1. Introduction
It is well known that the South China Sea summer monsoon (SCSSM) is a crucial component of the Asian summer monsoon because of its particular geographical location at the center of the Asian–Australian monsoon region and its connections to other Asian monsoon subsystems (Wang and Wu 1997; Wang et al. 2009; Ding et al. 2018). The SCSSM onset with abrupt climatological features is generally seen in pentad 28 (16–20 May), which is considered the precursor to the onset of the East Asian summer monsoons and marks the commencement of the rainy season in eastern Asia (Lau and Li 1984; Tao and Chen 1987; Ding 1992; Wang and Wu 1997; Ding and Chan 2005; Zhu et al. 2005; Tian and Wang 2010; Liu and Zhu 2016; Bombardi et al. 2019, 2020). In addition, the summer monsoon anomalies over the SCS can influence the weather and climate in other regions of the globe by atmospheric telecommunications (Wang and Chen 2018; Zheng and Huang 2019; Chen et al. 2022). The considerable increase in precipitation induced by SCSSM onset increases the latent heat release over the South China Sea and northwestern Pacific, which has implications for the rainfall over the Yangtze River basin due to changes in circulation. For instance, a late SCSSM onset is accompanied by above-normal rainfall in May (Hung et al. 2006; He and Zhu 2015; Jiang et al. 2018). In addition, the changes in circulation and moisture after SCSSM onset induce surface wind speed anomalies. Thus, latent heat flux anomalies are influenced, providing favorable conditions for intraseasonal oscillations propagating northward (Zheng and Huang 2019). In contrast, upward surface heat fluxes, including latent and sensible heat fluxes, contribute to declining land–sea thermal contrast, which has a marked impact on the variations in SCSSM onset (Li et al. 2020). To improve our understanding of the turbulent heat fluxes during the SCS onset, coordinated multiplatform air–sea observations covering the north-central SCS basin were conducted in 2021, as shown in Fig. 1.
Cruise route in the South China Sea (SCS) from 5 May to 8 Jun 2021 (magenta curve). The empty black circles indicate the ERA5 grids that are spatially matched to the ship-borne measurement trajectory. The colored background shows the mean air–sea turbulent heat flux differences (unit: W m−2) for the 10 days before and after the onset of the monsoon (29 May 2021) obtained from ERA5. The choices of 10-day averages are the same as those in Lau and Nath (2009). Positive values represent extra heat loss from the ocean to the atmosphere and vice versa. The black square is the location of the Bailong buoy (16.0°N, 115.4°E). (a) The A, B, C, and D buoys for observing air–sea variables are shown as filled black circles. The gray vectors represent the 10-day mean surface wind (a) before and (b) after SCSSM onset. Note that the cyan triangles and pentagram represent the time during the cruise course in (b).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Cruise route in the South China Sea (SCS) from 5 May to 8 Jun 2021 (magenta curve). The empty black circles indicate the ERA5 grids that are spatially matched to the ship-borne measurement trajectory. The colored background shows the mean air–sea turbulent heat flux differences (unit: W m−2) for the 10 days before and after the onset of the monsoon (29 May 2021) obtained from ERA5. The choices of 10-day averages are the same as those in Lau and Nath (2009). Positive values represent extra heat loss from the ocean to the atmosphere and vice versa. The black square is the location of the Bailong buoy (16.0°N, 115.4°E). (a) The A, B, C, and D buoys for observing air–sea variables are shown as filled black circles. The gray vectors represent the 10-day mean surface wind (a) before and (b) after SCSSM onset. Note that the cyan triangles and pentagram represent the time during the cruise course in (b).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Cruise route in the South China Sea (SCS) from 5 May to 8 Jun 2021 (magenta curve). The empty black circles indicate the ERA5 grids that are spatially matched to the ship-borne measurement trajectory. The colored background shows the mean air–sea turbulent heat flux differences (unit: W m−2) for the 10 days before and after the onset of the monsoon (29 May 2021) obtained from ERA5. The choices of 10-day averages are the same as those in Lau and Nath (2009). Positive values represent extra heat loss from the ocean to the atmosphere and vice versa. The black square is the location of the Bailong buoy (16.0°N, 115.4°E). (a) The A, B, C, and D buoys for observing air–sea variables are shown as filled black circles. The gray vectors represent the 10-day mean surface wind (a) before and (b) after SCSSM onset. Note that the cyan triangles and pentagram represent the time during the cruise course in (b).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
There are many factors reflecting the circulation changes before and after the onset of the South China Sea summer monsoon (SCSSM). Since the seasonal reversal of the lower-level wind field and rainfall bursts are the most significant features of the monsoon, generally the shift of the lower-level (850 hPa) wind field and/or precipitation (convection) in the South China Sea (SCS) are used to determine the SCSSM onset. Approximately 20 different definitions have been proposed for the SCSSM from different perspectives. A convenient index that is defined as 850-hPa zonal wind averaging over 5°–15°N, 110°–120°E (Wang et al. 2004) was used to determine the SCSSM onset pentads. Based on Wang (2004), stricter criteria were proposed to refine the definition further from pentad to day (Ding et al. 2016; Kajikawa and Wang 2012). In this study, the definition of SCSSM onset follows the operational provisions of the National Climate Center, China Meteorological Administration (CMA). The SCSSM onset date is defined as the first pentad when the regionally averaged 850-hPa zonal wind switches from easterlies to westerlies and the regionally averaged equivalent potential temperature is higher than 340 K over 10°–20°N, 110°–120°E (the north-central SCS), and these features persist for at least three consecutive pentads (Jiang et al. 2018). During the onset in 2021, the SCS is controlled by the western Pacific subtropical high pressure (WPSH), which may result in the later onset of SCSSM (Chen and Wang 2023). As shown in Fig. 2, the WPSH did not retreat eastward until early June 2021. In this process, the southeasterlies decayed with westward flow invading the SCS. In addition, the negative-to-positive shift of low-level zonal wind indicated the stable establishment of SCSSM and the transition of atmospheric circulation type from winter to summer. Prior studies (Shen and Lau 1995; Chen and Wang 1998; Wu 2002; Lau and Nath 2009) indicated that the tendency of sea surface temperature (SST) and the precipitation pattern in the SCS and western North Pacific (WNP) are closely linked in terms of significant air–sea interactions during the onset of the SCSSM. For example, the SST in the early summer in the SCS is higher than that in the WNP; however, the east–west SST gradient is reversed in July due to the prevalent dry, clear-sky conditions and, hence, stronger solar radiation in the WNP. As the precipitation center is accompanied by the site of maximum SST, understanding how SST varies in association with the air–sea heat fluxes is key to understanding air–sea coupling processes and the potential prediction of climate signals in terms of the onset of the SCSSM.
The evolution of 850-hPa wind (vectors; unit: m s−1), wind speed (shading; unit: m s−1), and geopotential height contours (black contours; unit: gpm) from 28 May to 7 Jun 2021. Note that the bold 5880-gpm contour has been widely adopted to determine the boundary of the WPSH (e.g., Yamaguchi and Majumdar 2010; Ren et al. 2013). The 10-m buoys A, B, C, and D are shown as filled magenta circles. The indices USCS and EQT marked on the upper left represent regionally averaged 850-hPa zonal wind and equivalent potential temperature over 10°–20°N, 110°–120°E, which refer to the onset definition of the CMA.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The evolution of 850-hPa wind (vectors; unit: m s−1), wind speed (shading; unit: m s−1), and geopotential height contours (black contours; unit: gpm) from 28 May to 7 Jun 2021. Note that the bold 5880-gpm contour has been widely adopted to determine the boundary of the WPSH (e.g., Yamaguchi and Majumdar 2010; Ren et al. 2013). The 10-m buoys A, B, C, and D are shown as filled magenta circles. The indices USCS and EQT marked on the upper left represent regionally averaged 850-hPa zonal wind and equivalent potential temperature over 10°–20°N, 110°–120°E, which refer to the onset definition of the CMA.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The evolution of 850-hPa wind (vectors; unit: m s−1), wind speed (shading; unit: m s−1), and geopotential height contours (black contours; unit: gpm) from 28 May to 7 Jun 2021. Note that the bold 5880-gpm contour has been widely adopted to determine the boundary of the WPSH (e.g., Yamaguchi and Majumdar 2010; Ren et al. 2013). The 10-m buoys A, B, C, and D are shown as filled magenta circles. The indices USCS and EQT marked on the upper left represent regionally averaged 850-hPa zonal wind and equivalent potential temperature over 10°–20°N, 110°–120°E, which refer to the onset definition of the CMA.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The variations in air–sea turbulent heat fluxes are closely associated with multiscale air–sea coupled processes. The onset of SCSSM is a typical weather-scale issue, indicating the seasonal transitions in wind and associated meteorological variables. Lau and Nath (2009) suggested that the increased cloud cover and surface wind speed (WS) during the onset of the monsoon over the SCS in May–June led to a reduction in incoming shortwave radiation and an enhancement of the air–sea upward latent heat (LH) flux, which thereby helps cool the local SST. The results based on the South China Sea Monsoon Experiment (SCSMEX) during 1996–2001 (reviewed by Ding et al. 2004) have revealed the balancing process between the latent heat flux and oceanic heating and the relationship among the turbulent heat fluxes and SST and moisture. Obvious spatial differences in air–sea heat fluxes exist between the southern and northern SCS. However, these differences have not been quantitatively estimated using coordinated fixed and moving observational platforms. Sporadic studies based on tower measurements on Yongxing Island of Xisha reported the relationship between changes in sea surface temperature and latent heat flux (Yan et al. 2003, 2005; Sun et al. 2010; Hung et al. 2012) and evaluated reanalysis products to improve simulations (Zhou et al. 2018). Nevertheless, direct in situ measurements of meteorological variables at the air–sea interface in recent years are still unfortunately rare in studies of air–sea heat fluxes due to the lack of observational facilities in the SCS.
