**4. Response of UEH to typhoon passage**

## **4.1 Statistics of typhoon**

Previous investigators have addressed that the hydrological characteristics and dynamic structure of the upwelling may have dramatic changes after typhoon passage [28, 38, 47]. The UEH is located in the pathway of typhoons formed both in the western Pacific Ocean and the SCS [47, 48]. In order to investigate the sudden impact of typhoon forcing on the UEH, Xie et al. [31] analyzed the statistics of typhoons passing the UEH zone during the past 34 years from 1982 to 2015.

As shown in **Figure 11**, there were a total of 42 tropical cyclones passing the UEH between April and September from 1982 to 2015, of which 24 cyclones originated from the western Pacific and 18 from the SCS. The cyclones originating from the Pacific passed over the UEH mostly in July, while the locally generated cyclones in the SCS appeared mostly in August. Most cyclones moved northwestward into the research area, while several cases passed this region north-/northwestward or even parallel to the coastline. **Figure 11g** shows the incidence angles of the cyclones, which is defined as the angle between the pathway of cyclone when entering the upwelling region and the trend of coastline, i.e., 0° for the passage parallel to the coastline and 90° for perpendicular. Twenty-five cyclones entered this area with

*Response of Coastal Upwelling East of Hainan Island in the South China Sea to Sudden Impact… DOI: http://dx.doi.org/10.5772/intechopen.88828*

#### **Figure 11.**

*Tracks and incidence angles of cyclones passing over the UEH zone (red box) during the period from April to September of 1982 to 2015. Black lines in (a)–(f) are tracks of tropical cyclones originated from the Pacific and blue lines from the SCS. Red boxes represent the UEH zone for statistics (cited from [31]).*

incidence angle between 70° and 90°, among which seventeen passages were nearly perpendicular to the coastline (80°–90° of the incidence angle). Only in two cases the incidence angles were smaller than 40°.

As for the monthly distribution, the passages of tropical cyclones were prevailing in summer from June to August, when upwelling had been well developed. There are 36 cases in summer, accounting for 86% of all passing cyclones, in which 18 cases occurred in July. As of the cyclone intensity, the most frequent category was severe tropical storm (STS) with 12 cases during 34 years, followed by the tropical depression (TD) and typhoon (TY) of 10 and 9 cases, respectively. The three most prevailing types account for 74% of all passing tropical cyclones. The severest super typhoon (super TY) occurred in July 2014.

#### **4.2 Statistics of SST variation**

The SST variations in the UEH induced by cyclone passages during 1982 and 2015 are shown in **Figure 12**. The SST difference (∆*SST*) between the post-cyclone and pre-cyclone values is used to specify the response of UEH to typhoon passages. Here the pre- and post-SST are averaged values of the whole region within 7 days before cyclones enter or after passing the upwelling zone (outlined by red dashed lines in **Figure 11**), respectively. In most cases, the SST in the upwelling zone decreased (∆SST < 0) after the typhoon passage, with the greatest decrease of −2.4°C induced by the severe typhoon (STY) Zeke in 1991. The SST increase occurred in 1985, 1989, and 2001, with the greatest increase of 3.4°C after the passage of a nameless weak

**Figure 12.**

*SST changes induced by the 42 tropical cyclones passing over the UEH from 1982 to 2015. (a) Yearly distribution. (b) Number of typhoons in different cases. (c) Monthly distribution. Warming cases are in red, cooling cases in blue, and no-significant-change cases in gray (cited from [31]).*

tropical depression (WTD). The warming of sea surface after typhoon passages is distinguished from the prevailing cooling in the open ocean.

All 42 cases are categorized into three types according to the SST variation ∆*SST*: cooling with ∆*SST* < −0.5°C, no-significant-change with −0.5 ≤ ∆*SST* ≤ 0.5°C, and warming with ∆*SST* > 0.5°C. The statistical result is listed in **Figure 12b**. Nineteen cases of forty-two cyclones triggered surface cooling, with ∆*SST* concentrated mainly between −2 and −1.5°C and averaged to −1.5°C; twenty cases were in the category of no-significant-change with a mean value of 0°C, accounting for 48% of all cases and the number of slightly warming is larger than that of slightly cooling; three warming cases were found with a mean value of 2.1°C. The magnitude of averaged warming is greater than that of cooling.

In monthly distribution (**Figure 12c**), most warming cases occurred from June to September, with a maximum in June. The largest magnitude and most frequent cooling were in July, preceding that in June and August.

#### **4.3 SST changes vs. typhoon parameters**

It is usually assumed that the intensity and moving speed of tropical cyclone are two predominant factors influencing SST decrease in the open ocean. The stronger or slower-moving cyclones are supposed to induce more SST decrease [49]. Other studies suggested that the incidence angle and path of tropical cyclone may also play an important role in SST variation in coastal regions [29, 50]. In the UEH, all the four mentioned parameters are analyzed in 42 cases from 1981 to 2015.

