FIFTH INTERNATIONAL WORKSHOP on TROPICAL CYCLONES
Topic 4 TROPICAL CYCLONE FORMATION AND EXTRATROPICAL TRANSITION
Topic Chair: Dr. Patrick A. Harr
Naval Postgraduate School
589 Dyer Rd., Room 254
Monterey CA 93943-5114, USA
In this section, an overview of the topic on tropical cyclone formation and extratropical transition is given. Initially, issues associated with the definition (Topic 4.1), observation (Topic 4.2), environmental influences (Topic 4.3), and numerical prediction (Topic 4.4) of tropical cyclone formation are presented. In a more general sense, Topic 4.5 defines issues that influence seasonal prediction of tropical cyclone formation. A final topic (Topic 4.6) addresses transition of a tropical cyclone into an extratropical cyclone.
In many World Meteorological Organization (WMO) Regions that contain tropical cyclones, the process of defining the formation a tropical cyclone has evolved to address regional considerations. The discussion in Topic 4.1 identifies issues associated with the variety of practices used to define tropical cyclone formation. Related to the definition of tropical cyclone formation, is the ability to observe when the defining characteristics have occurred. In Topic 4.2, the utility of traditional and specialized observations are documented for understanding and accurate forecasting of tropical cyclone formation and extratropical transition. A common theme associated with environmental influences, numerical prediction, and seasonal prediction is the complexity associated with interactions among a variety of time and space scales. In Topic 4.3, environmental influences on tropical cyclone formation are defined based on sources that vary from intraseasonal scales to the mesoscale. In the numerical prediction topic (4.4), the multi-scale nature of tropical cyclone formation is examined in relation to a hierarchy of impacts that range from synoptic scale, mesoscale, to cloud scale (i.e., ~10-50 km). Seasonal prediction (Topic 4.5) is discussed in relation to the interactions among global-scale phenomenon. The transition of a tropical cyclone into an extratropical cyclone is often associated with the extension of tropical cyclone force winds, heavy rain, and high seas into the midlatitudes. Forecast and research issues associated with these fast-moving, intense cyclonic circulations are discussed in Topic 4.6.
Since a tropical cyclone may form and move a long distance in five days, the interest and demand for extension of tropical cyclone forecasts to five days requires increased understanding and forecast accuracy of tropical cyclone formation. Alternatively, a tropical cyclone may experience extratropical transition during the five day forecast period. Because these cyclones remain a great threat to maritime or coastal regions, the forecast should include this extratropical transition period. Recommendations for resolving issues associated with the research and forecasting of tropical cyclone formation and extratropical transition are included in each topic.
4.0.2 Formation Definitions
Topic 4.1 addresses the complexities associated with defining exactly when a tropical cyclone has formed. In a general sense, a tropical cyclone is defined as a non-frontal, synoptic-scale cyclone that originated over tropical or subtropical water and contains organized convection and a definite cyclonic wind circulation. Although several well-defined structural characteristics are associated with a tropical cyclone, the declaration of exactly when formation has occurred depends on many important operational and forecast considerations that must be resolved with a strict scientific definition.
All five WMO Regions have developed a system for identifying the formation of a tropical cyclone that best suits the needs of their area of responsibility in terms of protection of life and property. Often, scenarios arise in which an area of disturbed weather may not possess strict dynamical and/or physical characteristics common to a tropical cyclone but still contain areas of significant wind, precipitation, or high seas. Usually, there are separate mechanisms (e.g., high seas alerts) by which the general public may be warned of potentially hazardous conditions associated with these types of systems. However, instances may occur when the best course of action is to declare a tropical cyclone formation in anticipation of the attainment of the full suite of the relevant physical characteristics. Typically, these scenarios arise in association with subtropical cyclones, monsoon depressions, and monsoon gyres.
Although the definition of a tropical cyclone formation under the above circumstances may provide for optimal protection of life and property, it must be resolved with the strict scientific definition of tropical cyclone formation. In the examination of histories of tropical cyclone formation, (e.g., in studies of global climate change), formation counts based arbitrary definitions will cause inconsistencies in a data base from a scientific point of view. Therefore, to optimize scientific utility, the definition of tropical cyclone formation may need to be based on stringent physical attributes that have been traditionally associated with tropical cyclones.
