Interest in understanding how and why tropical cyclone (TC) activity in various ocean basins is modulated by climate variability has grown substantially over the last four decades. This interest stems from the fact that the TC is one of the most destructive natural catastrophes and causes loss of lives and enormous property damage on a global scale. The large-scale circulation patterns conducive to TC activity during an extreme climate mode differ profoundly from those of an opposite climate mode. In this book, climate modes encompass a suite of time scales, ranging from the shortest ones within a season (90 days) such as the quasi-biweekly oscillation (QBW) and the Madden–Julian Oscillation (MJO), to the interannual variability such as the El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation, and Atlantic and Pacific Meridional Mode, to the multidecadal variability such as the Pacific Decadal Oscillation and the Atlantic Multidecadal Oscillation. TC activity includes formation location, frequency of storm occurrence, life span, tracks, landfall rate, and/or storm intensity.
The ocean basins considered here cover the western North Pacific (WNP) and the South China Sea, eastern and central North Pacific, South Pacific, and North Atlantic. Because reliable TC records for the Indian Ocean are relatively short, TC activity in the North and South Indian Oceans are excluded. This book is organized as follows: description of the intraseasonal oscillation in Chapter 2; interannual to interdecadal variability in Chapter 3; modulation of TC activity in each ocean basin in Chapter 4; discussion on the subseasonal to seasonal TC prediction in Chapter 5; typhoon rainfall variations under changing climate in Chapter 6; followed by Chapter 7 for future TC projections.
This book begins with the subject of intraseasonal oscillation (ISO), a time scale that is longer than 10 days but shorter than 90 days (Chapter 2). For the ISO, two distinct modes stand out clearly: the Madden–Julian Oscillation (MJO) and quasi-biweekly oscillation. The MJO phenomenon generally spans a time window of 30–60 days while the latter has a typical 14-day time scale. The MJO is fascinating and characterized by a planetary zonal wavenumber one structure in the global tropics. It exhibits a baroclinic vertical structure in winds and pronounced seasonality in its propagation direction. In the boreal winter, the MJO convection propagates eastward fast from the Indian Ocean to the equatorial central Pacific. Associated with the MJO deep convection are fluctuations in surface pressure and zonal winds at both lower and upper troposphere. In the boreal summer, the MJO system takes a different route, tracking northward in the Indian monsoon region and northwestward over the western North Pacific and South China Sea. Major aspects of the MJO to be discussed are the initiation process of the convection, periodicity, mechanisms for eastward propagation, role of the Maritime Continent in perturbing the eastward propagation, and mechanisms for the northward propagation in the boreal summer. We will also discuss three major theories to account for many aspects of the MJO. They include the coupled Kelvin–Rossby wave theory, moisture-mode theory, and unified dynamic moisture-mode theory. The recent moisture-mode theory that emphasizes the asymmetry of the column-integrated moist static energy tendency and related processes that are instrumental to the propagation and intensification of MJO convection will be discussed.
In Chapter 3, several climate feedback mechanisms involving the tropical ocean and atmosphere system are first introduced. Such feedbacks include the wind-evaporation process, Bjerknes hypothesis, footprinting mechanism, sea surface temperature–cloud feedback, and sea surface temperature–water vapor feedback. This is followed by the El Niño phenomenon, which is the leading mode of tropical climate variability with a recurring period of three to seven years. El Niño is manifested by the anomalous warming of the eastern and central tropical Pacific. La Niña is the opposite of El Niño and refers to anomalous cooling and steady and strong easterly trade winds in the tropical Pacific. Coupled to the oceanic El Niño and La Niña events is the atmospheric Southern Oscillation, a large-scale pressure seesaw between the tropical eastern and western Pacific. Taken together, the term El Niño–Southern Oscillation (ENSO) describes the coupled atmosphere–ocean interaction in the vast tropical Pacific basin. Two prevailing theories that provide foundation for understanding ENSO dynamics are presented. They are the delayed oscillation theory and recharge–discharge mechanism. Studies over the last 15 years show that there are at least two types of El Niño: Eastern Pacific and Central Pacific events, which modulate regional TC activity in a different manner. ENSO forecasts and a new approach called the Bayesian model averaging are followed. Other climate modes on the interannual time scale of interest include: North Atlantic Oscillation (NAO); Pacific Meridional Mode (PMM); and Atlantic Meridional Mode (AMM). On the decadal to interdecadal time scale well-known climate modes include the Pacific Decadal Oscillation (PDO) and the Atlantic Multidecadal Oscillation (AMO). The Aleutian low, atmospheric teleconnections from the tropics, and midlatitude ocean dynamics appear to play key roles in driving the PDO variability. The AMO variability is attributed to oceanic meridional overturning circulation in the Atlantic, stochastic forcing from the atmosphere, cloud feedback, and aerosol effects.
