582 WEATHER AND FORECASTING VOLUME 29 The Influence of Tropical Cyclone Size on Its Intensification CRISTINA ALEXANDRA CARRASCO North Carolina Agricultural and Technical State University, Greensboro, North Carolina CHRISTOPHER WILLIAM LANDSEA NOAA/NWS/NCEP/National Hurricane Center, Miami, Florida YUH-LANG LIN North Carolina Agricultural and Technical State University, Greensboro, North Carolina (Manuscript received 29 July 2013, in final form 29 January 2014) ABSTRACT This study investigates tropical cyclones of the past two decades (1990–2010) and the connection, if any, between their size and their ability to subsequently undergo rapid intensification (RI). Three different pa- rameters are chosen to define the size of a tropical cyclone: radius of maximum wind (RMW), the average 2 34-knot (kt; 1 kt 5 0.51 m s 1) radius (AR34), and the radius of the outermost closed isobar (ROCI). The data for this study, coming from the North Atlantic hurricane database second generation (HURDAT2), as well as the extended best-track dataset, are organized into 24-h intervals of either RI or slow intensification/constant intensity periods (non-RI periods). Each interval includes the intensity (maximum sustained surface wind speed), RMW, AR34, and ROCI at the beginning of the period and the change of intensity during the sub- sequent 24-h period. Results indicate that the ability to undergo RI shows significant sensitivity to initial size. Comparisons between RI and non-RI cyclones confirm that tropical cyclones that undergo RI are more likely to be smaller initially than those that do not. Analyses show that the RMW and AR34 have the strongest negative correlation with the change of intensity. Scatterplots imply there is a general maximum size threshold for RMW and AR34, above which RI is extremely rare. In contrast, the overall size of the tropical cyclones, as measured by ROCI, appears to have little to no relationship with subsequent intensification. The results of this work suggest that intensity forecasts and RI predictions in particular may be aided by the use of the initial size as measured by RMW and AR34. 1. Introduction forecasting have proved to be much more challenging (Gall et al. 2013). The operational prediction of rapid The analysis and prediction of tropical cyclones in the intensification [RI; defined as 30 kt1 or greater intensity Atlantic basin (including the North Atlantic Ocean, the gain over 24 h; Kaplan and DeMaria (2003)] continues to Gulf of Mexico, and the Caribbean Sea) have evolved be identified by NHC as their number one priority for significantly over the last few decades (Sheets 1990; improvement (Rappaport et al. 2012). Rappaport et al. 2009). Track predictions issued by the RI has proved difficult to forecast because of a general National Hurricane Center (NHC) have improved dra- lack of understanding of the physical mechanisms that matically due to more accurate numerical models and are responsible for these rare events. Previous work has more satellite-based, open-ocean observations. However, associated the ability of a tropical cyclone to undergo RI in recent years, making improvements to operational with the following: low tropospheric vertical wind shear, tropical cyclone intensity (maximum 1-min, 10-m wind) a very warm ocean with a deep mixed layer, a moist Corresponding author address: Yuh-Lang Lin, North Carolina 2 A&T State University, 302H, Gibbs Hall, EES/ISET, 1601 1 Knots (kt; 1 kt 5 0.51 m s 1) will be the metric of wind speed for E. Market St., Greensboro, NC 27411. the remainder of the paper, as this is what is used for both the E-mail: [email protected] tropical cyclone forecasting and database. DOI: 10.1175/WAF-D-13-00092.1 Ó 2014 American Meteorological Society JUNE 2014 C A R R A S C O E T A L . 583 troposphere, and inner-core processes (such as concen- and intensification that was statistically significant at 12-, tric eyewall cycles and vortex Rossby waves) (Kaplan 24-, and 36-h lead times. However, somewhat contradic- et al. 2010). However, little research has been done on tory to that, DeMaria and Kaplan (1994) also uncovered whether the size of a tropical cyclone plays a role in its that the tropical cyclones that most rapidly intensified subsequent intensity change. The idea that the initial over a 48-h period tended to have smaller than average size of a tropical cyclone can help dictate intensification outer circulation strength. They ascribed this behavior to to follow is a concept that is applied qualitatively by the tendency for the outer circulation of tropical cyclones some hurricane forecasters: ‘‘Strengthening is forecast to spin up later in the life cycle, generally after peak in- [for Tropical Storm Leslie] to begin around the time the tensity has taken place. This finding is consistent with shear relaxes, but the rate of intensification could ini- both observational climatological studies in the Atlantic tially be slow due to the large size of