A new regional observational program was implemented in the early summer of 2021 (Fig. 1) to investigate the anomalies in air–sea turbulent heat fluxes during the onset of the SCSSM and their spatial patterns. The observational network (Fig. 3) includes ship-borne air–sea measurements based on the R/V TAN KAH KEE, which is affiliated with Xiamen University, over the SCS basin, four air–sea 10-m buoys (Song 2020) that are operated and aintained by the State Oceanic Administration (SOA) of China, and one air–sea Bailong buoy (Song et al. 2021) in the central SCS. Unfortunately, the meteorological module of the Bailong buoy was sabotaged 3 days after its deployment. The ship-borne measurements plus the five buoys (shown in Fig. 1 and summarized in Table 1) are used to quantify the turbulent heat flux and study the dominance of sea–air variables under different marine atmospheric boundary (MABL) stability conditions during the onset of the SCSSM.
Schematic illustration of the regional coordinated multiplatform air–sea observations, including the R/V TAN KAH KEE and 10- and 3-m Bailong air–sea buoys, during the onset of the SCSSM.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Schematic illustration of the regional coordinated multiplatform air–sea observations, including the R/V TAN KAH KEE and 10- and 3-m Bailong air–sea buoys, during the onset of the SCSSM.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Schematic illustration of the regional coordinated multiplatform air–sea observations, including the R/V TAN KAH KEE and 10- and 3-m Bailong air–sea buoys, during the onset of the SCSSM.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Information on the buoys and ship-borne air–sea observations in the SCS used in this study. The numbers in parentheses following the variables are the observational heights (unit: m) with respect to the sea surface. The negative values represent the underwater depth.
Given the existence of significant uncertainties in air–sea turbulent heat fluxes globally (Weare 1989; Gleckler and Weare 1997; Josey 2001; Brunke et al. 2011; Song and Yu 2013; Yu 2019), this study will provide an estimate of air–sea turbulent heat fluxes and their anomalies during the onset of the SCSSM based on valuable in situ air–sea observations. In addition to improving the understanding of air–sea interactions during summer monsoon bursts, such studies also contribute to synoptic-scale heat flux anomalies, focusing on air–sea measurements (Cronin et al. 2019) and the problem of heat flux simulations in reanalysis, which provide insight into the global/regional sea surface heat budget imbalance (Yu 2019). This study focuses on the investigation of turbulent heat flux anomalies in the SCS based on comprehensive air–sea measurements. In the following section, we describe the field program and ship-borne observations; we also describe the bulk algorithms used to quantify the air–sea turbulent heat fluxes. Section 3 presents the observed heat flux anomalies in association with the onset of the summer monsoon using comprehensive air–sea observations in the basin, focusing on the northern part. Section 4 shows the discrepancies in the turbulent heat fluxes and associated air–sea variables between the reanalysis and the observations. Finally, a summary and discussion are provided in section 5.
2. Data and methods
a. High-resolution air–sea variables in the SCS
The major obstacles for satellite-based estimates of air–sea heat fluxes or reanalysis products are the unresolved diurnal variations in air–sea variables, particularly the SST. It is almost impossible to obtain high-resolution air–sea variables using satellites. Diurnal variations and other high-frequency variations (e.g., subdaily variations) significantly contribute to estimates of synoptic mean air–sea latent heat fluxes (Yan et al. 2021). Thus, to investigate the air–sea heat flux anomalies during the onset of the SCSSM, comprehensive multiplatform air–sea observations (Fig. 3) were conducted from May to June 2021 in the SCS, including ship-borne measurements, four 10-m operational air–sea buoys, and one 3-m Bailong buoy in the central SCS. The 10-m buoys are operated and maintained by the southern branch of the SOA of China, and the 3-m buoys are deployed by the First Institute of Oceanography, which is affiliated with the SOA. The temporal resolution of the 10-m buoy is hourly, while it is 10 min for the Bailong buoy. More details about the sensors attached to the 10- and 3-m buoys can be seen in Song (2020) and Song et al. (2021) and are not presented here to avoid repetition. The high-resolution ship-borne observations at 10-min intervals provide the opportunity to not only estimate the basin-scale heat fluxes but also quantify the localized latent heat flux anomalies induced by meso- and submesoscale eddies through cross-center cruise transects toward cyclonic eddies (Song et al. 2022). As the onset of the summer monsoon is on a synoptic weather scale, the daily mean air–sea heat fluxes are obtained by averaging high-resolution measurements. Information on the high-resolution air–sea variables based on these platforms is summarized in Table 1.
Regarding the monsoon observations, the SCSMEX should be mentioned first. As a multinational atmospheric and oceanic observation plan, the SCSMEX has already accumulated valuable air–sea measurements with efforts from scientists around the globe. Our observations in this study are an important supplement for the SCSMEX, with a special focus on air–sea turbulent heat fluxes. Our program provides synchronous air–sea variable measurements based on fixed buoys and moving platforms covering the basin to determine how many anomalies in turbulent heat fluxes are induced and associated with the onset of the monsoon.
Given the limited number of buoy sites and cruises, two new operational atmospheric reanalysis datasets are also used to analyze the heat flux anomalies during the onset of the SCSSM. One is the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis, version 5 (ERA5), which provides hourly outputs of atmospheric, land, and oceanic variables from 1950 to the present (Hersbach et al. 2020; Bell et al. 2021). The other is the Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA2; Gelaro et al. 2017), which provides hourly air–sea variables from 1980 to the present. The spatial resolutions of ERA5 and MERRA2 are 0.25° × 0.25° and 0.625° (longitude) × 0.5° (latitude), respectively. ERA5 and MERRA2 employ the Louis scheme (Louis 1979; ECMWF 2021) to calculate air–sea turbulent heat fluxes. MERRA2 has an updated parameterization based on Helfand and Schubert (1995). To achieve a faster forecasting system, the turbulent heat fluxes are estimated based on simple analytic solutions to couple the ocean (SST) and atmosphere. Thus, it should be mentioned that these algorithms and schemes are different from those of bulk formulas (Brodeau et al. 2017; Bonino et al. 2022), as shown in the following subsection. This induces an approximately 10% difference compared to the bulk algorithms (Brodeau et al. 2017). The differences between the direct outputs of turbulent heat fluxes and recalculated heat fluxes using air–sea variables in the reanalysis that are based on bulk formulas have been compared in appendix A. In turn, these heat flux products and associated air–sea variables are compared with the collocated buoy observations to help examine the uncertainties in operational simulations and propose suggestions to improve their prediction skills.
b. Bulk formulas to estimate the air–sea turbulent heat fluxes and anomalies
c. The method to determine marine atmospheric boundary (MABL) stability conditions
3. Observational turbulent heat flux anomalies and mechanisms
a. Regional disparity of turbulent heat flux discrepancies in the SCS basin during monsoon onset
Previous studies (e.g., Lau and Nath 2009) have hypothesized and concluded that an increased WS can induce an increase in evaporative LH in the SCS basin during the onset of the summer monsoon, which plays a role in damping the local SST. However, the turbulent heat flux discrepancies (Fig. 1, 10-day mean difference between the dates before and after the onset in 2021) over the SCS basin show significant spatial differences rather than a uniform basin mode. A significant east–west dipole-like pattern of turbulent heat flux discrepancies can be found based on ERA5 data (Fig. 1). This may result from the joint effect of surface wind and sea–air differences in temperature (ΔT) and humidity (Δq) due to active sea–air convection (Fig. 6). The heat flux discrepancies range from −100 to 100 W m−2. The zonal heat flux gradient is approximately 200 W m−2 from 110° to 120°E. This indicates a potential effect of the interannual variability of climate modes, for example, the leading effect of El Niño–Southern Oscillation (ENSO) on the onset of monsoons (Zhu and Li 2017; Martin et al. 2019) or intraseasonal oscillations, according to Chen et al. (2022). However, the relationship between ENSO and SCSSM onset has weakened significantly in recent years (Liu and Zhu 2020; Hu et al. 2020, 2022). The interannual variability in the turbulent heat flux discrepancies during the SCSSM onset, as shown in Fig. 1, is presented and summarized in the discussion section.
The time series of the buoy-based and ship-borne air–sea measurements and estimated turbulent heat fluxes based on the COARE 3.5 algorithm are shown in Figs. 4 and 5, respectively. As the parameter of relative humidity (RH) at buoy D was not available during the observational period, the heat flux variations are not calculated. It is interesting that buoys A, B, and C are located in areas of marginally zero, positive, and negative turbulent heat flux discrepancies (Fig. 1). Evident variations in air–sea variables can be found between the dates prior to and after the monsoon onset on 29 May 2021, with the differences summarized in Table 2.
