Terrence A. McCloskey and Jason T. Knowles
Abstract This paper proposes that a combination of short and long term atmospheric oscillations have resulted in latitudinal movement of the tropical cyclone (TC) zone and location of landfall through the Holocene. A GIS-based approach demonstrates that currently intensity changes of the Bermuda High (BH) result in a large latitudinal spread of TC track and landfall location across the western North Atlantic (NA), while a literature-based examination of paleoclimatic evidence supports the view that long-term changes in the pole-equator temperature gradient has resulted in significant latitudinal migration of the general NA atmospheric system throughout the Holocene, with a heightened (reduced) gradient moving the entire system southward (northward).
Our model suggests that the location of hurricane landfall since the mid Holo-cene is controlled by a millennial scale migration of the hurricane zone (paralleling latitudinal movement of the entire system), complicated by the superimposition of a higher frequency variation in track location, (controlled by intensity oscillations). The resulting millennial scale shifts in landfall location of major hurricanes are hindcast, and methods for testing this hypothesis are described.
The damage hurricanes inflicted upon the Caribbean and the United States during the 2004 and 2005 seasons dramatically demonstrate the societal importance of changes in tropical cyclone (TC) tracks and frequencies. The increase in coastal development that occurred during the relatively inactive TC regime that existed during the 1970s, 80s and early 90s has contributed to the mounting property losses and death toll that ensued following the return to a more active TC regime, beginning in 1995 (Pielke and Landsea 1998). Clearly, an increased understanding of the causes of these spatial/temporal oscillations is critical to achieving an effective response to this natural hazard. The proximate cause(s) of these shifts, which occur across a variety of scales, from interannual to millennial (Reading 1990; Walsh and Reading 1991; Liu and Fearn 2000; Elsner et al. 2000), are not
J.B. Elsner and T.H. Jagger (eds.), Hurricanes and Climate Change, 169
doi: 10.1007/978-0-387-09410-6, © Springer Science + Business Media, LLC 2009
well understood. This paper attempts to identify the average latitudinal position of the Intertropical Convergence Zone (ITCZ) as an important primary control over the location of the TC zone.
It should be noted that the correlation between the ITCZ and the TC zone is expected to manifest itself more clearly over longer (millennial) time scales, with the shorter term correspondence masked by "noisy" higher frequency atmospheric oscillations. Connecting the frequency/track pattern shifts to larger, better understood, and more easily tracked features of the general circulation system may lead not only to an improved understanding of the relationship between TC and climate, but also to improved coastal management.
In this paper we first explore the present relationship between various components of the North Atlantic (NA) atmospheric circulation system and NA TCs. We use a GIS-based approach to demonstrate a consistent relationship between circulation features and TCs, identifying TCs as a integrated feature of the larger system. We then use a literature-based approach to examine the structural stability of the NA circulation system over millennial timescales, focusing on the spatial relationship between the Bermuda High (BH) and the ITCZ, and examine evidence for long-term movement of this system. The effects of such movement on major hurricane landfall since the mid Holocene are explored.
Storm track data were downloaded from the National Oceanic and Atmospheric and Administration (NOAA) best-track dataset (HURDAT) (http://hurricane.csc.noaa. gov/hurricanes/index.html) and imported into a geographic information system (GIS). In order to minimize the use of less reliable (pre-aircraft reconnaissance) data, our investigation covers the period 1948-2003. Two North Atlantic Oscillation (NAO) indices were used. The standard index (http://www.cru.uea.ac.uk/cru/ data/nao.htm), which we refer to as NAO, is based on the normalized sea level pressures (SLP) between two fixed locations (normally southwest Iceland and the subtropical eastern NA). A second index, referred to here as NAO-mobile, calculates the NAO index values as the difference in normalized SLP anomalies at the locations of maximum negative correlation between the subtropical and subpolar North Atlantic (Portis et al. 2001). By being normalized both indices are dimen-sionless. Monthly and annual values from both NAO indices were added to the storm vectors database.
Three data sets were then created, 1. "Tropical Cyclones", including all vectors of all storms; 2. "Hurricanes", which included all vectors of all storms whose wind speed exceeded 74 mph at any point during the storm's lifetime, and 3. "Major Hurricanes", which included all vectors of all storms whose wind speed exceeded 111 mph at any point, corresponding to category 3 storms or greater on the Saffir-Simpson scale.
Using geoprocessing techniques the center of each 6 hr storm vector was converted to a point coverage in order to apply kernel density surface interpolation, which is a technique that generalizes individual point locations or events, si, to an entire area and provides density estimates, e" (s), at any location within the study region (Bailey and Gatrell, 1995). For a more detailed description of the methodology and different visualization results, see Knowles and Leitner (2007).
Current Seasonal Variations in the NA Circulation System
As is well known, the NA circulation system basically consists of a series of latitudinally adjacent belts, starting near the equator with the ITCZ, and proceeding poleward through the trade wind belt, the Subtropical High Pressure Ridge, characterized by the Bermuda High (BH), the zone of midlatitude westerlies, the high latitude low pressure belt characterized by the Icelandic Low (IL), and the Polar High. These components exhibit an annual latitudinal migration, following the apparent annual solar movement. In the boreal winter these components drift southward, moving the BH and its zone of subsiding air equatorward (Hastenrath 1966; Sahsamanoglou 1990; Davis et al. 1997; Machel et al. 1998; Portis et al. 2001). In the Caribbean and Central America this results in frequent atmospheric inversions, increased trade wind strength and generally dry conditions, (i.e. the annual December to May dry season) (Hastenrath 1966; Trewartha 1981). Around June, when the ITCZ approaches from the south, the BH and the associated zone of subsiding air moves north out of the Caribbean, resulting in uplift, condensation, precipitation, and the region's annual wet season (Hastenrath 1966; Trewartha 1981; Sahsamanoglou 1990; Davis et al. 1997; Portis et al. 2001).
From May to November TCs form between the ITCZ and the BH, with cyclo-genesis being dependent upon the same general conditions as regular rainfall, in addition to certain additional requirements, such as a threshold sea surface temperatures (SST), low vertical shear, an existing disturbance and high relative humidity in middle troposphere. TCs typically form in a narrow band off the west coast of Africa, with the Main Development Region (MDR), between 10 and 20°N, accounting for 60% of all TC and 85% of major hurricanes (Goldenberg and Shapiro 1996, Goldenberg et al. 2001). TCs then drift westward, spreading latitudi-nally, their track and eventual location of landfall (if any) controlled by a variety of transient meteorological factors. A spatially separate set of TCs forms in the western Caribbean and the Gulf of Mexico. The southern limit of TC activity is determined by a threshold level of vorticity generated by the Coriolis effect. This limit occurs around 8°N (Elsner and Kara 1999), meaning that the zone of TC formation consists of the area lying between 8°N and the subtropical high pressure ridge, although the latitude of track movement and landfall covers a much larger range.
Was this article helpful?