J. Meteor. Res.  2017, Vol. 31 Issue (5): 946-954 PDF
http://dx.doi.org/10.1007/s13351-017-7024-3
The Chinese Meteorological Society
0

#### Article Information

WANG, Yuxing, Yuan SUN, Qianfeng LIAO, et al., 2017.
Impact of Initial Storm Intensity and Size on the Simulation of Tropical Cyclone Track and Western Pacific Subtropical High Extent. 2017.
J. Meteor. Res., 31(5): 946-954
http://dx.doi.org/10.1007/s13351-017-7024-3

### Article History

in final form June 13, 2017
Impact of Initial Storm Intensity and Size on the Simulation of Tropical Cyclone Track and Western Pacific Subtropical High Extent
Yuxing WANG1, Yuan SUN1, Qianfeng LIAO1, Zhong ZHONG1,2, Yijia HU1,2, Kefeng LIU1
1. College of Meteorology and Oceanography, National University of Defense Technology, Nanjing 211101;
2. Jiangsu Collaborative Innovation Center for Climate Change, Nanjing University, Nanjing 210093
ABSTRACT: Typhoon Megi, the 13th typhoon of the 2010 typhoon season, was selected for case study by utilizing the Weather Research and Forecasting (WRF) model. Twelve sensitivity experiments with various initial tropical cyclone (TC) intensities and sizes were conducted to investigate their impacts on the simulation of typhoon track. Interaction between TC and the western Pacific subtropical high (WPSH) was also analyzed to explore the mechanism for the impact on TC track of the initial TC intensity and size. Numerical results indicate that the simulated TC size and TC track are sensitive to initial TC intensity and size. Stronger initial TC intensity and larger initial TC size often lead to larger simulated TC size and make TC turn northward earlier. Further analysis suggests that, with the increase of initial TC intensity and size, more air mass enters into the TC region, which subsequently reduces the extent of WPSH. As a result, the steering flow changes significantly and eventually causes the TC to turn northward earlier. The present study confirms that the initial TC intensity and size have certain influences on the TC track simulation, which demonstrates the importance of accurate initial condition for successful simulation of the TC intensity and TC track. Moreover, it also deepens our understanding of the interaction between TC and WPSH, provides helpful clues for the TC track change study, and discusses the future directions for improvement of TC track forecast.
Key words: tropical cyclone size     subtropical high     tropical cyclone track     WRF model
1 Introduction

Accurate prediction of tropical cyclone (TC) track has always been a challenging issue in operational weather forecasting. Many factors can affect TC tracks, e.g., the pressure gradient force, the Coriolis force, the interaction between TCs, the environmental background, and the temperature field (Murata et al., 2003). TC movement in the western North Pacific (WNP) is mainly under the control of the steering flow to the south of the western Pacific subtropical high (WPSH) (Chan and Gary, 1982). Advances and retreats of the WPSH and changes in its pattern could lead to abrupt changes in the steering flow that affects TC and subsequent changes in the TC track. Meanwhile, TCs also have great impacts on the WPSH. Ren et al. (2007) found that TCs with different tracks could lead to changes in the WPSH pattern. Zhong and Hu (2007) pointed out that TC activities in the WNP could weaken the WPSH intensity and significantly influence regional climate over East Asia. The above studies have revealed distinct interactions between TC and WPSH. However, mechanism studies for such interactions are still far less than enough, which is one of the reasons for the failure in simulating TC tracks and WPSH during active TC periods in the WNP. Apparently, the mechanism study for the interaction between TC and WPSH will be helpful for the improvement of TC and WPSH prediction.

The feedback of TC structure on TC track has not been fully recognized. Lee et al. (2010) statistically investigated tracks of 145 TCs over the WNP. They found that TCs with different sizes had different moving tracks and concluded that changes in the TC size might possibly affect its moving track. Sun et al. (2014) investigated the impacts of various cumulus parameterization schemes on Typhoon Megi based on numerical experiments. They found that TCs could feedback to large-scale environmental background through thermal effects, and thus modify the background circulation and change the TC track. The above study implies that changes in the TC structure may affect its track, and the study of the relationship between TC structure and TC track can provide insights for better understanding the mechanisms behind TC track changes.

