| Rapid increase in the rate of sea-level rise in the southeastern United States from 2011 to 2015
Arnoldo Valle-Levinson, Andrea Dutton, Jonathan B. Martin
University of Florida
This study documents a recent sea-level rise acceleration in the southeastern United States and proposes possible causes for the change. Tide gauge data from the eastern seaboard of the United States show that sea-level rise in the Southeastern United States (SEUSA) from 2011 to 2015 is 6-9 times the global rate of 2-3 mm/yr. This acceleration in sea-level rise followed a shorter-lived sea-level rise in the Mid-Atlantic Bight (MAB) from 2009-2010, which has since decayed. The recent sea-level rise in the SEUSA is between 18 and 20 mm/yr, which, together with the value of the mid 1940s, makes it the greatest 5-year trend that this region has seen. The restriction of the rise to the SEUSA resembles a similar pattern observed in the mid-1940s; however, there are no clear causative factors that account for the acceleration in both the 1940s and the 2010s. Potential drivers of the most recent sea-level change in the SAB include a combination of declining Atlantic Meridional Overturning Circulation, Gulf Stream transport, Florida Current transport, and wind velocities, a positive phase of the Atlantic Multidecadal Oscillation, and increasing regional water temperatures. It is still unclear whether the anomalous upsurge in the SAB will continue or whether it is short-lived and can be explained in the context of intra-decadal and/or multi-decadal variability in sea surface height.
Coastal counties of the United States are home to ~123 million people, who represent close to 39% of the total population (NOAA, 2013). These coastal communities are threatened by transient phenomena such as storms, storm surge, flooding, and increased erosion that capture most of the attention (NRC, 2007). Less consideration is given to the recently persistent threat of rising sea level and its risk to coastal water resources and infrastructure. Rates of sea-level rise vary through time as shown by sea-level reconstructions that span timescales of 10 to105 years (e.g. Lambeck et al., 2014; Rohling et al., 2009) and direct observations based on tide gauges and satellite altimetry (Church et al., 2013).
A recent acceleration in global (or eustatic) rates of sea-level rise has been observed in tide gauge and satellite altimetry data, which show an increase from a global mean rate of ~1.2-1.9 mm/yr between 1901 and 1990 to ~2.8-3.7 mm/yr for 1993 to 2010 (Haigh et al., 2014). But these rates vary spatially over hundreds of kilometers (Sallenger et al., 2012). For instance, tide gauge data show a rise of ~5 mm/yr along the Mid-Atlantic Bight (MAB) of the United States (north of Cape Hatteras) over the past two decades (Boon, 2012; Knight et al., 2005; Goddard et al., 2015), clearly larger than the global mean. These high rates have been attributed to changes in wind forcing, deceleration of the Gulf Stream, and decline in the Atlantic Meridional Overturning Circulation (Ezer et al., 2013; Woodworth et al., 2014; Thompson and Mitchum, 2014). The increased rates have also been related to the Atlantic Multi-decadal Oscillation and slowing of the Florida current (Frankcombe and Dijkstra, 2009; Park and Dusek, 2013).
Several of the above studies have pointed to a particular period of accelerated rise in the MAB, between Massachusetts and North Carolina, from 2009 to 2010 that was 3 to 4 times higher than the global mean rate (e.g., Sallenger et al., 2012; Goddard et al., 2015; Ezer, 2015). Those studies indicated negligible increase rates south of Cape Hatteras. However, cursory inspection at the National Oceanic and Atmospheric Administration (NOAA)’s website of interannual variation of mean sea level on the east Florida shelf (e.g., tidesandcurrents.noaa.gov/sltrends/residual.htm?stnid=8720030) and a recent analysis in southernmost Florida (Park and Sweet, 2015) revealed a marked upward trend in the period 2010-2015. The rapid rise in the MAB appears to have now shifted to the SEUSA, with tide gauges showing rates of ~18 to 20 mm/yr since 2010. The purpose of this study is to examine the magnitude and the geographic extent of the recent sea-level rise acceleration in the SEUSA and to propose possible causes.
Wind mean speeds and gust speeds were compiled from NOAA buoy 41009 from the National Data Buoy Center C station, 20 nautical miles east of Cape Canaveral, in Central Florida, to analyze other atmospheric forcings that could produce sea-level rise change. Sea level data were compiled for the east coast of the US from Florida to Maine (Fig. 1a) from two sea level data repositories: Hawaii Sea Level Center (uhslc.soest.hawaii.edu/) and NOAA’s tide stations (tidesandcurrents.noaa.gov). The Hawaii Sea Level Center provided hourly data from January 1, 1920 to December 31, 2012 (“long period”) at stations in black letters. NOAA provided data from January 1, 1996 to May 1, 2015 (“short period”) at stations in red numbers.. The short period represents the last 19 years, or one full nodal tidal cycle, following the convention adopted by the National Ocean Service to represent the time segment over which tide observations are taken to obtain mean values. The short period also spanned a shorter distance, from Florida to New York, relative to the long period, which extended to the Canada - United States border. A few stations report data beginning at later dates than those mentioned above.
