Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Properties and Drivers of Marine Heat Waves in the Northern South China Sea

Properties and Drivers of Marine Heat Waves in the Northern South China Sea MAY 2022 WA N G E T A L . 917 a,b b,c a,b b b a,b QIANG WANG, BO ZHANG, LILI ZENG, YUNKAI HE, ZEWEN WU, AND JU CHEN Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou, China State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China University of Chinese Academy of Sciences, Beijing, China (Manuscript received 21 October 2021, in final form 28 January 2022) ABSTRACT: The properties and heat budget of marine heat waves (MHWs) on the northern South China Sea (SCS) continental shelf are investigated. MHWs with warming amplitudes above 1.58C occur mainly along the coast, and their temperature anomaly decreases toward the open sea. MHWs with 18–1.58C warming and duration , 20 days dominate the northern SCS continental shelf. A heat budget analysis indicates that the main heat source is the sea surface net heat flux. Oceanic processes are dominated by the advection of mean temperature by the anomalous horizontal velocity (advha). The net contribution of advha always cools the upper layer of the ocean, resulting in the decay of MHWs. Active cross- slope water exchanges exist at the east and west sides of the northern SCS continental shelf edge, which makes the domi- nant contributions to the advha. In the MHW developing phase, the west (east) side makes a positive (negative) contribu- tion to the advha. In the decay phase, both sides make a negative contribution to the advha, resulting in the rapid decay of MHWs. Although the contribution of advha to the heat budget varies along the northern SCS continental shelf edge, its net effect always cools the MHWs over the shelf. These results provide new insight into the characteristics and formation mechanism of MHWs on the northern SCS continental shelf; in particular, they clarify the respective contributions of air–sea flux and oceanic processes to MHWs. SIGNIFICANCE STATEMENT: Marine heat waves (MHWs) are unusual warming events in oceans that heavily affect marine ecosystems and arouse great concern from citizens. MHWs are active in the northern South China Sea (SCS) continental shelf. On the northern SCS continental shelf, the sea surface net heat flux is the main heat source of MHWs, and ocean current anomalies always cool the upper layer of the ocean. Active cross-slope water exchange at the east and west sides of the northern SCS continental shelf edge is the main oceanic way that cools the water on the shelf, eventually resulting in the decay of MHWs. KEYWORDS: Advection; Extreme events; Mixed layer; Warm water volume; Surface temperature; Oceanic variability 1. Introduction 2017). Shortwave radiation and ocean advection anomalies contribute jointly to MHWs in the East China and South Yel- Marine heat waves (MHWs) are unusual warming events in low Seas (Tan and Cai 2018; Gao et al. 2020). In coastal oceans, and they have been occurring with increasing fre- waters off western Australia, MHW events are attributed to quency over the past few decades (Hobday et al. 2016; Oliver variations in heat advection by the Leeuwin Current and heat et al. 2017, 2018). Many MHW events have been recorded flux across the air–sea interface (Benthuysen et al. 2014, 2018; and are widely distributed on continental shelves and open Feng et al. 2013; Oliver et al. 2017). Local warming from the seas (Hobday et al. 2018; Oliver et al. 2021). MHW is not just atmosphere dominates the MHW events in the northwest a physical oceanographic phenomenon; it also exerts consid- Atlantic (Chen et al. 2014). Long-lasting MHWs in the tropi- erable influence on regional marine ecosystems (Genevier cal Indian Ocean were recorded, which were maintained by et al. 2019). Discrete MHWs can drive a gradual change in downwelling Rossby waves (Zhang et al. 2021). Strong sub- species distribution and affect the diversity and mortality of surface MHWs in the tropical western Pacific Ocean were commercial fisheries (Mills et al. 2013; Wernberg et al. 2013; observed, and anomalous sea surface convergence and Caputi et al. 2016). The increasing trend in MHW intensity Ekman downwelling played an important role in its formation and frequency has led to growing research interest (Collins (Hu et al. 2021). Reduced heat loss from the ocean to the et al. 2019). atmosphere and weak cold advection in the upper ocean can The contributions of atmospheric and oceanic processes to sometimes cause strongly positive temperature anomalies in MHW development vary in different oceans (Schlegel et al. the northeast Pacific(Bond et al. 2015; Di Lorenzo and Mantua 2016). Recent literature has provided a more detailed Denotes content that is immediately available upon publica- study of northeast Pacific MHWs, and double-peak and sin- tion as open access. gle-peak categories have been identified (Chen et al. 2021a). For the double-peak, the first peak is attributed to the surface heat flux, while the vertical entrainment and diffusion Corresponding authors: Qiang Wang, wqiang@scsio.ac.cn; Ju Chen, jchen@scsio.ac.cn are responsible for the second peak 5 months later. For the DOI: 10.1175/JPO-D-21-0236.1 Ó 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses). 918 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 FIG. 1. Composite properties of marine heat waves (MHWs). (a) Mean intensity and (b) duration of MHWs derived from the OISST dataset. (c),(d) As in (a) and (b), but from the OFES dataset. The black line indicates the southern boundary of the northern South China Sea continental shelf. Gray dots indicate where the composite results exceed 95% significance (Student’s t test). single-peak category, vertical entrainment and diffusion are budget on the northern SCS continental shelf. Although the the leading contributors. A persistent and intense MHW overall characteristics of the MHW properties in the SCS event in the northeast Pacific during 2019–20 was observed, have been discussed in previous studies (Cai et al. 2016; Li followed by a La Nina ˜ event, while other recorded double- et al. 2019; Yao et al. 2020; Yao and Wang 2021), details of peak events were associated with El Nino ˜ or neutral condi- MHWs in the northern SCS and their driving factors remain tions (Chen et al. 2021b). unclear, especially for the respective contributions of atmo- In the South China Sea (SCS), the occurrence frequency, spheric and oceanic processes to MHW development. intensity, and duration of MHW events have all increased The northern SCS continental shelf is close to the main- under global warming (Li et al. 2019; Yao and Wang 2021) land, and its thermal state greatly modulates the ecosystem and are projected to continue increasing in future decades and economic activities. The frequency and intensity of (Yao et al. 2020). El Nino ˜ –Southern Oscillation strongly regu- MHWs on the northern SCS continental shelf are continu- lates SCS MHWs (Liu et al. 2022). Both the sea surface net ously enhanced, attracting wide attention. Their drivers heat flux and ocean advection can cause extreme warming urgently need to be elucidated to better understand and pre- events at the basin scale (Xie et al. 2003; Xiao et al. 2018, dict their variability. Therefore, this paper focuses on the 2020; Wang et al. 2021). The western boundary current properties and drivers of the MHWs on the northern SCS (WBC) of the SCS is responsible for the advection of cooler continental shelf. water (Liu et al. 2004; Fang et al. 2013; Wei et al. 2016; Zhao The remainder of this paper is organized as follows. Data and methods are discussed in section 2. Section 3 describes and Zhu 2016; Zhao et al. 2017; Sun et al. 2020); weakening the properties of MHWs on the northern SCS continental or a complete shutdown of the WBC contributes greatly to shelf. The MHW heat budget is analyzed in section 4. Finally, the warming of the SCS upper layer (Wang et al. 2021). conclusions are given in section 5. The SCS MHW events occur mainly on the continental shelf, especially in the northern SCS (Cai et al. 2016; Li et al. 2019; Yao et al. 2020), where the SCS WBC divides the outer 2. Data and methods shelf edge from the open sea (Wang et al. 2013; Zhu et al. a. Data 2019). Multiple band-like currents flowing along the topogra- phy on the northern SCS continental shelf regulate an along- Observation data for sea surface temperature (SST) are topography distribution of climatological temperature (Shu used, together with model output, for a range of ocean varia- et al. 2018). Due to the instability of the SCS large-scale slope bles. Because of data availability, the period 1982–2017 is current, there are also active cross-slope water exchanges used in this study. between the continental shelf and the open sea in the north- The SST data are from NOAA OISST version 2 (Reynolds ern SCS (Wang et al. 2018, 2020; Liu and Gan 2020a). Due to et al. 2007), which is a daily and 0.258 3 0.258 gridded prod- the climatological temperature approximate belt-like distribu- uct from Advanced Very High Resolution Radiometer tion parallel to the slope (Chen et al. 2003), the cross-slope (AVHRR) satellite data, with bias correction using ship water exchanges imply potential contributions to the heat and buoy data. MAY 2022 WA N G E T A L . 919 FIG. 2. Comparisons of sea surface temperature (SST) between OISST and OFES. (a) 90-day low-pass SST. (b) SST anomaly derived from the SST minus the 90-day low-pass SST. (c) Composite series of SST anomalies for all marine heat waves (MHWs) in the northern SCS. The SSTs in (a) and (b) are averaged within the northern South China Sea continental shelf (bounded by the black line in Fig. 1); contours and colors represent the density of samples (red for high pair density and white for low pair density). The composite in (c) is produced as follows: the variable in each MHW is normalized to 0–1 in time, and then the variable of each MHW is composited. The results from the OGCM for the Earth Simulator OISST versus OFES validates the use of OFES on scales (OFES; Sasaki et al. 2007) are utilized in this study. OFES larger than or equal to seasonal time (Fig. 2a). A comparison was run at 0.18 3 0.18 spatial resolution and 54 vertical levels, of the SST anomaly, subtracting the 90-day low-pass SST, fur- forced by NCEP–NCAR reanalysis data. The ocean velocity, ther confirms the validation of OFES on an intraseasonal temperature, and surface net heat flux used in this study are scale (Fig. 2b). To investigate the differences in the life cycle supplied by OFES. OFES provides output in the form of of MHWs (see section 2b) expressed by OISST and OFES, a 3-day averages. The definition of an MHW is based on 11-day composite is produced from the MHW cases projected onto a window data (see section 2b). Thus, the 3-day averaged data time axis from 0 to 1 (Fig. 2c). Although the MHW of OFES can also be used to investigate the heat budget of MHWs. For is relatively weaker than that of OISST, it simulates a