Chapter 4 discusses how TC activity in various ocean basins is influenced by climate variability of the ocean–atmosphere system. Indeed, marked ISO, interannual and multidecadal variations are intrinsic to tropical climate and they exert a pronounced impact on TC changes. For example, the 2020 Atlantic hurricane season was extremely active and set numerous records for overall activity. It featured a record 30 named storms, of which 13 developed into hurricanes and 6 further intensified into major hurricanes. Twelve out of these 30 storms made landfall in the contiguous United States, in stark contrast to the mean rate of less than two landfalls per year during 1900–2000. Besides these staggering numbers, accumulated cyclone energy (ACE), which measures the strength and duration of tropical storms and hurricanes, was 75% above the long-term mean of 1981–2010. In 2020, La Niña, vigorous African easterly waves, a positive AMM state, and abnormally warm sea surface temperatures over the North Atlantic main development region all fueled the extremely active North Atlantic hurricane season. Also noteworthy is the consecutive above average Atlantic hurricane seasons since 2016. Whether this above normal activity over the last 5 years will continue in the next 5–10 years is of a great scientific and socio-economic interest.
In Chapter 4, background climatology for each basin is first introduced. Large-scale flow patterns and equatorial waves, which are regarded as TC precursors, are then followed. For the WNP, eastern North Pacific, and North Atlantic, the tropical depression-type disturbances appear to be the most common equatorial waves, trailed by the equatorial Rossby waves. For the South Pacific, it is just the opposite. Regional TC changes in each ocean basin as influenced by various climate modes are the major themes of this chapter. Central to TC changes are the genesis frequency, location of formation, tracks, and landfalls. Recent studies show that regional TC activity is modulated by a combination of two or three climate modes, such as the MJO and ENSO. Modulation of the Eastern Pacific El Niño, Central Pacific El Niño, PMM, and AMM on TC activity over the Pacific and Atlantic Oceans from recent studies is described. Subsequently, the attention is focused on decadal and interdecadal TC variability. The chapter concludes with a section on the observed variations in TC attributes such as frequency of occurrence, translation speed, intensity, and meridional migration of the latitude of lifetime maximum intensity based on historical or modern data.
Because of the huge socioeconomic repercussions incurred by TCs, developing a sound and modern method for predicting TC activity from weeks, months, or seasons in advance is becoming increasingly important. Better forecasts, advance warnings, and proper emergency management efforts would all lead to a reduction of loss of life and property damage. Chapter 5 provides a review of the current status regarding subseasonal to seasonal TC prediction and the corresponding methods used therein by various researchers and government agencies. Broadly speaking, prediction methods can be classified into three approaches: purely statistical; dynamical; and statistical-dynamical. Statistical methods utilize logistic regression, least absolute deviation regression, Poisson regression, and/or multiple linear regression techniques. A Poisson generalized regression model cast in the Bayesian framework is also applied to probabilistic forecast TC activity. Dynamical forecast methods rely on dynamical climate models. Dynamical seasonal TC forecasts were regularly issued by many organizations throughout the world in the last 20 years using either atmospheric general circulation models forced by sea surface temperature anomalies or by coupled atmosphere–ocean models. Many models were constantly upgraded with increasing higher horizontal resolution, improved model physics, and ensemble forecasting techniques. Statistical-dynamical methods refer to the use of forecast information (e.g., predictors) from dynamical models in predicting TC metrics using statistical methods. This hybrid approach takes advantage of the future, yet-to-be-observed values from dynamical models and statistical relationships between TC metrics and environmental parameters to forecast future TC changes. For the subseasonal to seasonal TC, an international effort was initiated in the last few years to develop the subseasonal to seasonal (S2S) prediction data set. Specifically, the S2S dataset contains dynamical subseasonal forecasts and reforecasts with leads up to 60 days.
Heavy rainfall and flooding resulting from TCs result in devastating consequences on society, human and animal life, and economics. Besides natural variability, anthropogenic climate change elevates the water vapor capacity in the atmosphere. A warmer atmosphere can hold more moisture in the air – about 7% more per 1°C of warming. With increasing moisture under global warming, TCs are expected to produce heavier rainfall, a notion consistent with climate models that show a projected increase in TC precipitation rate. Chapter 6 describes how TC rainfall may have changed over the last several decades. Taiwan is used as an example because of the availability of reliable and long-term hourly and daily rainfall records, its unique geographic location with regard to the prevailing typhoon tracks, and abundance of published studies relevant to the subject. Here, changes in rainfall frequency and intensity, storm duration and translation speed associated with typhoons and typhoon tracks are considered. This chapter concludes with a section on long-term variations in return levels during the typhoon season modulated by time and an ENSO index using a non-stationary generalized extreme value distribution.
Chapter 7 is aimed toward providing the state-of-the-art knowledge of future TC projections based on the numerical simulation from a suite of climate models. Historically, low and medium-resolution (commonly 120–300 km) global climate models have been used to simulate TC numbers in a warmer climate. Because the horizontal resolutions used in these models are not high enough to estimate intensity changes in TCs, a “dynamical downscaling” or “statistical-dynamical downscaling” method is used to yield finer details of TCs with a regional or hurricane model. Independently, a high-resolution (~20 km) global atmospheric model is used to improve fidelity of TC simulations. Modeling results commonly indicate a global decrease in TC numbers in the future. A few hypotheses relevant to the projected decrease in TC frequency are discussed including: weakening of tropical overturning circulation; increasing of entropy deficit; and increase in ventilation index. More recently, climate models with higher resolution also demonstrated an increase in storm numbers in the future, a stark contrast to the earlier view. Hypotheses are also advanced to help explain this new result. At the end of this chapter is an updated review of the expert opinions on the potential future changes in TC in the context of the annual frequency of TCs, annual frequency of category 4 and 5 TCs, storm intensity, and mean precipitation rate induced by TCs.
