Change in the Occurrence Frequency of Landfalling and Non-Landfalling Tropical Cyclones over the Northwest Pacific

Mingzhong Xiao aState Key Laboratory of Lunar and Planetary Sciences, Macau University of Science and Technology, Macau, China
bState Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China

Search for other papers by Mingzhong Xiao in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Understanding the tropical cyclone (TC) activity changes in response to climate change is of great importance for disaster mitigation and climate change adaptation. Change in the annual occurrence frequency of landfalling and non-landfalling weak, strong, and super TCs during 1980–2018 was analyzed. Results indicate that the super TCs have been more likely to make landfall in the northwest Pacific since 1980. Using an empirical orthogonal function–based method proposed to decompose the space–time field of TC occurrence into different patterns, the anthropogenic influence on the change in super TC occurrence was detected when the impacts of El Niño–Southern Oscillation (ENSO), the Pacific meridional mode (PMM), and the interdecadal Pacific oscillation (IPO) were separated. Results further show that TCs forming in the sea surface near land (6°–21°N, 130°–137°E) have been more likely to intensify to super TCs in recent years. These intensified TCs tend to favor subsequent landfall, which may be the reason for the increase in landfalling super TCs. The intensification of TC is mainly due to the increase in the intensification rate, which increases with increased sea surface temperature (SST), especially during the stronger wind periods. Along with the change in the occurrence of landfalling super TCs, the landfalling locations of super TCs also changed. For example, western South China, Southeast China, and Japan are facing an increase in landfalling super TCs. The destructiveness of super TCs to these economically developed and highly populated regions is great; more attention therefore should be paid to mitigate TC disasters.

© 2021 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Mingzhong Xiao, xmingzh@mail2.sysu.edu.cn

Abstract

Understanding the tropical cyclone (TC) activity changes in response to climate change is of great importance for disaster mitigation and climate change adaptation. Change in the annual occurrence frequency of landfalling and non-landfalling weak, strong, and super TCs during 1980–2018 was analyzed. Results indicate that the super TCs have been more likely to make landfall in the northwest Pacific since 1980. Using an empirical orthogonal function–based method proposed to decompose the space–time field of TC occurrence into different patterns, the anthropogenic influence on the change in super TC occurrence was detected when the impacts of El Niño–Southern Oscillation (ENSO), the Pacific meridional mode (PMM), and the interdecadal Pacific oscillation (IPO) were separated. Results further show that TCs forming in the sea surface near land (6°–21°N, 130°–137°E) have been more likely to intensify to super TCs in recent years. These intensified TCs tend to favor subsequent landfall, which may be the reason for the increase in landfalling super TCs. The intensification of TC is mainly due to the increase in the intensification rate, which increases with increased sea surface temperature (SST), especially during the stronger wind periods. Along with the change in the occurrence of landfalling super TCs, the landfalling locations of super TCs also changed. For example, western South China, Southeast China, and Japan are facing an increase in landfalling super TCs. The destructiveness of super TCs to these economically developed and highly populated regions is great; more attention therefore should be paid to mitigate TC disasters.

© 2021 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Mingzhong Xiao, xmingzh@mail2.sysu.edu.cn

1. Introduction

With global climate change, the annual variability of tropical cyclone (TC) activity has become a topic of great interest and importance (Emanuel 2017; Knutson et al. 2019; Murakami et al. 2020). One of the scientific interests in this issue is how the TC activity changes in response to climate change, and whether the anthropogenic influence on TC activity can be detected in observation (Knutson et al. 2019, 2020). The TCs are classified into landfalling and non-landfalling. The landfalling TCs, including the floods induced by TCs (Emanuel 2017; Lai et al. 2020; Liu and Wang 2020), are among the most destructive natural disasters, and the induced socioeconomic losses have shown a remarkable increase throughout the world in recent decades with economic development and population growth (Emanuel 2017; Li et al. 2017; Liu et al. 2020; Peduzzi et al. 2012; Zhang et al. 2009). The northwest Pacific is the most TC-active basin (Schreck et al. 2014). Along the northwest Pacific coastline, there are many economically developed and highly populated regions, such as Manila, Hong Kong, Shenzhen, Taipei, Gwangju, and Osaka. These regions are vulnerable to TCs; therefore, understanding change in the occurrence frequency of landfalling TCs over the northwest Pacific is essential for disaster mitigation and climate change adaptation.

Recently, Murakami et al. (2020) found that the spatial patterns of where TCs occur have changed since 1980 on the global scale. They found that the occurrence frequency of TCs substantially decreases in the northwest Pacific. However, whether landfalling TCs in the northwest Pacific have also decreased since 1980 is not clear. Besides, trends in the intensity of TCs have been observed; a slight increase in the intensity of TCs was identified on the global scale (Kang and Elsner 2015; Kossin et al. 2013; Zhao et al. 2018a). Zhan et al. (2017) found that the proportion of intense TCs relative to all TCs has increased by about 16%–20% since the late 1970s in the western part of the northwest Pacific. For the landfalling TCs over the northwest Pacific, Mei and Xie (2016) found that TC intensity has increased by 12%–15% since the late 1970s. In this study, the TCs were classified into weak, strong, and super types based on their lifetime maximum intensity (LMI). Thereby, a more specific question arises: what are the changes in the landfalling and non-landfalling TCs in different intensity categories?

In previous literature, the annual numbers of TCs were usually analyzed to detect the trend. However, no significant trend was found on a global scale since 1970 (Weinkle et al. 2012). Moreover, Knutson et al. (2019) summarized that no detectable anthropogenic influence has been identified to date in the observation of landfalling TCs. It has been well established that TC has multidecadal internal variability, which further enhances the challenges in the detection of anthropogenic influence in observations (Hsu et al. 2014; Knutson et al. 2019; Murakami et al. 2020). Zhang et al. (2012) and Gao et al. (2020) found that the landfalling TCs in East Asia are influenced by El Niño–Southern Oscillation (ENSO) and Pacific meridional mode (PMM) SST, respectively. Both Zhao et al. (2018b) and Zhang et al. (2018) found that the abrupt decrease of TC in the northwest Pacific around 1998 is attributed to the negative phase of the interdecadal Pacific oscillation (IPO) or Pacific decadal oscillation (PDO). Owing to the influences of climate indices on TCs, it is hard to detect the anthropogenic influence solely from the observations of landfalling or non-landfalling TCs. To solve this problem, an empirical orthogonal function (EOF)-based method to decompose the space–time field of TC occurrence into different patterns is proposed in this study. This method helps to separate the impact of climate index, and the anthropogenic influence on TCs may be detected in a pattern that is not affected by the climate index.