rela- convenience, the OFES outputs are all interpolated to daily tively real MHW evolution process. data using linear interpolation. The distribution of the mean Three mooring stations were located on the northern SCS intensity and duration of MHWs in the northern SCS derived slope. The sampling time interval was 1 h, and the vertical from OFES generally resembles that derived from OISST spatial resolution was 8 m at all stations. Velocity profiles (Fig. 1), which validates the use of OFES in the investigation at station DS01 (20.308N, 117.678E) were obtained from of MHWs. The 90-day low-pass SST scatter diagram of 18 September 2014 to 15 September 2015. Station DS02 920 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 FIG. 3. Taylor diagrams of the mean properties of each northern South China Sea marine heat wave (MHW) event derived from the (a) OISST dataset and (b) OFES dataset. The radius indicates the mean intensity of MHWs, and the angle indicates the duration. The percentage indicates the proportion of MHWs in that sector relative to the total number of MHWs. The color and size of each circle indicate the intensity and duration, respectively. Green: ,0.58C; yellow: between 0.58 and 18C; red: between 18 and 1.58C; purple: .1.58C. Small size: ,20 days; middle size: between 20 and 40 days; large size: .40 days. (19.718N, 116.998E) acquired data from 20 September 2014 to respectively. The term T is the daily SST on day d of year y,and 20 August 2018. Velocity profiles were obtained from 1 Janu- P (j) is the 90th percentile of X,where X ={T(y, d)|y # y # y , 90 s e ary 2014 to 12 June 2017 at station XS (17.298N, 111.368E). j 2 5 # d # j 1 5}. Two MHW metrics are used in the study: the duration Any velocity larger than 3 times the standard deviation at that (days between the start and end dates) and the intensity depth is defined as an invalid value. Short gaps caused by (average of daily intensity anomalies measured in 8C). invalid data were subsequently filled by linear interpolation. A 48-h low-pass filter was applied to the velocity data to c. Upper-ocean temperature budget remove high-frequency signals, and the data were then aver- The perturbation temperature budget equation, averaged aged over 24 h (daily). from the sea surface to a fixed depth h , can be written as fol- b. Definition of MHWs lows (Benthuysen et al. 2014; Feng et al. 2008; Oliver et al. 2017): Following Hobday et al. (2016), the climatology SST is defined relative to the time of year, using all data within an ­T Q () T 2 T m d 11-day window centered on the time of year from which the 2 u · =T 2 u · =T 2 w ­t rC h h p m m climatological mean is calculated, i.e., Eq. (1). advh advha Tend advv An MHW is defined as an anomalously warm, discrete, and Qnet prolonged event, which is a period when the daily SST is T 2 T () m d above a particular threshold for at least 5 days. The threshold 2 w , (3) on the day of the year is defined as the 90th percentile of the advva daily temperature within an 11-day window centered on this day in all years, i.e., Eq. (2): where T is the temperature and h is set as 50 m, with · rep- ye j15 resenting vertical integration dz . Parameter T is the Ty, d 2hm () T () j  , (1) average temperature over the upper 50 m; T is the water 11 y 2 y d () e s11 yy s dj25 temperature below h and is set to the value at 70 m. We also checked the values of h for 40 and 60 m and found that our results were not particularly sensitive to the choice of T () j  P () j , (2) 90 90 depthusedto calculate T .The term Q is the sea surface where T is the climatology SST and T is the 90th percentile net heat flux anomaly and is positive for heat input to the m 90 of the daily SST. Parameter j is the day of the year, and y and ocean; r = 1025 kg m is the reference density of seawater, 21 21 y are the start and end of the climatological base period, and C = 4007 J Kg K is the specific heat of seawater. e p MAY 2022 WA N G E T A L . 921 (a) SST anomaly of MHWs (d) Temperature advection by anomalous flo w 2 advha advha(South) 1.1 Developing Mature Decay advha(North) 0.9 −2 0.7 −4 0.5 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Period (b) Heat budget terms (e) Temperature advection by anomalous flow (South) 0.8 −1 Qnet advha −2 0.6 advh advwa −3 advw −4 0 0.2 0.4 0.6 0.8 1 0.4 (c) Ocean versus atmosphere 0.2 −2 −4 |Ocean/Ato|=0.65 |Ocean/Ato|=3.90 |Ocean/Ato|= 0.94 Qnet −6 112 113 114 115 116 117 118 119 Sum of ocean process Longitude −8 −7o [´ 10 C/s] 0 0.2 0.4 0.6 0.8 1 Period −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 FIG. 4. Composite heat budget terms for all marine heat waves (MHWs) in the northern SCS. (a) Composite series of SST anomalies. (b) Composite heat budget terms in Eq. (1). (c) Comparisons between the contribution of net heat flux (Qnet) and temperature advection (i.e., the sum of all terms except for Qnet). (d) Advection of mean temperature by the anomalous horizontal velocity (advha); south (north) indicates the contributions of advection at the south (north) boundary. (e) Composite profiles of advha at the south boundary. The composite is produced as follows: the variable in each MHW is normalized to 21 in time, and then the variable of each MHW is composited. Gray dots indi- cate where the composite results exceed 95% significance (Student’s t test). Vector u =(u, y) is the zonal and meridional velocity pair, advha  y () T 2 T dx dz south south and w is the vertical velocity; = =(­ , ­ ) is the horizontal t x y gradient operator. The overbar indicates the climatological south mean, and the prime indicates an anomaly. 2 y () T 2 T dx dz north north o The left-hand side of Eq. (3) is the rate of change or ten- dency of the upper 50-m averaged temperature (Tend). On north the right-hand side, Qnet is the sea surface net heat flux 1 u () T 2 T dy dz anomaly forcing, advh is the advection of anomalous tempera- west west o ture by the mean horizontal velocity, advha is the advection west of mean temperature by the anomalous horizontal velocity, advv is the advection of anomalous temperature by the mean 2 u () T 2 T dy dz , (4) east east o vertical velocity, and advva is the advection of mean tempera- east ture by the anomalous vertical velocity. In the northern SCS, the contribution of advection to the temperature anomalies can be represented by the tempera- where t is the volume of the box enclosed by the northern and ture flux at the southern and northern boundaries (Feng southern boundaries and additional eastern and western et al. 2008; Gao et al. 2020). Following Lee et al. (2004), boundaries. Parameter T is the daily average temperature in the contributions of advection at these northern and south- the northern SCS, which is used as the reference temperature ern boundaries can be further written (using advha as an in the heat advection. By including the T in the heat advec- example) as tion terms, unbalanced volume flux in one direction is assumed −7o −7o Temperature anomaly ( C) Terms ( ´ 10 C/s) Terms ( ´ 10 C/s) −7o Period Advha ( ´ 10 C/s) 922 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 FIG. 5. Distribution of composite heat budget terms for the (a) developing phase, (b) mature phase, and (c) decay phase. Qnet (8Cs ) is the net heat flux. The temperature anomaly (8C) is the temperature anomaly averaged over the upper 50 m. The Clim T and flow anom- aly indicates the climatological temperature and composite horizontal velocity anomaly for all marine heat wave events, averaged over the upper 50 m. The term advv 1 advva (8Cs ) indicates the vertical temperature advection, as in Eq. (1).The definition of each phase is shown in Fig. 4a. The heat budget terms have been multiplied by 1.0 3 10 . The purple vectors indicate the water flowing toward the deep sea. to carry waters with the same temperature as T so that it does OISST is 13.1 6 0.7 days, and OFES is 14.6 6 1.4 days. The not have a net contribution to the advection (Lee et al. 2004; spatial variation in duration is less than 2 days, and their evo- Feng et al. 2008). lution processes are similar (Fig. 2c), which suggests the lim- ited influence of the spatial duration difference between OFES and OISST on the validation of OFES to analyze 3. Properties of MHWs in the northern South China Sea MHW evolution. The small spatial variation also suggests that The mean properties of MHWs over the past four decades MHW occurrence and development are reasonably synoptic were calculated (Fig. 1). The intensity isolines run belt-like along the northern SCS continental shelf. parallel to the coastline, with decreasing values toward the The SST averaged over the northern SCS continental shelf open sea in the northern SCS (Figs. 1a,c). There are three bounded by the black line in Fig. 1 is used to identify the high-value centers in Zhanjiang Bay, the Pearl River estuary, northern SCS MHWs. The intensity–duration diagram of the and offshore Shantou, where the maximum SST anomaly can MHWs is shown in Fig. 3. Approximately 50% of the MHWs reach 28C. On the continental slope edge, the intensity have intensities in the range of 18–1.58C. One-third of MHWs decreases to ,1.58C. have intensities in the range 0.58–1.08C, and only ∼10% The mean MHW duration ranges from 12 to 15 days in the exceed 1.58C. In the OFES results, there are some MHWs northern SCS and is slightly longer on the western continental with intensities , 0.58C that do not appear in the OISST. This shelf than on the eastern shelf (Figs. 1b,d). In the northern inconsistency indicates the slightly underestimated SST anom- SCS continental shelf (bounded by line A in Fig. 1a), there is aly in the OFES. However, there are only six weak MHWs, a relatively significant spatial difference between OISST and ∼7% of the total number, which does not affect the composite OFES; however, the duration over the continental shelf of analysis of the heat budget in the next section. MAY 2022 WA N G E T A L . 