Decadal behaviors of tropical storm tracks in the North West Pacific Ocean

Highlights

• From July to October the TS tracks gradually increase their sinuosity in the NWP.

• Eastward shift in most of TS genesis positions is seen with increase in track sinuosity.

• Warm phase of ENSO is associated with a greater number of TS with higher SI values.

• Early cyclone season months dominance of sinuous TS tracks vice-versa in late months.

Abstract

Quantitative analysis on the shape of 959 tropical storm (TS) tracks in the North West Pacific (NWP) basin was carried out over the period from 1977 to 2016 by using International Best Track Archive for Climate Stewardship (IBTrACS) provided by National Oceanic and Atmospheric Administration (NOAA) based on an established track sinuosity metric. Track sinuosity is a measure of deviation of a storm track from its straight–line path between the cyclogenesis and cyclolysis locations. More sinuosity in the storms' tracks makes it rather challenging for the atmospheric models to accurately assess the storms' respective locations on the map and potentially causes higher damages due to lack of precise information about their movement. Statistical analysis was carried out on spatial and temporal trends (monthly to decadal) of the TS track shape and the obtained results were mapped based on sinuosity categories within a GIS environment. The sinuosity distribution results are normalized using a cube–root transformation function to reduce skewness and obtain sinuosity index (SI). Distinct enhancement of TS sinuosity was noticed from the months July to October (JASO). It is also detected that early months of TS season like June–August have dominance of more predictable straighter storm tracks over sinuous kind of tracks, and vice-versa in the case of late months of season like September–October. It is also evident that there is a one- to three-year cyclic pattern of changing sinuosity over the NWP basin. The 1987–1996 decade had the maximum dominance of sinuous tracks in comparison to the other three decades. Significant longitudinal eastward shift (from 110°–140° E to 130°–160° E) in majority of cyclogenesis locations is observed as sinuosity in storm track increases from straight to sinuous patterns. Similar shift for latitudinal track was not found. Finally, we investigate the sinuosity based on the warm/cold phase of the ENSO. Warm phase of ENSO is found to be associated with a greater number of TS with higher SI values in the NWP basin. These TS during warm phase mostly originate in the eastern part of the basin where a vast open area of warm sea surface temperature encourages their formation and intensification. Lastly, the study found a moderate positive relationship between SI and TS' longevity and distance coverage, which are crucial information for disaster risk assessment, mitigation and preparedness.

Introduction

There has been undoubtedly big improvement in the prediction of movement and cyclogenesis location of tropical storms (TS) and typhoons in the last few decades. The improvement in technology has decreased the average error by a large extent in forecasting models worldwide. An inter-comparison study of verification for tropical cyclones (TCs) in the North West Pacific (NWP) basin with three global models has successfully shown around 2.5–days lead–time improvement in the last quarter century (Yamaguchi et al., 2017). A similar study in the Atlantic basin also shows a reduction in track forecast error by 1.9% for TCs over the period 1970–1998 by subtracting the expected errors from the observed official errors (McAdie and Lawrence, 2000). Consequently, we have much better capability nowadays than past to investigate and predict the behavior of TS (Liou et al., 2019). Despite these facts, still millions of people around the globe remain vulnerable and suffer havoc induced by these storms every year (Terry and Etienne, 2010). Focusing to NWP ocean, for example, typhoon Haiyan in November 2013 devastated portions of Southeast Asia, particularly the Philippines, causing nearly 6300 fatalities in that country alone and damage estimated over USD$2.2 billion. Such similar tragic events happen worldwide every year, highlighting the importance of understanding TS' behavior more precisely for human preparedness and adaptation to these natural hazards (Nguyen et al., 2019; Nguyen and Liou, 2019a; Nguyen and Liou, 2019b). The ocean basins with more storms obviously have higher chance of causing tragic events, thus requiring more investigation.