Wu et al. (2015) found a westward shift in the location of TC genesis over the northwest Pacific, and they stated that this trend was associated with ocean warming. The increase in the intensity of TCs was also attributed to ocean warming in previous literature (Mei and Xie 2016; Zhan et al. 2017; Zhao et al. 2018a). The generation, intensification, and movement of TCs depend on the surrounding environment. Therefore, the influences of ocean warming and other atmospheric variables on the generation and intensification of TC were considered to investigate the possible physical mechanisms for the change in the landfalling and non-landfalling TC occurrence. Generally, the objectives of this study are 1) to identify the change in the occurrence frequency of landfalling and non-landfalling TCs in different intensity categories over the northwest Pacific, 2) to distinguish whether the detected trend is caused by the anthropogenic influence, and 3) to investigate the possible reasons responsible for this change.

2. Data

The TC data used in this study come from the Joint Typhoon Warning Center data collected by the International Best Track Archive for Climate Stewardship (IBTrACS, version 4; https://www.ncdc.noaa.gov/ibtracs/). The TC data include 3-hourly positions and intensity (measured as 1-min maximum sustained wind). The TC data before the modern era of satellite observations are usually considered to be questionable, so the TC data in the period 1980–2018 were used. Further, only TCs with LMI ≥ 35 knots (kt; 1 kt ≈ 0.51 m s−1) over the northwest Pacific (0°–60°N, 100°E–180°) were analyzed in this study. Here, a TC is judged to be in the northwest Pacific by checking that its first location (by date) with wind speed ≥ 35 kt is located in the northwest Pacific.

Monthly atmospheric data for the same period were derived from the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA5; Hersbach et al. 2020), including relative humidity (RH) at 600 hPa, sea surface temperature, vertical wind shear (the magnitude of the vertical wind shear between 850 and 200 hPa), and convective available potential energy (CAPE). The PMM (Chiang and Vimont 2004) and IPO (Henley et al. 2015) indices were obtained from the NOAA Physical Sciences Laboratory. As recommended by Henley et al. (2015), the tripole index (TPI) was used to represent the IPO phenomenon. The ENSO index in this study was obtained from the NOAA Climate Prediction Center. It was indicated by the Niño-3.4 indices, available at the website of https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php.

3. Methods

a. Kernel density-based counting of TC generation in space

In this study, we define TC genesis point as the first position where wind reaches the speed of 30 kt (about 15 m s−1) (Tippett et al. 2011). To analyze the trend of TC generation in space, the annual counting of TC generation in each space grid is required. A simple method is to count the number of TC genesis points located in each grid. However, this count is biased for the points located near the boundary of grids. It is unreasonable to count one grid as 1 and the other grid as 0 for a point located at the boundary of two grids. To solve this problem, a kernel density–based counting method was proposed in this study in which smooth counting data were obtained in each grid according to the kernel density (Lu and Xiong 2019). Based on all of the TC genesis points in a year, the kernel density in each grid was calculated. For the grid g(i, j), its kernel density is indicated by d(i, j), and its area is indicated as A(i, j); then its probability p(i, j) is defined as p(i, j) = d(i, j)A(i, j). This probability represents the ratio of the number of TCs generated in that grid to the total number of TCs. Thereby, the number of TC genesis points [N(i, j)] in g(i, j) can be estimated based on p(i, j):
N(i,j)=Tyearp(i,j)=Tyeard(i,j)A(i,j),
where Tyear indicates the total number of TCs in a specific year. For all of the grids, we have the following equation:
i=1mj=1nTyeard(i,j)A(i,j)=Tyear.
The grid areas are not the same; however, they were assumed to be the same in this study to simplify the estimation. Based on Eq. (2), their values are simplified:
A(i,j)=1i=1mj=1nd(i,j).
Substituting Eq. (3) into Eq. (1), the number of TC genesis points in a grid can then be estimated based on the kernel density:
N(i,j)=Tyeard(i,j)i=1mj=1nd(i,j).
In this study, the Gaussian kernel and Haversine distance were used to estimate the two-dimensional kernel density, where the Haversine distance measures the great-circle distance between two points on a sphere given the latitudes and longitudes (expressed in radians). The kernel density estimation was done based on the “KernelDensity” program in the Python module “scikit-learn 0.23.1” (Pedregosa et al. 2011), and its bandwidth was selected based on the cross-validation method where 80% of the data were used to fit the model while the remaining data were used to evaluate the performance of the model. In this study, the bandwidth was considered to be the same for each year, and the bandwidth was selected based on the TC genesis points in all years.

b. Identifying different patterns of TC generation in space and time

According to the method introduced in section 3a, the numbers of TC generation in space and time were obtained. For that regularized space–time field of TC occurrence, it can be decomposed into different patterns by the empirical orthogonal function. In general, the decomposition of TC occurrence space–time field X(t, s) by empirical orthogonal function is defined as follows (Hannachi et al. 2007; Xiao 2020; Xiao et al. 2015):
X(t,s)=k=1mEOFk(s)PCk(t),
where m indicates the number of patterns with the decreased explained variance of the field, EOF indicates the spatial pattern, and PC indicates the temporal pattern. In this study, the leading two patterns were analyzed, and the Python package “eofs” was used for the estimation of empirical orthogonal function (Dawson 2016).

c. Analysis of trend in intensification rate of TCs

In this study, the intensification period of TCs is defined as the period from the generation to the time first reaching the LMI. During that period, the intensification rate of TC (TCint_site) between two adjacent points in time is defined as
TCint_site=ΔWΔT,
where ΔW indicates the change in wind speed during the time interval ΔT. To analyze the change in intensification rate during the intensification period, three sub-intensification periods were considered: a weak wind period, strong wind period, and super wind period. Here the turning point of strong wind (TPstrong) is defined as the time when TC first reaches 64 kt, and the turning point of super wind (TPsuper) is defined as the time when the TC first reaches 96 kt. Thereby, the weak wind period is defined as the period from TC generation to TPstrong, the strong wind period is defined as the period between TPstrong and TPsuper, and the super wind period is defined as the period from TPsuper to the time when TC first reaches the LMI. Notably, only the super TCs have these three sub-intensification periods; the strong TCs just have weak and strong wind periods. The average TC intensification rate during each sub-intensification period (TCint_sub) was considered, and it was calculated as the average of TCint_site during that sub-intensification period. With the intensification rates in the sub-intensification period were calculated for each TC, the annual average TC intensification rate (TCint_sub_annual) for each sub-intensification period was estimated as the average of TCint_sub in the same year. In this study, the linear trend was used to represent the trend in TCint_sub_annual.