Time series of hourly air–sea variables from buoys A (red), B (black), C (blue), D (purple), and Bailong (green), including (a) sea surface temperature (SST; unit: °C), (b) surface air temperature (SAT; unit: °C), (c) ΔT (unit: °C), (d) relative humidity (RH; unit: %), (e) sea level pressure (SLP; unit: hPa), (f) Δq (unit: g kg−1), (g) wind speed (WS; unit: m s−1) at a height of 10 m following the Smith (1988) algorithm, (h) latent heat (LH; unit: W m−2), and (i) sensible heat (SH; unit: W m−2). Note that the RH of buoy D is not available during the onset of the summer monsoon. The vertical dashed lines indicate the date of onset of the SCSSM on 29 May 2021, while the horizontal dashes denote the zero values.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Time series of hourly air–sea variables from buoys A (red), B (black), C (blue), D (purple), and Bailong (green), including (a) sea surface temperature (SST; unit: °C), (b) surface air temperature (SAT; unit: °C), (c) ΔT (unit: °C), (d) relative humidity (RH; unit: %), (e) sea level pressure (SLP; unit: hPa), (f) Δq (unit: g kg−1), (g) wind speed (WS; unit: m s−1) at a height of 10 m following the Smith (1988) algorithm, (h) latent heat (LH; unit: W m−2), and (i) sensible heat (SH; unit: W m−2). Note that the RH of buoy D is not available during the onset of the summer monsoon. The vertical dashed lines indicate the date of onset of the SCSSM on 29 May 2021, while the horizontal dashes denote the zero values.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Time series of hourly air–sea variables from buoys A (red), B (black), C (blue), D (purple), and Bailong (green), including (a) sea surface temperature (SST; unit: °C), (b) surface air temperature (SAT; unit: °C), (c) ΔT (unit: °C), (d) relative humidity (RH; unit: %), (e) sea level pressure (SLP; unit: hPa), (f) Δq (unit: g kg−1), (g) wind speed (WS; unit: m s−1) at a height of 10 m following the Smith (1988) algorithm, (h) latent heat (LH; unit: W m−2), and (i) sensible heat (SH; unit: W m−2). Note that the RH of buoy D is not available during the onset of the summer monsoon. The vertical dashed lines indicate the date of onset of the SCSSM on 29 May 2021, while the horizontal dashes denote the zero values.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 4, but for cruise observations. Note that the neutral WS is adjusted from the observational height to 10 m (the same applies hereafter).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 4, but for cruise observations. Note that the neutral WS is adjusted from the observational height to 10 m (the same applies hereafter).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 4, but for cruise observations. Note that the neutral WS is adjusted from the observational height to 10 m (the same applies hereafter).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The statistics of the 10-day mean air–sea variables from different buoys before and after the onset of the SCSSM, including sea surface temperature (SST; unit: °C), surface air temperature (SAT; unit: °C), ΔT (unit: °C), sea level pressure (SLP; unit: hPa), relative humidity (RH; unit: %), wind speed (WS; unit: m s−1), Δq (unit: g kg−1), LH (unit: W m−2), and SH (unit: W m−2).
Three basic features can be found in the time series. First, significant regional disparities in air–sea LH and SH discrepancies exist during the onset of the SCSSM in the SCS. The buoy observations show a consistent tendency with those of ERA5. The evaporation and convection brought about by the establishment of a significant westerly wind and gradual easterly withdrawal of the WPSH (Fig. 2) may be a cause of the spatial discrepancies in WS, Δq, and ΔT presented in Fig. 6. In addition, station buoy B is located in an area with positive WS and turbulent heat flux discrepancies. The observations show increases in both LH and SH with magnitudes of 3.6 and 1.4 W m−2, respectively. In contrast, buoy C is located in the southern part of the SCS, where the discrepancies in WS, T, and turbulent heat flux are negative. Decreases in LH (∼3.9 W m−2) and SH (∼2.8 W m−2) are observed at buoy C. In terms of Station A which sits at the margin between areas of positive and negative WS, Δq, ΔT, and turbulent heat flux discrepancies, the LH after the monsoon onset experiences a decrease of 4.1 W m−2, while the SH experiences a slight increase of 1.5 W m−2. In total, the turbulent heat fluxes (LH+SH) show a decrease of 2.6 W m−2 at buoy A. Moreover, a significantly enhanced LH by 27.6 W m−2 is detected in the ship-borne observations, bearing in mind that the measurements are not fixed locations but cover a large portion of the northern SCS. The SH increases slightly by 1.7 W m−2, which is comparable to the buoy observations.
The discrepancies in (a) surface wind speed (WS; unit: m s−1), (b) sea–air humidity difference (Δq; unit: g kg−1), and (c) sea–air temperature difference (ΔT; unit: °C) for the 10-day average before and after the onset of the monsoon (29 May 2021) obtained from ERA5. Buoys A, B, C, and D for observing air–sea variables are shown as filled black circles.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The discrepancies in (a) surface wind speed (WS; unit: m s−1), (b) sea–air humidity difference (Δq; unit: g kg−1), and (c) sea–air temperature difference (ΔT; unit: °C) for the 10-day average before and after the onset of the monsoon (29 May 2021) obtained from ERA5. Buoys A, B, C, and D for observing air–sea variables are shown as filled black circles.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The discrepancies in (a) surface wind speed (WS; unit: m s−1), (b) sea–air humidity difference (Δq; unit: g kg−1), and (c) sea–air temperature difference (ΔT; unit: °C) for the 10-day average before and after the onset of the monsoon (29 May 2021) obtained from ERA5. Buoys A, B, C, and D for observing air–sea variables are shown as filled black circles.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Second, the variations in meteorological variables are also evident. Except for buoy station C, the other observations (from the cruise and buoys A and B) show an increase in RH but a decrease in both SST and surface air temperature (SAT). However, the sea–air temperature differences indicate few variations in association with the onset of the SCSSM. Due to the increasing RH and unchanged temperature gradient, Δq shows a slight decrease during the onset of the monsoon. The increased LH primarily results from the enhancement of the WS. The results are almost the opposite of those at station C, which is not presented here to avoid repetition. Third, there are significant high-frequency fluctuations in LH, SH, and associated air–sea variables, particularly at the buoy stations (Fig. 4). Considerable increases can be seen in some of the standard deviations (STDs) of the variables and heat fluxes; for example, the STDs of LH are nearly twice as high as before onset at buoy A. In addition, the STDs are evident compared to the mean magnitudes, as shown in Table 2. The SST, SAT, RH, WS, LH, and SH variables indicate a “white noise”–like spectrum, as the variability in these variables may be closely associated with intraseasonal oscillations, quasi-biweekly oscillations, synoptic weather processes, and diurnal variations. No specific energy peak is found in terms of the spectral analysis (not shown here).
b. Explanations of the turbulent heat flux anomalies under different MABL stability conditions
Before the mechanism of turbulent heat flux anomalies is investigated, the MABL stability is first calculated and categorized using Eq. (4). Figure 7 lists the percentiles of different MABL stability conditions during the onset of the SCSSM, totaling 20 days before and after the onset of the monsoon. The dominant near-neutral boundary condition is well understood (gray bars in Fig. 7). Thus, in this section, we show the high-resolution turbulent heat flux anomalies under continuous near-neutral and mixed (near-neutral plus unstable) boundary conditions during the onset of the monsoon at different locations.
The percentiles of different MABL stability conditions in the 10 days before and after the onset of the SCSSM (20 days in total).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The percentiles of different MABL stability conditions in the 10 days before and after the onset of the SCSSM (20 days in total).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The percentiles of different MABL stability conditions in the 10 days before and after the onset of the SCSSM (20 days in total).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Figure 8 shows the turbulent heat flux anomalies at station A in association with the effects of winds, sea–air difference and nonlinear terms, as shown in Eq. (3). Before the onset of the monsoon, the MABL stability is continuously near neutral, and sea–air humidity differences play roles in determining the LH anomalies, with a correlation coefficient (r) of 0.8 and root-mean-square (RMS) of 8.37 W m−2. The role of the WS anomalies is weak, and the nonlinear effect is negligible. Due to the gradual eastward withdrawal of the WPSH (Fig. 2), the local wind of station A intensifies and shifts from westerly to southwesterly at the margins of the WPSH. The slight thermal difference caused by significant wind shifts may be one of the reasons for the MABL stability remaining purely near-neutral after the onset of the SCSSM. With a mixed MABL stability (in particular after 5 June), the LH anomalies are still dominated by the humidity effect (Δq′) with a second contribution of the nonlinear effect, as shown in Table 3. Due to the significant variations in WS, the wind anomalies also have an obvious impact on LH anomalies during the quick change in MABL stability conditions. However, the anomalies in the sea–air temperature difference (ΔT′) almost entirely dominate the SH anomalies regardless of the MABL stability, with the most significant r and the lowest RMS (Table 3). Similar SH findings can also be obtained at other buoy stations but are not shown here to avoid repetition.
