1 Introduction

A certain percentage of typhoons undergo an extratropical transition (ET) when they move northward into higher latitudes. These typhoons bring strong winds and heavy precipitation near their center because of their interaction with the midlatitude baroclinic environment, leading to rapid deepening (Harr and Elsberry 2000; Jones et al. 2003; Anwender et al. 2010; Evans et al. 2017). They cause widespread damage, particularly in the areas where they make landfall. During the ET process, the structures of typhoons change dramatically, transforming from a symmetric, nonfrontal warm-core structure to an asymmetric, frontal cold-core structure (Klein et al. 2000; Quinting et al. 2014; Takamura and Wada 2020). In addition, the frontal features of typhoons can amplify the wind and precipitation distribution, not only around the typhoon center but also in broader areas (Nayak and Takemi 2020a). As a result, typhoons have substantial negative societal impacts around the areas where landfall occurs.

In 2018 and 2019, Japan experienced the devastating impact of three intense typhoons—Jebi (2018, T1821), Faxai (2019, T1915), and Hagibis (2019, T1919)—all of which made landfall and caused extensive damage to areas in their paths, including two megacities. These three typhoons were some of the strongest and most extensive typhoons to hit Japan, with lifetime maximum wind speeds of approximately 54, 44, and 54 m s−1 and lifetime minimum surface pressures of 915, 955, and 915 hPa for T1821, T1915, and T1919, respectively.

Takamura and Wada (2020) recently reported that the abnormally high intensity of typhoons that made landfall over northern Japan in August 2016 was caused by a swift structural transformation from a warm core to a cold core. By contrast, Kitabatake and Bessho (2008) demonstrated that the typhoons that made landfall over Japan in 2004–2006 mostly underwent an ET. Other studies have also examined the ET of several typhoons that made landfall over Japan (Kitabatake 2002, 2011). However, some powerful typhoons that struck northern Japan, such as Lionrock (2016), did not experience an ET (Nayak and Takemi 2020b). Therefore, not all strong typhoons moving northward undergo an ET.

In a review, Evans et al. (2017) highlighted the importance of understanding the structural evolution of cyclones during their ET transformation stage. Here, the term “stage” refers a specific period in the cyclone’s development and behavior. Each stage is characterized by unique changes in the cyclone’s structure, intensity, and behavior and represents a specific point in the cyclone’s lifecycle. The stages of a cyclone include the formation, mature, transformation, and dissipation stages. Each of these stages is defined by particular characteristics, such as the strength and position of the low-pressure center and the location and orientation of the warm and cold fronts.

According to Kitabatake (2011), approximately 40% of typhoons that occurred between 1979 and 2004 underwent an ET before dissipating. September is typically the month with the largest number of typhoons experiencing an ET, whereas both the maximum peak intensity and ET tend to occur in September and October (Klein et al. 2000; Kitabatake 2011; Evans et al. 2017). However, despite the fact that three powerful typhoons (T1821, T1915, and T1919) formed during these months, whether they experienced an ET and how their inner-core structure changed remain unclear. Previous research on these typhoons has mostly focused on their intensities over time and the surrounding environmental conditions (e.g., Takemi et al. 2019; Takemi and Unuma 2020; Wada and Chan 2021). However, their inner-core structural changes, which are critical to understanding their changes in intensity, have not been thoroughly studied. Therefore, a comprehensive analysis of these factors, along with other typhoon features, is necessary to attain a complete understanding of the ET processes of major typhoons. Specifically, the ET processes and the associated intensity changes depend heavily on the properties of typhoon inner-core structures and how the synoptic-scale environments affect these structures.

The present study aims to examine the structural changes of T1821, T1915, and T1919 during their northward movements across the Japanese islands. Using cyclone phase-space analysis, we investigate all the typhoons that approached Japan between 2018 and 2019. Our analysis focuses on three aspects: (1) the typhoons’ warm- and cold-core structures and whether they exhibited frontal characteristics; (2) their temporal evolutions in terms of central pressure and wind speed; and (3) the synoptic environments leading to typhoons undergoing an ET.

