Seasonal advance of intense tropical cyclones in a warming climate

Seasonal advance of intense tropical cyclones in a warming climate

Intense tropical cyclones, which often peak in autumn, have destructive impacts on life and property, making it crucial to determine whether any changes in intense tropical cyclones are likely to occur. Here, we identify a significant seasonal advance of intense tropical cyclones since the nineteen eighties in most tropical oceans, with earlier-shifting rates of three point seven and three point two days per decade for the Northern and Southern Hemispheres, respectively. This seasonal advance of intense tropical cyclones is closely related to the seasonal advance of rapid intensification events, favoured by the observed earlier onset of favourable oceanic conditions. Using simulations from multiple global climate models, large ensembles and individual forcing experiments, the earlier onset of favourable oceanic conditions is detectable and primarily driven by greenhouse gas forcing. The seasonal advance of intense tropical cyclones will increase the likelihood of intersecting with other extreme rainfall events, which usually peak in summer, thereby leading to disproportionate impacts.

Tropical cyclones are one of the most devastating natural disasters in the world. Understanding potential changes in tropical cyclone activity in response to global warming is important for tropical cyclone-related disaster prevention. Global warming caused by human activities is estimated to be about one degree Celsius above pre-industrial levels, with most of the warming occurring since the mid twentieth century. This warming may have already impacted tropical cyclone activity on the global scale. Although changes in the annual number of global tropical cyclones are controversial, an increasing trend in the number of global intense tropical cyclones has been noted, and significant efforts have been devoted to reducing uncertainties in data quality issues. Poleward migration of the locations of tropical cyclone activity was observed in most ocean basins in recent decades and is particularly robust in the western North Pacific basin, which is related to changes in tropical cyclone seasonality. A recent study identified an earlier onset of the beginning time of tropical cyclone season, in terms of the initial formation dates, in the North Atlantic basin associated with a warming ocean in spring, but there is no seasonal change in the date that Atlantic accumulated cyclone energy reaches its ten, fifty or ninety percent threshold. Overall, detection and attribution of changes in tropical cyclone activity are among the top priorities of tropical cyclone research.

Intense tropical cyclones, tropical cyclones with a lifetime maximum wind speed greater than one hundred ten knots, that pose the greatest threat require special attention. Most regions lack resilience to intense tropical cyclones, which could cause many deaths and damage to property through destructive winds, storm surges, heavy precipitation and inland flooding. For these reasons, it is crucial to understand changes in intense tropical cyclone characteristics. Although it is still challenging for high-resolution climate models to capture intense tropical cyclones, statistical analysis suggests that intense tropical cyclones are likely to be more susceptible to anthropogenic warming compared with other tropical cyclone categories. Changes in many characteristics of intense tropical cyclones under a warming climate, for example, the number, intensity,

and lifespan, are fairly well studied; however, little is known about changes in the seasonal cycle of intense tropical cyclones.

In general, intense tropical cyclones occur more frequently in autumn than in summer because they require sufficient heat from the ocean to develop. The seasonal cycle of intense tropical cyclones lags behind that of other high-impact weather events, for example, extreme rainfall events produced by the summer monsoon system, which often peak in summer and are largely determined by the seasonal cycle of atmospheric energy driven by solar radiation. Recently, the compound hazards of intense tropical cyclones and other high-impact weather events have attracted increasing attention. The devastating impact of the compound hazards is well beyond any one of these events individually. They can induce substantial inland rain and flooding associated with multiple weather systems, causing large-scale failures of power and transportation systems, straining emergency responses and depleting disaster preparation resources. As the preferred time of occurrence of intense tropical cyclones and other high-impact weather events is usually off by one season, the likelihood of their simultaneous occurrence is generally assumed to be small; however, given the seasonal advance of intense tropical cyclones, as shown in this study, its potential change should be considered.