(a) The time series of LH anomalies (blue), LH anomalies induced by wind anomalies (red), air–sea humidity difference (purple), and the nonlinear term (green) from buoy A under near-neutral MABL stability conditions. (b) As in (a), but for anomalies of the SH and its associated terms. (c),(d) As in (a) and (b), but for anomalies under mixed MABL stability. The black line represents the Monin–Obukhov stability parameter ζ = z/L, with ζ = 0.1 and −0.4 denoted by horizontal dashed lines. Note that ζ > 0.1, ζ < −0.4, and −0.4 < ζ < 0.1 indicate the stable, unstable, and near-neutral boundary conditions, respectively.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
(a) The time series of LH anomalies (blue), LH anomalies induced by wind anomalies (red), air–sea humidity difference (purple), and the nonlinear term (green) from buoy A under near-neutral MABL stability conditions. (b) As in (a), but for anomalies of the SH and its associated terms. (c),(d) As in (a) and (b), but for anomalies under mixed MABL stability. The black line represents the Monin–Obukhov stability parameter ζ = z/L, with ζ = 0.1 and −0.4 denoted by horizontal dashed lines. Note that ζ > 0.1, ζ < −0.4, and −0.4 < ζ < 0.1 indicate the stable, unstable, and near-neutral boundary conditions, respectively.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
(a) The time series of LH anomalies (blue), LH anomalies induced by wind anomalies (red), air–sea humidity difference (purple), and the nonlinear term (green) from buoy A under near-neutral MABL stability conditions. (b) As in (a), but for anomalies of the SH and its associated terms. (c),(d) As in (a) and (b), but for anomalies under mixed MABL stability. The black line represents the Monin–Obukhov stability parameter ζ = z/L, with ζ = 0.1 and −0.4 denoted by horizontal dashed lines. Note that ζ > 0.1, ζ < −0.4, and −0.4 < ζ < 0.1 indicate the stable, unstable, and near-neutral boundary conditions, respectively.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Correlation coefficients (r) and root-mean-squares (RMSs; unit: W m−2) of LH/SH anomalies and term B (wind anomalies), term C (sea–air difference in humidity/temperature), and term C (nonlinear effect) on the right-hand side of Eq. (3) under different MABL conditions from buoys A and B. Note that the r values that do not pass the 95% confidence test are shown in italic font. The most significant terms are shown in boldface font for readability.
Figure 9 shows similar information as shown in Fig. 8; however, it indicates the LH anomalies and their associated effects before and after the onset of the SCSSM at station B. At station B, the LH anomalies are mainly determined by anomalies in Δq′ under mixed MABL stability in the 10 days before and after the monsoon onset. Nevertheless, the WS anomalies and LH anomalies show good consistency when the MABL stability transitions from near-neutral to unstable conditions (Table 3, Figs. 9a,b). In addition, the LH anomalies are also controlled by Δq′ with an r of 0.94 and RMS of 9.37 W m−2 under continuous near-neutral boundary conditions even after the onset of the monsoon. This may result from the weak surface current, which contributes to relatively stable stratification. Similar results to those at stations A and B can also be found at station C but are not repeated here.
As in Fig. 8, but for LH at station B.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 8, but for LH at station B.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 8, but for LH at station B.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Thus, the LH anomalies strongly depend on the MABL stability. When the boundary condition shifts between near-neutral and unstable conditions, the high-resolution (hourly) variations in LH are determined by the WS. Fluctuations in WS tend to remove water vapor from the sea surface, which helps enhance evaporation and establish a new balance for LH. When the MABL remains unchanged, the hourly variations in LH during the onset of the monsoon are mainly determined by the anomalies in sea–air humidity differences. These results have been summarized based on three full-parameter air–sea buoys, indicating little dependence on the background LH discrepancies (buoys A–C are located at different locations with different LH discrepancies, see Fig. 1). However, the sea–air temperature differences mainly account for the high-resolution SH variations, which are independent of the MABL stability. The identification of the dominant term of the LH and SH anomalies is beneficial for improving the parameterization scheme for the model, with a view to better simulating turbulent heat fluxes during the monsoon.
4. Comparisons between reanalysis and air–sea measurements
This section compares the turbulent heat fluxes and associated air–sea parameters between the reanalysis and the observations during the onset of the SCSSM, which helps identify potential areas for improvement in operational forecasting systems for monsoon predictions and other predictions.
a. Turbulent heat flux
Figures 10 and 11 show the differences in LH and SH (including the fluxes recalculated using the air–sea variables based on the COARE algorithm) between the reanalysis and observations. The time series of the LH and SH differences (Fig. 10) from the beginning of May to the end of June indicate significant overestimates of turbulent heat fluxes during the onset of the SCSSM. Both the ERA5 and MERRA2 products show evident overestimates of LH during the 2-month observations. Before the monsoon onset, the overestimations of LH in both reanalysis datasets (45 W m−2 for ERA5 and 78 W m−2 for MERRA2) are lower than those (47 W m−2 for ERA5 and 80 W m−2 for MERRA2) after the monsoon. Similar results can also be found for the recalculated fluxes. All three stations A, B, and C and the cruise show overestimations, but they do not show consistent overestimation tendencies. The overestimation at buoy A is in a stable stage, and there is a sharp drop and then a sharp rise approximately 10 days after the onset. The overestimation at buoy B increases and then remains higher than before the onset. However, the overestimation of buoy C is relatively low compared to the periods before the onset. In addition, an obvious negative tendency is seen after the onset from cruise observations. For the SH, the overestimation is also slightly higher after the monsoon burst in the ERA5 data; however, the opposite results are found in the MERRA2 data. The SH differences in MERRA2 exhibit significant fluctuations and negative peaks after the onset of the monsoon, which eventually result in an average underestimation. The recalculated results are similar to those of the fluxes that are directly estimated in the reanalysis, which are analyzed in appendix A. The results from the other buoys and the cruise are the same as those at station B but are not shown here to avoid repetition.
The LH (SH) differences between the reanalysis (including direct output and recalculated fluxes in terms of the COARE algorithm) and observations at station B. (a),(c) The LH results and (b),(d) the SH differences. The results of ERA5 and MERRA2 are marked in blue and red, respectively. The vertical dashed lines represent the burst dates of the SCSSM.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The LH (SH) differences between the reanalysis (including direct output and recalculated fluxes in terms of the COARE algorithm) and observations at station B. (a),(c) The LH results and (b),(d) the SH differences. The results of ERA5 and MERRA2 are marked in blue and red, respectively. The vertical dashed lines represent the burst dates of the SCSSM.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The LH (SH) differences between the reanalysis (including direct output and recalculated fluxes in terms of the COARE algorithm) and observations at station B. (a),(c) The LH results and (b),(d) the SH differences. The results of ERA5 and MERRA2 are marked in blue and red, respectively. The vertical dashed lines represent the burst dates of the SCSSM.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The mean differences in (a),(c) LH and (b),(d) SH between the reanalysis and observations (x axis). The bars with solid (dashed) lines represent the mean differences averaged over 10 days before (after) the onset of the monsoon. Note the different -axes for LH and SH. Unfortunately, the Bailong buoy does not observe air–sea fluxes after the onset of the monsoon due to heavy fishery activities.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The mean differences in (a),(c) LH and (b),(d) SH between the reanalysis and observations (x axis). The bars with solid (dashed) lines represent the mean differences averaged over 10 days before (after) the onset of the monsoon. Note the different -axes for LH and SH. Unfortunately, the Bailong buoy does not observe air–sea fluxes after the onset of the monsoon due to heavy fishery activities.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The mean differences in (a),(c) LH and (b),(d) SH between the reanalysis and observations (x axis). The bars with solid (dashed) lines represent the mean differences averaged over 10 days before (after) the onset of the monsoon. Note the different -axes for LH and SH. Unfortunately, the Bailong buoy does not observe air–sea fluxes after the onset of the monsoon due to heavy fishery activities.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Figure 11 summarizes the mean results of the differences in LH and SH over the 10 days before and after the onset of the monsoon (29 May 2021) between both the reanalysis and the observations. Overestimation of LH (∼56 W m−2 for ERA5 and ∼54 W m−2 for MERRA2) is confirmed at all buoy locations with magnitudes ranging from 18 to 79 W m−2, even though they are sited in regions with different discrepancies in turbulent heat fluxes (Fig. 1). Some underestimation (up to 9 W m−2) can be seen in the cruise trajectory. At station B, the differences in LH are enhanced after the monsoon burst; however, at other stations, the results are the opposite. Similar overestimations can also be found for SH (∼4 W m−2 for ERA5 and ∼2 W m−2 for MERRA2), except in the cruise observations. Considering the impact of the dramatic oscillations in the differences and the insufficient duration of cruise observations of the mean differences following the onset, we suggest that there are overestimations of LH and SH over the entire SCS basin, regardless of the onset of the monsoon.
b. Comparison of associated air–sea variables
To better examine the sources of the overestimated LH and SH in the SCS basin, the associated air–sea variables are further compared. Using station B as an example (Fig. 12), higher SST and lower SAT result in a higher sea–air temperature difference (ΔT), with additional peak exceptions for MERRA2. As the sea–air temperature differences ΔT dominate the variations in SH (section 3b), an overestimation of ΔT (∼0.4°C for ERA5 and ∼0.1°C for MERRA2) results in a higher estimate of SH. It should also be emphasized that the differences in WS between the reanalysis and observations are not significant during the two-month observations, ranging from −7 to 7 m s−1, which plays a secondary role in determining the SH differences. Similarly, the higher near-surface specific humidity (qs) based on the Buck equation (Buck 1981) and the lower specific air humidity (qa) also cause overestimates of sea–air humidity differences Δq (2 g kg−1 for both ERA5 and MERRA2), which accounts for the higher estimate of LH in both reanalysis datasets. The underestimation of RH by ∼8% can be found over the 2-month observations for both reanalysis products, which highlights the importance of RH in accurately calculating air–sea heat fluxes (Yu 2019; Cronin et al. 2019). Similar results can also be obtained at the other buoy stations but are not shown here.