The change in the TC intensification rate depends on the surrounding environment such as RH, SST, CAPE, and vertical wind shear. In this study, the area-averaged RH, SST, CAPE, and vertical wind shear within 1° of longitude or latitude (about 100 km) around the TC tracks were considered. For simplicity, the monthly average RH, SST, CAPE, and vertical wind shear were used, and they were matched to each TC track by year and month. To investigate which variable is more important for the changes in TC intensification rate, a random forest regression between the average of TCint_site and area-averaged RH, SST, CAPE, vertical wind shear during the entire intensification period was built. The relative importance of variables was evaluated based on the permutation-based variable importance in a random forest method (Breiman 2001; Xiao et al. 2020). Here, 90%, 70%, and 50% of the data are randomly selected as training data while the rest are used as test data for the random forest regression, and the accuracies of training and testing were evaluated based on the coefficient of determination (R2).

4. Results

a. Change in the trends of annual numbers of landfalling and non-landfalling TCs

According to the LMI, the TCs were classified into weak (35 ≤ LMI < 64 kt), strong (64 ≤ LMI < 96 kt; Saffir–Simpson category 1–2), and super (LMI ≥ 96 kt; Saffir–Simpson category 3–5) TCs in this study. These types of TCs were further classified into landfalling and non-landfalling TCs, and an original landfalling TC is defined as having at least one point in the TC track located on land. Sometimes, the coastal regions were also affected by TCs that did not land. To consider these TCs that approached the coastal line, two other landfalling TC definitions were compared, with a landfalling TC defined having at least one point in the TC track located within 100 km offshore, or within 200 km offshore. According to these three landfalling definitions, the annual numbers of landfalling weak, strong, and super TCs were counted. Figure 1 shows that the annual counts of landfalling weak, strong, and super TCs are almost the same for these three definitions. Hence, the change in the results caused by the different landfalling TC definitions was ignored in this study, and the following results were based on the original landfalling TC definition.

Fig. 1.
Fig. 1.

Change in the annual counts of landfalling (a) weak, (b) strong, and (c) super TCs according to three different landfalling definitions.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

The annual numbers of TCs in the northwest Pacific were counted separately for the weak, strong, and super TCs. Figures 2a(2) and 2b(2) show that both the annual numbers of landfalling and non-landfalling strong TCs decreased significantly during the period 1980–2018 while no significant trends were found for weak TCs [Figs. 2a(1) and 2b(1)]. The landfalling super TCs increased significantly during the period 1980–2018 [Fig. 2a(3)] while the non-landfalling super TCs decreased significantly [Fig. 2b(3)]. Murakami et al. (2020) found that the occurrence frequency of TCs substantially decreased in the northwest Pacific. Here, our results indicate that this is mainly caused by a decrease in strong TCs, and an increase in the probability of landfalling super TCs was found [Fig. 2c(3)].

Fig. 2.
Fig. 2.

Trends in the annual numbers of landfalling and non-landfalling [a(1)],[b(1)] weak, [a(2)],[b(2)] strong, and [a(3)],[b(3)] super TCs, and the corresponding ratios of landfalling [c(1)] weak, [c(2)] strong, and [c(3)] super TCs. The solid trend lines indicate that the linear trend is significant at the 95% significance level.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

b. Evaluating the influence of climate index on the detected trend

Zhao et al. (2018b) found that the abrupt decrease of TC in the northwest Pacific around 1998 is attributed to the negative phase of IPO. Thereby, the trend detected in the annual counts of TCs may be impacted by the climate index. To separate the impact of climate index, the space–time field of TC occurrence was decomposed into different patterns by the empirical orthogonal function. Notably, here each TC track was represented by its genesis point as defined in section 3a. The original TC occurrence space–time field is irregular, so it was further processed into a regular gridded space–time field to meet the requirement of empirical orthogonal functions, and that was done based on the kernel density–based counting method as introduced in section 3a. The kernel density depends on the choice of bandwidth. In this study, the best bandwidth was estimated based on the cross-validation method; however, different best bandwidths were estimated when different ratios of data were used for testing in the cross-validation method (shown in Fig. 3). Thereby, arbitrariness in the choice of bandwidth is inevitable to some extent.

Fig. 3.
Fig. 3.

Difference in the estimated best bandwidths when different ratios of data were used for testing in the cross-validation method.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

According to Fig. 3, the maximum and minimum bandwidths were used to obtain the regularized super TC occurrence space–time field, and the differences in their decomposition were compared. Figure 4 shows that the differences in the leading first patterns of their decomposition are negligible. This indicates that the estimated bandwidths are in a reasonable range although a different ratio of data is used for testing in the cross-validation method. Hence, the change in the results caused by the different choices of bandwidth was ignored in this study, and the following bandwidth was selected based on the cross-validation method with 20% of data used for testing.

Fig. 4.
Fig. 4.

Difference in the leading first pattern of the decomposition of super TC when different choices of bandwidths were used to obtain the regularized super TC occurrence space–time field, where EOF and PC indicate the associated spatial and temporal patterns, respectively.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

The leading two patterns of the decomposition of landfalling and non-landfalling weak, strong, and super TCs were analyzed. According to the decomposition method introduced in section 3b, the associated spatial and temporal patterns were included in each leading pattern. For all types of TCs, the correlations between climate indices (i.e., PMM, ENSO, and IPO) and their temporal patterns were calculated. Table 1 shows that only the leading first patterns of landfalling and non-landfalling super TCs are significantly correlated with those climate indices. This indicated that the multidecadal internal variability of TC mainly exists in the leading first patterns of landfalling and non-landfalling super TCs. The spatial patterns of super TCs, illustrated in Figs. 5c(1) and 5c(2), show that the impacts of PMM, ENSO, and IPO on the occurrence of super TCs are different in space, the negative phases of these climate indices (corresponding to positive values in temporal patterns since negative correlations were shown in Table 1, and the value in a grid would be negative if its spatial pattern value is negative while temporal pattern value is positive) tend to decrease the occurrence of super TCs in the sea surface away from land but slightly increase it in the sea surface near land. Further, the decrease in the super TC occurrence was detected around 1998 based on Figs. 5c(1), 5c(2), and 5c(3), and this is consistent with the result of Zhao et al. (2018b) that the negative phase of IPO led to the abrupt decrease of TC in the northwest Pacific around 1998. The focus of this study is not on the multidecadal internal variability of TCs, and thus the possible physical mechanisms for the influences of PMM, ENSO, and IPO on TCs were not analyzed.