923 FIG. 6. Composite upper 50-m averaged temperature gradient and velocity for all marine heat waves (MHWs) in the northern SCS. (a) Composite upper 50-m averaged temperature gradient projected onto the direction perpendicular to line A (marked in Fig. 1a), and onshore is positive. (b) Composite profiles of velocity perpendicular to line A and onshore are positive. The compos- ite is produced as follows: the variable in each MHW is normalized to 0–1 in time, and then the variable of each MHW is composited. Gray dots indicate where the composite results exceed 95% significance (Student’s t test). More than 80% of MHWs in the northern SCS continental weakens, and the negative advha strengthens, leading to an shelf have a duration of less than 20 days, and ∼10% last approximate equilibrium between the oceanic and atmo- between 20 and 40 days. MHWs longer than 40 days are rare, spheric processes (Figs. 4b,c). In the decay phase, Qnet weak- and only one case occurred in the OISST data (Fig. 3a). ens sharply and becomes negative, and the advha continues The main features of the MHW intensity and duration on to strengthen rapidly (Fig. 4b). The upper ocean begins to the northern SCS continental shelf from OISST and OFES lose heat, which is mainly associated with oceanic processes are relatively similar, which validates the use of OFES output (Figs. 4b,c). Therefore, the Qnet warms the upper ocean and in the following heat budget analysis. induces an MHW outbreak in the northern SCS continental shelf; oceanic processes, which are dominated by the advha, cool the upper ocean, and cause the MHW to die out. 4. Air–sea heat flux versus ocean current for MHWs The continental shelf edge, indicated by the black line in To investigate the development of MHWs in the northern Fig. 1, and the Taiwan Strait are the two pathways for the SCS continental shelf, a composite is produced from the advha on the northern SCS continental shelf. The southern MHW cases projected onto a time axis from 0 to 1. The devel- boundary (i.e., the continental shelf edge) is the main temper- oping, mature, and decay phases can be identified using the ature advection (advha) pathway (Fig. 4d), which is concen- composite SST anomaly (Fig. 4a). In the developing phase, trated on the east and west sides of the shelf (Fig. 4e). In the the net heat flux (Qnet) is the main source of warming, developing phase, the east and west sides make opposite con- whereas the advection of mean temperature by the anoma- tributions to the advha. In the decay phase, the temperature lous horizontal velocity (advha) always opposes the contribu- advection (advha) is negative on both sides, leading to a rapid tion of Qnet, cooling the upper layer of the ocean (Fig. 4b). In enhancement of negative advha, which induces MHW decay. total, the oceanic process offsets 65% of the contribution of The spatial distributions of each term are shown in Fig. 5. the atmospheric process (Fig. 4c). In the mature phase, Qnet In the developing phase, a large positive Qnet is mainly 0 924 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 (a) Developing phase: windstress and its curl dominates both the east and west sides at the continental shelf edge (Figs. 5c and 6b), which leads to a strong negative advha, 23 2 0.1dyn/cm resulting in the rapid decay of the MHW. The vertical temper- ature advection contributes little during the developing and mature phases (Fig. 4b and Figs. 5a,b). However, the vertical process contributes significantly to the decay of MHWs near the coast (Fig. 5c). −9 3 [10 dyn/cm ] On the northern SCS continental shelf, the variations in 110 112 114 116 118 120 the flow anomaly are mainly associated with wind stress 1.5 anomalies (Fig. 7). In the developing and mature phases, the (b) Mature phase: windstress and its curl wind stress anomalies mainly blow toward the northeast (Figs. 7a,b), which induces an eastward flow anomaly along 0.5 the coast (Figs. 5a,b). In the decay phase, the wind stress turned toward the southwest (Fig. 7c), driving the continental shelf flow to turn direction accordingly (Fig. 5c). The wind −0.5 stress curl is one dominant factor modulating the cross-iso- −1 bath movement over a widened shelf in the northeastern SCS 110 112 114 116 118 120 (Gan et al. 2013; Liu et al. 2020b). During the developing and −1.5 decay phases, a significant negative wind stress curl occupied (c) Decay phase: windstress and its curl −2 the northeastern SCS (Figs. 7a,c). Responding to the negative wind stress vorticity input into the ocean, the ocean volume potential vorticity decreases. The decrease in potential vortic- ity will make the seawater tend to cross the isobath, flowing toward the deep sea (Fig. 6b). In the mature phase, there is a positive wind stress curl over the northeastern SCS (Fig. 7b), which weakens the offshore flow (Fig. 6b). On the western side of the continental shelf, the wind stress curl is mainly neg- 110 112 114 116 118 120 ative and gradually enhances from the developing phase to Longitude ( E) the decay phase (Fig. 7), which corresponds to the offshore FIG. 7. Distribution of the composite wind stress curl and wind flow between 1128 and 1138E(Fig. 6b). The direction of cross- stress vector for the (a) developing phase, (b) mature phase, and slope flow east of Hainan Island is mainly determined by the (c) decay phase. The blue lines indicate the topography isobaths. direction of the slope current, i.e., offshore (onshore) flow is Shading is the wind stress curl, and vectors are the wind stress. Red induced by the southwest (northeast) slope current due to the lines are the zero contours of the wind stress curl. block of Hainan Island. Three mooring stations were used to investigate the flow anomalies on the east and west sides of the northern SCS located in the center and east of the continental shelf, slope (Fig. 8b). The northern SCS continent shelf average although the entire northern SCS continental shelf warms SST anomaly during the mooring observation period is shown (Fig. 5a). Warming on the west side of the continental shelf is in Fig. 8a, and the MHW events are marked in purple. The attributed to the positive advha (Fig. 4e), where a northward velocities of the MHWs are shown in Figs. 8c, 8e, and 8g, flow anomaly carries high mean temperature water from the and composites of the MHWs projected onto a time axis south (Fig. 5a and Figs. 6a,b). Along the southern boundary from 0 to 1 are shown in Figs. 8d, 8f,and 8h. On the east of the continental shelf (i.e., line A in Fig. 1a), the tempera- side, the composite flow anomalies are mainly offshore and ture gradient perpendicular to line A is mainly negative (i.e., increase gradually with MHW development (Figs. 8d,f). On decreases northward; Fig. 6a). The onshore flow anomaly on the west side, the flow anomalies are mainly northeastward the western side (Fig. 6b), jointly with the negative tempera- and decrease during the MHW decay phase (Fig. 8h). The ture gradient, induces a positive advha (2 y ­T=­y , where y changes of OFES flow anomalies on both the east and west indicates the direction perpendicular to line A). On the east sides are generally consistent with the mooring observa- side, the strong offshore flow anomaly plus a negative temper- tions, which confirms the validation of the OFES oceanic ature gradient induces a large negative advha (Figs. 6a,b). In processes during MHWs. the mature phase, although Qnet decreases, the upper layer temperature anomaly reaches a maximum after early accumu- 5. Conclusions lation (Fig. 5b). The east side flow anomaly is still offshore and contributes negatively to the advha (Fig. 5b and Fig. 6b). Using the OISST and OFES datasets, the properties and The west side cross-slope flow anomaly weakens, which grad- heat budget of MHWs on the northern SCS continental shelf ually weakens its positive advha (Fig. 4e). In the decay phase, were analyzed. The intensity of MHWs is strong along the Qnet becomes negative and the upper layer temperature coast and decreases toward the open sea. MHWs with anomaly weakens (Fig. 5c). The offshore flow anomaly 18–1.58C warming and a duration of ,20 days dominate the Latitude ( N) MAY 2022 WA N G E T A L . 925 (a) SST anomaly ( C) (b) Topography and stations DS01 DS02 −1 18 XS 110 112 114 116 118 120 (c) DS01: Upper 100m velocity (d) DS01: composites of Upper 100m velocit y 0.8 0.2 0.1m/s 0.5m/s 0.4 0.1 0 0 −0.4 −0.1 −0.8 −0.2 (e) DS02: Upper 100m velocit y (f) DS02: composites of Upper 100m velocit y 0.8 0.2 0.1m/s 0.5m/s 0.4 0.1 0 0 −0.4 −0.1 −0.8 −0.2 (g) XS: Upper 100m velocit y (h) XS: composites of Upper 100m velocit y 0.8 0.2 0.1m/s 0.5m/s 0.4 0.1 0 0 −0.1 −0.4 −0.2 −0.8 0 0.25 0.5 0.75 1 Period Time FIG. 8. Observed velocity from mooring stations. (a) Time series of the OISST sea surface temperature (SST) anomaly in the northern South China Sea (SCS); purple indicates a marine heat wave (MHW) event. (b) Mooring stations and topography in the northern SCS. (c) Velocity vectors averaged in the upper 100 m at station DS01; red vectors indicate MHW events. (d) Composite DS01 velocity vector for MHW events, i.e., red vectors in (c). The red vectors for each MHW event have been normalized to 0 to 1 in time as in Fig. 4, and then all the events are composited as (d). (e),(g) As in (c), but for stations DS02 and XS. (f),(h) As in (d), but for stations DS02 and XS. northern SCS continental shelf. MHWs with warming of advection across the southern continental shelf edge in the 0.58–18C account for approximately one-third of the total, and northern SCS. The east and west sides of the continental shelf those .1.58C account for approximately one-tenth. Approxi- edge are the two major pathways; these make opposing contri- mately one-tenth of the MHWs persist for 20–40 days; those butions to the heat budget in the MHW developing phase (i.e., with durations of .40 days are very rare, and only one case negative and positive on the eastern and western sides, respec- was identified from the OISST dataset. tively) but are of the same sign in the decay phase (i.e., both The heat budget indicates that the net heat flux (Qnet) is negative). the main driver of the MHWs on the northern SCS continen- tal shelf. Oceanic processes, which are dominated by the Acknowledgments. This work is supported by the National advection of mean temperature by anomalous horizontal Key Research and Development Program (Grant 2017YFA velocity (advha), always cool the upper layer ocean, leading 0603201), Key Special Project for Introduced Talents Team of to the decay of MHWs. In the developing phase, Qnet is Southern Marine Science and Engineering Guangdong Labora- stronger than the negative oceanic process contribution to the tory (Guangzhou) (GML2019ZD0304), the National Natural heat budget, which results in warming of the upper layer. In Science Foundation of China (Grant 42076209), the Rising Star the mature phase, there is an equilibrium between the Qnet Foundation of the South China Sea Institute of Oceanology and oceanic process contributions. In the decay phase, both the (NHXX2019WL0101), and the Science and Technology Qnet and oceanic process contributions are negative, but the Planning Project of Guangzhou (202102080363, 20200 ocean process contribution is approximately 4 times that of Qnet. 2030490). The ADCP velocity profiles are supplied by the The oceanic process contribution to the heat budget is dom- Xisha Deep Sea Observatory, a member of the Network inated by advha, which is attributed mainly to the temperature of Field Observation and Research Stations of the Chinese Jan/14 Jun/14 Nov/14 Apr/15 Sep/15 Feb/16 Jul/16 Dec/16 May/17 Oct/17 Jan/14 Jun/14 Nov/14 Apr/15 Sep/15 Feb/16 Jul/16 Dec/16 May/17 Oct/17 Velocity (m/s) SST anomaly ( C) Velocity (m/s) 926 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 }}, M. J. McPhaden, S.