Structures and Environment of Extratropical Cyclones that Cause a Tornado Outbreak

Typical springtime tornado outbreaks in the United States occur in the warm sector of extratropical cyclones (hereafter, ECs). However, not all ECs produce tornado outbreaks. The differences in the structures and environment of ECs that cause tornado outbreaks and that do not, still remain poorly understood. In this paper, we investigate, by means of composite analyses of the newly-released Japanese reanalysis data (JRA-55) and idealized numerical simulations, the differences in the structure and environment of these ECs and physical mechanisms for the differences. In the composite analyses, we have categorized ECs which developed in the United States in April and May between 1995 and 2011 into two groups: ECs accompanied by 15 or more tornadoes (hereafter, referred to as outbreak cyclones (OCs)) and ECs accompanied by 5 or less tornadoes (non-outbreak cyclones (NOCs)). 55 OCs and 41 NOCs that are of similar strength as OCs are selected in this study. The composite analyses show significant differences in convective environmental parameters between OCs and NOCs. For OCs, convective available potential energy (CAPE) and storm relative environmental helicity (SREH) are larger and the areas in which these parameters have significant values are wider in the warm sector. The larger CAPE in OCs is due to larger amount of low-level water vapor, while the larger SREH in OCs due to stronger southerly wind at low levels. A piecewise potential vorticity (PV) diagnostics (Davis and Emanuel, 1991) indicates that low- to mid-level PV anomalies mainly contribute to the difference in the low-level winds between OCs and NOCs. On the other hand, the low-level winds associated with upper-level PV anomalies are not the major contributor to the difference. In the idealized numerical experiments, ECs are developed in zonally homogeneous basic states which are obtained by 5-day average of composite environments for OCs and NOCs, and their structures are examined. Weather Research and Forecasting (WRF) ver. 3.4 is used for the numerical experiments. Note that water vapor field in OCs is used for all the experiments to study dynamical influences of the basic states on the structure and intensity of ECs. The size of the calculation domain is 12000 km, 6000 km, and 25 km in the zonal, meridional and vertical directions, respectively. The horizontal grid size is 20 km. The boundary conditions are cyclic at the zonal boundaries, symmetric at the meridional boundaries. The lower boundary is assumed to be grassland. Kain-Fritsch scheme is used for the cumulus parameterization. WRF Single Moment 6 class (WSM6) for cloud physics, and the Mellor, Yamada, Nakanishi and Niino (MYNN) level 2.5 planetary boundary scheme for the boundary layer are used The results of the numerical experiments for the composite environments for OCs and NOCs (hereafter, referred to as OC-EXP and NOC-EXP, respectively) show that the characteristics of the low-level wind fields and SREH distributions for the simulated ECs in OC-EXP and NOC-EXP are similar to those for OCs and NOCs, respectively. In OC-EXP, SREH and low-level winds in the east-southeast region of the EC center is larger than those in NOC-EXP, respectively. It is suggested that these differences are due to the structures of jetstream. The structure of jetstream in OC-EXP has larger anticyclonic shear in the southern side of the jet axis than that in NOC-EXP. Larger horizontal anticyclonic shear of the jetstream in OC-EXP causes more meridionally- elongated structure of the EC, resulting in stronger low-level winds and larger SREH in the southeast quadrant of the cyclone center. These differences are also found in DRY experiments in which latent heating is excluded. On the other hand, there is no remarkable difference in CAPE and low-level water vapor between OC-EXP and NOC-EXP. This suggests that the larger thermodynamic parameters for OCs as found in the composite analyses are not explained by the cyclone structure alone. Thus, there is a possibility that the differences in environmental water vapor between OCs and NOCs, and the zonal inhomogeneity of environmental fields, which are not considered in the experiments, are important to the differences in the thermodynamic parameters between them.