Table 1.

Correlation between climate indices and the temporal patterns (PC) in the decomposition of landfalling and non-landfalling weak (35 ≤ LMI <64 kt), strong (64 ≤ LMI < 96 kt), and super (LMI ≥ 96 kt) TCs, where A indicates landfalling TCs, B indicates non-landfalling TCs, and boldface indicates the correlation coefficient is significant at the 95% significance level.

Table 1.
Fig. 5.
Fig. 5.

The leading first patterns of landfalling and non-landfalling (a) weak, (b) strong, and (c) super TCs, where EOF and PC indicate the associated spatial and temporal patterns, respectively. The bandwidth shown in the figure is used for the kernel density as introduced in section 3a. [a(3)],[b(3)],[c(3)] The solid trend line indicates that the linear trend of PC is significant at the 95% significance level, and this is the same for Fig. 6.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

No significant relationship with climate indices was found for the leading second patterns of landfalling and non-landfalling weak, strong, and super TCs (Table 1). Therefore, the trends detected in those patterns were considered to be not influenced by the multidecadal internal variability of TC. Figure 6 shows that significant trends in the temporal patterns were detected in the leading second patterns of strong and super TCs. The spatial pattern of landfalling strong TCs is dominated by the negative value [Fig. 6b(1)] and so the total number of landfalling strong TCs decreased as the value of its associated temporal pattern increased [Fig. 6b(3)]. Similarly, Figs. 6b(2) and 6b(3) show that the total number of non-landfalling strong TCs decreased, and the increase in landfalling super TCs and decrease in non-landfalling super TCs also can be detected based on Figs. 6c(1)–c(3). The increase in landfalling super TCs mainly occurs in the sea surface near land [Fig. 6c(1)], and this is consistent with the result of Wu et al. (2015) that a westward shift in the location of TC genesis was found. Besides, the same trends detected in the leading second patterns (Fig. 6) and annual counts (Fig. 2) of strong and super TCs indicate that these trends are not entirely caused by the multidecadal internal variability of TC.

Fig. 6.
Fig. 6.

The leading second patterns of landfalling and non-landfalling (a) weak, (b) strong, and (c) super TCs, where EOF and PC indicate the associated spatial and temporal patterns, respectively. [b(1)],[c(1)] The boxes represent the region of 6°–21°N, 130°–137°E. The bandwidth shown in the figure is used for the kernel density as introduced in section 3a. [a(3)],[b(3)],[c(3)] The solid trend lines are as in Fig. 5.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

c. The possible reason for the change in landfalling super TCs

Outlined as a box in Fig. 6, a location (6°–21°N, 130°–137°E) was found where the generation number of landfalling super TCs increases in a region where the generation number of strong TCs decreases. To explain this phenomenon, a hypothesis was posed in this study that TCs forms in the box shown in Fig. 6 are more likely to intensify to super TCs in recent years. Besides, Fig. 7a shows that the occurrence frequency of landfalling strong TCs during the period 2000–18 decreased in August and September when compared to that during the period 1980–99. Meanwhile, the occurrence frequency of landfalling super TCs during the period 2000–18 increased in August and September when compared to that during 1980–99 (Fig. 7b). These further suggest that the time of intensification to super TCs mainly occurs in August and September.

Fig. 7.
Fig. 7.

Change in the generation time of landfalling (a) strong and (b) super TCs for the periods 1980–99 and 2000–18.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

d. Increased intensification rate of landfalling super TCs

In this study, the change in TC intensity was decomposed into contributions from intensification rate and intensification duration. We found that there are no trends in the intensification durations for the landfalling weak, strong, and super TCs (results not shown), and this is consistent with the result of Mei and Xie (2016). As introduced in section 3c, the intensification rates during different sub-intensification periods were calculated for the landfalling strong and super TCs. Figure 8 shows that the TC intensification rate is generally larger in the stronger wind period, and both landfalling strong and super TCs show that their intensification rates have a more obvious increasing trend during the stronger wind periods.

Fig. 8.
Fig. 8.

Trend in the annual average intensification rates of landfalling (a) strong and (b) super TCs during different sub-intensification periods as introduced in section 3c, where the labels Wind < 64, 64 ≤ Wind < 96, and Wind ≥ 96 indicate the weak, strong, and super wind periods, respectively. The solid trend line indicates that the linear trend is significant at the 95% significance level.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

The change in the TC intensification rate depends on the surrounding environment such as RH, SST, CAPE, and vertical wind shear. In this study, the relative importance of SST, RH, CAPE, and vertical wind shear on super TC’s average intensification rates during the entire intensification period was analyzed based on the random forest method as introduced in section 3c. Results (Fig. 9) show that the SST has the most importance for TC intensification rates. This is consistent with the result in Mei and Xie (2016) that the increased intensification rates of landfalling TCs in the northwest Pacific are due to the locally enhanced SST.

Fig. 9.
Fig. 9.

The relative importance of SST, RH, CAPE, and vertical wind shear to the variation of super TCs’ average intensification rates during the entire intensification period. The different colors indicate the results of the random forest method using different training data.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

The relationships between intensification rate and SST in different sub-intensification periods were also analyzed for the landfalling super TCs. Figure 10 shows that the relationships between intensification rate and SST are more significant in the stronger wind periods. As the ocean warms, the energy is stored in the lower troposphere. Kang and Elsner (2016) found that those energies prefer to discharge in the upper portion of super TCs, and this may be the climate mechanism explaining why TC intensification rates are more significantly related to SST during the stronger wind periods.

Fig. 10.
Fig. 10.

The relationship between intensification rate and its surrounding SST during different sub-intensification periods of landfalling super TCs, where Wind < 64, 64 ≤ Wind < 96, and Wind ≥ 96 indicate the weak, strong, and super wind periods, respectively. The solid line indicates that the linear trend is significant at the 95% significance level.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

In the northwest Pacific, the TCs mainly occur during July–October. In this study, the trend in average SST during July–October was analyzed. Figure 11 shows that almost the entire northwest Pacific has had an increasing trend in SST from 1980 to 2018. The higher SST induced an increase in the TC intensification rate, especially during the stronger wind periods, and then a stronger intensity of TC was finally reached. This may be the reason why the TCs forming in the box shown in Fig. 6 are more likely to intensify to super TCs in recent years.

