P2.152 SPRING-NEAP TIDAL PATTERNS IN NORTH ATLANTIC TROPICAL CYCLONES Peter H. Yaukey* University of New Orleans, New Orleans, USA ABSTRACT spring-neap tidal cycle. Tropical cyclones are hypothesized to Atlantic best-track observations from reduce their own intensity by mixing of cool 1950-2007 displayed a pronounced peak of water upward, locally reducing the warmth of both hurricane occurrence and mean cyclone the sea surface. Little is known about how wind speed half way between the new and temperature variations in subsurface waters full moon. These patterns were consistent might moderate their influence on cyclones over time, being correlated between the first over sub-seasonal time scales. For instance, (1950-1978) and second (1979-2007) halves cold water might be mixed upward to depths of the study period. This pattern of hurricane accessible to storms by internal tidal waves occurrence was similar east and west of (bores) propagated along the thermocline. I 550W and in the Gulf of Mexico/Caribbean explored the possible influence of tidal mixing Sea, but appeared phase shifted by ~7 d on near-surface ocean temperature profiles north of 350N. Initiation dates of rapid and tropical cyclones by analyzing patterns of intensification events (≥15 kt increase in 24 h) cyclone frequency and intensity, and of were distributed non-uniformly across the subsurface water temperatures, across the synodic cycle, and showed peaks more ___________________________________ directly centered on the new (major peak) *Corresponding author address: Department and full moon (minor). of Geography, University of New Orleans, Water temperatures at 35-150m depth at New Orleans, LA, USA, 70148; e-mail: three long term moorings between 120N and [email protected]. 320N were each dominated by a peak near 1 the new moon, matching the timing of tracks, with especially large errors in the intensification events, although they varied in realm of rapid intensification (Rappaport et other details. Within 500 km of the 120N and al., 2009). The National Hurricane Center 150N moorings, cyclones that passed while has elevated improvements in intensity rapidly intensifying appeared to be associated forecasting to their top priority for tropical with warmer 35-150m waters than were other meteorology research. Intensification of passing cyclones. Evidence less consistent hurricanes is widely recognized to be affected with tidal influence include the presence of a by a variety of factors, including vertical wind relatively strong lunar pattern in hurricane shear, humidity, vorticity, and upper-level occurrence in the Gulf of Mexico and divergence. Ocean temperature is critical to Caribbean Sea where tides are weak, the storm intensification, including not just limited amplitude of spring-neap temperature surface temperature but also the depth of the variations (only 0.1-0.20C) and their surface layer of warm water (Hong et al., inconsistency across years at two of the 2000; Mao et al., 2000; Shay et al., 2000; mooring sites. Gino and Trinanes, 2003). The heat content of the uppermost KEY WORDS: Tropical cyclone; hurricane; ocean and variation of water temperature with rapid intensification; tidal; spring-neap. depth near the surface depend upon several factors, such as the presence and depth of 1. INTRODUCTION warm ocean currents and eddies, radiant Forecasting of hurricane intensity remains an fluxes, and wind. In recent years important challenge for tropical cyclone oceanographers have also become aware researchers. Accurate intensity forecasting is that much more tidal energy dissipation critical to warning coastal communities of the occurs away from the continental shelves level of threat posed by approaching storms, than was previously thought (Egbert and Ray, yet advances in forecasting intensity have not 2000; Garrett, 2003; Jayne, 2005; Thorpe, kept pace with those in predicting of storm 2005). The ability of tidal movements to 2 induce mixing of water near the ocean variations in the occurrence and intensity of surface is still not well understood, but mixing tropical cyclones across the spring-neap in the upper ocean has been reported from cycle. I then examine temperature modeling and empirical studies both far from fluctuations down to 150m at three long term (e.g., Martin, 2007; Schiller and Fiedler, moorings in the Atlantic, to assess whether 2007) and along continental shelves (e.g., sub-surface temperatures vary in a pattern Creswell and Badcock, 2000; Sharples et al., consistent with patterns exhibited by the 2006). Deep ocean tidal mixing is thought to cyclones. I focus on the Atlantic basin, where be generated at seafloor topographic the most complete and precise historical features, and it is not clear how far tidal database of storm