As in Fig. 10, but for (a) SST (unit: °C), (b) SAT (unit: °C), (c) sea–air temperature difference (unit: °C), (d) near-surface and (e) air specific humidity and the (f) Δq differences (unit: g kg−1), (g) WS (unit: m s−1), (h) SLP (unit: hPa), and (i) RH (unit: %).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 10, but for (a) SST (unit: °C), (b) SAT (unit: °C), (c) sea–air temperature difference (unit: °C), (d) near-surface and (e) air specific humidity and the (f) Δq differences (unit: g kg−1), (g) WS (unit: m s−1), (h) SLP (unit: hPa), and (i) RH (unit: %).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 10, but for (a) SST (unit: °C), (b) SAT (unit: °C), (c) sea–air temperature difference (unit: °C), (d) near-surface and (e) air specific humidity and the (f) Δq differences (unit: g kg−1), (g) WS (unit: m s−1), (h) SLP (unit: hPa), and (i) RH (unit: %).
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Figures 13 and 14 show the 10-day mean results of differences in associated air–sea variables at all the buoy stations, indicating why LH and SH are overestimated, as shown in the above results. In the mean states, a significantly higher sea–air temperature difference (∼0.7°C for ERA5 and ∼0.3°C for MERRA2) causes the overestimate of the SH shown in Fig. 9, as the SH anomalies are mainly induced by the anomalies in ΔT regardless of the MABL states and the time of the monsoon onset. The RH is significantly underestimated in the reanalysis (4.5% for ERA5 and 3.8% for MERRA2), which contributes to the overestimation of the sea–air differences. The overestimated near-surface specific humidity due to higher SST and reduced air specific humidity results in the overestimation of the sea–air humidity difference (∼2.0 g kg−1 for ERA5 and ∼1.5 g kg−1 for MERRA2), which leads to a higher estimate of LH, as the sea–air humidity difference also plays a central role in determining LH anomalies, particularly under near-neutral MABL conditions. The differences in WS and SLP do not show particular trends depending on the different buoy locations.
As in Fig. 11, but for SST, SAT, and sea–air temperature differences.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 11, but for SST, SAT, and sea–air temperature differences.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 11, but for SST, SAT, and sea–air temperature differences.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Figs. 11 and 13, but for RH, sea–air humidity differences, WS, and SLP.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Figs. 11 and 13, but for RH, sea–air humidity differences, WS, and SLP.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Figs. 11 and 13, but for RH, sea–air humidity differences, WS, and SLP.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
c. Taylor plots to identify the ability of model simulations
Figures 15 and 16 show the Taylor plots comprising standard deviations (STDs), root-mean-squares (RMSs), and correlation coefficients (CCs) between the time series of the daily mean reanalysis products and the air–sea measurements at different locations from May to June 2021. To compare the simulation ability at different locations, the magnitudes of the above statistical parameters are normalized by the STDs of the observational variables. Interestingly, the LH from ERA5 and MERRA2 indicate equivalent simulation abilities at all buoy and cruise locations (Fig. 15a), bearing in mind the overestimation for the turbulent heat fluxes in the reanalysis that is shown in section 4b. However, the time series of SH in MERRA2 are slightly more consistent with the observations than those in ERA5. For the recalculated SH-based reanalysis variables, the results are almost the same as those directly output by the operational systems. However, the recalculated LH for MERRA2 using the COARE 3.5 algorithm shows slight improvement over ERA5, particularly at station B. This indicates the effect of the different algorithms on the eventual heat fluxes, which mainly accounts for the global heat budget imbalance (Yu 2019).
Taylor diagrams showing three statistical properties of (a),(c) LH and (b),(d) SH comparisons: the standard deviations (STDs), root-mean-squares (RMSs), and correlation coefficients (CCs) of the differences between the two reanalysis products and air–sea buoy and cruise observations. Note that the reanalysis LH and SH in (c) and (d) were recalculated based on the COARE 3.5 algorithm. As the SST in the reanalysis does not resolve the diurnal cycles, the daily mean variables are used for comparison here. Note that the magnitudes are all normalized by the observational value itself.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Taylor diagrams showing three statistical properties of (a),(c) LH and (b),(d) SH comparisons: the standard deviations (STDs), root-mean-squares (RMSs), and correlation coefficients (CCs) of the differences between the two reanalysis products and air–sea buoy and cruise observations. Note that the reanalysis LH and SH in (c) and (d) were recalculated based on the COARE 3.5 algorithm. As the SST in the reanalysis does not resolve the diurnal cycles, the daily mean variables are used for comparison here. Note that the magnitudes are all normalized by the observational value itself.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Taylor diagrams showing three statistical properties of (a),(c) LH and (b),(d) SH comparisons: the standard deviations (STDs), root-mean-squares (RMSs), and correlation coefficients (CCs) of the differences between the two reanalysis products and air–sea buoy and cruise observations. Note that the reanalysis LH and SH in (c) and (d) were recalculated based on the COARE 3.5 algorithm. As the SST in the reanalysis does not resolve the diurnal cycles, the daily mean variables are used for comparison here. Note that the magnitudes are all normalized by the observational value itself.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 15, but for (a) SST, (b) SAT, (c) sea–air temperature difference, (d) RH, (e) WS, and (f) sea–air humidity difference.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 15, but for (a) SST, (b) SAT, (c) sea–air temperature difference, (d) RH, (e) WS, and (f) sea–air humidity difference.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. 15, but for (a) SST, (b) SAT, (c) sea–air temperature difference, (d) RH, (e) WS, and (f) sea–air humidity difference.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Figure 15 summarizes the Taylor comparisons for associated variables to better point out the potential aspects for improvement of operating systems. The two reanalysis datasets incorporate high-resolution daily mean SST products: ERA5 assimilates the Met Office Operational SST and Sea Ice Analysis (OSTIA), while MERRA2 uses 1/48 optimum interpolation SST (OISST) products from NOAA (Reynolds et al. 2007) as the boundary conditions. Similar to the LH, the SST products are consistent with the in situ observations (Fig. 16a), which can also be found in SAT, sea–air temperature differences, and other variables, such as RH. Thus, it can be concluded that both reanalysis products show equivalent abilities to reproduce synoptic processes in association with turbulent heat fluxes and can be used for studies of multiscale air–sea interactions.
5. Summary and discussion
On the basis of the SCSMEX during 1996–2001 (Ding et al. 2004), this study further examined the air–sea turbulent heat flux anomalies during monsoon onset using comprehensive air–sea measurements. In contrast to the previous observational program, our observations focus on the specific processes at the air–sea interface. The major findings can be summarized as follows: First, the buoy and cruise observations further confirm a significant regional variability in weather-scale turbulent heat flux discrepancies, rather than a uniform pattern with enhanced WS during the monsoon onset. Second, the high-resolution (hourly) variations in LH and SH are determined by different air–sea processes under different MABL conditions. SH variations are dominated by sea–air temperature differences regardless of the boundary conditions or the monsoon onset. The role of WS is minor, although its enhancement is a major characteristic of monsoon. However, the LH anomalies are determined by sea–air humidity differences under near-neutral MABL stability conditions with secondary wind effects, while they are mainly controlled by the wind anomalies under changing MABL conditions. It should be emphasized here that the SH values are much weaker than those of LH, with a Bowen ratio of ∼0.1. Third, the observed variations in LH and SH are compared to those in the high-resolution reanalysis products ERA5 and MERRA2, which can help identify simulation problems and improve prediction skills. Both ERA5 and MERRA2 show overestimates of LH and SH during the onset of the SCSSM. The sea–air humidity and temperature differences account for these higher estimates. The higher SST and lower SAT cause higher ΔT, inducing a higher SH. Higher near-surface humidity and lower specific air humidity in terms of lower estimates of RH and SAT contribute to higher Δq and LH values.
It should be noted that this study provides an observational case for 2021 only. The onset of the SCSSM experiences significant interannual variability (Chen et al. 2022), which depends on different climate modes. Figure 17 shows the turbulent heat flux discrepancies (as shown in Fig. 1) between the 10-day mean before and after the monsoon bursts from 2002 to 2021. In some years, such as the La Niña years (La Niña event in the previous winters) of 2006 and 2018, the positive turbulent heat flux discrepancies indicate a uniform basin mode. A similar basin mode can also be found in 2015 following an El Niño event. In other years, zonal or meridional dipole patterns exist, indicating year-to-year anomaly structures. Why the patterns of heat flux discrepancies are different from year to year will be examined in future studies, as sufficient air–sea measurements are accumulated in the SCS. However, this study explores the mechanism of turbulent heat flux discrepancies in association with variations in air–sea parameters against the background of monsoon bursts. This provides scientific evidence and opportunities to further investigate the role of air–sea interactions during the onset of the SCSSM.