-P. Xie, and J. Hafner, 2013: La Nina ˜ Academy of Sciences. The numerical calculation is sup- forces unprecedented Leeuwin Current warming in 2011. Sci. ported by the high performance computing division and Rep., 3, 1277, https://doi.org/10.1038/srep01277. Ms. Dandan Sui and Dr. Wei Zhou of the South China Gan, J. P.,H.S.Ho, andL. L.Liang,2013: Dynamics of intensi- Sea Institute of Oceanology. fied downwelling circulation.over a widened shelf in the northeastern South China Sea. J. Phys. Oceanogr., 43,80–94, Data availability statement. The OISST and OFES data https://doi.org/10.1175/JPO-D-12-02.1. for this paper are available at http://apdrc.soest.hawaii.edu/ Gao, G., and Coauthors, 2020: Drivers of marine heatwaves in the data/data.php. East China Sea and the South Yellow Sea in three consecu- tive summers during 2016–2018. J. Geophys. Res. Oceans, 125, e2020JC016518, https://doi.org/10.1029/2020JC016518. REFERENCES Genevier, L. G. C., T. Jamil, D. E. Raitsos, G. Krokos, and I. Hoteit, 2019: Marine heatwaves reveal coral reef zones Benthuysen, J., M. Feng, and L. Zhong, 2014: Spatial patterns susceptible to bleaching in the Red Sea. Global Change of warming off Western Australia during the 2011 Ninga- Biol., 25, 2338–2351, https://doi.org/10.1111/gcb.14652. loo Nino ˜ : Quantifying impacts of remote and local forcing. Hobday, A. J., and Coauthors, 2016: A hierarchical approach to Cont. Shelf Res., 91, 232–246, https://doi.org/10.1016/j.csr. defining marine heatwaves. Prog. Oceanogr., 141, 227–238, 2014.09.014. https://doi.org/10.1016/j.pocean.2015.12.014. Benthuysen, J. A., E. C. J. Oliver, M. Feng, and A. G. Marshall, }}, and Coauthors, 2018: Categorizing and naming marine heat- 2018: Extreme marine warming across tropical Australia dur- waves. Oceanography, 31, 162–173, https://doi.org/10.5670/ ing austral summer 2015–16. J. Geophys. Res. Oceans, 123, oceanog.2018.205. 1301–1326, https://doi.org/10.1002/2017JC013326. Hu, S., and Coauthors, 2021: Observed strong subsurface marine Bond, N. A., M. F. Cronin, H. Freeland, and N. Mantua, 2015: heatwaves in the tropical western PacificOcean. Environ. Causes and impacts of the 2014 warm anomaly in the NE Res. Lett., 16, 104024, https://doi.org/10.1088/1748-9326/ac26f2. Pacific. Geophys. Res. Lett., 42,3414–3420, https://doi.org/10. Lee, T., I. Fukumori, and B. Tang, 2004: Temperature advection: 1002/2015GL063306. Internal versus external processes. J. Phys. Oceanogr., 34, Cai, R., H. Tan, and Q. Qi, 2016: Impacts of and adaptation to 1936–1944, https://doi.org/10.1175/1520-0485(2004)034,1936: inter-decadal marine climate change in coastal China seas. Int. TAIVEP.2.0.CO;2. J. Climatol., 36,3770–3780, https://doi.org/10.1002/joc.4591. Li, Y., G. Ren, Q. Wang, and Q. You, 2019: More extreme Caputi, N., and Coauthors, 2016: Management adaptation of marine heatwaves in the China Seas during the global warm- invertebrate fisheries to an extreme marine heatwave event ing hiatus. Environ. Res. Lett., 14, 104010, https://doi.org/10. at a global warming hot spot. Ecol. Evol., 6,3583–3593, 1088/1748-9326/ab28bc. https://doi.org/10.1002/ece3.2137. Liu, K., K. Xu, C. Zhu, and B. Liu, 2022: Diversity of marine Chen, J.-M., C.-P. Chang, and T. Li, 2003: Annual cycle of the heatwaves in the South China Sea regulated by ENSO phase. South China Sea surface temperature using the NCEP/ J. Climate, 35, 877–893, https://doi.org/10.1175/JCLI-D-21- NCAR reanalysis. J. Meteor. Soc. Japan. Ser. II, 81, 879–884, 0309.1. https://doi.org/10.2151/jmsj.81.879. Liu, Q., X. Jiang, S.-P. Xie, and W. T. Liu, 2004: A gap in the Indo- Chen, K., G. G. Gawarkiewicz, S. J. Lentz, and J. M. Bane, 2014: Pacific warm pool over the South China Sea in boreal winter: Diagnosing the warming of the northeastern U.S. coastal Seasonal development and interannual variability. J. Geophys. ocean in 2012: A linkage between the atmospheric jet stream Res., 109, C07012, https://doi.org/10.1029/2003JC002179. variability and ocean response. J. Geophys. Res. Oceans, 119, Liu, Z., and J. Gan, 2020a: A modeling study of estuarine-shelf 218–227, https://doi.org/10.1002/2013JC009393. circulation using a composite tidal and subtidal open bound- Chen,Z., J. Shi,and C. Li,2021a:Two types of warmblobs in the ary condition. Ocean Modell., 147, 101563, https://doi.org/10. northeast Pacific and their potential effect on the El Nino. Int. 1016/j.ocemod.2019.101563. J. Climatol., 41,2810–2827, https://doi.org/10.1002/joc.6991. }}, T. Zu, and J. Gan, 2020b: Dynamics of cross-shelf water }}, }},Q.Liu,H.Chen, and C. Li, 2021b: A persistent and exchanges off Pearl River Estuary in summer. Prog. Ocean- intense marine heatwave in the Northeast Pacific during ogr., 189, 102465, https://doi.org/10.1016/j.pocean.2020.102465. 2019–2020. Geophys. Res. Lett., 48, e2021GL093239, https:// Mills, K. E., and Coauthors, 2013: Fisheries management in a doi.org/10.1029/2021GL093239. changing climate lessons from the 2012 ocean heat wave in Collins, M., and Coauthors, 2019: Extremes, abrupt changes and the Northwest Atlantic. Oceanography, 26, 191–195, https:// managing risk. IPCC Special Report on the Ocean and Cryo- ¨ doi.org/10.5670/oceanog.2013.27. sphere in a Changing Climate, H.-O. Portner et al., Eds., Oliver, E. C. J., and Coauthors, 2017: The unprecedented 2015/16 IPCC, 589–655, https://doi.org/10.1017/9781009157964.008. Tasman Sea marine heatwave. Nat. Commun., 8, 16101, Di Lorenzo, E., and N. Mantua, 2016: Multi-year persistence of https://doi.org/10.1038/ncomms16101. the 2014/15 North Pacific marine heatwave. Nat. Climate }}, and Coauthors, 2018: Longer and more frequent marine Change, 6, 1042–1047, https://doi.org/10.1038/nclimate3082. heatwaves over the past century. Nat. Commun., 9, 1324, Fang, Y., C. Tana, S. Sun, and B. Liu, 2013: Impact of the climate https://doi.org/10.1038/s41467-018-03732-9. change on the western boundary current in the South China }}, J. A. Benthuysen, S. Darmaraki, M. G. Donat, A. J. Hobday, Sea: An assessment based on numerical simulations. Malays. N. J. Holbrook, R. W. Schlegel, and A. S. Gupta, 2021: Marine J. Sci., 32,347–356. Feng,M., A. Biastoch,C.Boning, N. Caputi, and G. Meyers, heatwaves. Annu. Rev. Mar. Sci., 13, 313–342, https://doi.org/10. 2008: Seasonal and interannual variations of upper ocean 1146/annurev-marine-032720-095144. heat balance off the west coast of Australia. J. Geophys. Res., Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, 113, C12025, https://doi.org/10.1029/2008JC004908. and M. G. Schlax, 2007: Daily high-resolution-blended MAY 2022 WA N G E T A L . 927 analyses for sea surface temperature. J. Climate, 20, 5473– Wernberg, T., D. A. Smale, F. Tuya, M. S. Thomsen, T. J. Langlois, 5496, https://doi.org/10.1175/2007JCLI1824.1. T. deBettignies, S. Bennett, and C. S. Rousseaux, 2013: An Sasaki, H., M. Nonaka, Y. Masumoto, Y. Sasai, H. Uehara, and extreme climatic event alters marine ecosystem structure in a H. Sakuma, 2007: An eddy-resolving hindcast simulation of global biodiversity hotspot. Nat. Climate Change, 3,78–82, the quasi-global ocean from 1950 to 2003 on the Earth Simu- https://doi.org/10.1038/nclimate1627. lator. High Resolution Numerical Modeling of the Atmo- Xiao, F., L. Zeng, Q. Liu, W. Zhou, and D. Wang, 2018: Extreme sphere and Ocean, W. Ohfuchi and K. Hamilton, Eds., subsurface warm events in the South China Sea during 1998/ Springer, 157–185. 99 and 2006/07: Observations and mechanisms. Climate Dyn., Schlegel, R. W., E. C. J. Oliver, S. Perkins-Kirkpatrick, A. Kruger, 50,115–128, https://doi.org/10.1007/s00382-017-3588-y. and A. J. Smit, 2017: Predominant atmospheric and oceanic }}, D. Wang, and L. Yang, 2020: Can tropical Pacific winds patterns during coastal marine heatwaves. Front. Mar. Sci., 4, enhance the footprint of the interdecadal Pacificoscillation 323, https://doi.org/10.3389/fmars.2017.00323. on the upper-ocean heat content in the South China Sea? J. Shu, Y., Q. Wang, and T. Zu, 2018: Progress on shelf and slope Climate, 33,4419–4437, https://doi.org/10.1175/JCLI-D-19- circulation in the northern South China Sea. Sci. China Earth 0679.1. Sci., 61, 560–571, https://doi.org/10.1007/s11430-017-9152-y. Xie S.-P., Q. Xie, D. Wang, and W.-T. Liu, 2003: Summer upwell- Sun, S., Y. Fang, Y. Zu, and B. Liu, 2020: Seasonal characteristics ing in the South China Sea and its role in regional climate of mesoscalecouplingbetween thesea surfacetemperature variations. J. Geophys. Res., 108, 3261, https://doi.org/10.1029/ and wind speed in the South China Sea. J. Climate, 33, 625– 2003JC001867. 638, https://doi.org/10.1175/JCLI-D-19-0392.1. Yao, Y., and C. Wang, 2021: Variations in summer marine heat- Tan, H., and R. Cai, 2018: What caused the record-breaking waves in the South China Sea. J. Geophys. Res. Oceans, 126, warming in East China seas during August 2016? Atmos. Sci. Lett., 19, e853, https://doi.org/10.1002/asl.853. e2021JC017792, https://doi.org/10.1029/2021JC017792. Wang, D. X., and Coauthors, 2013: Progress of regional oceanog- }}, J. Wang, J. Yin, and X. Zou, 2020: Marine heatwaves in Chi- raphy study associated with western boundary current in the na’s marginal seas and adjacent offshore waters: Past, present, South China Sea. Chin. Sci. Bull., 58, 1205–1215, https://doi. and future. J. Geophys. Res. Oceans, 125, e2019JC015801, org/10.1007/s11434-012-5663-4. https://doi.org.10.1029/2019JC015801. Wang,Q., L. Zeng, J.Li, J. Chen,Y. K.He, J. L. Yao, D. X. Wang, Zhang, Y., Y. Du, M. Feng, and S. Hu, 2021: Long-lasting marine and W. D. Zhou, 2018: Observed cross-shelf flow induced by heatwaves instigated by ocean planetary waves in the tropical mesoscale eddies in the northern South China Sea. J. Phys. Oce- Indian Ocean during 2015–2016 and 2019–2020. Geophys. Res. anogr., 48, 1609–1628, https://doi.org/10.1175/JPO-D-17-0180.1. Lett., 48, e2021GL095350, https://doi.org/10.1029/2021GL095350. }}, W. D. Zhou, L. Zeng, J. Chen, Y. K. He, and D. X. Wang, Zhao, R. X., and X. H. Zhu, 2016: Weakest winter South China 2020: Intraseasonal variability of cross-slope flow in the Sea western boundary current caused by the 2015–2016 El northern South China Sea. J. Phys. Oceanogr., 50,2071–2084, Nino event. J. Geophys. Res. Oceans, 121,7673–7682, https:// https://doi.org/10.1175/JPO-D-19-0293.1. doi.org/10.1002/2016JC012252. }}, Y. X. Wang, J. P. Sui, W. D. Zhou, and D. N. Li, 2021: }}, }}, and X. Y. Guo, 2017: The impact of monsoon winds Effects of weak and strong winter currents on the thermal and mesoscale eddies on thermohaline structures and circula- state of the South China Sea. J. Climate, 34, 313–325, https:// tion patterns in the northern South China Sea. Cont. Shelf doi.org/10.1175/JCLI-D-19-0790.1. Res., 143, 240–256, https://doi.org/10.1016/j.csr.2016.06.009. Wei, J., M. Li, P. Malanotte-Rizzoli, A. L. Gordon, and D. Wang, Zhu, Y. H., J. C. Sun, Y. G. Wang, and S. J. Li, 2019: Overview 2016: Opposite variability of Indonesian throughflow and of the multi-layer circulation in the South China Sea. Prog. South China Sea throughflow in the Sulawesi Sea. J. Phys. Oceanogr., 175, 171–182, https://doi.org/10.1016/j.pocean.2019. Oceanogr., 46, 3165–3180, https://doi.org/10.1175/JPO-D-16- 0132.1. 04.001. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Physical Oceanography American Meteorological Society

Properties and Drivers of Marine Heat Waves in the Northern South China Sea

Download PDF
11 pages

Loading next page...
