Homogeneous record of Atlantic hurricane surge SEE COMMENTARY threat since 1923 Aslak Grinsteda,1,2, John C. Moorea,b,c, and Svetlana Jevrejevaa,d aState Key Laboratory of Earth Surface Processes and Resource Ecology, College of Global Change and Earth System Science, Beijing Normal University, Beijing 100875, China; bArctic Centre, University of Lapland, FI-96101 Rovaniemi, Finland; cDepartment of Earth Sciences, Uppsala University, SE-75236 Uppsala, Sweden; and dNational Oceanography Centre, Liverpool L3 5DA, United Kingdom Edited by Kerry A. Emanuel, Massachusetts Institute of Technology, Cambridge, MA, and approved August 31, 2012 (received for review June 7, 2012) Detection and attribution of past changes in cyclone activity are index reflects economic damage. Rather than a simple number hampered by biased cyclone records due to changes in observa- count of cyclones, we produce a yearly probability distribution tional capabilities. Here we construct an independent record of of storm surge intensity. We then apply a robust method of es- Atlantic tropical cyclone activity on the basis of storm surge statistics timating confidence intervals to the frequency of extreme events. from tide gauges. We demonstrate that the major events in our Finally we show that there is a difference in frequency of cy- surge index record can be attributed to landfalling tropical cyclo- clones between cold and warm years and that the effect is nes; these events also correspond with the most economically strongest for the larger cyclones and hurricanes. damaging Atlantic cyclones. We find that warm years in general were more active in all cyclone size ranges than cold years. The Results largest cyclones are most affected by warmer conditions and we We wish to produce a long-term, homogeneous record of storm detect a statistically significant trend in the frequency of large surge surge activity. Tide gauges are very suitable as they are simple events (roughly corresponding to tropical storm size) since 1923. In devices that have been used for hundreds of years to measure sea particular, we estimate that Katrina-magnitude events have been level. We define the region of interest to be the western Atlantic twice as frequent in warm years compared with cold years (P < 0.02). between 10°N and 40°N. This leaves us with the six tide gauges from the region of interest (Fig. 1, Inset and Fig. S1)withthe climate | extreme | hazard | risk | flood main criteria that we wanted to construct a homogeneous record that covered the great 1926 storm surge (13) and the general high he relationship between global warming and Atlantic hurricane cyclone activity of the 1930s. Here, we use data from the Research Tactivity is a controversial topic (1–3). Some have linked cyclone Quality Data Set (RQDS) (14). The RQDS records are extended activity to sea surface temperatures in the cyclogenesis region (2, to the present using fast delivery data from the global sea level 4). Other competing hypotheses include teleconnections with El observing system (14) and in a single instance (Mayport, FL) Niño–Southern Oscillation (5), North Atlantic Oscillation (6–8), using preliminary water-level data from the National Oceanic Atlantic Multidecadal Oscillation (6), tropical temperatures (3, 9), and Atmospheric Administration (NOAA) Center for Operational and Sahel drought (10). Whereas others note that bias in the ob- Oceanographic Products and Services (15). We manually screen servational record casts doubt on any statistical power from the the data quality of the Mayport preliminary water-level data before down-sampling. We then proceed to filter these records to enhance relationship (11). Estimating and correcting the historical bias the storm signal while reducing the signals due to instrument changes, hamper assessment of the links between tropical cyclone activity harbor development, and erroneous time shifts in the records. and climate change (1, 2, 11, 12). Hence, any discussion of ob- Tropical cyclones are highly localized. However, over time the servational links or causality between global mean temperatures sea-level disturbance will dissipate. Daily averages increase the and hurricane impacts relies on an unbiased estimate of hurricanes storm footprint to hundreds of kilometers, which means that as a function of time. There is a rising trend over the 20th century in relatively few tide-gauge stations provide adequate coverage. Large the observed numbers of Atlantic tropical cyclones (1, 12). How- storms can also produce extreme sea levels that can be seen in tide- ever, observational methods have improved over time, especially gauge records for several days. The potential energy stored in a sea- since the satellite era, but also after airborne observations became level perturbation is related to the square of the vertical displace- commonplace; therefore some cyclones were missed in the past. ment of the sea surface (16); hence