Optimally, a definition of tropical cyclone formation should serve to provide protection of life and property plus be consistent in a scientific context with specific physical characteristics that contribute to an internally consistent data base of tropical cyclone activity. Therefore, it is recommended that an effort be made to arrive on such a definition of tropical cyclone formation. It is also recommended that the comprehensive definition be applicable to formation from an incipient disturbance as well as a transformation from a separate pre-existing circulation (e.g., monsoon depression) to a tropical cyclone. Furthermore, it is recommended that the definition of the transformation of a tropical disturbance to a tropical cyclone be based on an objective diagnosis of relevant physical characteristics.
Related to the above recommendation is a determination of whether a single definition would act to satisfy operational/forecast criteria for protection of life and property and scientific consistency. If it is determined that this is not possible, then it is recommended that a reasonable process be defined such that operational/forecast considerations are met followed by an appropriate post-case examination for archival of the most scientific consistent determination of formation.
4.0.2 Observations in addition to satellites
The nature of tropical cyclone formation is that it typically occurs over regions of sparse non-satellite data coverage. Furthermore, many of the critical aspects of tropical cyclone formation may occur over periods of a few hours, or distances of a few hundred kilometers, which are much shorter and smaller than conventional observation practices. Direct observations over tropical cyclone formation regions are typically limited to specialized field programs, an operational program designed to provide data coverage during tropical cyclone events, or rare events of formation near land-based observing sites. Convectional observations have provided a basis for composite studies that have been instrumental in definition of the tropical cyclone structure.
Specialized field programs have provided snapshots of formation and extratropical transition cases over several tropical cyclone basins. Since formation may be dominated by mesoscale processes, it is rare to directly observe an entire tropical cyclone formation. However, careful examinations of field program data have provided bases for case studies associated with several large-scale and mesoscale aspects of tropical cyclone formation. Direct observations from field programs have also contributed to specialized numerical studies of tropical cyclone formation by providing a basis for an accurate representation of scales in the initial conditions that are often too small to be resolved with conventional observation distributions. Also, recent field programs have concentrated on collection of specialized observations of the extreme precipitation, wind, and ocean wave characteristics associated with the extratropical transition of a tropical cyclone.
Ideally, operational programs for direct observation of tropical cyclone characteristics, which would include formation, would be of great value to increased understanding and forecasts of these phenomenon. However, the logistic and financial commitments required for these programs are large. For many years, an operational field program has been conducted by the Hurricane Research Division of the U.S. National Oceanic and Atmospheric Administrations Atlantic Oceanographic and Meteorological Laboratory in Miami, Florida. Even with assets available to this program, direct observation of tropical cyclone formation has been rare due to the tendency for formation to occur over areas that are not within the range of field program aircraft, or due to the mesoscale aspect of formation such that the critical stages in the transformation of a disturbance might occur in a short time and over small space scales, which makes it difficult to plan the required flight operations. However, new developments in aircraft and instrument technology (e.g., GPS dropwindsondes) suggest a possibility of an increase in the coverage and quality of data gathered during future observational programs.
Over recent years, new technology associated with unmanned aircraft has been advanced for tropical cyclone reconnaissance. The advantage of these platforms is their ability to remain over a targeted area for extended periods.
In association with midlatitude synoptic meteorology, new technology has emerged to increase the utility of conventional observation data to numerical weather prediction systems. This application is based on the fact that small analysis errors in regions of large initial condition sensitivity will grow rapidly and degrade the numerical forecast. Therefore, regions of initial condition sensitivity may be targeted such that conventional observations may be augmented to provide a more accurate analysis over the sensitive region. Since tropical cyclone formation occurs over data sparse regions, the technology advanced in association with numerical forecasts of midlatitude synoptic conditions may be applicable to monitor sensitive areas over tropical areas that may impact forecasts of environmental conditions that could be favorable for tropical cyclone formation.
Because of tropical cyclone formation tends to occur over data sparse regions, and the current observational system is not projected to increase in the near future, it is recommended that the utility of current observational data to numerical weather prediction be increased as it applies to forecasts of conditions favorable for tropical cyclone genesis. This includes identification of regions associated with initial condition sensitivity during periods of enhanced potential of tropical cyclone formation. This may also include coordination with operational or special field programs to identify sensitive regions for targeted observations based on their potentially increased utility to the numerical model
Due to the potential benefit of sustained periods of observation, the technology associated with unmanned aircraft for tropical cyclone reconnaissance should continue to be explored.
4.0.3 Large-scale Control and Mesoscale Influences; Intraseasonal
Large-scale influences on tropical cyclone have been identified with a set of three thermodynamic and three dynamic parameters. While the thermodynamic parameters have been linked to seasonal indicators of the potential for tropical cyclone formation, the dynamic parameters are measures of daily formation potential. In a very general sense, these necessary but not sufficient criteria summarize the large-scale influence on tropical cyclone formation in terms of the capacity of the large-scale environment to support deep convection in an area of low-level cyclonic vorticity.