 
/lp/american-meteorological-society/properties-and-drivers-of-marine-heat-waves-in-the-northern-south-lquQj0Ms07

References (53)

Publisher
American Meteorological Society
Copyright
Copyright © American Meteorological Society
ISSN
1520-0485
eISSN
1520-0485
DOI
10.1175/jpo-d-21-0236.1
Publisher site
See Article on Publisher Site

Abstract

MAY 2022 WA N G E T A L . 917 a,b b,c a,b b b a,b QIANG WANG, BO ZHANG, LILI ZENG, YUNKAI HE, ZEWEN WU, AND JU CHEN Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou, China State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China University of Chinese Academy of Sciences, Beijing, China (Manuscript received 21 October 2021, in final form 28 January 2022) ABSTRACT: The properties and heat budget of marine heat waves (MHWs) on the northern South China Sea (SCS) continental shelf are investigated. MHWs with warming amplitudes above 1.58C occur mainly along the coast, and their temperature anomaly decreases toward the open sea. MHWs with 18–1.58C warming and duration , 20 days dominate the northern SCS continental shelf. A heat budget analysis indicates that the main heat source is the sea surface net heat flux. Oceanic processes are dominated by the advection of mean temperature by the anomalous horizontal velocity (advha). The net contribution of advha always cools the upper layer of the ocean, resulting in the decay of MHWs. Active cross- slope water exchanges exist at the east and west sides of the northern SCS continental shelf edge, which makes the domi- nant contributions to the advha. In the MHW developing phase, the west (east) side makes a positive (negative) contribu- tion to the advha. In the decay phase, both sides make a negative contribution to the advha, resulting in the rapid decay of MHWs. Although the contribution of advha to the heat budget varies along the northern SCS continental shelf edge, its net effect always cools the MHWs over the shelf. These results provide new insight into the characteristics and formation mechanism of MHWs on the northern SCS continental shelf; in particular, they clarify the respective contributions of air–sea flux and oceanic processes to MHWs. SIGNIFICANCE STATEMENT: Marine heat waves (MHWs) are unusual warming events in oceans that heavily affect marine ecosystems and arouse great concern from citizens. MHWs are active in the northern South China Sea (SCS) continental shelf. On the northern SCS continental shelf, the sea surface net heat flux is the main heat source of MHWs, and ocean current anomalies always cool the upper layer of the ocean. Active cross-slope water exchange at the east and west sides of the northern SCS continental shelf edge is the main oceanic way that cools the water on the shelf, eventually resulting in the decay of MHWs. KEYWORDS: Advection; Extreme events; Mixed layer; Warm water volume; Surface temperature; Oceanic variability 1. Introduction 2017). Shortwave radiation and ocean advection anomalies contribute jointly to MHWs in the East China and South Yel- Marine heat waves (MHWs) are unusual warming events in low Seas (Tan and Cai 2018; Gao et al. 2020). In coastal oceans, and they have been occurring with increasing fre- waters off western Australia, MHW events are attributed to quency over the past few decades (Hobday et al. 2016; Oliver variations in heat advection by the Leeuwin Current and heat et al. 2017, 2018). Many MHW events have been recorded flux across the air–sea interface (Benthuysen et al. 2014, 2018; and are widely distributed on continental shelves and open Feng et al. 2013; Oliver et al. 2017). Local warming from the seas (Hobday et al. 2018; Oliver et al. 2021). MHW is not just atmosphere dominates the MHW events in the northwest a physical oceanographic phenomenon; it also exerts consid- Atlantic (Chen et al. 2014). Long-lasting MHWs in the tropi- erable influence on regional marine ecosystems (Genevier cal Indian Ocean were recorded, which were maintained by et al. 2019). Discrete MHWs can drive a gradual change in downwelling Rossby waves (Zhang et al. 2021). Strong sub- species distribution and affect the diversity and mortality of surface MHWs in the tropical western Pacific Ocean were commercial fisheries (Mills et al. 2013; Wernberg et al. 2013; observed, and anomalous sea surface convergence and Caputi et al. 2016). The increasing trend in MHW intensity Ekman downwelling played an important role in its formation and frequency has led to growing research interest (Collins (Hu et al. 2021). Reduced heat loss from the ocean to the et al. 2019). atmosphere and weak cold advection in the upper ocean can The contributions of atmospheric and oceanic processes to sometimes cause strongly positive temperature anomalies in MHW development vary in different oceans (Schlegel et al. the northeast Pacific(Bond et al. 2015; Di Lorenzo and Mantua 2016). Recent literature has provided a more detailed Denotes content that is immediately available upon publica- study of northeast Pacific MHWs, and double-peak and sin- tion as open access. gle-peak categories have been identified (Chen et al. 2021a). For the double-peak, the first peak is attributed to the surface heat flux, while the vertical entrainment and diffusion Corresponding authors: Qiang Wang, wqiang@scsio.ac.cn; Ju Chen, jchen@scsio.ac.cn are responsible for the second peak 5 months later. For the DOI: 10.1175/JPO-D-21-0236.1 Ó 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses). 918 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 FIG. 1. Composite properties of marine heat waves (MHWs). (a) Mean intensity and (b) duration of MHWs derived from the OISST dataset. (c),(d) As in (a) and (b), but from the OFES dataset. The black line indicates the southern boundary of the northern South China Sea continental shelf. Gray dots indicate where the composite results exceed 95% significance (Student’s t test). single-peak category, vertical entrainment and diffusion are budget on the northern SCS continental shelf. Although the the leading contributors. A persistent and intense MHW overall characteristics of the MHW properties in the SCS event in the northeast Pacific during 2019–20 was observed, have been discussed in previous studies (Cai et al. 2016; Li followed by a La Nina ˜ event, while other recorded double- et al. 2019; Yao et al. 2020; Yao and Wang 2021), details of peak events were associated with El Nino ˜ or neutral condi- MHWs in the northern SCS and their driving factors remain tions (Chen et al. 2021b). unclear, especially for the respective contributions of atmo- In the South China Sea (SCS), the occurrence frequency, spheric and oceanic processes to MHW development. intensity, and duration of MHW events have all increased The northern SCS continental shelf is close to the main- under global warming (Li et al. 2019; Yao and Wang 2021) land, and its thermal state greatly modulates the ecosystem and are projected to continue increasing in future decades and economic activities. The frequency and intensity of (Yao et al. 2020). El Nino ˜ –Southern Oscillation strongly regu- MHWs on the northern SCS continental shelf are continu- lates SCS MHWs (Liu et al. 2022). Both the sea surface net ously enhanced, attracting wide attention. Their drivers heat flux and ocean advection can cause extreme warming urgently need to be elucidated to better understand and pre- events at the basin scale (Xie et al. 2003; Xiao et al. 2018, dict their variability. Therefore, this paper focuses on the 2020; Wang et al. 2021). The western boundary current properties and drivers of the MHWs on the northern SCS (WBC) of the SCS is responsible for the advection of cooler continental shelf. water (Liu et al. 2004; Fang et al. 2013; Wei et al. 2016; Zhao The remainder of this paper is organized as follows. Data and methods are discussed in section 2. Section 3 describes and Zhu 2016; Zhao et al. 2017; Sun et al. 2020); weakening the properties of MHWs on the northern SCS continental or a complete shutdown of the WBC contributes greatly to shelf. The MHW heat budget is analyzed in section 4. Finally, the warming of the SCS upper layer (Wang et al. 2021). conclusions are given in section 5. The SCS MHW events occur mainly on the continental shelf, especially in the northern SCS (Cai et al. 2016; Li et al. 2019; Yao et al. 2020), where the SCS WBC divides the outer 2. Data and methods shelf edge from the open sea (Wang et al. 2013; Zhu et al. a. Data 2019). Multiple band-like currents flowing along the topogra- phy on the northern SCS continental shelf regulate an along- Observation data for sea surface temperature (SST) are topography distribution of climatological temperature (Shu used, together with model output, for a range of ocean varia- et al. 2018). Due to the instability of the SCS large-scale slope bles. Because of data availability, the period 1982–2017 is current, there are also active cross-slope water exchanges used in this study. between the continental shelf and the open sea in the north- The SST data are from NOAA OISST version 2 (Reynolds ern SCS (Wang et al. 2018, 2020; Liu and Gan 2020a). Due to et al. 2007), which is a daily and 0.258 3 0.258 gridded prod- the climatological temperature approximate belt-like distribu- uct from Advanced Very High Resolution Radiometer tion parallel to the slope (Chen et al. 2003), the cross-slope (AVHRR) satellite data, with bias correction using ship water exchanges imply potential contributions to the heat and buoy data. MAY 2022 WA N G E T A L . 