TC intensity and size are two important components that affect the structure of the TC. The impact study of TC intensity and size on TC track will be helpful for the improvement of TC track forecast in operational weather service. In the present study, the Weather Research and Forecasting (WRF) model was applied to simulate Megi, the 13th typhoon of the 2010 typhoon season. Twelve sensitivity experiments with various initial intensities and sizes were conducted to investigate their impacts on TC track simulation. Interactions between TC and WPSH were also analyzed to explore the mechanism for the impacts of initial TC intensity and size on the TC track.

2 TC case and experimental design

Super Typhoon Megi formed at 2000 UTC 13 October 2010 over the oceanic area to the southwest of Guam with its center located at 11.8°N, 140.9°E. After its genesis, Megi moved northwestward and strengthened quickly, reaching typhoon intensity by 2100 UTC 14 October, when it was located to the southwest of the WPSH. Since then, it continued to move northwestward and reached the super-typhoon intensity by 0000 UTC 17 October over the ocean to the east of the Philippines. Megi then made landfall in northeastern Luzon of the Philippines, and entered the South China Sea at 1400 UTC 18 October. It moved westward steadily until 19 October before turning to the northwest and approaching the southern edge of the WPSH. The WPSH weakened and retreated eastward to the east of 130°E. Megi moved northward under the impact of the steering flow, then made landfall in Zhangpu of Fujian Province at 0400 UTC 23 October. It weakened significantly since then and became a tropical depression at 1500 UTC 23 October in Longhai, Fujian Province. Since 0200 UTC 24 October, it was no longer named and numbered. During the movement of Megi, interaction between WPSH and TC played a critical role (Yu and Fang, 2013, 2014; Sun et al., 2014, 2015a).

The numerical model utilized in our study was the WRF model (WRFV3.3) developed at the NCAR of US. Initial and boundary conditions were derived from the NCEP 1° × 1° gridded reanalysis product. The MODIS (Moderate-resolution Imaging Spectroradiometer) 30-s resolution topography data were used in this study. Two-way moving nest grids were utilized for the simulation. The two nesting domains both were initially centered at 22°N, 122°E with horizontal grid numbers of 160 × 180 and 150 × 150, respectively. The horizontal resolutions of the two domains are 20 and 4 km, respectively with a grid ratio of 1:5. There were 36 non-uniform σ levels in the vertical, and the time steps of the two domains are 90 and 18 s with a ratio of 1:5. The integration period of WRF was from 0000 UTC 16 to 0000 UTC 24 October 2010, corresponding to the time period before and after the turning of Megi. The model physics used in this study included the Grell–Dévényi cumulus parameterization scheme (Grell and Dévényi, 2002), the WSM 3-class microphysical scheme (Hong et al., 2004), the MYJ boundary later scheme (Mellor and Yamada, 1982; Janjić, 2001), the RRTM (Rapid Radiative Transfer Model) longwave radiation scheme (Mlawer et al., 1997), the Goddard shortwave radiation scheme (Chou and Suarez, 1994), the Eta surface layer scheme (Janjić, 2001), and the thermal diffusion scheme (Skamarock et al., 2008) for soil physics.

In order to study the impact of initial TC intensity and size on TC track, the bogus scheme in WRF was used in the 12 sensitivity experiments. The 12 experiments were divided into 3 groups by initial size with 4 sensitivity experiments in each group. The same parameterization schemes were adopted in the 12 experiments but the initial TC intensity and size were different. Among the 3 groups of experiments, the initial radius of maximum wind was set to 60 km and initial wind speed was set to 20, 30, 40, and 50 m s–1 in the experiments of the first group, which were denoted by ES20, ES30, ES40, and ES50. In the second group, the initial radius of maximum wind was set to 120 km and initial wind speed was set to 20, 30, 40, and 50 m s–1, respectively, and the experiments were denoted by EM20, EM30, EM40, and EM50. In the 4 experiments of the third group, the initial radius of maximum wind was set to 180 km and initial wind speed was set to 20, 30, 40, and 50 m s–1, respectively, and the experiments were denoted by EL20, EL30, EL40, and EL50.