2 Data and study method

The Japanese Reanalysis (JRA-55) data with 1.25° spatial and 6 h temporal resolution were used to analyze the ETs of 26 typhoons that approached Japan in 2018–2019 (Kobayashi et al. 2015). These typhoons were selected on the basis of whether they made landfall or whether their center passed within 100 km of a Japanese landmass before making landfall. The best-track datasets of the Regional Specialized Meteorological Center (RSMC) Tokyo were used to investigate the time evolutions of maximum wind speed and minimum central pressure.

Figure 1a shows the tracks of 26 typhoons that made landfall in 2018–2019. Figure 1b shows the tracks of typhoons of T1821, T1915, and T1919. Whereas T1915 and T1919 formed at similar locations and made landfall in the same regions of eastern Japan, T1821 formed at a different location and made landfall over western Japan (Fig. 1b). T1821 initially moved in a northwest direction, maintaining an almost uniform translation speed until it reached ~ 30°N, where it then moved relatively fast in the northward direction. Although T1915 initially traveled at the same speed as T1821, it moved relatively fast after reaching ~ 20°N. T1919 initially started westward with a comparatively higher translation speed than the other two typhoons until it reached ~ 20°N; it then traveled relatively slowly northward until reaching ~ 30°N, and then moved fast in the northeast direction.

Fig. 1
figure 1

Tracks of a 26 typhoons approaching Japan in 2018–2019 and b Typhoons Jebi (T1821), Faxai (T1915), and Hagibis (T1919). Marks along the tracks in a and b correspond to the location of the three typhoons in 12 h intervals. c Frontal nature and d warm—and cold-core structures of the 26 typhoons obtained through phase-space parameters in 6 h intervals. Blue (red) denotes the start (decay) of the typhoon. Parameter B in (c) and (d) represents the horizontal thermal symmetry; it indicates a frontal (asymmetric) structure when B >> 0 and a nonfrontal (symmetric) structure when B < 10

Full size image

Similar to Hart (2003), we computed three cyclone phase parameters (B, − VTL, and − VTU) to categorize the typhoons’ structures. Parameter B characterizes the strength and refers to the storm-motion-relative 900‒600 hPa thickness asymmetry across the tropical cyclone (TC) within a 500 km radius. Mathematically,

$$B=h\left(\overline{{Z }_{600 \mathrm{hPa}}-{Z}_{900 \mathrm{hPa}}}{|}_{\mathrm{R}}-\overline{{Z }_{600 \mathrm{hPa}}-{Z}_{900 \mathrm{hPa}}}{|}_{\mathrm{L}}\right),$$
(1)

where Z is the isobaric height, R indicates right of the current storm motion, and L indicates left of the storm motion. The overbar indicates the areal mean over a semicircle of radius 500 km. The integer h takes a value of + 1 (− 1) for the Northern (Southern) Hemisphere.

Parameters − VTL and − VTU represent the lower and upper atmospheric thermal winds and refer to the vertical derivatives of the horizontal height gradient at 900‒600 hPa and 600‒300 hPa, respectively. Mathematically,

$$- \left|{V}_{T}^{L}\right|={ \left.\frac{\partial \left(\Delta Z\right)}{\partial \mathrm{ln}p}\right|}_{900 \mathrm{hPa}}^{600 \mathrm{hPa}},$$
(2)
$$- \left|{V}_{T}^{U}\right|={\left.\frac{\partial \left(\Delta Z\right)}{\partial \mathrm{ln}p}\right|}_{600 \mathrm{hPa}}^{300 \mathrm{hPa}},$$
(3)

where ΔZ is the isobaric height perturbation (Zmax − Zmin) evaluated within a radius of 500 km. \(p\) is the pressure and the negative signs of the parameters − VTL and − VTU denote that the pressure decreases with height.

