Influence of the large-scale atmospheric environment on North Atlantic tropical cyclone size

Several of the most destructive tropical cyclones (TCs) within the historical record have been particularly large [e.g., TC Sandy (2012)]. In spite of the heightened damage potential associated with large TCs, our current understanding of the factors responsible for determining TC size remains incomplete. The size of TCs exhibits substantial variability as demonstrated by the order of magnitude difference between the radius of 34-kt surface wind speed for the smallest TC, 2011 TC Harvey (62 km), and the largest TC, 2012 TC Sandy (769 km), on record in the North Atlantic. Although previous studies have suggested that large TCs may require a relatively moist environment or relatively large angular momentum imports, a comprehensive study of the factors governing TC size is warranted given the potential benefits of improved forecasts of the size of a landfalling TC. Building upon the foundation of prior work, we will examine large-scale environmental factors governing the size of North Atlantic TCs.

We hypothesize that the occurrence of large TCs is dependent on three factors: a large precursor d¬¬isturbance with horizontal scales greater than ~1500 km, a moist lower-and-midtropospheric region with horizontal scales greater than ~1500 km, and an upper-tropospheric region of reduced inertial stability with horizontal scales greater than ~3000 km. To investigate the relative importance of these three factors, we will construct storm-relative composites of small, medium, and large North Atlantic TCs from the NCEP Climate Forecast System Reanalysis and diagnose the processes responsible for determining TC size. In this study, we consider small, medium, and large TCs to be in the first, third, and fifth quintiles of the Extended Best-Track radius of 34-kt surface wind speed for North Atlantic TCs from 1989 through 2012.

Results suggest that medium and large TCs initially grow at approximately the same rate following TC genesis before the growth rate of the medium TCs slows considerably. Storm-relative composites show that large TCs are embedded within a basin-scale region of negative lower-tropospheric geopotential height anomalies and on the northeastern flank of a region of anomalous westerly winds, both of which are statistically significant at the 95% confidence interval. The geopotential height anomalies may be associated with enhanced lower-and-midtropospheric moisture or enhanced lower-tropospheric convergence, while the anomalous equatorial westerlies may provide a source of cyclonic vorticity for the TC to grow. The location, and the space and time scales, of the geopotential height and westerly wind anomalies in the large TC composite appear to indicate that the anomalies are associated with the Madden– Julian Oscillation, a convectively coupled equatorial wave, or a combination of these phenomena. In contrast, the composites of medium TCs do not exhibit statistically significant basin-scale geopotential height and westerly wind anomalies, suggesting that the presence of these basin-scale anomalies is a necessary condition for the occurrence of a large TC. A similar analysis of the small TC composite is ongoing and the results will be presented at the conference.