3 Experimental results 3.1 TC structure

TC intensity and size are two important components that affect the structure of TC. In the present study, the TC size was defined as the total storm area in which the 10-m sustained winds exceed tropical cyclone strength (> 17 m s–1; A17); the TC intensity was defined as the maximum wind speed (MWS) at 10 m. To investigate the impact of initial TC intensity and size on the TC structure, Fig. 1 shows times series of the TC intensity (MWS), radius of maximum wind (RMW), and TC size (A17) in the sensitivity experiments. Note that it is not necessary to compare the simulated and observed TC structure in this study. Figure 1 shows clearly that the TC intensity was sensitive to its initial intensity, i.e., the TC with a stronger initial intensity was apt to be stronger during its development stage (Figs. 1a, b, c). The results of ES, EM, and EL experiments with the same initial TC intensity illustrate that the TC intensity was not sensitive to its initial size. As the initial TC size increased, the TC intensity was not stronger and even weaker in the later stage of its life cycle. For example, the simulated TC intensity in EL30 (EL50) was smaller than that in EM30 (EM50). This is probably because the TCs with a larger initial size were prone to turn northward and could reach higher latitudes, where the lower SST could not facilitate the development of TC intensity. Moreover, the RMW was not sensitive to initial TC intensity, especially in the ES and EM experiments, in which the simulated RMW basically remained 50 km in its life span (Figs. 1d, e, f). A comparison of ES, EM, and EL with the same initial TC intensity suggested that the RMW was sensitive to initial TC size and it increased as the initial TC size increased during the development stage. As is shown in Fig. 1g, the TC with stronger initial intensity was apt to become larger in the later stage of its life cycle. In the experiments of EM and EL, the trend that TC size increased as the initial intensity increased is significant (Figs. 1h, i). Note that the TC size simulated in the EL40 and EL50 decreased at 0000 UTC 19 October (Fig. 1i), which may be attributed to early northward turning of the TC. Comparisons of Figs. 1g, 1h, and 1iwith the same initial TC intensity indicate that the TC size was sensitive to its initial size, and TCs with larger initial sizes were prone to become larger later in its development.

 Figure 1 Times series of MWS (m s–1), RMW (km), and A17 (105 km2) in the sensitivity experiments (a, d, g) ES, (b, e, h) EM, and (c, f, i) EL.

To further analyze the impact of initial TC intensity and size on TC structure, Fig. 2 depicts time averages of the simulated MWS, RMW, and A17 before the occurrence of the large difference in TC center position among different sensitivity experiments (from 0000 UTC 16 to 0000 UTC 18 October, see Table 1). As is shown, the MWS increased with the increase in initial TC intensity, whereas it did not increase with the increase of initial TC size, especially in the EM and EL experiments (Fig. 2a). In contrast, the RMW increased with the increase of initial size but not with the increase in initial intensity (Fig. 2b). Moreover, under the joint effects of MWS and RMW, TC size increased with the increase in initial TC intensity and size (Fig. 2c).

 Figure 2 Time-average of the simulated MWS (m s–1), RMW (km), and A17 (105 km2) before the occurrence of the large difference in TC center position among different sensitivity experiments (from 0000 UTC 16 to 0000 UTC 18 October, see Table 1).
Table 1 Time and position when the TC turned northward in the sensitivity experiments
 Case Time (month:day:hour) Position OBS 10:18:18 16.80°N, 119.30°E ES20 10:19:12 14.94°N, 116.05°E ES30 10:19:00 15.21°N, 117.05°E ES40 10:18:18 15.76°N, 119.03°E ES50 10:18:12 16.57°N, 120.11°E EM20 10:19:00 15.30°N, 117.95°E EM30 10:18:18 16.84°N, 119.93°E EM40 10:18:12 17.56°N, 121.64°E EM50 10:18:12 18.55°N, 122.18°E EL20 10:18:18 16.75°N, 119.75°E EL30 10:18:12 17.74°N, 121.82°E EL40 10:18:06 19.72°N, 123.17°E EL50 10:18:00 20.71°N, 124.26°E
3.2 TC track