The B and − VTL values for the structure of a conventional tropical cyclone are typically B ≈ 0 and − VTL > 0. When a tropical cyclone experiences an ET, its structure corresponds to B > 10 and − VTL < 0. Parameter B represents the horizontal thermal symmetry. It indicates a frontal (asymmetric) structure when B >  > 0 and a nonfrontal (symmetric) structure when B < 10 (Hart 2003). Table 1 shows the time evolution of these parameters for T1821, T1915, and T1919 in 6 h intervals, which we used to examine their frontal and warm—and cold-core changes. We also analyzed low-level convergence, upper-level divergence, atmospheric temperature, and wind for these three typhoons to understand the synoptic environments before they underwent ETs.

Table 1 Classification of the structural characteristics based on the phase space parameter
Full size table

3 Results and discussion

3.1 Typhoon track and phase-space parameters

Figure 1 depicts the tracks of 26 typhoons, including T1821, T1915, and T1919, and their corresponding phase-space diagrams obtained through phase-space parameters. The majority of the typhoons made landfall in Japan and moved northward, far from 30°N (Fig. 1a). Figures 1c and d illustrate the structural transformation of the 26 typhoons in 6 h intervals. Most of the typhoons underwent a transitional phase from a nonfrontal warm core to a frontal cold core. A few typhoons, including T1919, demonstrated higher B values (> 50) before decay, indicating stronger fronts (Fig. 1c). However, most typhoons exhibited a deep warm-core structure in their early stage and underwent an ET with a deep cold-core structure. A few typhoons, including T1915 and T1919, displayed shallow or moderate warm cores during their lifetime (Fig. 1d). T1821 primarily had a deep warm core at all stages before their ET.

Figure 2 presents the phase diagrams of T1821, T1915, and T1919. The B values of all three typhoons were less than 10 in the typhoons’ early periods but increased to greater than 40 before the typhoons decayed. Therefore, these three typhoons formed as conventional TCs and underwent an ET before decay. T1919 exhibited a transitional phase (i.e., an asymmetric warm phase), whereas the other two did not. The values of − VTL and − VTU within the 500-km radius from the typhoon center were mostly positive until immediately before landfall. Here, a 500-km radius was selected on the basis of previous studies that showed that the values of cyclone phase parameters do not substantially differ (Hart 2003; Nayak and Takemi 2020b). T1821 exhibited a deep warm-core structure until undergoing an ET, T1915 displayed a moderate warm-core structure, and T1919 showed shallow and moderate warm-core structures at some stages before undergoing an ET. Studies related to other typhoons that underwent an ET have documented similar structural changes (Klein et al. 2000; Kitabatake 2002; Nayak and Takemi 2020b; Takamura and Wada 2020).

Fig. 2
figure 2

Phase-space diagram of a–b T1821, c–d T1915, and e–f T1919. The upper panels show the frontal nature and the lower panels show warm—and cold-core structures. Blue (red) denotes the start (decay) of the typhoon. Time \(T=0\) here corresponds to the onset of the ET. Parameter B in a, c, and e represents the horizontal thermal symmetry; it indicates a frontal (asymmetric) structure when B >> 0 and a nonfrontal (symmetric) structure when B < 10

Full size image

3.2 Temporal evolution of the phase-space parameters

Figure 3 displays the time series of the phase-space parameters, central pressure, and wind speed of the three typhoons from their initiation (T = 0) until their decay, in 6 h intervals. Because the magnitudes of these quantities differ greatly, they were normalized to enable a comparison in a single figure. T1821 displayed a symmetric deep warm-core structure from the initial time until approximately the fourth day, then transformed into a symmetric cold-core structure for approximately 1 day but returned to a symmetric deep warm-core structure until the eighth day. Afterward, it reverted to a symmetric deep cold-core structure until it became extratropical with an asymmetric deep cold-core structure.

