An observational study on the physics of tropical cyclone motion

Student thesis: Master's Thesis

View graph of relations


  • King Wai NGAN


Awarding Institution
Award date27 May 1997


The main theme of this research is to study the physics of tropical cyclone (TC) motion through observational data. Two approaches are chosen to study seven TCs that occurred during the international experiments conducted to study TC motion over the Western North Pacific in 1990. The first approach is to study the TC motion through the examination of the entire atmospheric wind field associated with the cyclone based on the Navy Operational Analysis and Prediction System (NOGAPS) analyses. The second approach is to study the different effects on the TC motion, including environmental steering and azimuthal wavenumbers one and two flows, based on the Final Analyses (FA) data. Elsberry (1990) hypothesized that the asymmetric flow of a TC is the main contributor to the deviation between the steering flow and the TC motion vectors which may be defined as the propagation vector. Thus, one of the main concerns of this study is to examine the relationship between this asymmetric flow and TC motion. The method used to study the NOGAPS data is similar to Chan (1986). The asymmetric flow associated with a TC is found to have a significant relationship with the TC direction of motion. Case studies of Typhoons Yancy, Dot and Ed suggest that when the centre of the cyclone is very close to and south of the axis of the subtropical ridge (STR), the asymmetric flow exhibits a (azimuthal) wavenumber- 1 pattern, similar to the ones obtained by Chan (1986). In the cases of Yancy and Dot, the line joining the centres of the counterrotating gyres is also perpendicular to the TC motion direction. However, not in all the cases does this pattern develop. This means that even when the STR persists around the TC, the gyres may not be present or only one gyre appears. The reasons may be due to the binary interaction between cyclones, a strong environmental vorticity gradient or higher wavenumber asymmetries induced in the cyclone. When the orientation of the counterrotating gyres is normal to the TC movement, the azimuthally-averaged asymmetric flow within the 3-7" latitude radial band matches the TC motion direction. This result is consistent with that of Chan and Gray (1982). There is also an important relationship between vorticity advection and the TC direction of motion. The contribution of the asymmetric advection of symmetric vorticity is the most significant term in the TC movement. It shows that the 'asymmetric flow' (counterrotating gyres) advects the entire TC forward. Consistent results can also be found in the divergence term and vorticity advection patterns at different layers. The second approach to study TC motion is using the FA data. The vortex circulations are first filtered from the total flow through the filtering technique proposed by Kurihara et al. (1993) and then the symmetric flow, wavenumber-1 and wavenumber-2 asymmetric flows are separated through a Fourier decomposition technique. However, the asymmetric flow associated with the TCs is found to have no unique relationship to the propagation vector. Besides, the wavenumber-1 vector changes from time to time and wavenumber-1 gyres can only be identified in some cases. On the other hand, wavenumber-2 gyres frequently appear and show a northwest-southeast and northeast-southwest orientation. These results suggest that the hypothesis proposed by Elsberry (1990) cannot be verified. Although a systematic relation between the asymmetric flow with the propagation vector cannot be found, propagation is generally small when the TC is embedded in a strong environmental flow. This implies that the TC motion is dominated by the environmental flow. Generally, a deviation of the environmental flow with the TC motion is always to the left of the motion with a larger magnitude, which is consistent with the results of previous studies. Straight tracks or stable tracks of TCs are found to be associated with steady outer wind profiles when the azimuthally-average tangential winds are examined. On the other hand, a time-varying outer wind profile is associated with zigzag or abnormal recurving tracks. Consistent results can also be found from the asymmetric flow of the TCs. No asymmetric gyres exist when the TCs have steady outer wind flow and vice versa. This implies if interaction between the environment and the vortex circulation is weak, the motion of the TC is almost purely due to steering, similar to the results found from NOGAPS data. In the overall cases, the TC motion can also be explained by the vorticity advection where a maximum area of vorticity tendency is always in the forward quadrant of the TC and a minimum on the back which form a dipole-like structure. A good TC motion direction approximation is found in this dipole-like structure either from the environmental vorticity advection or the total vorticity advection. In many cases, the symmetric advection of asymmetric vorticity offsets the asymmetric advection of symmetric vorticity terms in the vorticity advection. Thus the importance of the symmetric advection of asymmetric vorticity and asymmetric advection of symmetric vorticity will become relatively small. However, sometimes, the asymmetric advection of symmetric vorticity in the wavenumber-2 asymmetric circulation plays an important role especially when the vortex is recurving. The asymmetric advection of symmetric vorticity in the wavenumber-2 asymmetric circulation also plays an important role in the changing track especially when the environmental flow is weak.

    Research areas

  • Tropics, Cyclones, Cyclone tracks