we use squared day-to-day Most efforts have focused on estimating total Atlantic cyclone difference in local sea level. Daily data are insensitive to harbor activity rather than the number of land-falling storms. This is be- development and changes in measurement methods, which can cause relatively few storms make land, and small changes in storm strongly affect high-frequency variability such as significant wave tracks can make a difference between a landfall and a near miss. height. Day-to-day differencing further minimizes tidal influence However, from the economic damage perspective the hurricanes and slowly varying trends from, e.g., rising global sea levels (17). that remain far away from shore in the Atlantic are much less im- We observe that summer sea level is relatively calm except for portant than those closer to land. Hence in constructing an unbiased the sporadic and obvious influence from cyclones, whereas win- record of storms we need to ask what we want to measure. The ters have a higher degree of background variability. We therefore strong winds and intense low pressure associated with tropical cyclones generate storm surges. These storm surges are the most harmful aspect of tropical cyclones in the current climate (1, 12), Author contributions: A.G. designed research; A.G. performed research; A.G. analyzed and wherever tropical cyclones prevail they are the primary cause data; and A.G., J.C.M., and S.J. wrote the paper. of storm surges. A measure of storm surge intensity would therefore The authors declare no conflict of interest. be a good candidate measure of cyclone potential impact. This article is a PNAS Direct Submission. In this paper we construct such a record, using long-term tide- See Commentary on page 19513. gauge records from stations that have been operational for much 1To whom correspondence should be addressed. E-mail: [email protected]. longer than the satellite era. These provide a consistent dataset 2Present address: Centre for Ice and Climate, Niels Bohr Institute, University of Copenha- for examining hurricanes affecting the southeastern United gen, DK-2100 Copenhagen, Denmark. SCIENCES States. We also show that the index is actually dominated by This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. ENVIRONMENTAL land-falling hurricanes rather than winter storms and that the 1073/pnas.1209542109/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1209542109 PNAS | November 27, 2012 | vol. 109 | no. 48 | 19601–19605 Downloaded by guest on September 28, 2021 A (which we here define as July–November) and the second way is 5 a measure of the annual frequency of extreme events above a 4 threshold corresponding to five to six events per year. 3 Discussion In this section we demonstrate that the surge index primarily 2 records surge threat from tropical cyclones. We compare the new Surge Index (Jul−Nov) 1 surge index with several common measures of tropical cyclone B ) activity, most of them derived from the Atlantic Hurricane Da- −1 15 tabase (HURDAT; ref. 18) with wind speed corrections as de- tailed in ref. 2: tropical cyclone counts for various magnitude 10 thresholds on the Saffir–Simpson scale, accumulated cyclone energy (ACE) (19), net tropical cyclone activity (NTC) (20), and 5 power dissipation index (PDI). We also calculate several of these Large events (yr measures restricted to storms making US landfall. These we de- 0 note with a US prefix (e.g., US-ACE). Additionally, we compare C ) 150 2 the surge index with the “PL05” normalized hurricane damage kt 4 (NHD) from Pielke et al. (21). 100 As an example we show the daily surge index compared with ACE and US-ACE for the very active 2005 season (Fig. 2). The 50 ACE is traditionally defined as an annually integrated value. Here, US−ACE (10 we have made the natural extension of this definition to a daily in- 0 tegrated value. It is clear that the surge index is particularly sensitive D 1930 1940 1950 1960 1970 1980 1990 2000 0.6 to hurricanes making landfall, which supports the interpretation 0.4 that the surge index is closely related to the actual threat. Table 1 shows the correlation of the seasonal surge index with 0.2 other measures. The surge index is positively correlated with all T (K) 0 of the comparison measures. The best correlations are found −0.2 with measures that emphasize intensity (e.g., NTC, ACE, and −0.4 PDI) and measures that are restricted to US land-falling storms 1930 1940 1950 1960 1970 1980 1990 2000 only. Table 1 also shows that low-frequency correlation tends to Year be at a higher level than the year-to-year correlation. This is to be expected given that there are low-frequency driving agents Fig. 1. (A) Average surge index over the cyclone season. (B) Observed fre- related to various climate forcings (4, 8). One notable exception quency of surge events with surge index greater than 10 units/y (surge index > is NHD, which shows poor low-frequency correlations.