wind speeds is available. energy can spread from these. One possible mechanism of spread is via movements of 2. METHODS internal waves along the thermocline. Such 2.1 Tropical cyclone records internal waves create predictable upwelling I obtained tropical cyclone records at 6 h and surface cooling with fortnightly intervals from June-November between 1950 periodicity, at least at several locations on the and 2007 from the best track analyses of the Pacific Coast of North America (Pineda 1991, National Hurricane Center’s North Atlantic 1995). At present, tidal variations in Hurricane Database (HURDAT, subsurface water temperatures are not http://www.nhc.noaa.gov/pastatll.shtml). incorporated into cyclone forecasting. Records prior to 1950 were excluded The purpose of this paper is investigate because most lacked data from the early whether the subsurface waters within reach (tropical depression) phase of storm of wind-induced mixing by passing cyclones development, and observations made north exhibit meaningful temporal variation on the of 500N or coded as “extratropical” were spring-neap tidal cycle, which might indicate omitted. The onset of rapid intensification the potential to influence cyclone was identified for storms that experienced a intensification. I begin by examining Rapid Intensification Event (RIE) of ≥15 knots 3 (7.7 ms-1) over 24 h. Intensification individual storm accounts in the National calculations did not begin until a storm was Hurricane Center online archives, and storm classified as a tropical or subtropical system track maps on the UNISYS website of at least depression strength, and (www.weather.unisys.com/hurricane). consecutive (beginning 6 h after another ended) or overlapping RIE were considered a 2.2 Ocean temperature single event. When a storm had more than Three moorings that have especially one RIE, that with the largest total wind lengthy records of sub-surface temperatures speed increase was used; maximum 6 h in cyclone-prone areas of the Atlantic were increase within the RIE was the second tie selected for analysis (Fig. 1). Two were breaker. To exclude RIEs caused merely by adjacent components of the PIRATA array landfall-weakened storms intensifying as they located at 12.00N 38.00N and 15.00N 38.00N, returned to sea, post-landfall RIE did not 333 km apart (available at begin until storms had regained their pre- http://www.pmel.noaa.gov/tao/data_deliv/deli landfall strength. If this happened in between v-nojava-pir.html). Daily temperature observations, the storm was assumed to readings were available from these moorings have regained it at the midpoint in between back to 1999 and 1998 respectively, at 20 m those observations. Pre-landfall wind speed depth intervals down to 140 m and at 180m. was taken from the last observation before For days that were missing data from a single the official track came ashore. For purposes depth or two adjacent depths that were other than RIE, observations within 3 d after bounded by measurements 20m above and being over land (coded as 10 x 10 blocks that below, missing temperatures were filled in by intersected land) were excluded. For all interpolation between the next deeper and purposes, only contacts with a continental shallower values, based simply on the mean mainland or Greater Antillean island relative magnitude of the missing depth’s constituted landfall. Identification of RIE and temperature between its two bounding depths landfalls in HURDAT was assisted by 4 in the overall data set. Days with more HURDAT observations posed challenges for missing data were omitted testing for statistical significance. I dealt with The third mooring, the Bermuda Testbed this problem in two ways. First, I compared site (31.70N 64.20W), has been operating patterns of cyclone occurrence and wind since 1994 (data available at http://www.opl. speed or ocean water temperature on the ucsb.edu/btm.html). Sensor depths here lunar cycle between the years in the first and have shifted over time and this analysis will second halves of the study period (1950- focus on the four depths that offer the 1978, 1979-2007 for cyclones) using simple greatest number of observations since the correlation; correlated patterns were mooring began: 3, 35, 70, and 150 m. The considered unlikely to be spurious. For this sensor depths used actually varied around purpose, each synodic cycle was partitioned each of these targets by as much as 1m (for into 30 equal “days” (cf. mean length of 29.53 the 3m depth), 5m (for 35m and 70m), or 9m d) and the frequency of hurricane (for 150m). When more than one sensor was observations, mean wind speed, or mean available at a given depth, only one was water temperature was calculated for each used- generally the Tidbit sensor because of