Hurricanes and Typhoons - Will tropical cyclones become stronger and more frequent? [Past]

Figure 1: A) shows a 4500-year record of hurricane storm surges at Mullet Pond, Florida. Blue and red dots represent chronologies of small and large storm deposits as defined by the dashed blue and red threshold lines, respectively. The orange curve is the ratio of intense to total activity found by applying a 157-year sliding window to the chronologies of discrete events. B) shows a proxy record of El Niño frequency based on lake level inferred from sand content in the crater lake El Junco in the Galapágos (Conroy et al. 2008). C) shows a time series of foraminiferal δ18O (sea surface density) from the Bermuda Rise, inferred SST is shown on the y-axis, though a portion (estimated to be about one third) of the variability in δ18O is thought to be related to changes in salinity also (Keigwin 1996).

Paleohurricane reconstructions extend storm records further into the past to improve our understanding of the relationship between tropical cyclones and climate. Though several types of tropical cyclone proxies are under development, sediment-based records, which can span millennia, have thus far provided the longest storm reconstructions and have revealed the coarse centennial to millennial-scale features of hurricane climate (e.g. Donnelly and Woodruff 2007). New, high-resolution sediment records developed from coastal ponds along the Northeastern Gulf of Mexico and in the Northeastern USA document statistically-significant changes in storm activity in response to the modest climate variations of the late Holocene (Fig. 1A). These records provide evidence both for intervals with significantly elevated and depressed storm activity relative to the historic, instrumental period. The largest variability in these paleohurricane records occurs on multi-centennial and millennial timescales, which suggests that Atlantic hurricane activity is poorly constrained by the relatively short instrumental record.

Late Holocene variations in storm activity have been dominated by changes in the frequency of intense hurricanes rather than the overall number of landfalling tropical cyclones (e.g. Lane et al. 2011). A comparison between a 4500-year storm surge record from the Florida Panhandle (Fig. 1A) and reconstructions of SSTs and Loop Current migration within the northeastern Gulf (Richey et al. 2007) suggests that intense storms were most frequent in the region not when Gulf SSTs were warmest but rather when the high ocean heat content of the Loop Current was closest to the study site. Future, intense hurricane activity may similarly respond more sensitively to upper ocean thermal structure rather than SST. Larger-scale factors also may have driven basin-scale variability in Atlantic hurricane intensities, with more (less) intense events occurring more often during periods of reduced (increased) ENSO variability (Conroy et al. 2008; Fig. 1B) and warmer (cooler) SSTs in the western North Atlantic (Keigwin 1996; Fig. 1C). This is consistent with the idea that the relative warmth of the tropical North Atlantic may be a good, aggregate indicator of Atlantic hurricane activity on greater than inter-annual timescales.

Given the stochastic nature of hurricane landfalls at a given location, any trend in basin-wide hurricane activity during the late 20th century would not be detectable in a single-site paleohurricane record. Further, given the possible disconnect between landfalling and basin-wide activity as well as high-frequency regional variability in the occurrence of landfalling storms, multi-site compilations of paleohurricane records may also fail to capture centennial or shorter scale trends or variability. However, on long timescales, North Atlantic paleohurricane records are fairly coherent revealing multi-centennial to millennial-scale intervals with either frequent or few intense hurricanes.

Though the Earth’s climate state at the end of the 21st century may lack a Holocene analogue, hurricane proxies remain illustrative if not predictive. These records demonstrate that the climate system, on its own, can and has given rise to long-lived storm regimes much more active than anything experienced by vulnerable coastal cities and communities along the US Gulf and East Coasts. Paleo records of climate and hurricanes archive data from an experiment conducted in the laboratory of Earth’s climate system, and reproducing the findings of that experiment would improve our understanding of the dynamical controls on hurricane activity. Forcing statistical and dynamical models of tropical cyclone climate with the boundary conditions of past millennia and comparing the results with paleohurricane records may provide a pathway to evaluate the predictive power of these emerging techniques and to identify the climatic causes of both the extremely active and very quiet storm regimes of the late Holocene.

selected references

Full reference list online under: http://pastglobalchanges.org/products/newsletters/ref2012_1.pdf

Conroy JL et al. (2008) Quaternary Science Reviews 27: 1166-1180

Donnelly JP and Woodruff JD (2007) Nature 447: 465-468

Keigwin LD (1996) Science 274(5292): 1504-1508

Lane P, Donnelly JP, Woodruff JD and Hawkes AD (2011) Marine Geology 287: 14-30

Richey JN, Poore RZ, Flower BP and Quinn TM (2007) Geology 35(5): 423-426