In order to identify recent changes in sea-level rise, data from all stations, and for both the long and short periods, were detrended (linear trend) and filtered with a cosine Lanczos filter (e.g. Thomson and Emery, 2014) centered at 365 days. This filter smoothed out monthly, seasonal and semiannual variations in sea level (e.g. Fig. 1b for the short period at 3 stations in south Florida). Thus, reported values represented accelerations/decelerations rather than absolute elevations. One-year low-pass filtering resulted in loss of one half year at the beginning and end of the time series.
One-year low-pass filtered data of water level were arranged in Hovmöller (or phase) diagrams for most of the eastern seaboard of the United States at uniformly gridded values at intervals of 30 days and 50 km. Phase diagrams display the propagation of signals in space and time, in this case of sea level increase or decrease. These diagrams were constructed with the long and the short periods to identify spatial structures and magnitudes of the interannual variability. One-year filtered hourly data from recording stations were interpolated to this uniform space-time grid with Delaunay triangulations (e.g. Fang and Piegl, 1992; 1993).
One-year filtered time series of the long period, as portrayed in the phase diagram, were then decomposed into Empirical Orthogonal Functions (EOFs). The EOF analysis was performed only on the long period phase diagrams to ensure statistical reliability. These functions depict the spatial structure of sea level variability throughout the eastern United States and the temporal variations, from 1921 to 2011, of those spatial structures. The analysis also sheds light into whether recent observed acceleration in sea-level rise has occurred in the eastern United States during the analysis period.
In addition, linear trends in sea-level rise were determined for 5-year periods throughout the span of the long and short period observations. Five-year rates were determined from the monthly values gridded on the phase diagram. Most recent trends were compared to trends starting 1920, in the context of probability density functions. Also, wind mean speeds and gust speeds were compiled from NOAA buoy 41009 from the National Data Buoy Center C station, 20 nautical miles east of Cape Canaveral, in Central Florida, to analyze other atmospheric forcings that could be producing the recent sea-level rise change.
One-year filtered water levels in southeast Florida at 3 stations from Miami to Trident (located at Cape Canaveral in the middle of the peninsula) display a trend of rising sea levels from 1996 to 2010 of 4-6 mm/yr, which is about 33 to 100% greater than the rate of global mean sea-level rise of ~3 mm/yr over this period (Fig. X). A marked change is observed from 2011 to 2015, with sea-level rise increasing to 25 05 mm/yr (Fig. 1b). This increase is unprecedented for 5-year intervals during at least the last 19 years. Considering the change since 2010, the rate is between 18 and 20 mm/yr. A comparable rate of 10 to 20 mm/yr has been identified for 2 other sites in the Florida Keys in southernmost Florida (Park and Sweet, 2015). Including the 3 previously identified locations with accelerating sea-level rise makes at least five stations in Florida that have shown values of sea-level rise that are 6 to 9 times the global mean value of 2 to 3 mm/yr.
Phase or Hovmöller diagrams of the long and short period time series were generated to determine the spatial extent of this rapid change in the rate of sea-level rise and to assess whether it is anomalous in the context of past measurements of variability in sea level in this region. The Hovmöller diagram of the long period time series shows intervals of depressed and heightened water levels throughout the last 90+ years, relative to the linear trend (Fig. 2a). Intra-decadal variations may be attributed to wind (Woodworth et al., 2014). The highest water levels have occurred in the MAB in 2009-2010, in SEUSA in 1947-1948, and centered around Cape Hatteras in 1973. The long period time series ended in 2012 and fell short in documenting the recent sea-level rise in the SEUSA captured in the short period time series (Fig. 2b).
The Hovmöller diagram of the short period time series (Fig. 2b) accentuates ephemeral differences in sea level between the MAB and the SEUSA (i.e., north and south of Cape Hatteras). These one-year low-pass filtered data of sea level at tide gauges along the eastern U.S. show elevated sea levels during the period 2010-2015, although the nature of the regional response differs to the north and south of Cape Hatteras. A dramatic rise in sea level occurs in the MAB in 2009-2010 (Goddard, 2015; Ezer, 2015) but slowly tapers off over the next several years. In contrast, there is a sustained increase in sea level in the SEUSA after 2011 (Fig. 2b) that translates to a rate of sea-level rise of ~20 mm/yr over the 5-year period from 2011-2015. Such extreme values of sea-level rise in the SEUSA have not been observed at other times in the period starting in 1996 (Fig. 2b), although the long term period suggests a similar rise occurred around 1947 (Fig. 2a). An EOF analysis places these rates in the context of the long period time series.
