Structure of Tornado-spawning Tropical Cyclones: An Analysis by Energy Helicity Index Based on a CAPE Including the Effects of Entrainment

Previous statistical studies on tornadoes associated with tropical cyclones show that they occur most frequently in the right-front quadrant of the cyclones, where storm-relative environmental helicity (SREH) is large.  However, their occurrences are not necessarily consistent with the distribution of CAPE, which is large in the right-rear quadrant.  The present study examines the structure and environment of typhoons that spawned tornadoes in Japan, and shows that the distribution of CAPE including effects of entrainment explains the tornado occurrences well.

35 typhoons that spawned a total of 52 tornadoes in Japan between 1991 and 2012 are defined as tornadic typhoons (TTs).  199 non-tornadic typhoons (NTs) which did not spawn any tornadoes and had similar strength and locations as TTs are also selected.  CAPE and SREH are calculated using Japanese reanalysis dataset (JRA-55) and their distributions for composited TTs and NTs in which typhoon centers are superposed by aligning moving directions are investigated.  In the calculation of CAPE, effects of entrainment on lifted parcels are considered by the method of Romps and Kuang (2010).  Statistical significance in the differences of the CAPE distribution between TTs and NTs are evaluated through Welch's t-test.

CAPE calculated without considering entrainment (hereafter CAPE₀₀) is large on the right-rear quadrant of the composite typhoons and becomes large as the distance from the typhoon center is increased in the quadrant.  This distribution of CAPE₀₀ is not consistent with that of observed tornadoes, which are concentrated in the right-front quadrant.  In addition, difference of CAPE₀₀ on the right-front quadrant between TTs and NTs is not statistically significant even at the 10% level.   On the other hand, CAPE calculated with an entrainment rate of ε = 20% km^-1 (hereafter CAPE₂₀) is large in the right-front quadrant, which is consistent with locations of tornadoes.  This occurs because both the high equivalent potential temperature in the boundary layer and abundant moisture in the mid-troposphere contribute to the large CAPE₂₀ in the quadrant.  Furthermore, CAPE₂₀ in the right-front quadrant for TTs is larger than that for NTs, where the difference is statistically significant even at the 1% level.  This difference between TTs and NTs mainly reflects the fact that the temperature at a height of about 5 km in the right-front quadrant for TTs is slightly lower than that for NTs.

If CAPE₂₀ instead of CAPE₀₀ is used for calculating energy helicity index (hereafter EHI₂₀ and EHI₀₀, respectively), the distribution of EHI₂₀ is found to give much better agreement with the tornado occurrences than that of EHI₀₀ does in the case of TTs.  Since both SREH and CAPE₂₀ in the right-front quadrant for TTs are larger than those for NTs, EHI₂₀ for TTs is much larger than that for NTs.  As TTs move into the mid-latitude and undergo an extratropical transition, SREH tends to increase gradually with time.  On the other hand, CAPE₂₀ attains a maximum around the time of the tornado occurrences and declines immediately, and so does EHI₂₀.

These results imply that a CAPE including effects of entrainment is an appropriate parameter for measuring potential strength of convection in a tropical cyclone environment and a combination of veering shear and thermal instability work together to generate mini-supercells and associated tornadoes.