Figure 3 shows the best TC track issued by the US Joint Typhoon Warning Center (JTWC) and simulated TC tracks with different initial TC intensities and sizes. As is shown, except EL50, all other experiments reproduced the TC tendency to move westward first and then turned northward. Note that the TC with stronger initial intensity turned northward earlier. The EM experiments showed that, in the EM20, the TC moved westward steadily at the beginning, and then turned southwestward at Luzon. After reaching the South China Sea, the simulated TC turned northwestward and made landfall in Fujian of China. The EM20 experiment could well reproduce the observed track of Megi. In the EM30, the TC moved westward at the beginning, and then turned to northwest at Luzon. It eventually turned northeastward, bypassed around Taiwan Island, and made landfall in Zhejiang of China. The TC turned to the north too early in the EM40 and EM50, with large departures from the observation. The results of ES, EM, and EL experiments with the same initial intensity suggest that there were large differences in simulated TC tracks with different initial TC sizes, and the TC turned northward earlier with the increase in initial size. To further compare differences in TC track simulated by these experiments, Table 1 lists the time and position when the TC turned northward in the sensitivity experiments. The result shows clearly that the simulated TC with stronger initial intensity and larger initial size turned northward earlier.

 Figure 3 The model domain (sector area), the best track (OBS), and the simulated tracks of Megi from the sensitivity experiments.
4 Mechanism for the simulated TC track difference 4.1 Large-scale environmental background

TC track is largely dependent on the large-scale steering flow and the beta effect. However, based on our study, the beta effect is much smaller than the impact of the steering flow (figure omitted) and thus was not discussed here. In order to investigate the impact of initial TC intensity and size on the steering flow, Fig. 4 illustrates the vertical mean meridional steering flow (Vs, m s–1) in the sensitivity experiments. In the present study, Vs is the meridional wind averaged of 5°–7° latitudes away from the center of TC and vertically averaged between 850 and 300 hPa (Sun et al., 2015b). It shows clearly that Vs was not only sensitive to initial TC intensity but also sensitive to initial TC size, and increased with the increase of initial intensity and size. In addition, the difference in the simulated Vs was larger in the experiments of stronger initial TC intensity and larger initial TC size.

 Figure 4 Vertical mean meridional steering flow (Vs, m s–1) in the sensitivity experiments. Vs is the meridional wind averaged of 5°–7° latitudes away from the center of TC and vertically averaged between 850 and 300 hPa.

For TCs that affect China, their tracks mainly depend on the interaction between the WPSH and TC (Zhong, 2006). As suggested by Sun et al. (2015b), TCs could affect the WPSH intensity through thermal effects, and subsequently changed the environmental background and eventually influenced the TC track.

Figure 5 displays the 500-hPa geopotential height at 0000 UTC 18 October from FNL analysis and from simulations of the 12 experiments. At this time, there existed no significant differences in the central position of the simulated TC by the 12 experiments. In order to compare the differences in the simulated geopotential height by the 12 experiments, the contour of 5880-gpm was used to represent the WPSH extent. Figure 5 shows that initial TC intensity and size both had great impacts on the simulation of the WPSH extent. Compared to FNL analysis (Fig. 5a), the simulated WPSH extent in the experiments of stronger initial TC intensity and larger initial TC size was narrow and even broke in the middle (Figs. 5i, j, m, n, o) with its main body retreated eastward to the east of Taiwan Island. Figure 6 illustrates the average extent of 5880-gpm contour during the simulation period in the sensitivity experiments. As is shown, with increased initial TC size and intensity, the 5880-gpm contour extended further eastward.