Mean air–sea turbulent heat flux (LH + SH) differences (unit: W m−2) for the 10 days before and after monsoon onset from (a) 2002 to (t) 2021. Hourly heat flux data for these 20 years are obtained from ERA5. The mean SLP on the monsoon onset day for each year is incorporated by black contours. Triangles highlight El Niño events, and circles represent a La Niña event in the previous winter.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Mean air–sea turbulent heat flux (LH + SH) differences (unit: W m−2) for the 10 days before and after monsoon onset from (a) 2002 to (t) 2021. Hourly heat flux data for these 20 years are obtained from ERA5. The mean SLP on the monsoon onset day for each year is incorporated by black contours. Triangles highlight El Niño events, and circles represent a La Niña event in the previous winter.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Mean air–sea turbulent heat flux (LH + SH) differences (unit: W m−2) for the 10 days before and after monsoon onset from (a) 2002 to (t) 2021. Hourly heat flux data for these 20 years are obtained from ERA5. The mean SLP on the monsoon onset day for each year is incorporated by black contours. Triangles highlight El Niño events, and circles represent a La Niña event in the previous winter.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
This study is mainly focused on the air–sea turbulent heat flux anomalies from the perspective of atmospheric climate variations. During the onset of the SCSSM, the large-scale wind fields over the SCS intensify, which increases the momentum transfer between the atmosphere and the ocean through wind stress and generates significant mixing at the surface, leading to an increase in the turbulent heat flux (Lau and Nath 2009). Moreover, the increase in the humidity gradient can enhance the turbulent heat and moisture fluxes (Ding et al. 2004). Due to intrinsic ocean variability, the atmospheric stability and buoyancy over eddies are affected, modulating vertical eddy pumping and mixing and resulting in SST anomalies (Frenger et al. 2013; Hayes et al. 1989; Wallace et al. 1989). Variations in SST affect the SH and LH by influencing the air–sea temperature gradient and the surface saturation humidity, respectively, within mesoscale eddies (e.g., Villas Bôas et al. 2015). In addition, SST anomalies that induce MABL pressure and turbulence fields cause surface wind perturbations (O’Neill et al. 2012). Surface roughness features are affected and lead to changes in turbulent heat fluxes. However, the influence of eddies is a regional air–sea process compared to the large-scale atmosphere, so this study focuses on the contribution of the South China Sea monsoon to turbulent heat fluxes. Multiscale ocean dynamics (fronts and meso- and submesoscale processes) that impact the air–sea turbulent heat flux anomalies have already been analyzed using observations (Song et al. 2022, 2023) across a cyclonic eddy during the 2021 cruise, but the air–sea fluxes that occur on short temporal and spatial scales are still expected to be separated using long-term in situ and satellite observations in the future, as summarized by Cronin et al. (2019).
This study provides additional comparative evidence of the overestimation of turbulent heat flux, which shows the same findings as those of another observational study on extreme air–sea heat fluxes during the passage of tropical cyclones (Song et al. 2021). These results can help examine the operational simulating abilities in reanalysis, which help improve the heat flux simulations and their roles in climate predictions, for example, monsoon forecasting in the SCS. Finally, regional enhancement of the climate monitoring network in terms of international partnerships (Cronin et al. 2022) should be required for better ocean climate prediction and sustainable development in nearby countries.
Acknowledgments.
This study is supported by the National Natural Science Foundation of China (42122040 and 42076016). Special thanks to SOA for providing the in situ buoy observations. This study could not have been conducted without observational support from the SOA. The authors acknowledge the cruise data provided by the R/V TAN KAH KEE. The authors appreciate the constructive comments from the anonymous reviewers.
Data availability statement.
The new reanalysis of the ERA5 and MERRA2 heat flux data and their associated variables can be obtained at https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5 and https://gmao.gsfc.nasa.gov/, respectively. The air–sea observations in this study are available at http://www.ocean.iap.ac.cn/pages/dataService/dataService.html?navAnchor=dataService. The data can be easily found in the Air–sea Heat Flux folder.
APPENDIX A
Comparisons of the Turbulent Heat Fluxes between the Reanalysis and the Recalculated Fluxes Based on the COARE Algorithm
Using the observations at stations A and B, Fig. A1 shows the scatterplots of the differences in the turbulent heat fluxes between the direct output from the reanalysis (ERA5 and MERRA2) and the recalculated fluxes using air–sea variables in the reanalysis based on the COARE algorithm. For ERA5, the direct output fluxes and the recalculated fluxes are essentially equivalent, with a few larger values for the recalculated fluxes. However, the direct output from MERRA is slightly larger than the recalculated fluxes. Both the RMSE and STD are in a reasonable and acceptable range. We find a coincidence (Figs. A2, A3); that is, the difference in turbulent heat fluxes between reanalysis products and recalculations, in particular from MERRA2, is closely related to the variations in the MABL stability. The deviation between ERA5 and recalculated ERA5 tends to fluctuate correspondingly when the MABL switches, which addresses the large deviation values in Fig. A1. Significant consistency can be found in the difference between the MERRA2 and recalculations with the MABL stability. Larger deviation values occur under near-neutral boundary layer conditions. It should be emphasized that these algorithms and schemes are different from those of the bulk formulas (Brodeau et al. 2017; Bonino et al. 2022), inducing an approximately 6% (2% for ERA5 and 10% for MERRA2) difference compared to the bulk algorithms (∼10%, Brodeau et al. 2017) at buoy stations A and B.
The hourly scatterplots of (a),(c),(e),(g) LH and (b),(d),(f),(h) SH between the reanalysis products (a),(b),(e),(f) ERA5; (c),(d),(g),(h) MERRA2; and recalculated fluxes based on the COARE algorithm at (a)–(d) buoy A and (e)–(h) buoy B. The RMSE and STD results are incorporated into the panels. There are 1464 points in each subplot.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The hourly scatterplots of (a),(c),(e),(g) LH and (b),(d),(f),(h) SH between the reanalysis products (a),(b),(e),(f) ERA5; (c),(d),(g),(h) MERRA2; and recalculated fluxes based on the COARE algorithm at (a)–(d) buoy A and (e)–(h) buoy B. The RMSE and STD results are incorporated into the panels. There are 1464 points in each subplot.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
The hourly scatterplots of (a),(c),(e),(g) LH and (b),(d),(f),(h) SH between the reanalysis products (a),(b),(e),(f) ERA5; (c),(d),(g),(h) MERRA2; and recalculated fluxes based on the COARE algorithm at (a)–(d) buoy A and (e)–(h) buoy B. The RMSE and STD results are incorporated into the panels. There are 1464 points in each subplot.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Time series of differences in LH (blue) and SH (red) between ERA5 and recalculations based on the COARE algorithm from (a),(b) buoy A and (c),(d) buoy B. The Monin–Obukhov stability parameter ζ = z/L, with ζ = 0.1 and −0.4 denoted by horizontal dashed lines. The vertical black dashed line indicates the SCS monsoon onset (29 May 2021), while the top horizontal dashed line indicates zero values.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Time series of differences in LH (blue) and SH (red) between ERA5 and recalculations based on the COARE algorithm from (a),(b) buoy A and (c),(d) buoy B. The Monin–Obukhov stability parameter ζ = z/L, with ζ = 0.1 and −0.4 denoted by horizontal dashed lines. The vertical black dashed line indicates the SCS monsoon onset (29 May 2021), while the top horizontal dashed line indicates zero values.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
Time series of differences in LH (blue) and SH (red) between ERA5 and recalculations based on the COARE algorithm from (a),(b) buoy A and (c),(d) buoy B. The Monin–Obukhov stability parameter ζ = z/L, with ζ = 0.1 and −0.4 denoted by horizontal dashed lines. The vertical black dashed line indicates the SCS monsoon onset (29 May 2021), while the top horizontal dashed line indicates zero values.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. A2, but for MERRA2.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. A2, but for MERRA2.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
As in Fig. A2, but for MERRA2.
Citation: Monthly Weather Review 151, 9; 10.1175/MWR-D-22-0314.1
APPENDIX B
List of Abbreviations
| CCs |
Correlation coefficients |
| CMA |
China Meteorological Administration |
| ECMWF |
European Centre for Medium-Range Weather Forecasts |
| ENSO |
El Niño–Southern Oscillation |
| ERA5 |
Fifth major global reanalysis produced by ECMWF |
| LH |
Latent heat |
| MABL |
Marine atmospheric boundary layer |
| MERRA2 |
Modern-Era Retrospective Analysis for Research and Applications, version 2 |
| OISST |
Optimum interpolation SST |
| OSTIA |
Operational SST and Sea Ice Analysis |
| RH |
Relative humidity |
| RMS |
Root-mean-square |
| SAT |
Surface air temperature |
| SCS |
South China Sea |
| SCSMEX |
South China Sea Monsoon Experiment |
| SCSSM |
South China Sea summer monsoon |
| SH |
Sensible heat |
| SLP |
Sea level pressure |
| SOA |
State Oceanic Administration |
| SST |
Sea surface temperature |
| STD |
Standard deviation |
| WNP |
Western North Pacific |
| WPSH |
Western Pacific subtropical high |
| WS |
Wind speed |
| Δq |
Sea–air humidity difference |
| ΔT |
Sea–air temperature difference |
| Δq′ |
Anomalies in the sea–air humidity difference |
| ΔT′ |
Anomalies in the sea–air temperature difference |
| qs |
Near-surface specific humidity |
| qa |
Air specific humidity |
REFERENCES
Bell, B., and Coauthors, 2021: The ERA5 global reanalysis: Preliminary extension to 1950. Quart. J. Roy. Meteor. Soc., 147, 4186–4227, https://doi.org/10.1002/qj.4174.