Recent research and observations of tropical cyclone formation have pointed to the interactions across a variety of scales as important influences on the likelihood of tropical cyclone formation. Scale interactions occur among global-scale, intraseasonal characteristics associated with the Madden-Julian Oscillation (MJO), synoptic-scale tropical wave activity, and mesoscale convective systems (MCSs). Although many studies have identified periods of enhanced tropical convection during active phases of the MJO, it is the increased organization of convection during the active phases that may have the most significant impact on tropical cyclone formation. Increased tropical cyclone formation during active periods over the western North Pacific have been linked to the increased organization of convection that provides more disturbed regions that may transform into a tropical cyclone. The influence of the MJO on tropical cyclone formation has been identified over nearly every basin in which tropical cyclones occur. An interaction between the intraseasonal MJO and synoptic-scale tropical wave activity has been examined in terms of the contribution of the MJO to regions of preferred wave growth. The growth of synoptic-scale tropical waves then contributes to overall organization of convection, which contributes to the generation of an initial vortex that may develop into a tropical cyclone.
On the synoptic scale, convective organization may also be associated with the role of continuous convection within the regional Intertropical Convergence Zone (ITCZ) as a source of potential vorticity contributed to the breakdown of the linear-oriented ITCZ into a series of vortices. This mechanism has been identified in several tropical cyclone basins. Varying basic states associated with regional characteristics in ITCZ development have been identified as mechanisms by which synoptic-scale tropical wave activity may become modified to increase environmental conditions favorable for tropical cyclone formation. For example, equatorially-trapped waves in the central Pacific region during the Northern Hemisphere summer become modified as they move into the western North Pacific and move northwestward along the dynamic equator defined by ITCZ convection and increased generation of potential vorticity.
Although the origin of synoptic-scale disturbances varies (e.g., African easterly waves in the North Atlantic, monsoon depressions in the western North Pacific) in each tropical cyclone basin, the primary role of synoptic-scale disturbances in tropical cyclone formation is to provide an incipient vortex that acts to organize convection. Although many such disturbances may form during a tropical cyclone season, relatively few actually develop into a tropical cyclone. The vortex and associated convection (i.e., MCSs) must remain coherent until they becomes embedded in an environment that will be favorable for tropical cyclone formation. Often, the ability of an incipient vortex to develop is related to the interactions between the synoptic-scale and the MCS(s), which may be linked to the larger circulation. Recent research has concentrated on identification of the physical mechanisms responsible for the development of one or more MCSs into a tropical cyclone. During this process, a deep warm-core cyclone must be developed from a shallow or mid-level vortex that is typically cold in the lower layers due to the cooling effects of precipitation evaporation. Once the mid-level vortex has been able to extend to the surface and the low-level cold region has been modified, the potential for the vortex to be sustained independent of its environment is increased.
It has become clear that the large-scale influence on tropical cyclone formation is dependent on interactions among a variety of space and time scales. It is recommended that these interactions continue to be examined such that a coherent theory of tropical cyclone formation be identified. Additional aspects that should be examined include contributions to periods of enhanced and reduced tropical cyclone activity, basin dependencies, and seasonal dependencies.
It is recommended that advances in theoretical understanding and observational analysis of tropical cyclone formation be examined to define new diagnostics for genesis potential that may be applicable to operational forecast models. Transition of new theories into new forecast models is a challenge for both research and forecast communities.
4.0.4 Prediction of Tropical Cyclone Formation with Numerical Models
As discussed in relation to large-scale influences on tropical cyclone formation, interactions among a variety of scales have long been considered to be critical for the transformation of an incipient tropical vortex to a tropical cyclone. Whereas influences of smaller scale (i.e., ~10-50 km) deep convective cells have always been thought to be crucial in organization of the incipient disturbance, lack of critical observations has prevented explicit examination of these influences. Recent advances in computer technology have now made it possible to numerically simulate tropical cyclone formation at cloud-resolving scales, which provides a basis for examination of a more complete set of processes during the formation process.
Idealized numerical experiments have been conducted to examine the roles of episodic deep convection and related vortex-tube stretching in acting to strengthen an incipient tropical vortex. The processes of vortex merger and axisymmetrization act to organize individual convectively-induced cyclonic potential vorticity anomalies such that tropical cyclone characteristics are attained over realistic time scales.