According to the National Oceanic and Atmospheric Administration's (NOAA's) Hurricane Research Division, the NWP Ocean alone accounts for one–third of all TC activities on the planet, which makes it the most active basin in the world. Furthermore, mounting evidence from both climate models and trends analysis based on global data indicates the increasing strength and longevity of TCs for future (Knutson et al., 2008; Emanuel, 2005; Webster et al., 2005). A similar prediction of stronger TS in future due to global warming impact is also estimated by the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report in 2007. Recently, the year 2015 had 9 super typhoons (STYs), the second maximum number of STYs in the recorded history just after a tie of 11 STYs for years 1965 and 1997. In addition, years 2013–2016 have 7 STYs on average per year in comparison to years 2001–2004, 2005–2008 and 2009–2012 with 5.5, 4.25, and 3.5 STYs on average per year, respectively. Note that a storm is called a STY if its maximum (01-min mean) sustained wind ≥113 knots. Recent data supports the various predictions of increasing TS strength in the NWP region, which is an alarming motivation for continuous investigation on TS behavior in this region. In a few recent studies, the Southwest air flows in the summer season (April–September) and cold air masses from the North in the winter season (October–March) were identified as vital factors in converting a typhoon into a STY in the NWP basin (Lee et al., 2017; Liou et al., 2018).

Traditionally, most of cyclone researches aim on finding the distribution, intensity and frequency of TCs in the numerous ocean basins (Knutson et al., 2008; Emanuel, 2005; Webster et al., 2005). It is because these three aspects directly provide the idea of risk of the hazards associated with TCs. Yet, focus has not been put on examining the nature and patterns in tracks of TCs. However, having said that, attention of researchers for investigating the nature and patterns of TC tracks is increasing in the recent decade (Liou et al., 2018; Kelly et al., 2018; Gienko and Terry, 2012; Zhang et al., 2013). A track is the twisting path followed by any cyclone, with its investigation lucrative in many ways. The trajectory of any storm decides which place or coastline it can possibly affect by making landfall (Wang et al., 2011; Nakazawa and Rajendran, 2007). Also, the track's recurvature point of a storm sometimes can be associated with its maximum intensity (Knaff, 2009). A changing pattern in tracks of storms in any ocean basin may be correlated with global warming (Wu and Wang, 2004). Moreover, storms that tend to survive longer can cause more damages and rather difficult to predict by models due to their shifting track directions, and thus need more scientific attention. For example, typhoon Parma in 2009 took many turns causing three times landfall over the Luzan Island of Philippines. As a result, it survived around 14 days causing big flooding over a large area and nearly 500 fatalities in that island alone with an estimated economical loss up to USD$ 581 million. It was later found in investigation that typhoon Parma interacted with another typhoon Melor, which made Parma more devastating and unpredictable in nature. Interactions between typhoons are more common in the NWP basin than the other ocean basins as it possesses a greater number of typhoons every typhoon season. Thus, an improved understanding of interactions between tropical depressions and typhoons is important and of interest by the community (Liu et al., 2015). Later, a generalized empirical formula for quantifying threshold distance required between two typhoons for possible interaction between them was developed and tested in the NWP basin (Liou et al., 2016). Note that a storm ranges from a tropical disturbance (TD) to a STY, whereas a TC/typhoon is also one of storm categories with maximum (10-min mean) sustained wind ≥64 knots. Terry and Gienko provided a new track sinuosity metric to quantify the tracks of storms in terms of sinuous nature (Terry and Gienko, 2011). The present work utilizes this metric to measure cyclone track sinuosity. Also, it has been established that TS with more sinuous tracks have more tendency to survive longer and thus cause more damage (Terry and Gienko, 2011, Terry and Gienko, 2018; Terry and Kim, 2015; Terry et al., 2013). Hence, enhancing our current understanding about nature and changing patterns of storm tracks in numerous ocean basins is crucial for advancement of science along with human preparedness to these natural hazards.

The concept of track sinuosity originates from the stream channel sinuosity of fluvial geomorphology. Stream channel sinuosity is a fundamental geometric feature, which is mainly associated with stream flow behavior. In this paper, sinuosity is the measure of deviation of a storm track from its straight–line path between the cyclogenesis and cyclolysis locations. How sinuosity is explained in detail in section 4.2. In general, the NWP basin possesses three kinds of patterns in the tracks of storms. First, the tracks formed by storms moving westwards along with a little northward movement towards South China Sea. Second, the tracks formed by storms moving northeastwards almost parallel to Japan. Third, the tracks formed by storms earlier moving westwards with a little northwards movement then later shifts towards northeast again parallel to Japan taking a recurvature. Aside from these general patterns of tracks, every year many storms create atypical track shapes with sharp bending and even sometimes creating recurvature in clockwise and anti–clockwise directions. It makes them tougher to predict by models, potentially causing more damages to people due to lack of precise information about their movement. Within this context, investigating whether track sinuosity possesses any spatial or temporal variability across the NWP basin offers the enhancement of our current knowledge of storm migratory behavior. Detailed analysis is discussed in section 5.3 and section 5.4.