Fig. 11.
Fig. 11.

Linear trend in average SST during July–October in the northwest Pacific from 1980 to 2018.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

e. The changes in landfalling locations of super TCs

Overall, super TCs are more likely to make landfall in the northwest Pacific. However, the risk of landfalling super TCs may be different for regions along the northwest Pacific coastline. In this study, the changes in landfalling locations of super TCs were analyzed. The power of TCs gradually decreased after landfalling; however, we believe that the destructiveness of TCs is still terrible within 24 h after landfalling. Thereby, the locations of TCs within 24 h after landfalling were considered as landfalling locations in this study. Based on the kernel density–based counting method introduced in section 3a, the annual number of landfalling super TCs was calculated in each grid along the northwest coastline, and then their linear trend was estimated. Figure 12 shows that the threats of TC attack are different in the region along the northwest Pacific coastline, but more regions are facing an increase in landfalling super TCs.

Fig. 12.
Fig. 12.

Changes in the landfalling locations of super TCs, where the black dot indicates that the linear trend in the annual occurrence frequency of landfalling TCs over the grid is significant at the 90% significance level. The bandwidth shown in the figure is used for the kernel density as introduced in section 3a.

Citation: Journal of Climate 34, 8; 10.1175/JCLI-D-20-0647.1

Western South China, Southeast China, Taiwan, South Korea, and Japan are facing an increase in landfalling super TCs (Fig. 12). Meantime, landfalling super TCs are decreasing in northern Vietnam and eastern South China (Fig. 12). Significant poleward migration of TCs is observed in the northwest Pacific (Kossin et al. 2016), which may be the reason for the increase of landfalling super TCs in Japan. Usually, the potential destructiveness of TC is considered as the cube of the maximum sustained wind speed (Emanuel 2005; Li et al. 2017); thus, the destructiveness caused by these landfalling super TCs increases exponentially with their increases in wind speed. The increased exposure risk of super TCs is therefore very dangerous for these economically developed and highly populated regions, especially for regions that lack effective response measures. In 2017, super Typhoon Hato hit Macao, a city in western South China with very high population density. Due to the lack of effective measures to deal with super TCs in Macao, great damage and loss were caused by Typhoon Hato, including 12 deaths, an economic loss of 12.55 billion patacas, and longtime water and power failure in some areas (https://en.wikipedia.org/wiki/Typhoon_Hato, accessed 15 August 2020). The landfalling super TCs are found to be increased in western South China, indicating that the cities in western South China need to prepare for the next attacks of super TCs like Hato.

5. Conclusions

By focusing on the differences in the occurrence of landfalling and non-landfalling TCs, we found that the super TCs are more likely to landfall in the northwest Pacific. The reason for this change is that TCs forming in the sea surface near land (6°–21°N, 130°–137°E) are more likely to intensify to super TCs in recent years and they tend to favor subsequent landfall. The intensification of TC is mainly due to the increase in the intensification rate. Based on the permutation-based variable importance in a random forest method, results show that the SST has the most importance for TC intensification rates, and TC intensification rates increase with increased SST, especially during the stronger wind periods.

Along with the change in the occurrence frequency of landfalling super TCs, the landfalling locations of super TCs also changed. We found that western South China, Southeast China, Taiwan, South Korea, and Japan are facing an increase in landfalling super TCs, while landfalling super TCs are decreasing in northern Vietnam and eastern South China. The destructiveness of super TCs is great, especially for the economically developed and highly populated regions. Hence, results in this study would be helpful for the policymakers and residents to prepare for the next attack of super TCs.

Acknowledgments

This work is financially supported by a grant from the “Macao Young Scholars Program” (AM201903), the Macau Science and Technology Development Fund (0045/2018/AFJ), and the National Natural Science Foundation of China (51909057, 42071055). The author would like to thank three anonymous reviewers for their constructive comments and review of the manuscript, and Dr. Dongdong Kong for his constructive comments during the draft.

Data availability statement.

All data used in this paper are available online. The tropical cyclone data used in this study are collected from the International Best Track Archive for Climate Stewardship (IBTrACS, version 4, https://www.ncdc.noaa.gov/ibtracs/). The ERA5 data used in this study are downloaded from the Copernicus Climate Change Service (C3S) Climate Date Store (https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-land?tab=overview). The PMM, IPO, and ENSO indices are obtained from the NOAA Physical Sciences Laboratory and NOAA Climate Prediction Center, respectively.

REFERENCES

  • Breiman, L., 2001: Random forests. Mach. Learn., 45, 532, https://doi.org/10.1023/A:1010933404324.

  • Chiang, J. C. H., and D. J. Vimont, 2004: Analogous Pacific and Atlantic meridional modes of tropical atmosphere–ocean variability. J. Climate, 17, 41434158, https://doi.org/10.1175/JCLI4953.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dawson, A., 2016: eofs: A library for EOF analysis of meteorological, oceanographic, and climate data. J. Open Res. Software, 4, e14, https://doi.org/10.5334/jors.122.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K., 2005: Increasing destructiveness of tropical cyclones over the past 30 years. Nature, 436, 686688, https://doi.org/10.1038/nature03906.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K., 2017: Assessing the present and future probability of Hurricane Harvey’s rainfall. Proc. Natl. Acad. Sci. USA, 114, 12 68112 684, https://doi.org/10.1073/pnas.1716222114.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, S., L. Zhu, W. Zhang, and X. Shen, 2020: Impact of the Pacific meridional mode on landfalling tropical cyclone frequency in China. Quart. J. Roy. Meteor. Soc., 146, 24102420, https://doi.org/10.1002/qj.3799.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hannachi, A., I. T. Jolliffe, and D. B. Stephenson, 2007: Empirical orthogonal functions and related techniques in atmospheric science: A review. Int. J. Climatol., 27, 11191152, https://doi.org/10.1002/joc.1499.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Henley, B. J., J. Gergis, D. J. Karoly, S. Power, J. Kennedy, and C. K. Folland, 2015: A tripole index for the interdecadal Pacific oscillation. Climate Dyn., 45, 30773090, https://doi.org/10.1007/s00382-015-2525-1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