Significant progress has been made in examining the underlying organization of individual convective cells to organization of an incipient vortex. Fore example, a hierarchy of simulations (based on a 9 km grid with cumulus parameterization and 3 km grids with no explicit treatment of cumulus clouds) were used to examine tropical cyclone formation in which baroclinic forcing of the incipient vortex was stronger than typically associated with a formation over more tropical latitudes. Nevertheless, there may be a fundamental process of vortex generation and merger that acts in a variety of environments. Results demonstrated the generation and maintenance of localized low- and mid-level cyclonic vorticity anomalies associated with individual deep convective towers with horizontal scales near 20 km. The individual anomalies underwent merger and axisymmetrization as the incipient vortex attained tropical storm strength.
Based on recent capabilities to numerically examine cloud-scale influences tropical cyclone formation, it is recommended that idealized models continue to be used to identify basic dynamical processes and their interactions over a variety of scales.
Results of high-resolution idealized simulations should be used to identify how sensitive the various processes are to environmental influences. These types of determinations should be undertaken in conjunction with the use of operational models to identify potential weaknesses in general circulation model simulation of tropical cyclone formation.
Finally, results of numerical simulations must be compared with in-situ data. Therefore, it is recommended that increased emphasis be placed on obtaining critical observational data from a variety of platforms during the critical stages of tropical cyclone formation.
4.0.5 Seasonal Prediction of Tropical Cyclones
Accurate prediction of tropical cyclone activity on a seasonal scale would be a powerful tool in disaster preparedness and prevention. As discussed in Section 4.0.3, tropical cyclone formation is a multiscale process that may begin with a suitably favorable large-scale environment that is capable of sustaining various thermodynamic and dynamic factors. In terms of seasonal prediction, variability in large-scale, slowly-varying environmental factors may act to enhance or reduce the potential for tropical cyclone formation. Identification of the specific environmental factors and the mechanisms by which they may influence tropical cyclone formation has been a topic of active research. Specific environmental factors that have been linked to seasonal tropical cyclone prediction include the El Nino/Southern Oscillation (ENSO), quasi-biennial oscillation (QBO), and the MJO. However, there has been a lack of research on identifying how these factors may be interrelated or combined to influence tropical cyclone formation on seasonal scales.
Research on seasonal prediction of tropical cyclone formation has been conducted on nearly each ocean basin where tropical cyclones form. Although it has been shown that there is some relationship between tropical cyclone formation and ENSO in each basin, several additional factors have been identified to influence seasonal formation over the North Atlantic. Recent research has concentrated on variability of African easterly waves (AEWs) that may be attributed to the western Africa monsoon. The monsoon variability may also be related to other factors such as ENSO and the QBO. These types of teleconnections need to be investigated in more detail.
Over the western North Pacific, ENSO has not been identified to have a significant relationship with the number of tropical cyclone formations, but rather influences the location of preferred formation regions. There are significant shifts from formation in the southeastern portion of the western North Pacific during warm ENSO events to away from this region during cold events.
Over the eastern North Pacific, only a weak relationship has been identified between the number of tropical cyclone formations and ENSO but a more significant factor is that an increase in intense hurricanes occurs during warm ENSO events. Studies have identified the potential for AEWs to pass into the eastern North Pacific to serve as an incipient disturbance for tropical cyclone formation. Therefore, some relationship may exist between seasonal factors that influence Atlantic basin seasonal formation and seasonal variability in tropical cyclone formation over the eastern North Pacific.
Overall, the interactions among environmental factors must be examined more thoroughly. For example, ENSO and the QBO have been linked separately to seasonal variability in tropical cyclone formation over the North Atlantic, but identification of relations between ENSO and the QBO may increase the amount of variability in tropical cyclone formation that is explained by the two factors separately. Also, influences on the MJO and various monsoon systems will affect seasonal tropical cyclone formation over several basins. Many studies of seasonal prediction have concentrated on extreme warm/cold ENSO events and not on neutral years when other factors may play significant roles in seasonal tropical cyclone formation. Research has indicated that the state of the ENSO system explains a significant portion of the variability in tropical cyclone formation in nearly each tropical cyclone basin. However, there is still little skill in seasonal ENSO predictions beyond one season.
It is recommended that study of the physical basis for the prediction of seasonal tropical cyclone activity remain a fundamental for further improvement and development of skillful prediction models. Critical to this study is the examination of teleconnections between factors that influence seasonal variability in tropical cyclone formation.
The dependence of seasonal tropical cyclone formation on ENSO implies that the prediction of the global sea-surface temperature (SST) distribution has a direct impact on seasonal forecasting. It is recommended that development of a skillful ocean forecasting model that incorporates independent verification be investigated as an essential component for identifying factors associated with seasonal tropical cyclone formation.
