The Role of Equatorial Rossby Waves in Tropical Cyclogenesis

Open Access
- Author:
- Gall, Jeffrey
- Graduate Program:
- Meteorology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- March 19, 2009
- Committee Members:
- William Menaul Frank, Dissertation Advisor/Co-Advisor
William Menaul Frank, Committee Chair/Co-Chair
Sukyoung Lee, Committee Member
Dr David Stauffer, Committee Member
Diane Marie Henderson, Committee Member - Keywords:
- tropical cyclones
hurricanes
tropical meteorology
rossby wave - Abstract:
- A set of unique methodologies utilizing both idealized and real-data numerical simulations was employed to examine the links between the equatorial Rossby (ER) wave and tropical cyclogenesis. All of the recent research involving ER waves applied various wavenumber-frequency filtering algorithms on observational data sets to obtain statistical links between ER waves and tropical cyclogenesis. We contend that the methodology employed in this study removes many of the limitations inherent in the previous work on ER waves and tropical cyclogenesis. The idealized simulations performed in this dissertation were broken down into two different sets of experiments - forced and initial-value simulations. The forced simulations were the simpler of the two and utilized a forcing function centered on the equator in a no-flow background environment to excite ER waves of varying intensity. The anomalous circulations associated with the ER wave relevant to tropical cyclone (TC) genesis (e.g. low-level vorticity, low-level convergence, and vertical shear) were quantified for each of the ER waves. It was shown that the maximum low-level vorticity and low-level convergence associated with the ER wave were on the order of 2x10^-5 1/s and 1x10^-5 1/s, respectively. While portions of the ER wave were favorable for genesis, it was shown that when the ER wave anomalies were combined into a single genesis parameter (GP), the values of the parameter were near, but below, a threshold value commonly associated with genesis. Based on this result it was hypothesized that the magnitude of the anomalous circulations of the ER wave alone is usually insufficient to cause TC genesis; rather, only when an intense ER wave combines with a favorable background flow does the likelihood of genesis increase significantly. The selected background flow for this study was a monsoon trough (MT), as this large-scale feature has been shown to play a prominent role in many TC genesis cases, especially in the North West Pacific Ocean. A series of idealized MT structures were superimposed on the ER wave, and it was demonstrated that within certain portions of the wave, the GP exceeded the threshold value. Not surprisingly, the most favorable conditions for genesis were observed for the most intense (largest amplitude) ER wave in combination with the most intense (large background cyclonic vorticity) MT. The enhanced genesis region had a longitudinally-elongated, latitudinally-contracted, cigar-like shape within the area just equatorward of the ER wave gyre center. This region also lies on the poleward side of the MT, in a region of large background cyclonic vorticity. While the results from the forced idealized simulations demonstrated that the likelihood for genesis increased significantly for intense ER waves within a favorable background environment, owing to the experiential design, no genesis events within the ER wave were actually simulated. For this reason, a series of initial value idealized simulations were conducted to study the genesis mechanism within an ER wave. A three-dimensional (x,y,z), n=1, wavenumber 10 ER wave was inserted in the model initial condition within a no-flow background environment. It was shown that resulting properties of the ER wave (e.g. phase speed, vertical structure, etc.) agreed relatively well with recent observations of convectively-coupled ER waves. The quadrant located in the eastern half of the cyclonic gyre of the ER wave was the only quadrant associated with both cyclonic vorticity and low-level convergence. Further, the anomalous circulations associated with the resulting simulated ER wave were compared to the ER wave anomalies from the forced simulations. The magnitude of the low-level vorticity and low-level convergence fields agreed relatively well, while the vertical shear was larger in the ER waves from the forced experiments. This wavenumber 10 ER wave was then used to initialize a series of simulations with different background idealized monsoon troughs. Two different idealized MT flow configurations were simulated for a nine day period. The first (second) had a strip of relative vorticity of 1$ imes 10^{-5}$~s$^{-1}$ (2$ imes 10^{-5}$~s$^{-1}$) between about 4.5dg~N and 10dg~N, and