Fig. 3
figure 3

Temporal evolution of the central pressure and wind speed with a warm-core structure (upper panel) and a cold-core structure (lower panel) for Typhoons a, d T1821, b, e T1915, and c, f T1919. The results correspond to the normalized magnitude. The vertical shaded line indicates the extratropical transition time. The black dotted line indicates the transformation time from a cold core to a warm core

Full size image

T1915 exhibited a symmetric deep warm-core structure on the first day, a symmetric shallow warm-core structure on the second day, and then returned to a symmetric deep warm-core structure until the sixth day. It thereafter developed a symmetric cold-core structure with a deep warm core for some time before returning to a symmetric deep warm-core structure until the eighth day. It finally became extratropical.

T1919 displayed a symmetric deep warm-core structure for the first 18 h, then transitioned into a symmetric cold-core structure until the third day. Afterward, it exhibited a symmetric deep warm core until the seventh day but presented a symmetric cold core for a few hours on the fourth day. It exhibited an asymmetric deep warm-core structure on the eighth day until it became extratropical.

Notably, all three typhoons reached their maximum intensities (maximum wind speed and minimum pressure) immediately after transitioning from a cold-core structure to a warm-core structure, mostly exhibiting a symmetric deep warm core during this period. The maximum wind speed of T1821 was reached at approximately the fourth day from its initiation, ~ 5 days prior to its ET, whereas T1915 attained a maximum wind speed after the fifth day of its initiation, just 1 day before experiencing an ET. T1919 reached its maximum wind speed approximately the third day after its initiation, which was 5 days before becoming extratropical. T1919 maintained its maximum wind speed for 3 days, whereas the other two maintained their maximum wind speeds for 1–2 days.

To understand the strengths of T1821, T1919, and T1915 relative to those of other typhoons, we compared the temporal variations of the maximum wind speed of these three typhoons with all other typhoons that approached Japan in 2018–2019 (Fig. 4). The maximum wind speeds of T1821 and T1919 were in the top three among all 26 typhoons, whereas that of T1915 was relatively low. These observations indicate that the intensity of strong typhoons depends on the properties of the inner-core structures of the typhoons during the development of the ET.

Fig. 4
figure 4

Temporal evolution of the maximum wind speed of 26 typhoons at 6 h intervals (black dots). Solid lines represent the maximum wind speed of T1821, T1915, and T1919

Full size image

The overall analysis indicates that structural changes during the movement of these three typhoons played major roles in the intensification of their activity in terms of strong winds. T1821, T1915, and T1919 all exhibited interactions with mid-latitude baroclinic environments, which strongly influenced their behaviors and impacts. These interactions led to the storms producing strong cold fronts, becoming elongated and asymmetric, and undergoing rapid intensification immediately prior to their ET. In the case of T1821, the storm interacted with a mid-latitude baroclinic zone, which contributed to its rapid intensification immediately before landfall. The storm also generated a strong cold front as it traversed Japan, resulting in substantial wind effects. Similarly, T1915 interacted with a potent mid-latitude weather system, causing its elongation and asymmetry. This interaction also led to the storm’s atypical rapid intensification immediately before its ET. T1919 interacted with a low-pressure trough over Japan, leading to its asymmetric structure and displacement of the strongest winds to the northeast of its center. The storm also produced a robust cold front as it crossed Japan, which caused substantial wind effects.

3.3 Synoptic environment

To gain insight into the environments involved in the ET processes, we analyzed the composites of middle-level winds and temperatures, low-level convergence, and upper-level divergence at various times before and after the ET of each typhoon. In this context, \(T=0\) refers to the onset of the ET. Figure 5 depicts these characteristics for T1821, which had a nonfrontal thermal structure at \(T=-48\) h (Fig. 5a) and at \(T=-24\) h (Fig. 5b). However, during the transition, it lost its symmetry (Fig. 5c); it then demonstrated a frontal thermal structure after the ET (Fig. 5d). The wind fields transformed from concentric to nonconcentric and gradually spread over an area of low-level convergence from \(T=-48\) h until reaching their maximum during the ET at \(T=0\) over the landfall region (Figs. 5e–h). Moreover, the divergent field displayed a gradual increase of stronger and wider upper-level divergence and reached its maximum around the typhoon center during the ET (Figs. 5i–l).

