**6. Discussion**

The above analysis shows that Megi-induced NIKE was mainly concentrated in the deep SCS basin, which was caused by the reflection of NIWs at the supercritical continental slope around the SCS Basin. A similar phenomenon has been reported for the diurnal internal tides in the SCS [56]. According to the equation of topographic criticality:

$$\gamma = \frac{s\_{topo}}{s\_{wave}} = \frac{s\_{topo}}{\sqrt{\left(\omega^2 - f^2\right)/\left(N^2 - \omega^2\right)}}\tag{6}$$

where *stopo* and *swave* are the topographic slope and internal wave slope, because the frequency of Megi-induced NIWs (Figure 2) is smaller than that of diurnal internal tides, the *swave* for Megi-induced NIWs is smaller than that for diurnal internal tides. In other words, Megi-induced NIWs are more susceptible to reflection on the continental slope in the northern SCS than the diurnal internal tides.

The damping feature of Megi-induced NIWs was site-dependent: In the region near Megi's track, the e-folding time of NIWs was generally less than one week; whereas in two zones to the west of Luzon Island and the Luzon Strait, which are far away from Megi's track, the e-folding time could be longer than 20 days. This result emphasizes the correlation between the distance away from typhoon's track and the decay of NIWs and can partly answer the following questions: Why NIWs generated by different typhoons have different e-folding times at the same position [31] and why NIWs generated by the same typhoon have different e-folding times at different positions [37,57].

The modal decomposition results indicate that Megi-induced NIWs were dominated by the first three baroclinic modes. Along Megi's wake, Megi-induced NIWs quickly dampened after the passage of Megi. However, at point B which is far away from Megi's wake, higher modes (mode-4 to mode-7) appeared and gradually became dominant after

24 October. To explore the possible cause of these higher modes (mode-4 to mode-7), Figure 13 illustrates the zonal currents of NIWs at four points along 18.48◦ N. By simply counting the times of sign changing of zonal currents in the vertical direction, we can find that these higher modes (mode-4 to mode-7) mainly appeared at point B and 119.04◦ E, 18.48◦ N, whereas at point A and 118◦ E, 18.48◦ N, these higher modes are nearly invisible. In other words, the appearance of these higher modes (mode-4 to mode-7) was limited to a small region near point B. According to [58], higher modes can be generated when low-mode internal waves interact with mesoscale eddies. Figure 14 illustrates the HYCOM surface elevations from 18 October to 1 November with an interval of 2 days. From Figure 14, we can find that an anticyclonic eddy formed around 24 October and influenced the region around point B. As mentioned above, the NIWs initially generated at 118.8◦ E propagated eastward to point B after 24 October (Figure 11). This case is similar to that reported by [58]. Therefore, we speculated that the higher modes (mode-4 to mode-7) at point B after 24 October were likely caused by the interaction between NIWs and a mesoscale eddy. However, point A was also under the influence of a cyclonic eddy after 22 October. Why higher modes (mode-4 to mode-7) did not become significant at point A remains unclear. Therefore, the difference of modal content at points A and B still needs further exploration. *J. Mar. Sci. Eng.* **2021**, *9*, x FOR PEER REVIEW 14 of 17 Why higher modes (mode-4 to mode-7) did not become significant at point A remains unclear. Therefore, the difference of modal content at points A and B still needs further exploration. *J. Mar. Sci. Eng.* **2021**, *9*, x FOR PEER REVIEW 14 of 17 Why higher modes (mode-4 to mode-7) did not become significant at point A remains unclear. Therefore, the difference of modal content at points A and B still needs further exploration.

**Figure 13.** Zonal currents of NIWs (shading, unit: m/s) at (**a**) point A (117.04°E, 18.48°N), (**b**) 118°E, 18.48°N, (**c**) 119.04°E, 18.48°N and (**d**) point B (119.52°E, 18.48°N) from 16 to 30 October 2010. **Figure 13.** Zonal currents of NIWs (shading, unit: m/s) at (**a**) point A (117.04◦ E, 18.48◦ N), (**b**) 118◦ E, 18.48◦ N, (**c**) 119.04◦ E, 18.48◦ N and (**d**) point B (119.52◦ E, 18.48◦ N) from 16 to 30 October 2010. **Figure 13.** Zonal currents of NIWs (shading, unit: m/s) at (**a**) point A (117.04°E, 18.48°N), (**b**) 118°E, 18.48°N, (**c**) 119.04°E, 18.48°N and (**d**) point B (119.52°E, 18.48°N) from 16 to 30 October 2010.

**Figure 14.** HYCOM surface elevations (shading, unit: m) at 00:00 on (**a**) 18, (**b**) 20, (**c**) 22, (**d**) 24, (**e**) 26, (**f**) 28 and (**g**) 30 October and (**h**) 1 November 2010. **Figure 14.** HYCOM surface elevations (shading, unit: m) at 00:00 on (**a**) 18, (**b**) 20, (**c**) 22, (**d**) 24, (**e**) 26, (**f**) 28 and (**g**) 30 October and (**h**) 1 November 2010. **Figure 14.** HYCOM surface elevations (shading, unit: m) at 00:00 on (**a**) 18, (**b**) 20, (**c**) 22, (**d**) 24, (**e**) 26, (**f**) 28 and (**g**) 30 October and (**h**) 1 November 2010.

tions and provide us with an opportunity to better understand NIWs.

tions and provide us with an opportunity to better understand NIWs.

2010 in the SCS were investigated in this study. Through a comparison with in situ observations at mooring UIB6, we first showed that the HYCOM reanalysis results can reasonably reproduce typhoon-induced NIWs, which can act as a supplement to in situ observa-

Based on the HYCOM reanalysis results, the NIWs generated by typhoon Megi in 2010 in the SCS were investigated in this study. Through a comparison with in situ observations at mooring UIB6, we first showed that the HYCOM reanalysis results can reasonably reproduce typhoon-induced NIWs, which can act as a supplement to in situ observa-

The results indicate that Megi-induced NIWs showed temporal and spatial variations in the SCS. The NIKE in the SCS was rapidly enhancedin response to typhoon Megi. However, the strongest NIKE appeared several days after the passage of Megi, rather than under its influence. Moreover, it is interesting to note that Megi-induced NIKE was mainly concentrated in the deep SCS basin where the water depth is greater than 1000 m, although typhoon Megi passed over both the deep SCS basin and shallow continental shelf and slope in the northern SCS. Through analysis, it was found that the continental slope

The results indicate that Megi-induced NIWs showed temporal and spatial variations in the SCS. The NIKE in the SCS was rapidly enhancedin response to typhoon Megi. However, the strongest NIKE appeared several days after the passage of Megi, rather than under its influence. Moreover, it is interesting to note that Megi-induced NIKE was mainly concentrated in the deep SCS basin where the water depth is greater than 1000 m, although typhoon Megi passed over both the deep SCS basin and shallow continental shelf and slope in the northern SCS. Through analysis, it was found that the continental slope

**7. Conclusions**

**7. Conclusions**
