,
2013
:
Seasonality of interannual atmosphere–ocean interaction in the South China Sea
.
J. Oceanogr.
,
69
,
699
712
, doi:.
He
,
Z.
, and
R.
Wu
,
2014
:
Indo-Pacific remote forcing in summer rainfall variability over the South China Sea
.
Climate Dyn.
,
42
,
2323
2337
, doi:.
Hu
,
W.-T.
,
R.
Wu
, and
Y.
Liu
,
2014
:
Relation of the South China Sea precipitation variability to tropical Indo-Pacific SST anomalies during spring-to-summer transition
.
J. Climate
,
27
,
5451
5467
, doi:.
Huffman
,
G. J.
,
R. F.
Adler
,
D. T.
Bolvin
, and
G.
Gu
,
2009
:
Improving the global precipitation record: GPCP version 2.1
.
Geophys. Res. Lett.
,
36
,
L17808
, doi:.
Kanamitsu
,
M.
,
W.
Ebisuzaki
,
J.
Woollen
,
S.-K.
Yang
,
J. J.
Hnilo
,
M.
Fiorino
, and
G. L.
Potter
,
2002
:
NCEP–DOE AMIP-II reanalysis (R-2)
.
Bull. Amer. Meteor. Soc.
,
83
,
1631
1643
, doi:.
Lau
,
N.-C.
, and
M. J.
Nath
,
2000
:
Impact of ENSO on the variability of the Asian–Australian monsoons as simulated in GCM experiments
.
J. Climate
,
13
,
4287
4309
, doi:.
Reynolds
,
R. W.
,
N. A.
Rayner
,
T. M.
Smith
,
D. C.
Stokes
, and
W.
Wang
,
2002
:
An improved in situ and satellite SST analysis for climate
.
J. Climate
,
15
,
1609
1625
, doi:.
Roxy
,
M.
,
2014
:
Sensitivity of precipitation to sea surface temperature over the tropical summer monsoon region—and its quantification
.
Climate Dyn.
,
43
,
1159
1169
, doi:.
Roxy
,
M.
, and
Y.
Tanimoto
,
2007
:
Role of SST over the Indian Ocean in influencing the intraseasonal variability of the Indian summer monsoon
.
J. Meteor. Soc. Japan
,
85
,
349
358
, doi:.
Roxy
,
M.
, and
Y.
Tanimoto
,
2012
:
Influence of sea surface temperature on the intraseasonal variability of the South China Sea summer monsoon
.
Climate Dyn.
,
39
,
1209
1218
, doi:.
Roxy
,
M.
,
Y.
Tanimoto
,
B.
Preethi
,
T.
Pascal
, and
R.
Krishnan
,
2013
:
Intraseasonal SST–precipitation relationship and its spatial variability over the tropical summer monsoon region
.
Climate Dyn.
,
41
,
45
61
, doi:.
Wang
,
B.
,
R.
Wu
, and
X.
Fu
,
2000
:
Pacific–East Asian teleconnection: How does ENSO affect East Asian climate?
J. Climate
,
13
,
1517
1536
, doi:.
Wang
,
B.
,
R.
Wu
, and
K.-M.
Lau
,
2001
:
Interannual variability of the Asian summer monsoon: Contrasts between the Indian and the western North Pacific–East Asian monsoons
.
J. Climate
,
14
,
4073
4090
, doi:.
Wang
,
B.
,
R.
Wu
, and
T.
Li
,
2003
:
Atmosphere–warm ocean interaction and its impacts on the Asian–Australian monsoon variation
.
J. Climate
,
16
,
1195
1211
, doi:.
Wu
,
R.
,
2009
:
Possible role of the Indian Ocean in the out-of-phase transition of the Australian to Indian summer monsoon
.
J. Climate
,
22
,
1834
1849
, doi:.
Wu
,
R.
,
2010
:
Subseasonal variability during the South China Sea summer monsoon onset
.
Climate Dyn.
,
34
,
629
642
, doi:.
Wu
,
R.
, and
B.
Wang
,
2000
:
Interannual variability of summer monsoon onset over the western North Pacific and the underlying processes
.
J. Climate
,
13
,
2483
2501
, doi:.
Wu
,
R.
, and
B. P.
Kirtman
,
2005
:
Roles of Indian and Pacific Ocean air–sea coupling in tropical atmospheric variability
.
Climate Dyn.
,
25
,
155
170
, doi:.
Wu
,
R.
, and
B. P.
Kirtman
,
2007
:
Regimes of local air–sea interactions and implications for performance of forced simulations
.
Climate Dyn.
,
29
,
393
410
, doi:.
Wu
,
R.
, and
S.-W.
Yeh
,
2010
:
A further study of the tropical Indian Ocean asymmetric mode in boreal spring
.
J. Geophys. Res.
,
115
,
D08101
, doi:.
Wu
,
R.
,
Z.-Z.
Hu
, and
B. P.
Kirtman
,
2003
:
Evolution of ENSO-related rainfall anomalies in East Asia
.
J. Climate
,
16
,
3742
3758
, doi:.
Wu
,
R.
,
B. P.
Kirtman
, and
K.
Pegion
,
2006
:
Local air–sea relationship in observations and model simulations
.
J. Climate
,
19
,
4914
4932
, doi:.
Wu
,
R.
,
B. P.
Kirtman
, and
V.
Krishnamurthy
,
2008
:
An asymmetric mode of tropical Indian Ocean rainfall variability in boreal spring
.
J. Geophys. Res.
,
113
,
D05104
, doi:.
Wu
,
R.
,
S.
Yang
,
Z.-P.
Wen
,
G.
Huang
, and
K.
Hu
,
2012
:
Interdecadal change in the relationship of southern China summer rainfall with tropical Indo-Pacific SST
.
Theor. Appl. Climatol.
,
108
,
119
133
, doi:.
Wu
,
R.
,
G.
Huang
,
Z.
Du
, and
K.
Hu
,
2014
:
Cross-season relation of the South China Sea precipitation variability between winter and summer
.
Climate Dyn.
,
43
,
193
207
, doi:.
Xie
,
S.-P.
,
C.-H.
Chang
,
Q.
Xie
, and
D.
Wang
,
2007
:
Intraseasonal variability in the summer South China Sea: Wind jet, cold filament, and recirculations
.
J. Geophys. Res.
,
112
,
C10008
, doi:.