 Figure 5 Geopotential height at 500 hPa from (a, f, k) FNL analysis and (b, c, d, e, g, h, i, j, l, m, n, o) simulations of sensitivity experiments at 0000 UTC 18 October 2010. The contours of 5880 gpm are highlighted in red. Note: (f) and (k) are duplicated from (a) the convenience of comparision with panels on the same row.
 Figure 6 Average extent of the 5880-gpm contour during the simulation period in the sensitivity experiments.

The comparison of Figs. 3, 5, and 6 shows that the decreasing degree of the WPSH extent could greatly affect the position of the northward turning of TC because of the changes of the steering flow that depend on the WPSH. With the increase of TC size, the WPSH extent gradually reduced, and this phenomenon was more significant in the experiments of stronger initial intensity and larger initial size.

To further analyze the influence of TC size on the WPSH extent, Fig. 7 presents radius–time cross-sections of simulated 500-hPa geopotential height in the peripheral region of TC before the TC turned northward in the sensitivity experiments. The peripheral region was defined as the area 400–1200 km outward from the storm center. The simulated 500-hPa geopotential height displayed a regular semi-diurnal variation in all the 12 experiments, which is probably due to the impact of the semi-diurnal atmospheric tide. More importantly, the contours of the simulated geopotential height (e.g., the 5860-gpm contour) extended outward notably and decreased significantly with the increase in the initial TC intensity (Figs. 7a, b, c, d). Similar results could be found in the sensitivity experiments with various initial TC sizes (Figs. 7a, e, i). This suggests that initial TC intensity and size could affect the WPSH extent in its peripheral region at certain degrees. Such feedbacks between the withdraw of the WPSH extent and the northward moving of the storm when it approached the WPSH contributed to the large differences in the simulated storm tracks among these sensitivity experiments. Those TCs with stronger initial intensities and larger initial sizes showed stronger effects on the shrinking of the WPSH extent, which led to significant changes in the large-scale environmental fields and eventually resulted in earlier northward turning of the TCs.

 Figure 7 Radius–time cross-sections of simulated 500-hPa geopotential height in the peripheral region of TC before TC turned northward in the sensitivity experiments.

Recent studies suggest that, when the pressure gradient force (PGF) becomes larger in the peripheral region of a TC, inward radial velocity would increase correspondingly, and the inflow mass flux (IMF) to inner TC would intensify (Gopalakrishnan et al., 2011; Sun et al., 2015b). When a TC moves closer to the WPSH, part of the air mass in the WPSH region is transported to the TC region, contributing to the reduction of the WPSH extent.

The IMF is defined by

 ${\rm IMF} = \int\limits_0^{{Z_{850}}} {\int\limits_0^{2\pi } {\int\limits_0^R {{u_r}{\rm{(}}z{\rm{, }}\theta {\rm{, }}r{\rm{)}}} } } {\rm{ }}\rho {\rm{(}}z{\rm{, }}\theta {\rm{, }}r{\rm{) }}{\rm d}z{\rm d}\theta {\rm d}r,$ (1)

where ${u_r}\,\,{\rm{and}}\,\,\rho$ represent the radial velocity and density, r represents the radial distance to TC center, z represents the distance from the sea level, $\theta$ represents the azimuthal angle on the TC center, and R is 800 km.