The distributions of ∆*SST* on two joint parameters are illustrated in **Figure 13**. One can see that all three warming cases occurred after passage of low-wind-speed tropical cyclones, with duration period of 5–9 h, and the magnitude of SST variation increases with the incidence angle. For cooling cases, the relation between SST variation and maximal wind speed is not significant, with correlation coefficient R2 = 0.29. The magnitude of SST decrease is not dependent on the duration period of tropical cyclone. Most cooling cases occurred after tropical cyclones with the

*Response of Coastal Upwelling East of Hainan Island in the South China Sea to Sudden Impact… DOI: http://dx.doi.org/10.5772/intechopen.88828*

#### **Figure 13.**

*2-D distribution of SST variation with tropical cyclone parameters (a) SST variation vs. maximum wind speed and residence time. (b) SST variation vs. maximum wind speed and incident angle. (c) SST variation vs. incident angle and residence time (cited from [31]).*

duration time smaller than 8 h. All cooling cases occurred with the incidence angles greater than 50°, and there is an increasing trend of ∆*SST* magnitude with the incidence angle. The SST variation in no-significant-change is independent of the wind speed or the duration time of tropical cyclone. The incidence angles were all larger than 50°. It seems that the maximal wind speed and the incidence angle contribute more to the SST response, and the duration time might have slight influence.

#### **4.4 Mechanisms of SST change**

The change in SST induced by passage of tropical cyclone depends on the net heat flux in the research area. As a severe local disturbance, strong vertical convection and turbulent mixing during the momentum transfer from the cyclone to the sea water are generally regarded as the primary mechanism for sea surface cooling [47]. The magnitude of cooling increases as the strength of the cyclone increases or the movement speed decreases, due to the enhancement of vertical pumping and mixing. Generally, the heat variations of the upper ocean due to vertical convection and mixing induced by tropical cyclone, denoted as *Qv* and *Qm*, respectively, are negative. However, 20 no-significantchanges and even 3 warming cases appear in the above statistical analysis, indicating that there are other mechanisms inducing the positive heat flux via tropical cyclones.

As illustrated in **Figure 14**, the typhoon approaching the upwelling zone might decrease or increase the SST in the offshore water due to onshore or offshore Ekman transport induced by the typhoon winds. The heat flux due to the typhoon Ekman flow is denoted by *QE*, and its sign depends on the relative locations of the typhoon track and the upwelling zone [29]. For three SST warming cases mentioned above, the locations of tropical cyclone tracks relative to the upwelling zone are variant: one was south to the upwelling core, one passed the central upwelling zone, and one veered from the north to the south. The local typhoon-induced heat advection *QEa* by Ekman transport is merely one factor influencing the SST variation.

Zheng et al. [50] suggested that typhoon could force a single sea-level soliton in the deep offshore water with the positive amplitude. Warm water was thus transported onshore with the shoreward propagation of the soliton due to nonlinearity and resulted in net positive heat flux shoreward. The onshore amplitude of

**Figure 14.** *Sketch of typhoon-induced upwelling-favorable/upwelling-unfavorable wind and shoreward propagating soliton.*

**Figure 15.** *SST variation vs. incidence angle of tropical cyclone for three SST warming cases in UEH (cited from [31]).*

soliton induced by typhoon (thus the heat advection marked as *Qra*) depends on the typhoon incidence angle, i.e.,

$$\beta\_d \cong \alpha\_1 \sin^{1/2}\theta \exp\left(-\frac{a\_2}{\sin\theta}\right) \tag{1}$$

in which *βd* is the amplitude of soliton, *θ* the typhoon incidence angle, and *α*1 and *α*2 are parameters related to the continental shelf width, the turbulent viscosity and the propagation velocity of the soliton (refer to [50] for detail).

The normalized SST variation vs. the incidence angle of tropical cyclone in three warming cases is shown in **Figure 15**. One can see that the normalized ∆*SST* increases with the incidence angle and conforms to the idealized curve suggested by Zheng et al. [50]. It shows that the heat transport by the offshore soliton is a significant mechanism of SST increase in UEH after typhoon passages.

To sum up, the heat flux induced by tropical cyclone in UEH consists of two main categories: nonlocal heat transport *Q*r through shoreward transport of the offshore warm water by the tropical cyclone induced soliton, here *Q*r = *Q*ra and is positive; local heat flux by the momentum input of tropical cyclone *Q*1, including the heat change due to vertical convection/mixing *Q*vm (negative) and Ekman flow *Q*Ea (positive/negative), here *Q*1 = *Q*vm + *Q*Ea. Thus, the net heat flux by tropical cyclone is summarized as

$$\begin{aligned} \text{4. 11uus, une net meat nux by toppcuan eyonne is summarzzeu as} \\\\ \text{Q}\_{\text{net}} = \text{Q}\_{r} + \text{Q}\_{1} \begin{cases} > 0 & \text{if } \text{Q}\_{1} > 0 \text{ or } \text{Q}\_{r} > \\ = 0 & \text{if } \text{Q}\_{1} < 0 \text{ and } \text{Q}\_{r} = -\text{Q}\_{1} \\ < 0 & \text{if } \text{Q}\_{1} < \text{and } \text{Q}\_{r} < -\text{Q}\_{1} \end{cases} \end{aligned} \tag{2}$$

*Response of Coastal Upwelling East of Hainan Island in the South China Sea to Sudden Impact… DOI: http://dx.doi.org/10.5772/intechopen.88828*

The above statistics of 42 tropical cyclone passages indicate that the SST response in the UEH could be warming, no-significant-change, or cooling, accounting for 7, 48, and 45%, respectively. The heat flux is affected by the intensity, the moving speed, the incidence angle, and the relative location of tropical cyclone. The combination of these factors modifies the relative magnitudes of *Q*r and *Q*1, thus influences the SST response signatures.