Fig. 5
figure 5

a–d Composite analyses of winds (m s−1 in arrows) and temperature (°C in shaded areas) at 600 hPa, e–h low-level convergent wind at 850 hPa, and i–l upper-level divergent wind at 300 hPa at different times before and after the ET of T1821. Black circles indicate the 500 km radius from the typhoon center. Time \(T=0\) here means the onset of ET

Full size image

Figure 6 illustrates the evolution of the frontal thermal structure, low-level convergence, and upper-level divergence in the vicinity of T1915’s landfall. At \(T=-48\) h, the thermal front was weak; it slightly deepened by \(T=-24\) h (Fig. 6a, b). However, a much more pronounced thermal front developed before and after the storm’s ET at \(T=0\) h (Fig. 6c‒d). Low-level convergence and upper-level divergence formed over the landfall areas before the ET. However, the low-level convergent and upper-level divergent winds near the typhoon center were relatively weak at \(T=-48\) h and \(T=-24\) h compared with those near the center of T1821. During the ET, areas of strong convergence and divergence were observed only in limited regions, suggesting that the low-level flow and rising motion of air around the landfall area could not develop over wider regions.

Fig. 6
figure 6

Same as Fig. 5, but for T1915

Full size image

Figure 7 shows the synoptic environments that occurred during development of the ET of T1919. Before the ET process, T1919 displayed a frontal thermal structure. At \(T=-48\) h, deep thermal fronts were present; however, the tendency of the thermal fronts decreased at \(T=-24\) h and during the ET process, after which they became deeper (Fig. 7a‒d). Compared with the other two typhoons, T1919 exhibited more prominent low-level convergence and upper-level divergence over wider regions around the typhoon center before undergoing ET development.

Fig. 7
figure 7

Same as Fig. 5, but for T1919

Full size image

To further understand the synoptic environments of these three typhoons during the ET processes, we extended the analysis to include the vertical wind shear between 850 and 500 hPa and the total column water vapor available for moisture supply. Vertical wind shear, in this context, refers to the change in wind speed and/or direction with height in the atmosphere. To calculate the vertical wind shear, we subtracted the wind vector at the lower level (850 hPa) from that at the higher level (500 hPa). The selection of the 500 hPa altitude for calculating the wind shear was based on meteorological conventions and practices. The 500 hPa level is commonly used as a standard reference level in meteorology because it is a good indicator of the mid-levels of the atmosphere and is situated at approximately one-half the atmospheric pressure and one-half the altitude of the Earth’s surface. By contrast, for calculating upper-level divergent winds, we used 300 hPa because it is located higher in the atmosphere and is typically above the jet stream. The jet-stream level is typically between 250 and 300 hPa, and this level is known for its strong winds and high variability in wind speed and direction. By selecting the 300-hPa level for calculating upper-level divergent winds, we can better capture the strong winds and variability that occur in the upper atmosphere.

Figure 8 shows the vertical wind shear and precipitable water for T1821. At \(T=-48\) h, the conditions of wind shear were favorable for rapid intensification, with values of 2 m s−1 and lower near the eye region. These favorable conditions became much wider at \(T=-24\) h. However, at \(T=0\) h, the wind-shear values near the eye region increased to ≥ 10 m s−1, indicating a change from intensification to dissipation. The precipitable water showed a similar tendency, with substantial amounts observed 24 h before landfall, flowing toward coastal regions, indicating that the atmosphere had sufficient moisture to sustain the typhoon’s intensity.

Fig. 8
figure 8

a–d Vertical wind shear between 850 and 500 hPa and e–h precipitable water at different times before and after the ET of T1821. Black circles indicate the 500 km radius from the typhoon center. Time \(T=0\) here corresponds to the onset of ET

Full size image

Figure 9 illustrates the vertical wind shear and precipitable water during the ET development of T1915. From \(T=-48\) h until the onset of the ET, T1915 experienced lower wind-shear values, indicating rapid intensification. The amount of precipitable water gradually increased from \(T=-48\) h and reached a maximum at \(T=0\) h. Notably, the amount of precipitable water remained relatively constant at \(T=+12\) h compared with that at \(T=0\) h, indicating spatial redistribution of atmospheric moisture.