Figure 8 illustrates the simulated IMFs transported to the TC region from areas 800 km away from TC center and below 850 hPa in the 12 experiments. As is shown, the IMFs exhibited a distinct diurnal variation, which may be attributed to the diurnal cycling of wind and temperature. Stronger initial TC intensity corresponded to larger IMF transport from the WPSH to TC region. Take sensitivity experiments of EM, for example, in the EM20 and EM30 experiments with weaker initial TC intensity, TC sizes at its mature stage were smaller (Fig. 1a). As a result, the radial IMF transport simulated in EM20 and EM30 was smaller than that simulated in other experiments. In contrast, in the EM40 and EM50 experiments with stronger initial TC intensity, more IMFs were transported from the WPSH region to TC region, the extent of WPSH was withdrawn, resulting in the division of the WPSH (Figs. 5i, j). A comparison of Figs. 8a, b, and c with same initial TC intensity suggests that the simulated IMF transport increased with initial TC size. A similar conclusion was also found in Sun et al. (2015b). The above results suggest that the IMF simulated in various experiments were highly sensitive to the initial TC intensity and size. Stronger initial TC intensity and size corresponded to larger IMF transport from the WPSH to TC region and more severe withdraw of the WPSH extent. This is one of the reasons for the break of the WPSH, and it eventually led to early northward turning of TCs in the EM40, EL40, and other experiments.

 Figure 8 Simulated IMF (107 kg m s–1) transported to the TC region in the sensitivity experiments.
5 Conclusions

In this study, the 13th typhoon of the 2010 typhoon season was selected for sensitivity case study by utilizing the WRF model. Twelve sensitivity experiments with various initial TC intensities and sizes were conducted to explore the influence of initial intensity and size on the simulation of TC track. Interaction between TC and WPSH was explored to explain the differences in TC track simulation caused by different initial TC intensities and sizes. Major conclusions are as follows.

(1) TC size, during its development, was highly sensitive to initial TC intensity and size. TC size increased with the increase of initial TC intensity and size.

(2) The initial TC intensity and size both had great impacts on its future track. TCs with stronger initial intensity and larger initial size would turn northward earlier. Further analysis showed that those TCs with stronger initial intensity and larger initial size were apt to larger size later during its development.

(3) Significant interaction existed between TC and the WPSH. As TC approached the WPSH, it would lead to the shrinking of the WPSH extent as well as changes of the steering flow and eventually influence the track of TC. The weakening impact of TC on the WPSH extent varied with various initial TC intensities and sizes. Further analysis indicated that the TC size at its mature stage was different due to different initial TC intensities and sizes, which led to large differences in the IMF transport from the WPSH to TC. This partly explains why the reduction of the WPSH extent was different for different initial TC sizes and intensities, and this eventually led to large differences in the simulated TC track.

(4) When TC approached the WPSH, more IMFs were transported from the WPSH to TC, and the extent of WPSH reduced more severely in response to the influence of larger TC than to that of smaller TC. Changes of WPSH subsequently changed the steering flow, leading to earlier northward turning of TC. As a result, the TC no longer moved westward persistently as shown in the case of a smaller TC.

The present study compared the influences of initial TC intensity and size on simulations of TC structure, track, and large-scale environment, confirmed that the initial TC intensity and size have certain effects on the WPSH and TC track. It deepens our understanding of the interaction between TC and WPSH, and provides helpful clues for TC track study. This study emphasized the importance of accurate and appropriate initial condition for successful simulation of the intensity and track of TC, which may help to improve the TC track forecast. Note that conclusions of this study are based on the single case of Typhoon Megi. A study of more cases would be necessary to further validate the above results. This will be one of our future research topics.

