Sedimentary evidence indicates that the control the BH currently exerts over TCs operates over longer scales as well. Liu and Fearn (2000) have identified a period of hyperactivity on the northern Gulf coast (Louisiana-Florida), for the period 3400-1000 14C yr BP., with periods of reduced activity before and after. They suggest millennial-scale positional movement of the BH as the proximate cause of this oscillation in frequency of landfall, with positions to the southwest funneling storms into the Gulf of Mexico and positions to the northeast pushing the tracks along the Atlantic Coast. This implies an anti-phase relationship between the frequency of strike activity between the Gulf and Atlantic Coasts, supported by results from an Atlantic Coast paleotempestology study (Scott et al. 2003). The timing of the posited movements of the BH is supported by proxy paleoenviron-mental data (Hodell et al. 1991), based on the premise that long-term residency of the BH over an area results in increased aridity.
The long-term (centennial to millennial-scale) relationship between these two features may be more direct than the short-term, with the longer-term average annual latitude of both less affected by "noisy" short-term conditions. For both features the principal latitudinal control seems to be the pole-equator temperature gradient, with a steeper gradient resulting in a southern movement. For the BH this relationship is demonstrated by Fig. 5 (modified from Flohn 1984), based on the
monthly average positions of the subtropical highs and the pole-equator temperature difference at the 300/700 mb layer. Since this gradient is primarily dependent upon polar temperatures (there being less variation in equatorial temperatures), warm (cold) periods tend to move the BH to the north (south). Flohn (1984) estimates that an increase in the average annual Arctic temperature of 7°C moves the average BH latitude 100-200 km northward in summer and 800 km northward in winter. Empirical evidence for this relationship also exists; as early as Lamb's (1977) estimation of a paleolatitudinal record for the BH based on palynological and vegetational boundary displacement and marine microfaunal analysis evidence found significant northward (southward) shifts paralleling overall hemispheric heating (cooling).
Much evidence, on a variety of time scales, supports temperature-driven movement of the ITCZ. Modeling studies simulating polar ice cover at the Last Glacial Maximum (LGM) support a southern movement of the ITCZ (Chiang et al. 2003; Chiang and Bitz 2005), with the increased pole-equator temperature gradient resulting in a 6° southern displacement of the ITCZ (Broccoli et al. 2006). Similar long-term temperature-driven movement of the ITCZ has occurred in both the equatorial Indian Ocean (Tiwari et al. 2006 and the eastern Pacific (Koutavasa and Lynch-Stieglitz 2004). Paleoclimatic evidence includes shifts in South American precipitation paralleling ITCZ movement, as recorded by Andean ice cores (Thompson et al. 2000), spleothems and travertine deposits in northeastern Brazil (Wang et al. 2004), and riverine discharge (Peterson et al. 2000; Haug et al. 2001). Drought records from the western United States support northward migration of the BH during the Medieval Warm Period (Seager et al., 2007).
In addition to the average long-term latitude of the two features being controlled by a single primary factor, Flohn (1984) suggests a more direct physical connection, arguing that the northern displacement of the BH resulting from a 7°C increase of average annual north polar temperature alone is enough to move the ITCZ 3-4° northward. Theoretically, therefore, coordination of the low frequency movements of the two features seems likely.
Coordinated movement between the ITCZ and the subtropical highs implies antiphase rainfall anomalies across the ITCZ; i.e. if the subtropical high moves in parallel with the ITCZ, southerly migration of the ITCZ resulting in positive Amazonian rainfall anomalies should correlate with negative rainfall anomalies for the Caribbean, which would be increasingly influenced by the dry subsiding air associated with the BH. Numerous studies support this relationship during the instrumental record for the tropical Atlantic for both the Caribbean-Central American region and the Sahel to the north and northeastern Brazil to the south of the ITCZ (Hastenrath 1976, 1985, 2000a,b; Lamb 1978; Kapala et al. 1998; Curtis and Hastenrath 1999). On both sides of the ITCZ, movement of the ITCZ toward (away from) the location tends to result in positive (negative) rainfall anomalies on an interannual basis, indicating coordinated movement between the subtropical highs and the ITCZ. Paleoenvironmental evidence (Baker et al. 2001; Mayle et al. 2000, Maslin and Burns 2000, Poore et al. 2003; Tedesco and Thunell 2003) for such coupling has also been demonstrated. Koutavasa and Lynch-Stieglitz (2004) include a review of a large number of studies, based on several different proxies supporting anti-phase precipitation anomalies across the ITCZ. Since trade wind-driven up-welling provides direct evidence for the proximity of the subtropical high, anti-phase rainfall and upwelling records from the Caricao basin provides strong support for parallel movement between the ITCZ and the BH (Haug et al. 2001).
Marine cores from the coast of Venezuela (Haug et al. 2001) indicate that, driven by latitudinal movement of the ITCZ, the region has been alternately subject to either ITCZ-induced rainfall or BH-driven trade winds for the last 14,000 years. This suggests that the TC zone, locked in at the northern edge of the ITCZ, south of the zone of intense trade winds and upwelling, has experienced a parallel migration. Based on sedimentary evidence from Saint-Martin in the French West Indies, Bertran et al. (2004) have suggested a millennial-scale ITCZ-driven latitudinal movement in the zone of TC activity.
The influence of El Nino-Southern Oscillation (ENSO) on NA TC activity is well known, with El Nino (La Nina) periods reducing (increasing) overall activity, with some regional variation (Gray 1984; Richards and O'Brien 1996; Bove et al. 1998; Elsner and Kara 1999: Pielke and Landsea 1998; Bengtsson 2001; Tartaglione et al. 2003). The Quasi-Biennial Oscillation (QBO) also influences NA TC, with the westerly phase corresponding to increases in frequency of both TCs and major hurricanes (Gray and Shaeffer 1991; Elsner et al. 1999; Elsner and Kara 1999; Landsea et al. 1999; Goldenberg et al. 2001). In this paper, however, we ignore both cycles, as in effect, both of these high-frequency oscillations become noise superimposed on the underlying system at the time scales (centennial to millennial) under consideration. The only exception is the possibility of significant long-term changes in ENSO frequency (Haug et al. 2001; Koutavas and Olive 2006), which potentially could affect TC frequency on the time scales of interest.
Was this article helpful?