  • Hsu, P.-C., P.-S. Chu, H. Murakami, and X. Zhao, 2014: An abrupt decrease in the late-season typhoon activity over the western North Pacific. J. Climate, 27, 42964312, https://doi.org/10.1175/JCLI-D-13-00417.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kang, N.-Y., and J. B. Elsner, 2015: Trade-off between intensity and frequency of global tropical cyclones. Nat. Climate Change, 5, 661664, https://doi.org/10.1038/nclimate2646.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kang, N.-Y., and J. B. Elsner, 2016: Climate mechanism for stronger typhoons in a warmer world. J. Climate, 29, 10511057, https://doi.org/10.1175/JCLI-D-15-0585.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knutson, T., and Coauthors, 2019: Tropical cyclones and climate change assessment: Part I: Detection and attribution. Bull. Amer. Meteor. Soc., 100, 19872007, https://doi.org/10.1175/BAMS-D-18-0189.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knutson, T., and Coauthors, 2020: Tropical cyclones and climate change assessment: Part II: Projected response to anthropogenic warming. Bull. Amer. Meteor. Soc., 101, E303E322, https://doi.org/10.1175/BAMS-D-18-0194.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., T. L. Olander, and K. R. Knapp, 2013: Trend analysis with a new global record of tropical cyclone intensity. J. Climate, 26, 99609976, https://doi.org/10.1175/JCLI-D-13-00262.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., K. A. Emanuel, and S. J. Camargo, 2016: Past and projected changes in western North Pacific tropical cyclone exposure. J. Climate, 29, 57255739, https://doi.org/10.1175/JCLI-D-16-0076.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lai, Y., and Coauthors, 2020: Greater flood risks in response to slowdown of tropical cyclones over the coast of China. Proc. Natl. Acad. Sci. USA, 117, 14 75114 755, https://doi.org/10.1073/pnas.1918987117.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, R. C. Y., W. Zhou, C. M. Shun, and T. C. Lee, 2017: Change in destructiveness of landfalling tropical cyclones over China in recent decades. J. Climate, 30, 33673379, https://doi.org/10.1175/JCLI-D-16-0258.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, L., and Y. Wang, 2020: Trends in landfalling tropical cyclone–induced precipitation over China. J. Climate, 33, 22232235, https://doi.org/10.1175/JCLI-D-19-0693.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, L., Y. Wang, R. Zhan, J. Xu, and Y. Duan, 2020: Increasing destructive potential of landfalling tropical cyclones over China. J. Climate, 33, 37313743, https://doi.org/10.1175/JCLI-D-19-0451.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lu, M., and R. Xiong, 2019: Spatiotemporal profiling of tropical cyclones genesis and favorable environmental conditions in the western Pacific basin. Geophys. Res. Lett., 46, 11 54811 558, https://doi.org/10.1029/2019GL084995.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mei, W., and S.-P. Xie, 2016: Intensification of landfalling typhoons over the northwest Pacific since the late 1970s. Nat. Geosci., 9, 753757, https://doi.org/10.1038/ngeo2792.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murakami, H., T. L. Delworth, W. F. Cooke, M. Zhao, B. Xiang, and P.-C. Hsu, 2020: Detected climatic change in global distribution of tropical cyclones. Proc. Natl. Acad. Sci. USA, 117, 10 70610 714, https://doi.org/10.1073/pnas.1922500117.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pedregosa, F., and Coauthors, 2011: Scikit-learn: Machine learning in Python. J. Mach. Learn. Res., 12, 28252830, https://dl.acm.org/doi/10.5555/1953048.2078195.

    • Search Google Scholar
    • Export Citation
  • Peduzzi, P., B. Chatenoux, H. Dao, A. De Bono, C. Herold, J. Kossin, F. Mouton, and O. Nordbeck, 2012: Global trends in tropical cyclone risk. Nat. Climate Change, 2, 289294, https://doi.org/10.1038/nclimate1410.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schreck, C. J., III, K. R. Knapp, and J. P. Kossin, 2014: The impact of best track discrepancies on global tropical cyclone climatologies using IBTrACS. Mon. Wea. Rev., 142, 38813899, https://doi.org/10.1175/MWR-D-14-00021.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tippett, M. K., S. J. Camargo, and A. H. Sobel, 2011: A Poisson regression index for tropical cyclone genesis and the role of large-scale vorticity in genesis. J. Climate, 24, 23352357, https://doi.org/10.1175/2010JCLI3811.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weinkle, J., R. Maue, and R. Pielke Jr., 2012: Historical global tropical cyclone landfalls. J. Climate, 25, 47294735, https://doi.org/10.1175/JCLI-D-11-00719.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, L., C. Wang, and B. Wang, 2015: Westward shift of western North Pacific tropical cyclogenesis. Geophys. Res. Lett., 42, 15371542, https://doi.org/10.1002/2015GL063450.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xiao, M., 2020: Quantifying spatiotemporal influences of climate index on seasonal extreme precipitation based on hierarchical Bayesian method. Int. J. Climatol., 40, 30873098, https://doi.org/10.1002/joc.6384.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xiao, M., Q. Zhang, and V. P. Singh, 2015: Influences of ENSO, NAO, IOD and PDO on seasonal precipitation regimes in the Yangtze River basin, China. Int. J. Climatol., 35, 35563567, https://doi.org/10.1002/joc.4228.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xiao, M., and Coauthors, 2020: Stomatal response to decreased relative humidity constrains the acceleration of terrestrial evapotranspiration. Environ. Res. Lett., 15, 094066, https://doi.org/10.1088/1748-9326/ab9967.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhan, R., Y. Wang, and J. Zhao, 2017: Intensified mega-ENSO has increased the proportion of intense tropical cyclones over the western northwest Pacific since the late 1970s. Geophys. Res. Lett., 44, 11 95911 966, https://doi.org/10.1002/2017GL075916.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, Q., L. Wu, and Q. Liu, 2009: Tropical cyclone damages in China 1983–2006. Bull. Amer. Meteor. Soc., 90, 489496, https://doi.org/10.1175/2008BAMS2631.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, W., H.-F. Graf, Y. Leung, and M. Herzog, 2012: Different El Niño types and tropical cyclone landfall in East Asia. J. Climate, 25, 65106523, https://doi.org/10.1175/JCLI-D-11-00488.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, W., G. A. Vecchi, H. Murakami, G. Villarini, T. L. Delworth, X. Yang, and L. Jia, 2018: Dominant role of Atlantic multidecadal oscillation in the recent decadal changes in western North Pacific tropical cyclone activity. Geophys. Res. Lett., 45, 354362, https://doi.org/10.1002/2017GL076397.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhao, J., R. Zhan, and Y. Wang, 2018a: Global warming hiatus contributed to the increased occurrence of intense tropical cyclones in the coastal regions along East Asia. Sci. Rep., 8, 6023, https://doi.org/10.1038/s41598-018-24402-2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhao, J., R. Zhan, Y. Wang, and H. Xu, 2018b: Contribution of the interdecadal Pacific oscillation to the recent abrupt decrease in tropical cyclone genesis frequency over the western North Pacific since 1998. J. Climate, 31, 82118224, https://doi.org/10.1175/JCLI-D-18-0202.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save
  • Breiman, L., 2001: Random forests. Mach. Learn., 45, 532, https://doi.org/10.1023/A:1010933404324.