Seasonal advance of intense tropical cyclones

We begin with a broad view of the change in the seasonal cycle of intense tropical cyclone occurrence. The occurrence time of an intense tropical cyclone is defined by the date on which the tropical cyclone first achieves its lifetime maximum intensity. Tropical cyclone data during nineteen eighty-one to twenty seventeen are taken from the advanced Dvorak Technique-Hurricane Satellite dataset, tropical cyclone observations. Figure one a, b shows the seasonal distribution of intense tropical cyclone numbers, solid line, and the linear trend, bar, for each month in two hemispheres. There is an obvious seasonal advance of intense tropical cyclones in both the Northern Hemisphere and the Southern Hemisphere that manifests as increased intense tropical cyclone occurrence in the early season and decreased intense tropical cyclone occurrence in the late season, occurrence time.

To depict the spatial pattern of the seasonal advance of intense tropical cyclones, we estimate the linear trend of the interseasonal difference, early season minus late season, in the number of intense tropical cyclones, Figure one c. The increasing trends, that is, shifts towards earlier onset, are apparent over the western North Pacific basin, east of one hundred thirty degrees west in the eastern North Pacific, over the Gulf of Mexico and the western part of the North Atlantic, west of eighty degrees east of the South Indian, near the northern coast of Australia and east of one hundred eighty degrees of the South Pacific basin. Note that the significant earlier-shift trend of intense tropical cyclone occurrence in the North Atlantic basin features large earlier shifts in the Gulf of Mexico and western part of the North Atlantic basin and nearly no change in the eastern part, which is associated with changes in oceanic conditions and the rarity of intense tropical cyclone occurrence in the eastern part, Extended Data Figure one. The decreasing trends, that is, shifts towards later onset, are observed over the central Pacific and near the northeastern coast of Australia. The consistently earlier-shifting trends of intense tropical cyclones over most of the tropical oceans indicate the robustness of this phenomenon.

The time series of the median value of intense tropical cyclone occurrence time during the active season in two hemispheres based on the ADT-HURSAT

from the ADT-HURSAT dataset. The early season is defined as June to August in the Northern Hemisphere and December to February in the Southern Hemisphere. The late season is defined as September to November in the Northern Hemisphere and March to April in the Southern Hemisphere. The black crosses indicate significance at the ninety-five percent confidence level. The annual number of intense tropical cyclones is calculated over five degrees by five degrees boxes prior to computing trends. The basemap in c was plotted using the Matplotlib basemap toolkit with the geographical coordinate system World Geodetic system nineteen eighty-four generated by the Global Positioning System.

dataset are shown in Figure two A, D. In both the NH and the SH, the occurrence time of intense TCs shows a significant trend towards earlier onset at the ninety-five percent confidence level based on a non-parametric statistical test, with rate of three point seven and three point two days per decade, respectively. The earlier-shifting trend of intense TC occurrence can also be identified in five major ocean basins, and it is significant at the ninety-five percent confidence level for each basin except the eastern North Pacific. It is most evident in the western North Pacific basin, with a seasonal advance of eight point one days per decade, which may have profound impacts on the surrounding regions. This earlier-shifting trend does not depend on how it is estimated or which dataset is used.

To confirm whether the seasonal advance only occurs in the intense TCs, we quantify the shifting rates of TCs with different intensities. Using the ADT-HURSAT dataset and best-track dataset, the results consistently show a significant earlier-shifting trend for intense TCs in two hemispheres based on a non-parametric statistical test but not for less intense TC events. Note that the earlier-shifting trend in the seasonal occurrence of TCs with intensity larger than one hundred ten knots in the NH obtained from the ADT-HURSAT dataset is statistically significant on the basis of two different statistical tests, whereas the earlier-shifting trend in the SH is statistically significant based on a non-parametric statistical test but with large ninety-five percent confidence intervals. The earlier-shifting trends in the NH and the SH obtained from the best-track dataset are statistically significant based on different statistical tests. These results suggest that the seasonal advance is more evident for more intense TCs. In contrast, no significant change is observed in either hemisphere if all TCs are considered. This contrast between intense TCs and less intense ones naturally draws our attention to the essential differences between them.