Fig. 9
figure 9

Same as Fig. 8, but for T1915

Full size image

Figure 10 shows the vertical wind shear and precipitable water vapor during the ET process of T1919. At \(T=-48\) h and \(T=-24\) h, the area around the eye region exhibited lower wind-shear values, indicating rapid intensification. However, the wind-shear values increased to ≥ 10 m s−1 around the eye region at \(T=0\) h and after, indicating a transition from intensification to dissipation after the ET. Prior to the onset of the ET, higher values of precipitable water were observed at \(T=-48\) h and \(T=-24\) h compared with the corresponding values for the other two typhoons, suggesting a potential relationship between moisture content and ET development.

Fig. 10
figure 10

Same as Fig. 8, but for T1919

Full size image

When we compared our results with those of similar studies, we found that Quinting et al. (2014) also reported the strongest upper-level divergence and low-level convergences with a thermal trough in the case of Typhoon Sinlaku (2008), which underwent deep convection. Park and Lee (2007) reported similar results for Typhoon Rusa (2002). Furthermore, several other studies have documented a strong vertical wind shear before a typhoon undergoes an ET (Ritchie and Elsberry 2001; Paterson et al. 2005). Evans et al. (2017) described several theories associated with ET development. Another reason for the intensification of typhoons before transforming into extratropical systems might be the change in eyewall distribution over time. The development of a new eyewall around an old eyewall results in a temporary period of weakening around the latter. After this stage, when the tangential wind around the new eyewall exceeds that of the old eyewall, the typhoon re-intensifies (Sitkowski et al. 2011; Fischer et al. 2020).

Our comparative analysis showed that T1919 was linked to deep thermal fronts, high levels of atmospheric water vapor, and favorable atmospheric conditions that led to strong convergence in the days before and after landfall. By contrast, T1821 was associated with favorable environmental conditions only a few hours before landfall. The atmospheric conditions necessary to enhance strong convergence during the ET of T1915 were not as favorable as those of the other two typhoons. Compared with the cyclone phase-space parameters of the other two typhoons, those of T1919 also indicate a relatively stronger frontal warm core after it underwent an ET. T1915 had lower B values after the phase transition, indicating a decreased tendency to maintain frontal characteristics. The frontal characteristics of T1821 showed B values higher than those of T1915 but smaller than those of T1919. The magnitudes of the other two phase-space parameters (− VTL and − VTU) followed the same tendency as the B values. In addition, the maximum wind intensities followed the order T1915 < T1821 < T1919. Hence, the phase-space parameters provide a reasonable explanation for the changes in intensity during the movements of strong typhoons and play a major role in the intensification of their activities in terms of strong winds.

Our findings confirm that the synoptic environments control the typhoon inner-core structures, which determine the strengths of the typhoons during the onset of their ETs. The stronger typhoons (T1821 and T1919) have asymmetric structures with either deep warm or deep cold cores, whereas the relatively weaker T1915 has a symmetric deep warm-core structure. Notably, however, a high-resolution weather model could potentially provide more accurate information about the structural changes of typhoons. The aforementioned results were obtained using JRA-55 at a resolution of 1.25°, which is considered coarse. Nevertheless, studies based on reanalysis have demonstrated robust results in capturing typhoon tracks and the phase-space parameters (Zarzycki et al. 2017; Hodges and Cobb 2017). Despite a deeper understanding of extratropical development using the phase-space parameters, predicting the structural changes of typhoons using weather models remains a challenge, as highlighted by several studies (e.g., Jones et al. 2003; Evans et al. 2017).

3.4 What factors differentiate these typhoons from becoming extratropical?

A TC typically experiences ET when it interacts with mid-latitude weather systems such as cold fronts or jet streams. These interactions cause changes in cyclone structure, dynamics, and intensity, leading to a transformation from a symmetric warm core system to an asymmetric cold core system resembling mid-latitude TCs. There are various factors that contribute to the ET process of each typhoon.