919 FIG. 2. Comparisons of sea surface temperature (SST) between OISST and OFES. (a) 90-day low-pass SST. (b) SST anomaly derived from the SST minus the 90-day low-pass SST. (c) Composite series of SST anomalies for all marine heat waves (MHWs) in the northern SCS. The SSTs in (a) and (b) are averaged within the northern South China Sea continental shelf (bounded by the black line in Fig. 1); contours and colors represent the density of samples (red for high pair density and white for low pair density). The composite in (c) is produced as follows: the variable in each MHW is normalized to 0–1 in time, and then the variable of each MHW is composited. The results from the OGCM for the Earth Simulator OISST versus OFES validates the use of OFES on scales (OFES; Sasaki et al. 2007) are utilized in this study. OFES larger than or equal to seasonal time (Fig. 2a). A comparison was run at 0.18 3 0.18 spatial resolution and 54 vertical levels, of the SST anomaly, subtracting the 90-day low-pass SST, fur- forced by NCEP–NCAR reanalysis data. The ocean velocity, ther confirms the validation of OFES on an intraseasonal temperature, and surface net heat flux used in this study are scale (Fig. 2b). To investigate the differences in the life cycle supplied by OFES. OFES provides output in the form of of MHWs (see section 2b) expressed by OISST and OFES, a 3-day averages. The definition of an MHW is based on 11-day composite is produced from the MHW cases projected onto a window data (see section 2b). Thus, the 3-day averaged data time axis from 0 to 1 (Fig. 2c). Although the MHW of OFES can also be used to investigate the heat budget of MHWs. For is relatively weaker than that of OISST, it simulates a rela- convenience, the OFES outputs are all interpolated to daily tively real MHW evolution process. data using linear interpolation. The distribution of the mean Three mooring stations were located on the northern SCS intensity and duration of MHWs in the northern SCS derived slope. The sampling time interval was 1 h, and the vertical from OFES generally resembles that derived from OISST spatial resolution was 8 m at all stations. Velocity profiles (Fig. 1), which validates the use of OFES in the investigation at station DS01 (20.308N, 117.678E) were obtained from of MHWs. The 90-day low-pass SST scatter diagram of 18 September 2014 to 15 September 2015. Station DS02 920 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 FIG. 3. Taylor diagrams of the mean properties of each northern South China Sea marine heat wave (MHW) event derived from the (a) OISST dataset and (b) OFES dataset. The radius indicates the mean intensity of MHWs, and the angle indicates the duration. The percentage indicates the proportion of MHWs in that sector relative to the total number of MHWs. The color and size of each circle indicate the intensity and duration, respectively. Green: ,0.58C; yellow: between 0.58 and 18C; red: between 18 and 1.58C; purple: .1.58C. Small size: ,20 days; middle size: between 20 and 40 days; large size: .40 days. (19.718N, 116.998E) acquired data from 20 September 2014 to respectively. The term T is the daily SST on day d of year y,and 20 August 2018. Velocity profiles were obtained from 1 Janu- P (j) is the 90th percentile of X,where X ={T(y, d)|y # y # y , 90 s e ary 2014 to 12 June 2017 at station XS (17.298N, 111.368E). j 2 5 # d # j 1 5}. Two MHW metrics are used in the study: the duration Any velocity larger than 3 times the standard deviation at that (days between the start and end dates) and the intensity depth is defined as an invalid value. Short gaps caused by (average of daily intensity anomalies measured in 8C). invalid data were subsequently filled by linear interpolation. A 48-h low-pass filter was applied to the velocity data to c. Upper-ocean temperature budget remove high-frequency signals, and the data were then aver- The perturbation temperature budget equation, averaged aged over 24 h (daily). from the sea surface to a fixed depth h , can be written as fol- b. Definition of MHWs lows (Benthuysen et al. 2014; Feng et al. 2008; Oliver et al. 2017): Following Hobday et al. (2016), the climatology SST is defined relative to the time of year, using all data within an ­T Q () T 2 T m d 11-day window centered on the time of year from which the 2 u · =T 2 u · =T 2 w ­t rC h h p m m climatological mean is calculated, i.e., Eq. (1). advh advha Tend advv An MHW is defined as an anomalously warm, discrete, and Qnet prolonged event, which is a period when the daily SST is T 2 T () m d above a particular threshold for at least 5 days. The threshold 2 w , (3) on the day of the year is defined as the 90th percentile of the advva daily temperature within an 11-day window centered on this day in all years, i.e., Eq. (2): where T is the temperature and h is set as 50 m, with · rep- ye j15 resenting vertical integration dz . Parameter T is the Ty, d 2hm () T () j  , (1) average temperature over the upper 50 m; T is the water 11 y 2 y d () e s11 yy s dj25 temperature below h and is set to the value at 70 m. We also checked the values of h for 40 and 60 m and found that our results were not particularly sensitive to the choice of T () j  P () j , (2) 90 90 depthusedto calculate T .The term Q is the sea surface where T is the climatology SST and T is the 90th percentile net heat flux anomaly and is positive for heat input to the m 90 of the daily SST. Parameter j is the day of the year, and y and ocean; r = 1025 kg m is the reference density of seawater, 21 21 y are the start and end of the climatological base period, and C = 4007 J Kg K is the specific heat of seawater. e p MAY 2022 WA N G E T A L . 921 (a) SST anomaly of MHWs (d) Temperature advection by anomalous flo w 2 advha advha(South) 1.1 Developing Mature Decay advha(North) 0.9 −2 0.7 −4 0.5 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Period (b) Heat budget terms (e) Temperature advection by anomalous flow (South) 0.8 −1 Qnet advha −2 0.6 advh advwa −3 advw −4 0 0.2 0.4 0.6 0.8 1 0.4 (c) Ocean versus atmosphere 0.2 −2 −4 |Ocean/Ato|=0.65 |Ocean/Ato|=3.90 |Ocean/Ato|= 0.94 Qnet −6 112 113 114 115 116 117 118 119 Sum of ocean process Longitude −8 −7o [´ 10 C/s] 0 0.2 0.4 0.6 0.8 1 Period −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 FIG. 4. Composite heat budget terms for all marine heat waves (MHWs) in the northern SCS. (a) Composite series of SST anomalies. (b) Composite heat budget terms in Eq. (1). (c) Comparisons between the contribution of net heat flux (Qnet) and temperature advection (i.e., the sum of all terms except for Qnet). (d) Advection of mean temperature by the anomalous horizontal velocity (advha); south (north) indicates the contributions of advection at the south (north) boundary. (e) Composite profiles of advha at the south boundary. The composite is produced as follows: the variable in each MHW is normalized to 21 in time, and then the variable of each MHW is composited. Gray dots indi- cate where the composite results exceed 95% significance (Student’s t test). Vector u =(u, y) is the zonal and meridional velocity pair, advha  y () T 2 T dx dz south south and w is the vertical velocity; = =(­ , ­ ) is the horizontal t x y gradient operator. The overbar indicates the climatological south mean, and the prime indicates an anomaly. 2 y () T 2 T dx dz north north o The left-hand side of Eq. (3) is the rate of change or ten- dency of the upper 50-m averaged temperature (Tend). On north the right-hand side, Qnet is the sea surface net heat flux 1 u () T 2 T dy dz anomaly forcing, advh is the advection of anomalous tempera- west west o ture by the mean horizontal velocity, advha is the advection west of mean temperature by the anomalous horizontal velocity, advv is the advection of anomalous temperature by the mean 2 u () T 2 T dy dz , (4) east east o vertical velocity, and advva is the advection of mean tempera- east ture by the anomalous vertical velocity. In the northern SCS, the contribution of advection to the temperature anomalies can be represented by the tempera- where t is the volume of the box enclosed by the northern and ture flux at the southern and northern boundaries (Feng southern boundaries and additional eastern and western et al. 2008; Gao et al. 2020). Following Lee et al. (2004), boundaries. Parameter T is the daily average temperature in the contributions of advection at these northern and south- the northern SCS, which is used as the reference temperature ern boundaries can be further written (using advha as an in the heat advection. By including the T in the heat advec- example) as tion terms, unbalanced volume flux in one direction is assumed −7o −7o Temperature anomaly ( C) Terms ( ´ 10 C/s) Terms ( ´ 10 C/s) −7o Period Advha ( ´ 10 C/s) 922 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 FIG. 5. Distribution of composite heat budget terms for the (a) developing phase, (b) mature phase, and (c) decay phase. Qnet (8Cs ) is the net heat flux. The temperature anomaly (8C) is the temperature anomaly averaged over the upper 50 m. The Clim T and flow anom- aly indicates the climatological temperature and composite horizontal velocity anomaly for all marine heat wave events, averaged over the upper 50 m. The term advv 1 advva (8Cs ) indicates the vertical temperature advection, as in Eq. (1).The definition of each phase is shown in Fig. 4a. The heat budget terms have been multiplied by 1.0 3 10 . The purple vectors indicate the water flowing toward the deep sea. to carry waters with the same temperature as T so that it does OISST is 13.1 6 0.7 days, and OFES is 14.6 6 1.4 days. The not have a net contribution to the advection (Lee et al. 2004; spatial variation in duration is less than 2 days, and their evo- Feng et al. 2008). lution processes are similar (Fig. 2c), which suggests the lim- ited influence of the spatial duration difference between OFES and OISST on the validation of OFES to analyze 3. Properties of MHWs in the northern South China Sea MHW evolution. The small spatial variation also suggests that The mean properties of MHWs over the past four decades MHW occurrence and development are reasonably synoptic were calculated (Fig. 1). The intensity isolines run belt-like along the northern SCS continental shelf. parallel to the coastline, with decreasing values toward the The SST averaged over the northern SCS continental shelf open sea in the northern SCS (Figs. 1a,c). There are three bounded by the black line in Fig. 1 is used to identify the high-value centers in Zhanjiang Bay, the Pearl River estuary, northern SCS MHWs. The intensity–duration diagram of the and offshore Shantou, where the maximum SST anomaly can MHWs is shown in Fig. 3. Approximately 50% of the MHWs reach 28C. On the continental slope edge, the intensity have intensities in the range of 18–1.58C. One-third of MHWs decreases to ,1.58C. have intensities in the range 0.58–1.08C, and only ∼10% The mean MHW duration ranges from 12 to 15 days in the exceed 1.58C. In the OFES results, there are some MHWs northern SCS and is slightly longer on the western continental with intensities , 0.58C that do not appear in the OISST. This shelf than on the eastern shelf (Figs. 1b,d). In the northern inconsistency indicates the slightly underestimated SST anom- SCS continental shelf (bounded by line A in Fig. 1a), there is aly in the OFES. However, there are only six weak MHWs, a relatively significant spatial difference between OISST and ∼7% of the total number, which does not affect the composite OFES; however, the duration over the continental shelf of analysis of the heat budget in the next section. MAY 2022 WA N G E T A L . 