  • Chiang, J. C. H., and D. J. Vimont, 2004: Analogous Pacific and Atlantic meridional modes of tropical atmosphere–ocean variability. J. Climate, 17, 41434158, https://doi.org/10.1175/JCLI4953.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dawson, A., 2016: eofs: A library for EOF analysis of meteorological, oceanographic, and climate data. J. Open Res. Software, 4, e14, https://doi.org/10.5334/jors.122.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K., 2005: Increasing destructiveness of tropical cyclones over the past 30 years. Nature, 436, 686688, https://doi.org/10.1038/nature03906.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K., 2017: Assessing the present and future probability of Hurricane Harvey’s rainfall. Proc. Natl. Acad. Sci. USA, 114, 12 68112 684, https://doi.org/10.1073/pnas.1716222114.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gao, S., L. Zhu, W. Zhang, and X. Shen, 2020: Impact of the Pacific meridional mode on landfalling tropical cyclone frequency in China. Quart. J. Roy. Meteor. Soc., 146, 24102420, https://doi.org/10.1002/qj.3799.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hannachi, A., I. T. Jolliffe, and D. B. Stephenson, 2007: Empirical orthogonal functions and related techniques in atmospheric science: A review. Int. J. Climatol., 27, 11191152, https://doi.org/10.1002/joc.1499.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Henley, B. J., J. Gergis, D. J. Karoly, S. Power, J. Kennedy, and C. K. Folland, 2015: A tripole index for the interdecadal Pacific oscillation. Climate Dyn., 45, 30773090, https://doi.org/10.1007/s00382-015-2525-1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 19992049, https://doi.org/10.1002/qj.3803.

  • Hsu, P.-C., P.-S. Chu, H. Murakami, and X. Zhao, 2014: An abrupt decrease in the late-season typhoon activity over the western North Pacific. J. Climate, 27, 42964312, https://doi.org/10.1175/JCLI-D-13-00417.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kang, N.-Y., and J. B. Elsner, 2015: Trade-off between intensity and frequency of global tropical cyclones. Nat. Climate Change, 5, 661664, https://doi.org/10.1038/nclimate2646.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kang, N.-Y., and J. B. Elsner, 2016: Climate mechanism for stronger typhoons in a warmer world. J. Climate, 29, 10511057, https://doi.org/10.1175/JCLI-D-15-0585.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knutson, T., and Coauthors, 2019: Tropical cyclones and climate change assessment: Part I: Detection and attribution. Bull. Amer. Meteor. Soc., 100, 19872007, https://doi.org/10.1175/BAMS-D-18-0189.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knutson, T., and Coauthors, 2020: Tropical cyclones and climate change assessment: Part II: Projected response to anthropogenic warming. Bull. Amer. Meteor. Soc., 101, E303E322, https://doi.org/10.1175/BAMS-D-18-0194.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., T. L. Olander, and K. R. Knapp, 2013: Trend analysis with a new global record of tropical cyclone intensity. J. Climate, 26, 99609976, https://doi.org/10.1175/JCLI-D-13-00262.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., K. A. Emanuel, and S. J. Camargo, 2016: Past and projected changes in western North Pacific tropical cyclone exposure. J. Climate, 29, 57255739, https://doi.org/10.1175/JCLI-D-16-0076.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lai, Y., and Coauthors, 2020: Greater flood risks in response to slowdown of tropical cyclones over the coast of China. Proc. Natl. Acad. Sci. USA, 117, 14 75114 755, https://doi.org/10.1073/pnas.1918987117.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, R. C. Y., W. Zhou, C. M. Shun, and T. C. Lee, 2017: Change in destructiveness of landfalling tropical cyclones over China in recent decades. J. Climate, 30, 33673379, https://doi.org/10.1175/JCLI-D-16-0258.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, L., and Y. Wang, 2020: Trends in landfalling tropical cyclone–induced precipitation over China. J. Climate, 33, 22232235, https://doi.org/10.1175/JCLI-D-19-0693.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Liu, L., Y. Wang, R. Zhan, J. Xu, and Y. Duan, 2020: Increasing destructive potential of landfalling tropical cyclones over China. J. Climate, 33, 37313743, https://doi.org/10.1175/JCLI-D-19-0451.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lu, M., and R. Xiong, 2019: Spatiotemporal profiling of tropical cyclones genesis and favorable environmental conditions in the western Pacific basin. Geophys. Res. Lett., 46, 11 54811 558, https://doi.org/10.1029/2019GL084995.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mei, W., and S.-P. Xie, 2016: Intensification of landfalling typhoons over the northwest Pacific since the late 1970s. Nat. Geosci., 9, 753757, https://doi.org/10.1038/ngeo2792.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murakami, H., T. L. Delworth, W. F. Cooke, M. Zhao, B. Xiang, and P.-C. Hsu, 2020: Detected climatic change in global distribution of tropical cyclones. Proc. Natl. Acad. Sci. USA, 117, 10 70610 714, https://doi.org/10.1073/pnas.1922500117.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pedregosa, F., and Coauthors, 2011: Scikit-learn: Machine learning in Python. J. Mach. Learn. Res., 12, 28252830, https://dl.acm.org/doi/10.5555/1953048.2078195.