Although most typhoons follow a typical TC path, T1821 experienced an ET as it approached the Japanese archipelago, making it different from normal typhoons in several ways. The interaction of T1821 with a mid-latitude trough or a low-pressure system associated with a cold front (Fig. 5g) might have weakened the storm’s core as a result of the convection separating from the low-level circulation. The mid-latitude trough might have carried cold, and dry air from the north that mixed with the warm, and moist air of the typhoon’s core, resulting a decrease in the sea surface temperature (Fig. 11) and possibly greater atmospheric stability, which further weakened the storm. When T1821 interacted with the mid-latitude trough, it encountered strong vertical wind shear, resulting the convection to shift eastward from the low-level circulation (Fig. 8c). This wind shear further weakened the core of the storm.

Fig. 11
figure 11

Time evolution of surface brightness temperature (in K) at 3 h intervals extracted from JRA55 dataset

Full size image

During the ET, T1915 underwent substantial changes in both structure and intensity. The storm increased in size and became more asymmetric, with its strongest winds shifting to the northeast of the center (Fig. 6c). In addition, T1915 underwent a rapid intensification phase immediately before making landfall over Japan, which is unusual for a TC in transition (Miyamoto et al. 2022). Several factors contributed to T1915’s ET process, including the interaction between the typhoon and mid-latitude weather systems. The strong upper-level winds associated with these systems might have contributed to shearing of the storm and disruption of its circulation, diminishing its tropical characteristics (Fig. 6k). The relatively low sea surface temperatures over the western Pacific might have also contributed to the decrease in the storm’s intensity and eventual transition to an extratropical cyclone (Fig. 11).

During the ET of T1919, the storm interacted with a powerful mid-latitude weather system, resulting in it becoming more elongated and asymmetric (Fig. 7c). Unlike typical typhoons, T1919 had an irregular shape rather than a circular and symmetric one. Moreover, T1919 underwent a rapid intensification phase immediately before transitioning into an extratropical cyclone, which is not typical for TCs during the transition process. This intensification was attributed to the storm’s interaction with a nearby atmospheric disturbance (Fig. 7a, b). T1919’s path over the ocean was also a factor in its ET process. Unlike most typhoons, which move from east to west across the Pacific Ocean, T1919 moved northward towards Japan before turning eastward, allowing it to interact with a larger and more powerful mid-latitude weather system.

Notably, T1821, T1915, and T1919 were atypical typhoons because of their ETs and the unique factors that contributed to their characteristics. However, ET is a natural process that occurs in some tropical cyclones, and understanding the factors that influence ET can help improve our ability to predict the behavior and impact of these storms.

4 Conclusion and recommendations

We studied the structure and synoptic environments of three typhoons (T1821, T1915, and T1919) that were exceptionally strong and intense throughout their lifecycles. Our aim was to gain a general understanding of the ET processes and the associated intensity changes, which are largely influenced by the inner-core structures of typhoons and the synoptic-scale environments that govern the changes. Our investigation revealed that, during the onset of the ET, stronger typhoons tend to have an asymmetric nature with either a deep warm-core or deep cold-core structure. Our analysis of 26 typhoons that approached Japan in 2018–2019 also revealed that most underwent phase changes during their lifecycles. Prior to its ET, T1821 had a deep warm-core structure, T1915 had a moderate warm-core structure, and T1919 exhibited shallow to moderate warm-core structures. We also examined the temporal evolution of the central pressure and wind speed during the lifecycles of these three typhoons and found that all of them showed strong intensification after a structural change from a cold to warm core. Further analysis of the synoptic environments during ET development indicated that, compared with the other two typhoons, T1919 was associated with deep thermal fronts and favorable atmospheric conditions for strong convergence. Overall, our analysis suggests that structural changes in strong typhoons play an important role in the intensification of strong winds.