References
 Chan, J. C. L., and W. M. Gray, 1982: Tropical cyclone movement and surrounding flow relationships. Mon. Wea. Rev., 110, 1354–1374. DOI:10.1175/1520-0493(1982)110<1354:TCMASF>2.0.CO;2 Chou, M. D., and M. J. Suarez, 1994: An Efficient Thermal Infrared Radiation Parameterization for Use in General Circulation Models. NASA Technical Memorandum 104606, Greenbelt, Maryland, Goddard Space Flight Center, 85 pp. Gopalakrishnan, S. G., F. Marks, X. J. Zhang, et al., 2011: The experimental HWRF system: A study on the influence of horizontal resolution on the structure and intensity changes in tropical cyclones using an idealized framework. Mon. Wea. Rev., 139, 1762–1784. DOI:10.1175/2010MWR3535.1 Grell, G. A., and D. Dévényi, 2002: A generalized approach to parameterizing convection combining ensemble and data assimilation techniques. Geophys. Res. Lett., 29, 38-1–38-4. DOI:10.1029/2002GL0153 Hong, S.-Y., J. Dudhia, and S.-H. Chen, 2004: A revised approach to ice microphysical processes for the bulk parameterization of clouds and precipitation. Mon. Wea. Rev., 132, 103–120. DOI:10.1175/1520-0493(2004)132<0103:ARATIM>2.0.CO;2 Janjić, Z. I., 2001: Nonsingular Implementation of the Mellor–Yamada Level 2.5 Scheme in the NCEP Meso Model. NCEP Office Note No. 437, 61 pp. Lee, C.-S., K. K. W. Cheung, W.-T. Fang, et al., 2010: Initial maintenance of tropical cyclone size in the western North Pacific. Mon. Wea. Rev., 138, 3207–3223. DOI:10.1175/2010MWR3023.1 Mellor, G. L., and T. Yamada, 1982: Development of a turbulence closure model for geophysical fluid problems. Rev. Geophys., 20, 851–875. DOI:10.1029/RG020i004p00851 Mlawer, E. J., S. J. Taubman, P. D. Brown, et al., 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102(D14), 16663–16682. DOI:10.1029/97JD00237 Murata, A., K. Saito, and M. Ueno, 2003: The effects of precipitation schemes and horizontal resolution on the major rainband in Typhoon Flo (1990) predicted by the MRI mesoscale nonhydrostatic model. Meteor. Atmos. Phys., 82, 55–73. DOI:10.1007/s00703-001-0591-x Ren, S. L., Y. M. Liu, and G. X. Wu, 2007: Interactions between typhoon and subtropical anticyclone over western pacific revealed by numerical experiments. Acta Meteor. Sinica, 65, 329–340. DOI:10.11676/qxxb2007.032 Skamarock, W. C., J. B. Klemp, J. Dudhia, et al., 2008: A description of the advanced research WRF version 3. NCAR Technical Note NCAR/TN-475+STR, Boulder, CO, USA, 7–25. Sun, Y., Z. Zhong, W. Lu, et al., 2014: Why are tropical cyclone tracks over the western North Pacific sensitive to the cumulus parameterization scheme in regional climate modeling? A case study for Megi (2010).. Mon. Wea. Rev., 142, 1240–1249. DOI:10.1175/MWR-D-13-00232.1 Sun, Y., Z. Zhong, and W. Lu, 2015a: Sensitivity of tropical cyclone feedback on the intensity of the western Pacific subtropical high to microphysics schemes. J. Atmos. Sci., 72, 1346–1368. DOI:10.1175/JAS-D-14-0051.1 Sun, Y., Z. Zhong, L. Yi, et al., 2015b: Dependence of the relationship between the tropical cyclone track and western Pacific subtropical high intensity on initial storm size: A numerical investigation. J. Geophys. Res., 120, 11451–11467. DOI:10.1002/2015JD023716 Yu, L. Y., and J. Fang, 2013: An analysis on the impact system of Typhoon MEGI (2010). J. Nanjing Univ. (Nat. Sci.), 49, 285–299. DOI:10.13232/j.cnki.jnju.2013.03.008 Yu, L. Y., and J. Fang, 2014: Environmental effects on the intensity of Typhoon Megi (2010). J. Nanjing Univ. (Nat. Sci.), 50, 113–128. DOI:10.13232/j.cnki.jnju.2014.02.001 Zhong, Z., 2006: A possible cause of a regional climate model’s failure in simulating the East Asian summer monsoon. Geophys. Res. Lett., 33, L24707. DOI:10.1029/2006GL027654 Zhong, Z., and Y. J. Hu, 2007: Impacts of tropical cyclones on the regional climate: An East Asian summer monsoon case. Atmos. Sci. Lett., 8, 93–99. DOI:10.1002/(ISSN)1530-261X