    • Search Google Scholar
    • Export Citation
  • Peduzzi, P., B. Chatenoux, H. Dao, A. De Bono, C. Herold, J. Kossin, F. Mouton, and O. Nordbeck, 2012: Global trends in tropical cyclone risk. Nat. Climate Change, 2, 289294, https://doi.org/10.1038/nclimate1410.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schreck, C. J., III, K. R. Knapp, and J. P. Kossin, 2014: The impact of best track discrepancies on global tropical cyclone climatologies using IBTrACS. Mon. Wea. Rev., 142, 38813899, https://doi.org/10.1175/MWR-D-14-00021.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tippett, M. K., S. J. Camargo, and A. H. Sobel, 2011: A Poisson regression index for tropical cyclone genesis and the role of large-scale vorticity in genesis. J. Climate, 24, 23352357, https://doi.org/10.1175/2010JCLI3811.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weinkle, J., R. Maue, and R. Pielke Jr., 2012: Historical global tropical cyclone landfalls. J. Climate, 25, 47294735, https://doi.org/10.1175/JCLI-D-11-00719.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wu, L., C. Wang, and B. Wang, 2015: Westward shift of western North Pacific tropical cyclogenesis. Geophys. Res. Lett., 42, 15371542, https://doi.org/10.1002/2015GL063450.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xiao, M., 2020: Quantifying spatiotemporal influences of climate index on seasonal extreme precipitation based on hierarchical Bayesian method. Int. J. Climatol., 40, 30873098, https://doi.org/10.1002/joc.6384.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xiao, M., Q. Zhang, and V. P. Singh, 2015: Influences of ENSO, NAO, IOD and PDO on seasonal precipitation regimes in the Yangtze River basin, China. Int. J. Climatol., 35, 35563567, https://doi.org/10.1002/joc.4228.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Xiao, M., and Coauthors, 2020: Stomatal response to decreased relative humidity constrains the acceleration of terrestrial evapotranspiration. Environ. Res. Lett., 15, 094066, https://doi.org/10.1088/1748-9326/ab9967.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhan, R., Y. Wang, and J. Zhao, 2017: Intensified mega-ENSO has increased the proportion of intense tropical cyclones over the western northwest Pacific since the late 1970s. Geophys. Res. Lett., 44, 11 95911 966, https://doi.org/10.1002/2017GL075916.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, Q., L. Wu, and Q. Liu, 2009: Tropical cyclone damages in China 1983–2006. Bull. Amer. Meteor. Soc., 90, 489496, https://doi.org/10.1175/2008BAMS2631.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, W., H.-F. Graf, Y. Leung, and M. Herzog, 2012: Different El Niño types and tropical cyclone landfall in East Asia. J. Climate, 25, 65106523, https://doi.org/10.1175/JCLI-D-11-00488.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhang, W., G. A. Vecchi, H. Murakami, G. Villarini, T. L. Delworth, X. Yang, and L. Jia, 2018: Dominant role of Atlantic multidecadal oscillation in the recent decadal changes in western North Pacific tropical cyclone activity. Geophys. Res. Lett., 45, 354362, https://doi.org/10.1002/2017GL076397.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhao, J., R. Zhan, and Y. Wang, 2018a: Global warming hiatus contributed to the increased occurrence of intense tropical cyclones in the coastal regions along East Asia. Sci. Rep., 8, 6023, https://doi.org/10.1038/s41598-018-24402-2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zhao, J., R. Zhan, Y. Wang, and H. Xu, 2018b: Contribution of the interdecadal Pacific oscillation to the recent abrupt decrease in tropical cyclone genesis frequency over the western North Pacific since 1998. J. Climate, 31, 82118224, https://doi.org/10.1175/JCLI-D-18-0202.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Change in the annual counts of landfalling (a) weak, (b) strong, and (c) super TCs according to three different landfalling definitions.

  • Fig. 2.

    Trends in the annual numbers of landfalling and non-landfalling [a(1)],[b(1)] weak, [a(2)],[b(2)] strong, and [a(3)],[b(3)] super TCs, and the corresponding ratios of landfalling [c(1)] weak, [c(2)] strong, and [c(3)] super TCs. The solid trend lines indicate that the linear trend is significant at the 95% significance level.

  • Fig. 3.

    Difference in the estimated best bandwidths when different ratios of data were used for testing in the cross-validation method.

  • Fig. 4.

    Difference in the leading first pattern of the decomposition of super TC when different choices of bandwidths were used to obtain the regularized super TC occurrence space–time field, where EOF and PC indicate the associated spatial and temporal patterns, respectively.

  • Fig. 5.

    The leading first patterns of landfalling and non-landfalling (a) weak, (b) strong, and (c) super TCs, where EOF and PC indicate the associated spatial and temporal patterns, respectively. The bandwidth shown in the figure is used for the kernel density as introduced in section 3a. [a(3)],[b(3)],[c(3)] The solid trend line indicates that the linear trend of PC is significant at the 95% significance level, and this is the same for Fig. 6.

  • Fig. 6.

    The leading second patterns of landfalling and non-landfalling (a) weak, (b) strong, and (c) super TCs, where EOF and PC indicate the associated spatial and temporal patterns, respectively. [b(1)],[c(1)] The boxes represent the region of 6°–21°N, 130°–137°E. The bandwidth shown in the figure is used for the kernel density as introduced in section 3a. [a(3)],[b(3)],[c(3)] The solid trend lines are as in Fig. 5.

  • Fig. 7.

    Change in the generation time of landfalling (a) strong and (b) super TCs for the periods 1980–99 and 2000–18.

  • Fig. 8.

    Trend in the annual average intensification rates of landfalling (a) strong and (b) super TCs during different sub-intensification periods as introduced in section 3c, where the labels Wind < 64, 64 ≤ Wind < 96, and Wind ≥ 96 indicate the weak, strong, and super wind periods, respectively. The solid trend line indicates that the linear trend is significant at the 95% significance level.

  • Fig. 9.

    The relative importance of SST, RH, CAPE, and vertical wind shear to the variation of super TCs’ average intensification rates during the entire intensification period. The different colors indicate the results of the random forest method using different training data.

  • Fig. 10.

    The relationship between intensification rate and its surrounding SST during different sub-intensification periods of landfalling super TCs, where Wind < 64, 64 ≤ Wind < 96, and Wind ≥ 96 indicate the weak, strong, and super wind periods, respectively. The solid line indicates that the linear trend is significant at the 95% significance level.

  • Fig. 11.

    Linear trend in average SST during July–October in the northwest Pacific from 1980 to 2018.

  • Fig. 12.

    Changes in the landfalling locations of super TCs, where the black dot indicates that the linear trend in the annual occurrence frequency of landfalling TCs over the grid is significant at the 90% significance level. The bandwidth shown in the figure is used for the kernel density as introduced in section 3a.

All Time Past Year Past 30 Days
Abstract Views 977 0 0
Full Text Views 939 238 84
PDF Downloads 612 113 20