was zero elsewhere. Both MT flows were shown to be relatively stable. That is, while they did not breakdown, their relative vorticity structure remained quasi-steady over the course of the simulations. The idealized MT flow configurations were then added to the ER wave. For both MT cases, the ER wave structure quickly deformed within the background horizontal shear region of the MT. A smaller-scale cyclonic circulation formed as a result of the wave-breaking process and was shown to have a horizontal length scale comparable to a TC. The wave-breaking process was accompanied with a shift of the location of enhanced convection from east of the ER wave cyclonic gyre center to a region co-located with the TC-scale circulation. The TC-scale circulation formed near the initial critical latitude in both cases. This process was very similar to theoretical studies of the wave-breaking of Rossby waves in a horizontally-sheared flow, i.e. the formation of the so-called ``cat's eyes'. The primary difference between the two simulations was that the wave-breaking process occurred much faster in the stronger MT case. Thus, a TC-scale circulation of sufficient intensity capable of intensifying via its own air-sea interaction formed on a much faster timescale in the stronger idealized MT case. By nine days, a TC of tropical storm intensity had formed in the strong MT flow plus ER wave, while only a weak tropical depression-strength disturbance formed in the weaker MT plus ER wave simulation. The initial value idealized simulations were repeated, but with moisture and diabatic effects turned off. The resulting ER wave propagated at a phase speed about 1~m~s$^{-1}$ faster when compared to the moist ER wave. While no TCs formed when the idealized MT environments were added to the dry ER wave, the ER wave was observed to break regardless of moisture and diabatic effects in the vicinity of the critical latitude. As was the case in the moist simulations, the TC-scale circulation was stronger for the case of the stronger MT flow configuration. In the final part of the study, real-data simulations were performed in which an ER wave was inserted into a more realistic background environment. Two TC genesis cases from the real-data simulations were examined in much more detail; one case (Case 1) in which an ER wave promoted TC genesis and the other (Case 2) in which the ER wave suppressed TC formation relative to the control simulation. For both cases, the local conditions prior to and at the time of genesis were documented. The large-scale environment in Case 1 was associated with anomalous cyclonic low-level vorticity, anomalous low-level convergence, and weaker vertical shear relative to the CON simulation. In Case 2, the large-scale environment featured much larger vertical shear and anticyclonic relative vorticity, owing to the circulations of the ER wave. Results from the various numerical experiments demonstrated that ER waves have the ability to both enhance and suppress TC formation. In the case of ER wave-enhanced genesis, the anomalous low-level vorticity and convergence were the dominant factors. For the case of ER wave-suppressed activity, vertical shear was much more important. It appears that an anomalously intense ER wave is usually not sufficient on its own to initiate genesis. However, an anomalously intense ER wave placed within a favorable background environment (e.g. monsoon trough) makes genesis much more likely. The wave-breaking of the ER wave is a viable mechanism for the initial formation of the TC scale circulation. The timescale of the formation of this circulation occurred over a much faster timescale for the stronger MT flow configuration when compared with the weaker MT. Owing to the uniqueness of the methodology of both the idealized and real-data simulations, there remains a plethora of unanswered questions that may provide numerous avenues for future research. First, in the initial value idealized simulations, only a wavenumber 10 ER wave was considered. We want to perform a suite of sensitivity studies in which the wavenumber, initial equivalent depth, and wave type are varied and examine how the phase speed as well as the convectively-coupled structure of the wave changes as a result of the varied parameters. Second, we want to examine the effect of other background flow configurations on the structure and evolution of the ER wave. Finally, we plan to map out the relationship of the two parameters - the critical latitude of the ER wave (if it exists) and the meridional absolute vorticity profile prior to genesis for ER wave-TC genesis events. This observational study will utilize a large data set (e.g. Frank and Roundy 2006) of TC genesis events in which ER waves played a significant role.