Climatology of Tropical Cyclones in the Caribbean Region and Analysis of Storm Characteristics and Their Relationship to ENSO

The majority of tropical cyclone studies focus on storms that directly impact the US mainland; however, this scope is limiting in that it overlooks more vulnerable areas such as the Caribbean Region. Therefore, this study uses the HURDAT2 dataset to create a climatology of tropical cyclones that have impacted the Caribbean Region from 1945 to 2017. In addition to the full period of analysis, we decided to establish a second time period from 1976 to 2015, also referred to below as the Satellite Era. During the hurricane seasons in which satellites have been operational, the data is believed to be more complete and accurate, whereas before their use, there might have been storms that were completely missed along with other storms that were incorrectly categorized based on intensity. The purpose of the Satellite Era is to be used for comparisons with the full climatology to establish a means of checking trends found in the full climatology, where these data errors would be present. As another means of attempting to account for data errors, we checked our list of landfalls against the list of landfalls provided by the Hurricane Re-Analysis Project. Furthermore, by looking at aspects such as storm origin, maximum intensity, category at landfall, and El Niño Southern Oscillations (ENSO) conditions at the time of each storm, we were able to identify possible relations between such factors. It is important to note that the storms included in this study range from tropical storms to category five hurricanes. Additionally, we decided to include storms that did not make direct landfall but approached our area of interest. These cases are included in analyses of “impacts”, yet not in analyses of “landfalls”.

Upon looking at trends seen in the full period and in the Satellite Era, there were cases in which both periods suggest similar trends, along with other cases suggesting opposing trends. For example, both time periods indicated an overall increase in the number of storms that impacted the Caribbean Region over the course of their respective time frames. However, when looking at the total number of minor hurricanes (categories one and two) that made landfall in the Bahamas per year, the full period suggested a decreasing trend while the Satellite Era suggested an increasing trend. The reasons for such occurrences are outside the scope of this paper; however, one can speculate it could be related to potentially missed storms before the use of satellites.

Origin analyses allowed us to identify whether there were trends in the locations of storms forming based on where they made landfall and in what month they made landfall. For example, an analysis showed that the vast majority of storms that impacted the Lesser Antilles formed over the Atlantic Ocean. When looking at the origins based on the month in which the storms made landfall, there was a shift in origin concentrations throughout peak months. In August and September, most tropical cyclones formed over the Atlantic Ocean; yet in October, the concentration was more over the Western Caribbean Sea. In addition to this shift in location, the intensities of the storms seemed to increase from August to September and decrease from September to October.

Lastly, by analyzing storms based on the ENSO conditions under which they formed, it became clear that less storms form under El Niño conditions. However, there was nothing to support that these same storms forming under El Niño conditions were weaker in magnitude than those forming under neutral or La Niña conditions.