923 FIG. 6. Composite upper 50-m averaged temperature gradient and velocity for all marine heat waves (MHWs) in the northern SCS. (a) Composite upper 50-m averaged temperature gradient projected onto the direction perpendicular to line A (marked in Fig. 1a), and onshore is positive. (b) Composite profiles of velocity perpendicular to line A and onshore are positive. The compos- ite is produced as follows: the variable in each MHW is normalized to 0–1 in time, and then the variable of each MHW is composited. Gray dots indicate where the composite results exceed 95% significance (Student’s t test). More than 80% of MHWs in the northern SCS continental weakens, and the negative advha strengthens, leading to an shelf have a duration of less than 20 days, and ∼10% last approximate equilibrium between the oceanic and atmo- between 20 and 40 days. MHWs longer than 40 days are rare, spheric processes (Figs. 4b,c). In the decay phase, Qnet weak- and only one case occurred in the OISST data (Fig. 3a). ens sharply and becomes negative, and the advha continues The main features of the MHW intensity and duration on to strengthen rapidly (Fig. 4b). The upper ocean begins to the northern SCS continental shelf from OISST and OFES lose heat, which is mainly associated with oceanic processes are relatively similar, which validates the use of OFES output (Figs. 4b,c). Therefore, the Qnet warms the upper ocean and in the following heat budget analysis. induces an MHW outbreak in the northern SCS continental shelf; oceanic processes, which are dominated by the advha, cool the upper ocean, and cause the MHW to die out. 4. Air–sea heat flux versus ocean current for MHWs The continental shelf edge, indicated by the black line in To investigate the development of MHWs in the northern Fig. 1, and the Taiwan Strait are the two pathways for the SCS continental shelf, a composite is produced from the advha on the northern SCS continental shelf. The southern MHW cases projected onto a time axis from 0 to 1. The devel- boundary (i.e., the continental shelf edge) is the main temper- oping, mature, and decay phases can be identified using the ature advection (advha) pathway (Fig. 4d), which is concen- composite SST anomaly (Fig. 4a). In the developing phase, trated on the east and west sides of the shelf (Fig. 4e). In the the net heat flux (Qnet) is the main source of warming, developing phase, the east and west sides make opposite con- whereas the advection of mean temperature by the anoma- tributions to the advha. In the decay phase, the temperature lous horizontal velocity (advha) always opposes the contribu- advection (advha) is negative on both sides, leading to a rapid tion of Qnet, cooling the upper layer of the ocean (Fig. 4b). In enhancement of negative advha, which induces MHW decay. total, the oceanic process offsets 65% of the contribution of The spatial distributions of each term are shown in Fig. 5. the atmospheric process (Fig. 4c). In the mature phase, Qnet In the developing phase, a large positive Qnet is mainly 0 924 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 (a) Developing phase: windstress and its curl dominates both the east and west sides at the continental shelf edge (Figs. 5c and 6b), which leads to a strong negative advha, 23 2 0.1dyn/cm resulting in the rapid decay of the MHW. The vertical temper- ature advection contributes little during the developing and mature phases (Fig. 4b and Figs. 5a,b). However, the vertical process contributes significantly to the decay of MHWs near the coast (Fig. 5c). −9 3 [10 dyn/cm ] On the northern SCS continental shelf, the variations in 110 112 114 116 118 120 the flow anomaly are mainly associated with wind stress 1.5 anomalies (Fig. 7). In the developing and mature phases, the (b) Mature phase: windstress and its curl wind stress anomalies mainly blow toward the northeast (Figs. 7a,b), which induces an eastward flow anomaly along 0.5 the coast (Figs. 5a,b). In the decay phase, the wind stress turned toward the southwest (Fig. 7c), driving the continental shelf flow to turn direction accordingly (Fig. 5c). The wind −0.5 stress curl is one dominant factor modulating the cross-iso- −1 bath movement over a widened shelf in the northeastern SCS 110 112 114 116 118 120 (Gan et al. 2013; Liu et al. 2020b). During the developing and −1.5 decay phases, a significant negative wind stress curl occupied (c) Decay phase: windstress and its curl −2 the northeastern SCS (Figs. 7a,c). Responding to the negative wind stress vorticity input into the ocean, the ocean volume potential vorticity decreases. The decrease in potential vortic- ity will make the seawater tend to cross the isobath, flowing toward the deep sea (Fig. 6b). In the mature phase, there is a positive wind stress curl over the northeastern SCS (Fig. 7b), which weakens the offshore flow (Fig. 6b). On the western side of the continental shelf, the wind stress curl is mainly neg- 110 112 114 116 118 120 ative and gradually enhances from the developing phase to Longitude ( E) the decay phase (Fig. 7), which corresponds to the offshore FIG. 7. Distribution of the composite wind stress curl and wind flow between 1128 and 1138E(Fig. 6b). The direction of cross- stress vector for the (a) developing phase, (b) mature phase, and slope flow east of Hainan Island is mainly determined by the (c) decay phase. The blue lines indicate the topography isobaths. direction of the slope current, i.e., offshore (onshore) flow is Shading is the wind stress curl, and vectors are the wind stress. Red induced by the southwest (northeast) slope current due to the lines are the zero contours of the wind stress curl. block of Hainan Island. Three mooring stations were used to investigate the flow anomalies on the east and west sides of the northern SCS located in the center and east of the continental shelf, slope (Fig. 8b). The northern SCS continent shelf average although the entire northern SCS continental shelf warms SST anomaly during the mooring observation period is shown (Fig. 5a). Warming on the west side of the continental shelf is in Fig. 8a, and the MHW events are marked in purple. The attributed to the positive advha (Fig. 4e), where a northward velocities of the MHWs are shown in Figs. 8c, 8e, and 8g, flow anomaly carries high mean temperature water from the and composites of the MHWs projected onto a time axis south (Fig. 5a and Figs. 6a,b). Along the southern boundary from 0 to 1 are shown in Figs. 8d, 8f,and 8h. On the east of the continental shelf (i.e., line A in Fig. 1a), the tempera- side, the composite flow anomalies are mainly offshore and ture gradient perpendicular to line A is mainly negative (i.e., increase gradually with MHW development (Figs. 8d,f). On decreases northward; Fig. 6a). The onshore flow anomaly on the west side, the flow anomalies are mainly northeastward the western side (Fig. 6b), jointly with the negative tempera- and decrease during the MHW decay phase (Fig. 8h). The ture gradient, induces a positive advha (2 y ­T=­y , where y changes of OFES flow anomalies on both the east and west indicates the direction perpendicular to line A). On the east sides are generally consistent with the mooring observa- side, the strong offshore flow anomaly plus a negative temper- tions, which confirms the validation of the OFES oceanic ature gradient induces a large negative advha (Figs. 6a,b). In processes during MHWs. the mature phase, although Qnet decreases, the upper layer temperature anomaly reaches a maximum after early accumu- 5. Conclusions lation (Fig. 5b). The east side flow anomaly is still offshore and contributes negatively to the advha (Fig. 5b and Fig. 6b). Using the OISST and OFES datasets, the properties and The west side cross-slope flow anomaly weakens, which grad- heat budget of MHWs on the northern SCS continental shelf ually weakens its positive advha (Fig. 4e). In the decay phase, were analyzed. The intensity of MHWs is strong along the Qnet becomes negative and the upper layer temperature coast and decreases toward the open sea. MHWs with anomaly weakens (Fig. 5c). The offshore flow anomaly 18–1.58C warming and a duration of ,20 days dominate the Latitude ( N) MAY 2022 WA N G E T A L . 925 (a) SST anomaly ( C) (b) Topography and stations DS01 DS02 −1 18 XS 110 112 114 116 118 120 (c) DS01: Upper 100m velocity (d) DS01: composites of Upper 100m velocit y 0.8 0.2 0.1m/s 0.5m/s 0.4 0.1 0 0 −0.4 −0.1 −0.8 −0.2 (e) DS02: Upper 100m velocit y (f) DS02: composites of Upper 100m velocit y 0.8 0.2 0.1m/s 0.5m/s 0.4 0.1 0 0 −0.4 −0.1 −0.8 −0.2 (g) XS: Upper 100m velocit y (h) XS: composites of Upper 100m velocit y 0.8 0.2 0.1m/s 0.5m/s 0.4 0.1 0 0 −0.1 −0.4 −0.2 −0.8 0 0.25 0.5 0.75 1 Period Time FIG. 8. Observed velocity from mooring stations. (a) Time series of the OISST sea surface temperature (SST) anomaly in the northern South China Sea (SCS); purple indicates a marine heat wave (MHW) event. (b) Mooring stations and topography in the northern SCS. (c) Velocity vectors averaged in the upper 100 m at station DS01; red vectors indicate MHW events. (d) Composite DS01 velocity vector for MHW events, i.e., red vectors in (c). The red vectors for each MHW event have been normalized to 0 to 1 in time as in Fig. 4, and then all the events are composited as (d). (e),(g) As in (c), but for stations DS02 and XS. (f),(h) As in (d), but for stations DS02 and XS. northern SCS continental shelf. MHWs with warming of advection across the southern continental shelf edge in the 0.58–18C account for approximately one-third of the total, and northern SCS. The east and west sides of the continental shelf those .1.58C account for approximately one-tenth. Approxi- edge are the two major pathways; these make opposing contri- mately one-tenth of the MHWs persist for 20–40 days; those butions to the heat budget in the MHW developing phase (i.e., with durations of .40 days are very rare, and only one case negative and positive on the eastern and western sides, respec- was identified from the OISST dataset. tively) but are of the same sign in the decay phase (i.e., both The heat budget indicates that the net heat flux (Qnet) is negative). the main driver of the MHWs on the northern SCS continen- tal shelf. Oceanic processes, which are dominated by the Acknowledgments. This work is supported by the National advection of mean temperature by anomalous horizontal Key Research and Development Program (Grant 2017YFA velocity (advha), always cool the upper layer ocean, leading 0603201), Key Special Project for Introduced Talents Team of to the decay of MHWs. In the developing phase, Qnet is Southern Marine Science and Engineering Guangdong Labora- stronger than the negative oceanic process contribution to the tory (Guangzhou) (GML2019ZD0304), the National Natural heat budget, which results in warming of the upper layer. In Science Foundation of China (Grant 42076209), the Rising Star the mature phase, there is an equilibrium between the Qnet Foundation of the South China Sea Institute of Oceanology and oceanic process contributions. In the decay phase, both the (NHXX2019WL0101), and the Science and Technology Qnet and oceanic process contributions are negative, but the Planning Project of Guangzhou (202102080363, 20200 ocean process contribution is approximately 4 times that of Qnet. 2030490). The ADCP velocity profiles are supplied by the The oceanic process contribution to the heat budget is dom- Xisha Deep Sea Observatory, a member of the Network inated by advha, which is attributed mainly to the temperature of Field Observation and Research Stations of the Chinese Jan/14 Jun/14 Nov/14 Apr/15 Sep/15 Feb/16 Jul/16 Dec/16 May/17 Oct/17 Jan/14 Jun/14 Nov/14 Apr/15 Sep/15 Feb/16 Jul/16 Dec/16 May/17 Oct/17 Velocity (m/s) SST anomaly ( C) Velocity (m/s) 926 J OUR N A L O F P HY SI C A L O C E A N OGR A P HY VOLUME 52 }}, M. J. McPhaden, S.