Bombardi, R. J., J. L. Kinter, and O. W. Frauenfeld, 2019: A global gridded dataset of the characteristics of the rainy and dry seasons. Bull. Amer. Meteor. Soc., 100, 1315–1328, https://doi.org/10.1175/BAMS-D-18-0177.1.
Bombardi, R. J., V. Moron, and J. S. Goodnight, 2020: Detection, variability, and predictability of monsoon onset and withdrawal dates: A review. Int. J. Climatol., 40, 641–667, https://doi.org/10.1002/joc.6264.
Bonino, G., D. Iovino, L. Brodeau, and S. Masina, 2022: The bulk parameterizations of turbulent air–sea fluxes in NEMO4: The origin of sea surface temperature differences in a global model study. Geosci. Model Dev., 15, 6873–6889, https://doi.org/10.5194/gmd-15-6873-2022.
Brodeau, L., B. Barnier, S. K. Gulev, and C. Woods, 2017: Climatologically significant effects of some approximations in the bulk parameterizations of turbulent air–sea fluxes. J. Phys. Oceanogr., 47, 5–28, https://doi.org/10.1175/JPO-D-16-0169.1.
Brunke, M. A., Z. Wang, X. Zeng, M. Bosilovich, and C.-L. Shie, 2011: An assessment of the uncertainties in ocean surface turbulent fluxes in 11 reanalysis, satellite-derived, and combined global datasets. J. Climate, 24, 5469–5493, https://doi.org/10.1175/2011JCLI4223.1.
Buck, A. L., 1981: New equations for computing vapor pressure and enhancement factor. J. Appl. Meteor., 20, 1527–1532, https://doi.org/10.1175/1520-0450(1981)020<1527:NEFCVP>2.0.CO;2.
Chen, H., and L. Wang, 2023: Mechanism of the late summer monsoon onset in Bay of Bengal and South China Sea during 2021: Impacts of intraseasonal oscillation and local thermal contrast. Dyn. Atmos. Oceans, 101, 101348, https://doi.org/10.1016/j.dynatmoce.2022.101348.
Chen, L., and Y. Wang, 1998: Numerical experiment of impact of sea surface temperature in South China Sea and warm pool of West Pacific on East Asian summer monsoon. Acta Meteor. Sin., 12, 162–176.
Chen, W., P. Hu, and J. Huangfu, 2022: Multi-scale climate variations and mechanisms of the onset and withdrawal of the South China Sea summer monsoon. Sci. China Earth Sci., 65, 1030–1046, https://doi.org/10.1007/s11430-021-9902-5.
Cronin, M. F., and Coauthors, 2019: Air-sea fluxes with a focus on heat and momentum. Front. Mar. Sci., 6, 430, https://doi.org/10.3389/fmars.2019.00430.
Cronin, M. F., and Coauthors, 2022: Developing an Observing Air-Sea Interactions Strategy (OASIS) for the global ocean. ICES J. Mar. Sci., 80, 367–373, https://doi.org/10.1093/icesjms/fsac149.
Ding, S. Y., Z. Wen, and W. Chen, 2016: Interdecadal change in the relationship between the South China Sea summer monsoon onset and two types of Pacific Sea surface temperature anomaly. Chin. J. Atmos. Sci., 40, 243–256, https://doi.org/10.3878/j.issn.1006-9895.1507.15102.
Ding, Y., 1992: Summer monsoon rainfalls in China. J. Meteor. Soc. Japan, 70, 373–396, https://doi.org/10.2151/jmsj1965.70.1B_373.
Ding, Y., and J. C. L. Chan, 2005: The East Asian summer monsoon: An overview. Meteor. Atmos. Phys., 89, 117–142, https://doi.org/10.1007/s00703-005-0125-z.
Ding, Y., C. Li, and Y. Liu, 2004: Overview of the South China Sea Monsoon Experiment. Adv. Atmos. Sci., 21, 343–360, https://doi.org/10.1007/BF02915563.
Ding, Y., D. Si, Y. Liu, Z. Wang, Y. Li, L. Zhao, and Y. Song, 2018: On the characteristics, driving forces and inter-decadal variability of the East Asian summer monsoon. Chin. J. Atmos. Sci., 42, 533–558, https://doi.org/10.3878/j.issn.1006-9895.1712.17261.
ECMWF, 2021: IFS Documentation Cy47R3—Part IV: Physical processes. ECMWF Rep. 4, 232 pp., https://doi.org/10.21957/eyrpir4vj.
Edson, J. B., and Coauthors, 2013: On the exchange of momentum over the open ocean. J. Phys. Oceanogr., 43, 1589–1610, https://doi.org/10.1175/JPO-D-12-0173.1.
Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young, 1996: Bulk parameterization of air–sea fluxes for tropical ocean global atmosphere Coupled-Ocean Atmosphere Response Experiment. J. Geophys. Res., 101, 3747–3764, https://doi.org/10.1029/95JC03205.
Fairall, C. W., E. F. Bradley, J. E. Hare, A. A. Grachev, and J. B. Edson, 2003: Bulk parameterization of air–sea fluxes: Updates and verification for the COARE algorithm. J. Climate, 16, 571–591, https://doi.org/10.1175/1520-0442(2003)016<0571:BPOASF>2.0.CO;2.
Frenger, I., N. Gruber, R. Knutti, and M. Münnich, 2013: Imprint of Southern Ocean eddies on winds, clouds and rainfall. Nat. Geosci., 6, 608–612, https://doi.org/10.1038/ngeo1863.
Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-0758.1.
Gleckler, P. J., and B. C. Weare, 1997: Uncertainties in global ocean surface heat flux climatologies derived from ship observations. J. Climate, 10, 2764–2781, https://doi.org/10.1175/1520-0442(1997)010<2764:UIGOSH>2.0.CO;2.
Hayes, S. P., M. J. McPhaden, and J. M. Wallace, 1989: The influence of sea surface temperature on surface wind in the eastern equatorial Pacific: Weekly to monthly variability. J. Climate, 2, 1500–1506, https://doi.org/10.1175/1520-0442(1989)002<1500:TIOSST>2.0.CO;2.
He, J., and Z. Zhu, 2015: The relation of South China Sea monsoon onset with the subsequent rainfall over the subtropical East Asia. Int. J. Climatol., 35, 4547–4556, https://doi.org/10.1002/joc.4305.
Helfand, H. M., and S. D. Schubert, 1995: Climatology of the simulated Great Plains low-level jet and its contribution to the continental moisture budget of the United States. J. Climate, 8, 784–806, https://doi.org/10.1175/1520-0442(1995)008<0784:COTSGP>2.0.CO;2.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hu, P., W. Chen, S. Chen, Y. Liu, and R. Huang, 2020: Extremely early summer monsoon onset in the South China Sea in 2019 following an El Niño event. Mon. Wea. Rev., 148, 1877–1890, https://doi.org/10.1175/MWR-D-19-0317.1.
Hu, P., W. Chen, S. Chen, L. Wang, and Y. Liu, 2022: The weakening relationship between ENSO and the South China Sea summer monsoon onset in recent decades. Adv. Atmos. Sci., 39, 443–455, https://doi.org/10.1007/s00376-021-1208-6.
Hung, L., C. Wang, J. Yan, Q. Sun, H. Yao, X. Zhao, and C. Chen, 2012: Air–sea fluxes exchange and heat budget over the SCS Xisha seas during the period of 2008 summer monsoon. Acta Meteor. Sin., 70, 492–505, https://doi.org/10.11676/qxxb2012.041.
Hung, R., L. Gu, L. Zhou, and S. Wu, 2006: Impact of the thermal state of the tropical western Pacific on onset date and process of the South China Sea summer monsoon. Adv. Atmos. Sci., 23, 909–924, https://doi.org/10.1007/s00376-006-0909-1.
Jiang, X., Z. Wang, and Z. Li, 2018: Signature of the South China Sea summer monsoon onset on spring-to-summer transition of rainfall in the middle and lower reaches of the Yangtze River basin. Climate Dyn., 51, 3785–3796, https://doi.org/10.1007/s00382-018-4110-x.
Josey, S. A., 2001: A comparison of ECMWF, NCEP-NCAR, and SOC surface heat fluxes with moored buoy measurements in the subduction region of the Northeast Atlantic. J. Climate, 14, 1780–1789, https://doi.org/10.1175/1520-0442(2001)014<1780:ACOENN>2.0.CO;2.
Kajikawa, Y., and B. Wang, 2012: Interdecadal change of the South China Sea summer monsoon onset. J. Climate, 25, 3207–3218, https://doi.org/10.1175/JCLI-D-11-00207.1.
Lau, K.-M., and M.-T. Li, 1984: The monsoon of East Asia and its global associations—A survey. Bull. Amer. Meteor. Soc., 65, 114–125, https://doi.org/10.1175/1520-0477(1984)065<0114:TMOEAA>2.0.CO;2.
Lau, N.-C., and M. J. Nath, 2009: A model investigation of the role of air–sea interaction in the climatological evolution and ENSO-related variability of the summer monsoon over the South China Sea and western North Pacific. J. Climate, 22, 4771–4792, https://doi.org/10.1175/2009JCLI2758.1.
Li, Y., S. Yang, Y. Deng, and B. Zheng, 2020: Signals of spring thermal contrast related to the interannual variations in the onset of the South China Sea summer monsoon. J. Climate, 33, 27–38, https://doi.org/10.1175/JCLI-D-19-0174.1.