It is recommended that dynamical-statistical models be investigated for seasonal tropical cyclone prediction. Dynamical model prediction at seasonal time scales using ensemble prediction techniques has shown increasing accuracy. Seasonal tropical cyclone forecast aids may be developed by combining the ensemble-based seasonal predictions and statistical techniques to infer seasonal tropical cyclone activity.
It is recommended that emphasis be placed on increasing the utility of seasonal forecasts by including information on the track, timing, and landfall characteristics on seasonal time scales. Research into environmental factors and combinations of factors that affect these types of parameters must be increased.
Finally, accurate best-track data sets are essential for consistent development of seasonal forecast techniques. Due to many issues as discussed in Topic 4.1, consistent data sets are lacking in even the best-observed basins. It is recommended that best-track data sets be reanalyzed using current understanding and techniques.
4.0.6 Extratropical Transition
Often a tropical cyclone recurves into the midlatitudes and transforms into a fast-moving, rapidly-intensifying extratropical cyclone. Extratropical transition (ET) may result in a larger and more powerful storm than the original tropical cyclone. During ET, the translation speed of the cyclone increases from typical speeds of 5 m s-1 to more than 20 m s-1. The resultant extratropical cyclone may bring winds and waves that are more indicative of severe cold season cyclones, but during a season when such systems do not normally exist. While precipitation totals are usually limited by the fast motion of the storm, large amounts are a threat with decelerating systems, such as those undergoing rapid occlusion. These severe impacts usually occur after the characteristic tropical cyclone structure has weakened and typical forecast tools used to classify storm intensity are no longer valid. All of these factors contribute to a very difficult forecast problem.
Extratropical transition occurs in nearly every ocean basin that experiences tropical cyclones. The exception is the eastern North Pacific where synoptic conditions are not conducive to the ET of tropical cyclones. When a tropical cyclone interacts with the midlatitude flow, the nearly symmetric wind and precipitation distributions that are concentrated near the circulation center expand to a broad asymmetric distribution of high winds and waves, and heavy precipitation.
Recent research has examined the types of transitions that may occur and the relative roles of the tropical cyclone structure and the midlatitude circulation into which the tropical cyclone is moving. Environmental factors that influence the ET process include increased vertical wind shear, frontogenesis, and energy transfers between the decaying tropical circulation and the midlatitude circulation.
Typically, numerical forecast errors increase during an ET event. This is usually attributed to uncertainties in the initial conditions due to a lack of conventional observations over regions where ET typically occurs. A number of recent studies have demonstrated this sensitivity by using special observation data sets.
Extratropical transition (ET) is a complex four-dimensional process that involves interactions over a variety of space and time scales. Although there is yet much to be learned about the characteristics of a mature tropical cyclone, viable research programs exist to examine these issues, as well as to examine processes responsible for extratropical cyclogenesis. However, the transition from a tropical cyclone into an extratropical cyclone is poorly understood and incompletely researched.
The evolutionary nature of ET and the time between successive observations makes it difficult to define a precise time at which a tropical cyclone has become extratropical. It is recommended that techniques similar to those used to assign tropical cyclone characteristics from satellite data be investigated as a benefit to operations such that an objective assessment of ET can be made. A framework for this type of definition could be the two-stage description of ET in terms of transformation from a tropical cyclone and re-intensification as an extratropical cyclone. Also, recent research into cyclone classification by thermal and environmental characteristics could be expanded from being tied to numerical model analyses to include observed characteristics. Development of these types of diagnostic measures would provide forecasters with conceptual models of ET that would lead to more effective advisories and warnings.
Because the physical impacts associated with ET can be extreme over large coastal regions, research is needed to improve understanding of the evolution and distribution of high impact variables such as extreme precipitation, an expanded wind distribution, and significant wave heights.
A significant amount of new data sources are being made available, especially from remote sensing platforms. New techniques are being developed to utilize these data to diagnose tropical cyclone characteristics. It is recommended that the full suite of recent techniques and observation capabilities also be applied to the examination of decaying tropical cyclones and ET. Improved knowledge of the ET characteristics requires increased observations during critical stages of the transformation and re-intensification stages.
In conjunction with the development of improved observational capabilities of ET, it is recommended that the impact of data assimilation and potential use of modified synthetic observations be investigated to provide improved numerical guidance of ET characteristics. This should also include investigation of predictability limits based on ensemble prediction techniques that may define an envelope of solutions for determination of resultant extratropical cyclone characteristics and potential threats to coastal regions.