-P. Xie, and J. Hafner, 2013: La Nina ˜ Academy of Sciences. The numerical calculation is sup- forces unprecedented Leeuwin Current warming in 2011. Sci. ported by the high performance computing division and Rep., 3, 1277, https://doi.org/10.1038/srep01277. Ms. Dandan Sui and Dr. Wei Zhou of the South China Gan, J. P.,H.S.Ho, andL. L.Liang,2013: Dynamics of intensi- Sea Institute of Oceanology. fied downwelling circulation.over a widened shelf in the northeastern South China Sea. J. Phys. Oceanogr., 43,80–94, Data availability statement. The OISST and OFES data https://doi.org/10.1175/JPO-D-12-02.1. for this paper are available at http://apdrc.soest.hawaii.edu/ Gao, G., and Coauthors, 2020: Drivers of marine heatwaves in the data/data.php. East China Sea and the South Yellow Sea in three consecu- tive summers during 2016–2018. J. Geophys. Res. Oceans, 125, e2020JC016518, https://doi.org/10.1029/2020JC016518. REFERENCES Genevier, L. G. C., T. Jamil, D. E. Raitsos, G. Krokos, and I. Hoteit, 2019: Marine heatwaves reveal coral reef zones Benthuysen, J., M. Feng, and L. Zhong, 2014: Spatial patterns susceptible to bleaching in the Red Sea. Global Change of warming off Western Australia during the 2011 Ninga- Biol., 25, 2338–2351, https://doi.org/10.1111/gcb.14652. loo Nino ˜ : Quantifying impacts of remote and local forcing. Hobday, A. J., and Coauthors, 2016: A hierarchical approach to Cont. Shelf Res., 91, 232–246, https://doi.org/10.1016/j.csr. defining marine heatwaves. Prog. Oceanogr., 141, 227–238, 2014.09.014. https://doi.org/10.1016/j.pocean.2015.12.014. Benthuysen, J. A., E. C. J. Oliver, M. Feng, and A. G. Marshall, }}, and Coauthors, 2018: Categorizing and naming marine heat- 2018: Extreme marine warming across tropical Australia dur- waves. Oceanography, 31, 162–173, https://doi.org/10.5670/ ing austral summer 2015–16. J. Geophys. Res. Oceans, 123, oceanog.2018.205. 1301–1326, https://doi.org/10.1002/2017JC013326. Hu, S., and Coauthors, 2021: Observed strong subsurface marine Bond, N. A., M. F. Cronin, H. Freeland, and N. Mantua, 2015: heatwaves in the tropical western PacificOcean. Environ. Causes and impacts of the 2014 warm anomaly in the NE Res. Lett., 16, 104024, https://doi.org/10.1088/1748-9326/ac26f2. Pacific. Geophys. Res. Lett., 42,3414–3420, https://doi.org/10. Lee, T., I. Fukumori, and B. Tang, 2004: Temperature advection: 1002/2015GL063306. Internal versus external processes. J. Phys. Oceanogr., 34, Cai, R., H. Tan, and Q. Qi, 2016: Impacts of and adaptation to 1936–1944, https://doi.org/10.1175/1520-0485(2004)034,1936: inter-decadal marine climate change in coastal China seas. Int. TAIVEP.2.0.CO;2. J. Climatol., 36,3770–3780, https://doi.org/10.1002/joc.4591. Li, Y., G. Ren, Q. Wang, and Q. You, 2019: More extreme Caputi, N., and Coauthors, 2016: Management adaptation of marine heatwaves in the China Seas during the global warm- invertebrate fisheries to an extreme marine heatwave event ing hiatus. Environ. Res. Lett., 14, 104010, https://doi.org/10. at a global warming hot spot. Ecol. Evol., 6,3583–3593, 1088/1748-9326/ab28bc. https://doi.org/10.1002/ece3.2137. Liu, K., K. Xu, C. Zhu, and B. Liu, 2022: Diversity of marine Chen, J.-M., C.-P. Chang, and T. Li, 2003: Annual cycle of the heatwaves in the South China Sea regulated by ENSO phase. South China Sea surface temperature using the NCEP/ J. Climate, 35, 877–893, https://doi.org/10.1175/JCLI-D-21- NCAR reanalysis. J. Meteor. Soc. Japan. Ser. II, 81, 879–884, 0309.1. https://doi.org/10.2151/jmsj.81.879. Liu, Q., X. Jiang, S.-P. Xie, and W. T. Liu, 2004: A gap in the Indo- Chen, K., G. G. Gawarkiewicz, S. J. Lentz, and J. M. Bane, 2014: Pacific warm pool over the South China Sea in boreal winter: Diagnosing the warming of the northeastern U.S. coastal Seasonal development and interannual variability. J. Geophys. ocean in 2012: A linkage between the atmospheric jet stream Res., 109, C07012, https://doi.org/10.1029/2003JC002179. variability and ocean response. J. Geophys. Res. Oceans, 119, Liu, Z., and J. Gan, 2020a: A modeling study of estuarine-shelf 218–227, https://doi.org/10.1002/2013JC009393. circulation using a composite tidal and subtidal open bound- Chen,Z., J. Shi,and C. Li,2021a:Two types of warmblobs in the ary condition. Ocean Modell., 147, 101563, https://doi.org/10. northeast Pacific and their potential effect on the El Nino. Int. 1016/j.ocemod.2019.101563. J. Climatol., 41,2810–2827, https://doi.org/10.1002/joc.6991. }}, T. Zu, and J. Gan, 2020b: Dynamics of cross-shelf water }}, }},Q.Liu,H.Chen, and C. Li, 2021b: A persistent and exchanges off Pearl River Estuary in summer. Prog. Ocean- intense marine heatwave in the Northeast Pacific during ogr., 189, 102465, https://doi.org/10.1016/j.pocean.2020.102465. 2019–2020. Geophys. Res. Lett., 48, e2021GL093239, https:// Mills, K. E., and Coauthors, 2013: Fisheries management in a doi.org/10.1029/2021GL093239. changing climate lessons from the 2012 ocean heat wave in Collins, M., and Coauthors, 2019: Extremes, abrupt changes and the Northwest Atlantic. Oceanography, 26, 191–195, https:// managing risk. IPCC Special Report on the Ocean and Cryo- ¨ doi.org/10.5670/oceanog.2013.27. sphere in a Changing Climate, H.-O. Portner et al., Eds., Oliver, E. C. J., and Coauthors, 2017: The unprecedented 2015/16 IPCC, 589–655, https://doi.org/10.1017/9781009157964.008. Tasman Sea marine heatwave. Nat. Commun., 8, 16101, Di Lorenzo, E., and N. Mantua, 2016: Multi-year persistence of https://doi.org/10.1038/ncomms16101. the 2014/15 North Pacific marine heatwave. Nat. Climate }}, and Coauthors, 2018: Longer and more frequent marine Change, 6, 1042–1047, https://doi.org/10.1038/nclimate3082. heatwaves over the past century. Nat. Commun., 9, 1324, Fang, Y., C. Tana, S. Sun, and B. Liu, 2013: Impact of the climate https://doi.org/10.1038/s41467-018-03732-9. change on the western boundary current in the South China }}, J. A. Benthuysen, S. Darmaraki, M. G. Donat, A. J. Hobday, Sea: An assessment based on numerical simulations. Malays. N. J. Holbrook, R. W. Schlegel, and A. S. Gupta, 2021: Marine J. Sci., 32,347–356. Feng,M., A. Biastoch,C.Boning, N. Caputi, and G. Meyers, heatwaves. Annu. Rev. Mar. Sci., 13, 313–342, https://doi.org/10. 2008: Seasonal and interannual variations of upper ocean 1146/annurev-marine-032720-095144. heat balance off the west coast of Australia. J. Geophys. Res., Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, 113, C12025, https://doi.org/10.1029/2008JC004908. and M. G. Schlax, 2007: Daily high-resolution-blended MAY 2022 WA N G E T A L . 927 analyses for sea surface temperature. J. Climate, 20, 5473– Wernberg, T., D. A. Smale, F. Tuya, M. S. Thomsen, T. J. Langlois, 5496, https://doi.org/10.1175/2007JCLI1824.1. T. deBettignies, S. Bennett, and C. S. Rousseaux, 2013: An Sasaki, H., M. Nonaka, Y. Masumoto, Y. Sasai, H. Uehara, and extreme climatic event alters marine ecosystem structure in a H. Sakuma, 2007: An eddy-resolving hindcast simulation of global biodiversity hotspot. Nat. Climate Change, 3,78–82, the quasi-global ocean from 1950 to 2003 on the Earth Simu- https://doi.org/10.1038/nclimate1627. lator. High Resolution Numerical Modeling of the Atmo- Xiao, F., L. Zeng, Q. Liu, W. Zhou, and D. Wang, 2018: Extreme sphere and Ocean, W. Ohfuchi and K. Hamilton, Eds., subsurface warm events in the South China Sea during 1998/ Springer, 157–185. 99 and 2006/07: Observations and mechanisms. Climate Dyn., Schlegel, R. W., E. C. J. Oliver, S. Perkins-Kirkpatrick, A. Kruger, 50,115–128, https://doi.org/10.1007/s00382-017-3588-y. and A. J. Smit, 2017: Predominant atmospheric and oceanic }}, D. Wang, and L. Yang, 2020: Can tropical Pacific winds patterns during coastal marine heatwaves. Front. Mar. Sci., 4, enhance the footprint of the interdecadal Pacificoscillation 323, https://doi.org/10.3389/fmars.2017.00323. on the upper-ocean heat content in the South China Sea? J. Shu, Y., Q. Wang, and T. Zu, 2018: Progress on shelf and slope Climate, 33,4419–4437, https://doi.org/10.1175/JCLI-D-19- circulation in the northern South China Sea. Sci. China Earth 0679.1. Sci., 61, 560–571, https://doi.org/10.1007/s11430-017-9152-y. Xie S.-P., Q. Xie, D. Wang, and W.-T. Liu, 2003: Summer upwell- Sun, S., Y. Fang, Y. Zu, and B. Liu, 2020: Seasonal characteristics ing in the South China Sea and its role in regional climate of mesoscalecouplingbetween thesea surfacetemperature variations. J. Geophys. Res., 108, 3261, https://doi.org/10.1029/ and wind speed in the South China Sea. J. Climate, 33, 625– 2003JC001867. 638, https://doi.org/10.1175/JCLI-D-19-0392.1. Yao, Y., and C. Wang, 2021: Variations in summer marine heat- Tan, H., and R. Cai, 2018: What caused the record-breaking waves in the South China Sea. J. Geophys. Res. Oceans, 126, warming in East China seas during August 2016? Atmos. Sci. Lett., 19, e853, https://doi.org/10.1002/asl.853. e2021JC017792, https://doi.org/10.1029/2021JC017792. Wang, D. X., and Coauthors, 2013: Progress of regional oceanog- }}, J. Wang, J. Yin, and X. Zou, 2020: Marine heatwaves in Chi- raphy study associated with western boundary current in the na’s marginal seas and adjacent offshore waters: Past, present, South China Sea. Chin. Sci. Bull., 58, 1205–1215, https://doi. and future. J. Geophys. Res. Oceans, 125, e2019JC015801, org/10.1007/s11434-012-5663-4. https://doi.org.10.1029/2019JC015801. Wang,Q., L. Zeng, J.Li, J. Chen,Y. K.He, J. L. Yao, D. X. Wang, Zhang, Y., Y. Du, M. Feng, and S. Hu, 2021: Long-lasting marine and W. D. Zhou, 2018: Observed cross-shelf flow induced by heatwaves instigated by ocean planetary waves in the tropical mesoscale eddies in the northern South China Sea. J. Phys. Oce- Indian Ocean during 2015–2016 and 2019–2020. Geophys. Res. anogr., 48, 1609–1628, https://doi.org/10.1175/JPO-D-17-0180.1. Lett., 48, e2021GL095350, https://doi.org/10.1029/2021GL095350. }}, W. D. Zhou, L. Zeng, J. Chen, Y. K. He, and D. X. Wang, Zhao, R. X., and X. H. Zhu, 2016: Weakest winter South China 2020: Intraseasonal variability of cross-slope flow in the Sea western boundary current caused by the 2015–2016 El northern South China Sea. J. Phys. Oceanogr., 50,2071–2084, Nino event. J. Geophys. Res. Oceans, 121,7673–7682, https:// https://doi.org/10.1175/JPO-D-19-0293.1. doi.org/10.1002/2016JC012252. }}, Y. X. Wang, J. P. Sui, W. D. Zhou, and D. N. Li, 2021: }}, }}, and X. Y. Guo, 2017: The impact of monsoon winds Effects of weak and strong winter currents on the thermal and mesoscale eddies on thermohaline structures and circula- state of the South China Sea. J. Climate, 34, 313–325, https:// tion patterns in the northern South China Sea. Cont. Shelf doi.org/10.1175/JCLI-D-19-0790.1. Res., 143, 240–256, https://doi.org/10.1016/j.csr.2016.06.009. Wei, J., M. Li, P. Malanotte-Rizzoli, A. L. Gordon, and D. Wang, Zhu, Y. H., J. C. Sun, Y. G. Wang, and S. J. Li, 2019: Overview 2016: Opposite variability of Indonesian throughflow and of the multi-layer circulation in the South China Sea. Prog. South China Sea throughflow in the Sulawesi Sea. J. Phys. Oceanogr., 175, 171–182, https://doi.org/10.1016/j.pocean.2019. Oceanogr., 46, 3165–3180, https://doi.org/10.1175/JPO-D-16- 0132.1. 04.001.

Journal

Journal of Physical Oceanography – American Meteorological Society

Published: May 27, 2022

There are no references for this article.