Liu, B., and C. Zhu, 2016: A possible precursor of the South China Sea summer monsoon onset: Effect of the South Asian high. Geophys. Res. Lett., 43, 11 072–11 079, https://doi.org/10.1002/2016GL071083.
Liu, B., and C. Zhu, 2020: Boosting effect of tropical cyclone “Fani” on the onset of the South China Sea summer monsoon in 2019. J. Geophys. Res. Atmos., 125, e2019JD031891, https://doi.org/10.1029/2019JD031891.
Liu, W. T., K. B. Katsaros, and J. A. Businger, 1979: Bulk parameterization of air–sea exchange of heat and water vapor including the molecular constraints at the interface. J. Atmos. Sci., 36, 1722–1735, https://doi.org/10.1175/1520-0469(1979)036<1722:BPOASE>2.0.CO;2.
Louis, J. F., 1979: A parametric model of vertical eddy fluxes in the atmosphere. Bound.-Layer Meteor., 17, 187–202, https://doi.org/10.1007/BF00117978.
Martin, G. M., A. Chevuturi, R. E. Comer, N. J. Dunstone, A. A. Scaife, and D. Zhang, 2019: Predictability of South China Sea summer monsoon onset. Adv. Atmos. Sci., 36, 253–260, https://doi.org/10.1007/s00376-018-8100-z.
Monin, A. S., and A. M. Obukhov, 1954: Basic laws of turbulent mixing in the atmosphere near the ground. Tr. Geofiz. Inst., Akad. Nauk SSSR, 24, 163–187.
O’Neill, L. W., D. B. Chelton, and S. K. Esbensen, 2012: Covariability of surface wind and stress responses to sea surface temperature fronts. J. Climate, 25, 5916–5942, https://doi.org/10.1175/JCLI-D-11-00230.1.
Ren, X., X.-Q. Yang, and X. Sun, 2013: Zonal oscillation of western Pacific subtropical high and subseasonal SST variations during Yangtze persistent heavy rainfall events. J. Climate, 26, 8929–8946, https://doi.org/10.1175/JCLI-D-12-00861.1.
Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax, 2007: Daily high-resolution-blended analyses for sea surface temperature. J. Climate, 20, 5473–5496, https://doi.org/10.1175/2007JCLI1824.1.
Shen, S., and K.-M. Lau, 1995: Biennial oscillation associated with the East Asian summer monsoon and tropical sea surface temperatures. J. Meteor. Soc. Japan, 73, 105–124, https://doi.org/10.2151/jmsj1965.73.1_105.
Smith, S. D., 1988: Coefficients for sea surface wind stress, heat flux, and wind profiles as a function of wind speed and temperature. J. Geophys. Res., 93, 15 467–15 472, https://doi.org/10.1029/JC093iC12p15467.
Song, X., 2020: The importance of relative wind speed in estimating air–sea turbulent heat fluxes in bulk formulas: Examples in the Bohai Sea. J. Atmos. Oceanic Technol., 37, 589–603, https://doi.org/10.1175/JTECH-D-19-0091.1.
Song, X., and L. Yu, 2013: How much net surface heat flux should go into the western Pacific warm pool. J. Geophys. Res. Oceans, 118, 3569–3585, https://doi.org/10.1002/jgrc.20246.
Song, X., C. Ning, Y. Duan, H. Wang, C. Li, Y. Yang, J. Liu, and W. Yu, 2021: Observed extreme air–sea heat flux variations during three tropical cyclones in the tropical southeast Indian Ocean. J. Climate, 34, 3683–3705, https://doi.org/10.1175/JCLI-D-20-0170.1.
Song, X., X. Xie, B. Qiu, H. Cao, S.-P. Xie, Z. Chen, and W. Yu, 2022: Air–sea latent heat flux anomalies induced by oceanic submesoscale processes: An observational case study. Front. Mar. Sci., 9, 850207, https://doi.org/10.3389/fmars.2022.850207.
Song, X., H. Cao, B. Qiu, W. Wang, and W. Yu, 2023: Subsurface imbalance stimulated in a mesoscale eddy. Part I: Observations. Deep-Sea Res. I, 196, 104001, https://doi.org/10.1016/j.dsr.2023.104001.
Sun, Q., and Coauthors, 2010: The variation characteristics of air–sea fluxes over the Xisha area before and after the onset of the South China Sea monsoon in 2008. Acta Oceanol. Sin., 32, 12–23.
Tao, S. Y., and L. X. Chen, 1987: A review of recent research on the East Asian summer monsoon in China. Monsoon Meteorology, C.-P. Chang and T. N. Krishnamurti, Eds., Oxford University Press, 60–92.
Tian, Y., and Q. Wang, 2010: Definition of the South China Sea summer monsoon onset. Chin. J. Oceanol. Limnol., 28, 1281–1289, https://doi.org/10.1007/s00343-010-9950-0.
Villas Bôas, A. B., O. T. Sato, A. Chaigneau, and G. P. Castelão, 2015: The signature of mesoscale eddies on the air-sea turbulent heat fluxes in the South Atlantic Ocean. Geophys. Res. Lett., 42, 1856–1862, https://doi.org/10.1002/2015GL063105.
Wallace, J. M., T. P. Mitchell, and C. Deser, 1989: The influence of sea-surface temperature on surface wind in the eastern equatorial Pacific: Seasonal and interannual variability. J. Climate, 2, 1492–1499, https://doi.org/10.1175/1520-0442(1989)002<1492:TIOSST>2.0.CO;2.
Wang, B., and R. Wu, 1997: Peculiar temporal structure of the South China Sea summer monsoon. Adv. Atmos. Sci., 14, 177–194, https://doi.org/10.1007/s00376-997-0018-9.
Wang, B., LinHo, Y. Zhang, and M.-M. Lu, 2004: Definition of South China Sea monsoon onset and commencement of the East Asia summer monsoon. J. Climate, 17, 699–710, https://doi.org/10.1175/2932.1.
Wang, B., F. Huang, Z. Wu, J. Yang, X. Fu, and K. Kikuchi, 2009: Multi-scale climate variability of the South China Sea monsoon: A review. Dyn. Atmos. Oceans, 47, 15–37, https://doi.org/10.1016/j.dynatmoce.2008.09.004.
Wang, L., and G. Chen, 2018: Relationship between South China Sea summer monsoon onset and landfalling tropical cyclone frequency in China. Int. J. Climatol., 38, 3209–3214, https://doi.org/10.1002/joc.5485.
Weare, B. C., 1989: Uncertainties in estimates of surface heat fluxes derived from marine reports over the tropical and subtropical oceans. Tellus, 41A, 357–370, https://doi.org/10.1111/j.1600-0870.1989.tb00388.x.
Wu, R., 2002: Processes for the northeastward advance of the summer monsoon over the western North Pacific. J. Meteor. Soc. Japan, 80, 67–83, https://doi.org/10.2151/jmsj.80.67.
Yamaguchi, M., and S. J. Majumdar, 2010: Using TIGGE data to diagnose initial perturbations and their growth for tropical cyclone ensemble forecasts. Mon. Wea. Rev., 138, 3634–3655, https://doi.org/10.1175/2010MWR3176.1
Yan, J., H. Yao, J. Li, Z. Tang, G. Jiang, W. Sha, X. Li, and Y. Xiao, 2003: Air–sea heat flux exchange over the South China Sea under different weather conditions before and after southwest monsoon onset in 2000. Acta Oceanol. Sin., 22, 369–383.
Yan, J., Z. Tang, H. Yao, J. Li, Y. Xiao, and Y. Chen, 2005: Air–sea flux exchange and variation over the Xisha Sea area before and after the onset of southwest monsoon in 2002. Chin. J. Geophys., 48, 1078–1090, https://doi.org/10.1002/cjg2.751.
Yan, Y., L. Zhang, X. Song, G. Wang, and C. Chen, 2021: Diurnal variation in surface latent heat flux and the effect of diurnal variability on the climatological latent heat flux over the tropical oceans. J. Phys. Oceanogr., 51, 3401–3415, https://doi.org/10.1175/JPO-D-21-0128.1.
Yu, L., 2019: Global air–sea fluxes of heat, fresh water, and momentum: Energy budget closure and unanswered questions. Annu. Rev. Mar. Sci., 11, 227–248, https://doi.org/10.1146/annurev-marine-010816-060704.
Zheng, B., and Y. Huang, 2019: Mechanisms of northward-propagating intraseasonal oscillation over the South China Sea during the pre-monsoon period. J. Climate, 32, 3297–3311, https://doi.org/10.1175/JCLI-D-18-0391.1.
Zhou, F., R. Zhang, R. Shi, J. Chen, Y. He, D. Wang, and Q. Xie, 2018: Evaluation of OAFlux datasets based on in situ air–sea flux tower observations over Yongxing Island in 2016. Atmos. Meas. Tech., 11, 6091–6106, https://doi.org/10.5194/amt-11-6091-2018.
Zhu, C., W.-S. Lee, H. Kang, and C.-K. Park, 2005: A proper monsoon index for seasonal and interannual variations of the East Asian monsoon. Geophys. Res. Lett., 32, L02811, https://doi.org/10.1029/2004GL021295.
Zhu, Z., and T. Li, 2017: Empirical prediction of the onset dates of South China Sea summer monsoon. Climate Dyn., 48, 1633–1645, https://doi.org/10.1007/s00382-016-3164-x.