the circulation’’ (Merrill 1984) and idealized modeling work (Ooyama (T. Kimberlain, NHC Tropical Cyclone Discussion, 1969). Subsequent SHIPS updates (DeMaria and Kaplan Tropical Storm Leslie, 0300 UTC 4 September 2012, 1999; DeMaria et al. 2005) still employ an outer circulation personal communication). If indeed there exists a robust metric, though this now is represented by the 0–1000-km connection, such knowledge should be better quantified radii, 850-mb relative vorticity. This large-scale vorticity and used objectively to help forecasters better predict RI. has a moderately skillful, positive association with in- Therefore, this study analyzes the sizes of tropical cyclones tensity change in their 12–120-h prediction scheme. It is that underwent RI versus those that are steady state or noted, however, that DeMaria et al. (2005) consider this slowly intensifying over a 24-h period from Atlantic basin metric to be more indicative of the synoptic environ- tropical cyclones during 1990–2010. The goal of this study ment around the tropical cyclone, rather than a direct is to investigate if there is an association between a tropical measure of the storm’s size itself. cyclone’s size and its subsequent intensification. The rapid intensification index (RII) scheme (Kaplan and DeMaria 2003), a method for probabilistically pre- dicting RI, was tested to see whether 0–1000-km, 850-mb 2. Previous research on size and subsequent relative vorticity aided these predictions. However, this intensity change particular metric was not a skillful predictor of RI and The effect of tropical cyclone size on subsequent in- thus was not included within the model, nor is it included tensity change has been the subject of a few investi- in the most recent version of RII (Kaplan et al. 2010). gations. From a theoretical perspective, Emanuel (1989), Kimball and Mulekar (2004) established a climatol- building off of the results from Rotunno and Emanuel ogy of multiple tropical cyclone size parameters for the (1987), concluded that the initial size of the vortex, as North Atlantic basin with data from 1988 through 2002. measured by the radius of maximum wind (RMW), They showed that the RMW decreases from an average played a substantial role in its subsequent intensification of about 55 n mi (1 n mi 5 1.852 km) for tropical storms, rate. If the initial vortex was too large, then no sub- 40 n mi for category 3 storms, and 30 n mi for category 5 sequent intensification occurred. But as the initial RMW hurricanes. In contrast, the average radii of 34-, 50-, and was progressively reduced in size in the model, the in- 64-kt winds, as well as the radius of the outermost closed tensification rate increased significantly with the smallest isobar (ROCI), increase in size with increasing intensity. vortex having the fastest intensification. It is of note that Kimball and Mulekar (2004) also stratified their dataset the small- to medium-sized vortices in his study all even- by intensifying, steady-state, and weakening tropical tually reached the same peak intensity, even though their cyclones over the next 6 h. This showed a tendency for rates of intensification differed. the radii of 34-, 50-, and 64-kt winds to be smaller in size There have been additional papers published on the for intensifiers compared to weakeners, while the RMW topic from an observational perspective. DeMaria and and ROCI showed no significant differences. Kaplan (1994) developed a statistical model—the Statis- Chen et al. (2011) compared the 24-h intensification for tical Hurricane Intensity Prediction Scheme (SHIPS)— western North Pacific typhoons that were compact versus for predicting intensity changes of Atlantic tropical those that were incompact. Their ‘‘compactness’’ size cyclones at 12, 24, 36, 48, and 72 h. This model used a parameter is based upon both the RMW and tangential standard multiple regression technique with climatolog- winds at a radius of twice the RMW compared with cli- ical, persistence, and synoptic predictors, including one matological values. They found that compact tropical based upon the outer circulation strength (850-mb rela- cyclones (either small RMW, weak winds at twice the tive angular momentum measured between 400- and RMW radius, or both) had a substantially higher rate 800-km radii from the center). Their results showed a weak of intensification and more frequent RI relative to in- positive association between outer circulation strength compact systems. 584 WEATHER AND FORECASTING VOLUME 29 One limitation to these studies is that any possible ef- a sample size that is over twice as large outweighs these fects of inner- and outer-core size parameters upon sub- concerns. For RMW, which is an operationally estimated sequent intensification may be made ambiguous because parameter throughout the time period, it is likely that the of the general life cycle tendency for tropical cyclones to value is more uncertain farther back in time.