9C.2 Characteristics of Tropical Cyclone Rapid Intensification in Environments of Upper-Tropospheric Troughs

Wednesday, 18 April 2018: 10:45 AM
Champions ABC (Sawgrass Marriott)
Michael S. Fischer, Univ. at Albany, SUNY, Albany, NY; and B. H. Tang and K. L. Corbosiero
Manuscript (2.2 MB)

Predicting tropical cyclone (TC) rapid intensification remains a challenging forecast problem. For those TCs that interact with upper-tropospheric troughs, the intensity forecasts are further complicated. Upper-tropospheric troughs may affect TC intensity through, among other influences, quasigeostrophic (QG) forcing for ascent, eddy flux convergence of angular momentum, and vertical wind shear.

This work explores the environmental and convective characteristics of overwater TC–trough interaction events occurring in the North Atlantic basin between 1989 and 2016. TCs are placed into one of four intensity change groups, referred to as rapid intensification (RI), slow intensification (SI), neutral (N), and weakening (W) episodes, based on the 24-h change in maximum sustained 10-m best-track wind speed. TCs are also binned into one of three environmental regimes, termed high-potential vorticity (PV), mid-PV, and low-PV, based on the maximum PV anomaly magnitude on the 350-K isentropic surface within a 250–1000-km TC-centered annulus, as analyzed by the ERA-Interim reanalysis. This study focuses on those TCs that reside in high-PV environments, which are characterized by coherent upper-tropospheric troughs.

High-PV RI TCs are associated with upper-tropospheric PV anomalies of weaker magnitudes than PV anomalies in other intensity change groups; however, high-PV environments are composed of a variety of upper-tropospheric trough morphologies. In order to identify similar TC–trough configurations, a dimensionality-reduction machine learning technique, t-Distributed Stochastic Neighbor Embedding (t-SNE), was implemented. Using the results of the t-SNE technique in conjunction with k-means clustering, three clusters of TC–trough interactions were analyzed. Within each cluster, high-PV RI episodes are associated with lower vertical wind shear magnitudes than TCs in the other intensity change groups. In two of the three clusters, high-PV RI episodes are associated with a secondary PV anomaly located on the equatorward side of TC and opposite of the poleward, primary upper-tropospheric PV anomaly. It is hypothesized that for these two clusters, the cyclonic flow associated with the secondary PV anomaly acts to oppose the cyclonic flow associated with the primary upper-tropospheric PV anomaly, thereby reducing the magnitude of the upper-tropospheric winds, and consequently, reducing the vertical wind shear.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner