Proceedings of the National Conference On Undergraduate Research (NCUR) 2006 The University of North Carolina at Asheville Asheville, North Carolina April 6 8, 2006 Tropical Cyclone Hyperactivity in the Eastern and Central Caribbean Sea During the 2005 Atlantic Hurricane Season Amy Harless Department of Atmospheric Sciences University of North Carolina at Asheville One University Heights Asheville, NC 28804. USA Faculty Advisor: Dr. Christopher Hennon Abstract The 2005 hurricane season produced an unprecedented number of hurricanes in the Atlantic Ocean. With twentyseven named storms, the pre-determined list of storm names was exhausted; and for the first time since recordkeeping, an auxiliary list was implemented. In addition, a second record was broken when Hurricane Wilma s minimum pressure hit 882 millibars, exceeding the record low of Hurricane Gilbert (1988). One aspect of the overall hyperactivity was the elevated number of tropical cyclone formations in the eastern and central Caribbean Sea. This research focuses on the tropical cyclone activity in this area and examines the reasons contributing to this increased frequency. The North Atlantic covers a vast amount of area, and different regions of this area have distinctive atmospheric conditions that make the creation and development of hurricanes more favorable. Using reanalysis data from the National Centers for Environmental Prediction (NCEP)-National Center for Atmospheric Research (NCAR), results are presented that examine these specific conditions. The driving forces investigated are mean sea level pressure, sea surface temperatures, specific humidity, and relative vorticity. Specifically the months in which cyclogenesis occurred in the area of study are examined in this research. The uniqueness of this season, which made this region more conducive to tropical storm formation and intensification, was compared to the past climatology of this area of focus. Throughout this research, differences between the 2005 monthly averages and the 1985-2004 monthly averages were obtained and used to determine the reason for a hyperactive hurricane season in this region of the Caribbean Sea. The data and calculations revealed that during the months in which cyclogenesis occurred during 2005, the environment was more favorable for tropical storm formation. Specifically, low sea level pressure, warm sea surface temperatures, and greater values for relative vorticity and moisture content were found. The data partially explains why there was an increase in tropical storm activity in the region being studied compared to other hurricane seasons dating back to 1985. 1. Introduction June 1, 2005 was the beginning of a historic hurricane season. Over the Atlantic basin during the following seven months, twenty-seven named storms formed, fifteen of which were hurricanes, and seven of these a category three or higher on the Saffir-Simpson Scale. This intensity scale, based on wind speed of the storm, ranges from a category one to category five in which a category five is the most intense. Specifically, compared to the past twenty years, the eastern and central Caribbean Sea was an active region as well. Dating back to 1985, a total of ten storms originated in the region between latitudes 10 and 20 North and between longitudes 60 and 70 West. This number is approximately 3.4 % of the total number of tropical cyclogenesis events in the Atlantic Ocean. Three of these storms, Hurricane Dennis and Tropical Storms Alpha and Gamma, were just from the 2005 hurricane season accounting for near 10% of the 2005 cyclogenesis events. This is three times the climatological value for this area.
Hurricane Dennis formed on July 5 th in the southwestern region of the Lesser Antilles. During its eight-day track, Dennis propagated northwestward cutting across northwest Cuba, moving into the Gulf of Mexico, and then reaching the contiguous United States at the Florida panhandle. On October 22, Tropical Storm Alpha formed just south of the Dominican Republic and moved northwestward across Haiti. Taking on a more north-north-easterly course, the storm crossed the Bahamas and moved farther into the north Atlantic where it weakened to below tropical depression status. Tropical Storm Gamma also formed near the Lesser Antilles. Forming on the 14 th of November, Gamma moved westward, remaining south of the Caribbean islands. It weakened below tropical depression status south of Haiti on the 16 th of November and reformed near Central America two days later. Table 1 provides the genesis points of these three storms. The following section will discuss the data used and methodology followed out during this study. Section 3 contains the results of the study followed by the discussion and conclusions in section 4. Finally, future work on this project will be discussed in section 5. Table 1. genesis events in the area of study 1 Storm mm/dd/time Latitude Longitude (UTC) ( N) ( W) Dennis 07/04/1800 12.5 63.1 Alpha 10/22/1200 16.1 68.0 Gamma 11/14/0000 13.5 62.7 2. Data and Methodology Genesis location points for the storms in the Atlantic Ocean from 1985 through 2005 were obtained from the National Hurricane Center. Figure 1 is a map of the Atlantic Ocean with all of the genesis points for the storms during this 21-year span. Storms developing during 1985-2004 are in blue; the 2005 genesis points are in red. This map illustrates where majority of storms develop on average compared to where they formed during the 2005 season. Figure 2 is a closer view of the 10 x 10 area that is under examination in this research. This map shows that through the 20-year period prior to 2005, only seven storms originated in this area. During the 2005 season, three storms formed here. The atmospheric data retrieved for this study include sea level pressure (SLP), specific humidity, meridional wind speed, and zonal wind speed data from the NCEP-NCAR Reanalysis 2. NCEP-NCAR Reanalysis is the product of data assimilation of numerical models and observations around the world from1948 to the present. Taking a subset of the data fields, monthly means were calculated. Sea surface temperature (SST) was a parameter also examined. The monthly SST data fields were derived through optimum interpolation (OI) analysis. The OI version 2, described in Reynolds et al. 3, is the improved version of the OI version 1 SST analysis. Monthly averages for SLP, specific humidity, SST, and wind speeds (used to calculate relative vorticity) were calculated for the years 1985 through 2004. Monthly averages of the same parameters were calculated for the months of 2005. After taking the averages of each of these parameters, the difference between the 1985-2004 values and 2005 values were calculated. This process allowed for any anomalies in sea level pressure, sea surface temperatures, humidity levels, and vorticity to be seen. Figure 1. View of genesis points for 2005 (red) and for 1985-2004 (blue).
Figure 2. Closer view of genesis points within region of study. 3. Results 3.1 sea surface temperatures Warm SSTs are a necessary ingredient for the formation of tropical cyclones as they are a proxy for the energy available to the system. An environment with SSTs of 26 C or more is critical in providing the power to fuel to storm. The storm will use the sensible heat and water vapor from the warm ocean to enable cumulonimbus convection. During the months of genesis examined in this study (July, October, and November), temperatures were well above this requirement with SSTs in all three months of 29 C or higher. SSTs during July (shown in Figure 3 A), when Dennis formed, were 0.6-0.9 C warmer than the average for this month. Similarly, October s SST anomalies for this region were about 1.8 C. Finally, November SSTs were near 0.8 warmer than the 20-year average for this month. a. b. c. Figure 3. Sea surface temperature differences in a) July b) October, and c) November.
3.2 sea level pressure SLP is another atmospheric parameter that is necessary for the formation and maintenance of tropical cyclones. In a study conducted by Knaff 4, an analysis on the inverse relationship between hurricane activity in the tropical Atlantic and the SLP was performed. The research was centered around the question as to why only slight anomalies in sea level pressure have a significant effect on the development of tropical cyclones. Results of Knaff s research 4 yielded that lower SLP values and greater sea level pressure anomalies heavily influence the tropical cyclone activity in different regions. This is due in part that higher pressure allows for an increase in vertical wind shear (a cyclogenesis inhibitor) especially near the Caribbean, as the westerlies become more significant at the 200 mb level. High-pressure areas also decrease the midlevel moisture content, which in turn allows for more radiative cooling and less cumulonimbus convection. Looking at the results of this research and comparing the 2005 season to the climatology for 1985 to 2004, all three months had lower values of SLP during this past season. In the area that cyclogenesis occurred, there was a difference in about 0.5 to 1 mb in pressure with the 2005 season having lower values. Hurricane Dennis formed in the southwestern region of the Lesser Antilles during the month of July. During this month in this region, the average SLP was 1014-1014.5 mb which was approximately 0.5 mb less than the 20-year average for this particular area. Dennis formed in a region wedged between an area of lower pressure values around 1011 mb to the southwest and high pressure values around 1025 mb to the northeast. Figure 4a illustrates the sea level pressure differences surrounding the location of Dennis formation during July. In October, the area in which Alpha formed had an average sea level pressure near 1015 mb. Similar to the environment of Dennis, low pressure values of 1009 mb were located to the west of the genesis location and high pressure values centered at 1020 mb were to the northwest of where Alpha formed. In this area, the SLP was approximately 1 mb less than the 20-year average (Figure 4b). The lowest pressure values were significantly less (about 4 mb) than the climatology values. Gamma formed in mid-november in an area where the monthly average SLP was also approximately 1 mb lower than the 20-year average (Figure 4c). The average value for this area was around 1011 mb. Northwest of gamma is a region in which values were up to 3.5 mb lower than the climatology value. Just north of this, SLP differences increase rapidly in the positive direction approaching a positive difference of 5 mb. a. b. c. Figure 4. SLP differences (mb) for a) July, b) October, and c) November. Genesis points indicated by red stars.
3.3 vertical wind shear and relative vorticity Past studies 5,6,7,8 have indicated that low-level relative vorticity with convergence along with upper-level negative relative vorticity and divergence is a contributor to the formation of tropical cyclones. In a study done by Gray 5, he concluded that low levels of vertical wind shear along with large values of low level vorticity are pertinent for the formation of tropical cyclones. Gray concluded in another study 6 that the absence of strong vertical wind shear will allow the moisture and heat to build up into a vertical system allowing the formation of a tropical cyclone to proceed. McBride and Zehr 7 concluded that there are certain primary factors that distinguish an environment conducive to hurricane formation from one that is not. In place must be an area of high values of low level relative vorticity, zero vertical wind shear close to the center of the developing storm, the need for large positive zonal shear to the north with negative zonal shear to the south of the developing storm, and northerly shear east of the system and southerly shear to the west. McBride and Zehr developed an equation that quantified these key parameters by looking at the vertical gradient component of the relative vorticity at the 900 mb and 200 mb levels. To derive the daily genesis potential (DGP), the relative vorticity at 200 mb is subtracted from the relative vorticity at 900mb. ζ 900 mb - ζ 200 mb = DGP (1) Areas for greater potential are characterized by larger values, which result from large negative values of vorticity at 200 mb (anti-cyclonic) and large values of positive relative vorticity values at 900 mb (cyclonic). In this research, calculations were made using the wind data at the 850 mb level because the reanalysis datasets did not contain wind data at the 900 mb level. McBride and Zehr used this quantity when studying developing and non-developing cloud clusters. They concluded that DGP values evaluated at 6 radii was the best for composite fields. For this particular research, the scope at this parameter is bound within a 10 x 10 box. In research done by Hennon and Hobgood 8 on forecasting cyclone genesis in the Atlantic basin, it was concluded that DGP was a significant indicator of possible genesis. The units of DGP are 10-5 s -1. For July, the DGP value where Dennis formed was around 0.8. Greater values of DGP were located to the east and west of region of cyclogenesis, 2.4 and 3.2, respectively, and lower values were located to the north and south of the area, -2.4 and -1.4, respectively. Compared to the 20-year average for July, the area of where Dennis formed had a DGP value of approximately 0.5 greater during the 2005 season (Figure 5 A). To the east and west of this area, values were 1 to 1.5 x 10-5 s -1 greater for 2005. Areas to the north showed little to no difference in DGP values while to the South, DGP values for July of 2005 were approximately 1.5 x 10-5 s -1 lower than the 20-year climatology values. The month of October saw a similar trend. The area where tropical storm Alpha formed had an average DGP value of approximately 1 1.5. Higher values of DGP were located to the north-northwest region of the Caribbean. South and east of the formation area had relatively lower values of DGP. Furthermore, the areas of higher values of DGP were in fact up to 2 x 10-5 s -1 higher than the 20-year average for October (Figure 5 B). In the south where values were lower relative to where Alpha formed, these values were about 1.5 x 10-5 s -1 lower than the 20-year climatology values. Gamma developed in a region where the average DGP value for November was roughly 0.6. This value is between relatively higher values to the west and lower values to the east. When examining the difference between November of 2005 and the 20-year average value (Figure 5C), the DGP is about 0.6 x 10-5 s -1 greater during November 2005. 3.4 specific humidity Moisture in the atmosphere is another significant factor in the formation of tropical storms. Moisture (high humidity) allows for strong convection as the water droplets in the air condense releasing latent heat. During this process, the air becomes lighter and continues to rise. Air near the surface moves in to replace the rising air. As an indicator of the moisture content present over the region where tropical cyclone genesis occurred, specific humidity was analyzed. The monthly average for July 2005 at 850 mb near that region in which Hurricane Dennis formed was near 9.5 g/kg. This value is approximately 1.2 g/kg greater than the 20-year average for the month of July.
a. b. c. Figure 5. As in Figure 4 except for Daily Genesis Potential. Compared to the other months in which genesis occurred, this is a significant increase in moisture over this area. During October, the monthly mean was 9.9 g/kg in the region of genesis of tropical storm Alpha. This value is approximately 0.6 g/kg greater than the 20-year average for the October. In the region where Gamma formed in November, specific humidity level was lower than the other two months with an average value of ~9.2. However, this month is similar to October with deviation from the climatology value being approxiametly 0.3 g/kg (figure not shown). 4. Discussion and conclusion The 2005 Atlantic hurricane season set many records. The Caribbean Sea, likewise, had the right ingredients at the right times to invoke tropical cyclogenesis. Data revealed that throughout the 2005 season SSTs were above the threshold that is conducive for cyclone development. Furthermore, SSTs were significantly warmer than the climatogical mean. The moisture content in this region was also greater all three months of genesis compared to the climatology values. At the same time, SLPs were lower than average. These lower pressure values allowed the warm, moist air over the warm seas to rise and condense above the pre-cyclone system. Keeping this warm air from being ventilated, high values of vertical wind shear were absent. In agreement with McBride and Zehr, the genesis parameter values for large-scale vorticity fields quantified into a DGP value were on average 0.5 to 1.0 x 10 5 s -1 greater for the 2005 months than the climatological. This indicated that this region had stronger vorticity fields and lower low-level wind shear. During the months in which genesis occurred, DGP values were greater than the average from 1985-2004. The other months were not characterized by larger values; in fact, the values were either near zero or lower than the average (not shown). Dennis formed in a region in between two areas of high DGP values (east and west of genesis locale) and low DGP values (north and south of the region). In the study done by McBride and Zehr, they found that based on the mean DGP values for their composite data sets, even the weak developing systems had a value of 1.6 x 10-5 s -1. This value is larger than both the July of 2005 and climatology values. One thing to note is that this study uses the 850 mb level to measure vorticity instead of the 900 mb as used by McBride and Zehr. By using wind data closer to the surface, larger values of relative vorticity may have been obtained thus producing a larger
DGP value. However, moisture content, low air pressure, and warm SSTs were favorable to induce genesis. At the time Dennis formed, Hurricane Cindy was swirling in the Gulf of Mexico. Hurricane Cindy may have impacted the DGP values in this area. In further research, the impact of this preexisting storm on the environment surrounding the genesis locale of Dennis will be investigated. October was a very active month for the Caribbean Sea. In comparison to the other months under examination, the environment was more conducive for genesis and intensification as seen with Hurricane Wilma. The DGP numbers show that high values of low-level vorticity were present. The average sea level pressure was near 1011mb and specific humidity was around 10.0 g/kg. During the time that Alpha was forming, Wilma was northwest of the region pumping in warm, moist air from the gulf. It is possible that Wilma enabled favorable environment to sustain Alpha s formation. November conditions looked similar to July in that DGP values were lower than the developing cloud clusters in the study by McBride and Zehr--but higher than the climatology values. One thing different during this month and the formation of Gamma is that, unlike when Dennis and Alpha formed, there is not a storm already spinning in and around the Caribbean Sea or the Gulf of Mexico. In agreement with Knaff 4, the results show that the 2005 SLP values were lower than the climatogical values. The results presented confirm that there is a possible correlation between the 1 to 0.5 mb SLP difference, sea surface temperatures, and relative vorticity and low vertical wind shear (synthesized into the DGP value). In all three cases, as the SLP difference became more negative over the area in which genesis occurred, the DGP difference became greater. However, this trend is not similar across the entire Atlantic. October stands apart from July and November in that positive DGP values, negative sea level pressure differences, and positive sea surface temperature differences were greater over the area of genesis than both of the other months. Alpha formed in the northwestern corner of the studied area while both Gamma and Dennis formed in the southeastern region. When Alpha formed, Wilma was on a track just northwest of Alpha. The rapid intensification and strength of Wilma indicates the favorable environment for cyclogenesis. 5. Further Research Further work will involve a statistical analysis of the results presented here. Correlation factors, as done in Knaff s research, will be used to examine how closely related the parameters examined in the research are. The interaction between preexisting storms and the storms that formed in the area of study will be examined further. Knaff s research included a study as to how preexisting storms play a role on sea level pressure anomalies. He concluded that on a long-term basis, preexisting storms have no effect on the anomalies. It would be beneficial to continue along these lines by studying the situations that were present during and and following genesis events on both longterm and short-term timescales. 6. References 1. NOAA-NHC-TPC, 2005 Tropical Cyclone Archive, http://www.nhc.noaa.gov/archive/2005/index.shtml. 2. NOAA-CIRES ESRL/PSD Climate Diagnostics Branch The NCEP/NCAR Reanalysis Project http://www.cdc.noaa.gov/. 3. Richard W. Reynolds, Nick A. Rayner, Thomas M. Smith, Diane C. Stokes, and Wanqiu Wang, An Improved In Situ and Satellite SST Analysis for Climate, Journal of Climate 15 (July 2002): 1609-1625. 4. John A. Knaff, Implications of Summertime Sea Level Pressure Anomalies in the Tropical Atlantic Region, Journal of Climate 10 (April 1997): 789-804. 5. W.M. Gray, Tropical Cyclone Genesis in the Western North Pacific, Environment Prediction Research Facility Tech. Paper (1975): 16-75, 66. 6. W.M. Gray, Global View of the Origin of Tropical Disturbances and Storms, Monthly Weather Review 96: 669-700. 7. John L. McBride and Raymond Zehr, Observational Analysis of Tropical Cyclone Formation. Part II: Comparison of Non-Developing versus Developing Systems, Journal of the Atmospheric Sciences 38 (June 1981):1141-1151. 8. Christopher C. Hennon and Jay S. Hobgood, Forecasting Tropical Cyclogenesis Over the Atlantic Basin Using Large-Scale Data, Monthly Weather Review 131 (December 2003): 2927-2940.