Decadal variability of tropical cyclogenesis and decay in the southwest Pacific

Abstract

Tropical cyclones (TCs) represent a significant hazard to the southwest Pacific (SWP) region, impacting on properties, lives and infrastructure, accounting for ~76% of reported disasters within the region. A particular challenge that island nations face is the inherent degree of variability in TC risk from season to season, which hampers adaptation planning due to the difficulty in predicting how many TCs will form and the subsequent areas of impact. Therefore, this study aims to assess the historical variability of TC genesis and decay (i.e., the start and end of a TC life cycle) for the SWP over the last 70 years. This is achieved through a spatio‐temporal analysis of TC Best Track Data from South Pacific Enhanced Archive of Tropical Cyclones (SPEArTC) dataset from 1948 to 2017 for TCs that originate within 0°–35°S and 135°E–120°W. Results highlight substantial decadal variability in dominant regions of genesis and decay over the last seven decades. In particular, we observe a statistically significant displacement of TC genesis (decay) by 898 km (909 km) in an east‐northeast (east‐southeast) direction from 1948 to 2017. We also demonstrate that TCs undergoing extratropical transition have significantly decreased in terms of both frequency and duration during the last four decades, however, the length of the extratropical cyclone tracks has increased. We attribute these observations to conducive environmental conditions, including warmer sea surface temperatures in the eastern region of TC genesis, sufficient moisture content (relative humidity) and favourable vertical wind shear. Further, the observed ocean warming and decreased wind shear during recent decades across the south of the SWP region favour the east‐southeasterly extension of extratropical cyclones. The findings of this study increase our understanding of decadal to multidecadal TC risk for SWP island nations and may assist in improving seasonal TC outlooks.

Variability of southwest Pacific tropical cyclone track geometry over the last 70 years

“…Consistent with Sharma et al . (2020), this study analyses TC tracks from 1948 to 2017, available at 6‐hour intervals for the whole TC lifetime where the first (last) point of each unique track in the SPEArTC data set is considered as the point of TC genesis (TC decay). While it is acknowledged that this dataset may have some TC data missing (Sharma et al ., 2020) along with an error of ±12 hours for reasons outlined in Diamond et al .…”

“…The cluster analysis was applied to TC tracks from SPEArTC dataset from 1948 to 2017. The analysis was also repeated for the shorter 1970–2017 period to account for any potential impacts of changing observational technologies (see Sharma et al ., 2020).…”

“…This study uses the TC Best Track Data from SPEArTC (Diamond et al, 2012) for the SWP region ( Figure 1). Consistent with Sharma et al (2020), this study analyses TC tracks from 1948 to 2017, available at 6-hour intervals for the whole TC lifetime where the first (last) point of each unique track in the SPEArTC data set is considered as the point of TC genesis (TC decay). While it is acknowledged that this dataset may have some TC data missing (Sharma et al, 2020) along with an error of ±12 hours for reasons outlined in Diamond et al (2012), the authors in the latter study have also emphasised this as not a significant issue when constructing TC track morphologies.…”

“…Consistent with Sharma et al (2020), this study analyses TC tracks from 1948 to 2017, available at 6-hour intervals for the whole TC lifetime where the first (last) point of each unique track in the SPEArTC data set is considered as the point of TC genesis (TC decay). While it is acknowledged that this dataset may have some TC data missing (Sharma et al, 2020) along with an error of ±12 hours for reasons outlined in Diamond et al (2012), the authors in the latter study have also emphasised this as not a significant issue when constructing TC track morphologies. Furthermore, only TCs that occur within the SWP TC season, i.e., from November to April (862 TC events) and those that originate in the SWP region but subsequently move outside the SWP domain are included in this analysis.…”

“…This was mainly due to the issues surrounding the reliability of TC data from the pre-satellite era, i.e., before 1970 (Holland, 1981). However, the emergence of the TC Best Track Data from South Pacific Enhanced Archive of Tropical Cyclones (SPEArTC) dataset (Diamond et al, 2012), and findings of Sharma et al (2020), suggest that TC track data may be analysed from 1948-onwards (with some caution), which will improve our climatological understanding of SWP TC tracks. Analysing track characteristics over an extended time period will improve the robustness of the probabilistic track cluster due to increased sample size and may potentially identify additional trajectories and provide an appropriate baseline for temporal trend analysis.…”