The spatial structure of EOF mode 1 explains variations in the same direction (filtered sea level increases or decreases) throughout the eastern seaboard of the United States between 1920 and 2012 (Fig. 3a). Unidirectional variations throughout the eastern seaboard have occurred 64% of the time in 9+ decades. The greatest increases, as illustrated by temporal variations of mode 1 (Fig. 3b), occurred in 1949, 1973 and 2009, while the largest decreases occurred in 1977, 1981, and 1988-1990. The spatial structure of EOF mode 2 (Fig. 3a) describes opposite changes of sea level in the MAB relative to the SEUSA; positive values in Figure 3c represent increases in the north relative to the south, while negative values denote the opposite. The greatest increases in the MAB with negligible, or even negative, change in the SEUSA (maxima in positive values of EOF mode 2) occurred in 1970-1972 and in 2010. On the other hand, the greatest increase in the SEUSA with negligible, or even negative, change in the MAB has occurred in 1947-1949 and 1995 (greatest negative values). The type of variability depicted by Mode 2 has occurred 20% of the time in the span 1920-2012. The 2009-2010 increase in the MAB appears in this EOF mode 2. Finally, the spatial structure of EOF Mode 3 defines a change in the middle of the eastern seaboard, around Cape Hatteras, which was in a different direction from changes in the MAB and SEUSA (Fig. 3a). The greatest influence of Mode 3 (Fig. 3d) has occurred in 1955 (positive) and 1939 (negative), representing 6% of the variability of one-year filtered water levels between 1920 and 2012.
Prevailing variabilities of the time series of the EOFs, which in turn represent temporal variability of one-year filtered sea level variations, were characterized by their spectra (Fig. 3e). Modes 1 and 2 exhibit intra-decadal and quasi-decadal variability. Intra-decadal variations are dominated by oscillations between 4 and 5 years, and between 6 and 7 years. Quasi-decadal oscillations are around 12 years for Mode 1 and 14 years for Mode 2. In addition, Mode 2 exhibits broad-banded multi-decadal variability centered at ~40 years. Mode 3 variations have been mostly associated with multi-decadal oscillations. The relevance of the variability suggested by EOF analysis is now explored with analysis of 5-year trends in sea level.
Linear 5-yr trends for the long period time series show positive and negative variations with magnitudes that exceed the global mean rate of 2-3 mm/yr by severalfold (Fig. 4a). Maximum rates of sea-level rise occurred in the SEUSA in the mid-1940s. In the middle of the eastern seaboard of the United States, maximum rates occurred in the early 1970s. The MAB has repeatedly had extended periods (~5-10 yrs) of sea-level rise throughout this time interval, separated by shorter periods (<5 yrs) of sea-level fall. The Probability Density Function for the rates of sea-level rise in the span 1920-2012 showed that the most common rates were highest (around 3 mm/yr) between New York (Montauk) and Cape Hatteras (dotted line on Fig. 4a). Outside the MAB, the most common rates are near zero or 1 mm/yr. The last (2008-2012) 5-yr trend for the long period dataset (denoted by white dots) displayed values of sea-level rise up to 10 mm/yr in the MAB, even leaking to North Carolina (Wilmington record). This increase was mainly linked to the extremely high sea levels recorded in the MAB during 2009-2010 (Goddard et al., 2015; Ezer et a., 2015). In contrast, during 2008-2012, the trend was 0-1 mm/yr in most of the SEUSA.
Linear trends from the short period data also showed values that surpass rates of global mean sea-level rise by severalfold (Fig. 4b). Since 1996, trends of sea-level rise have shown spatially variable structures. In the MAB there have been two periods of rapid rise centered in late 2002, and 2009 through 2010. The rapid rise in the latest period in the MAB leaked into North Carolina but not farther south. In the SEUSA there has been only one span of rapid increasing trend, where most dramatic increases have been seen in Florida, Georgia, and South Carolina (up to 1200 km in Fig. 4b). Examination of the Probability Density Function of the 5-yr sea-level rise rates indicated that, since 1996, the most frequently observed sea-level rise trend in the MAB has been near 10 mm/yr, in agreement with results from the long period data. In contrast, the SEUSA has shown most frequent values around 2 mm/yr, except in South Carolina where the most frequent trend has been slightly negative. The most recent 5-yr interval (denoted by white dots on the probability figure) indicated a slightly negative tendency in the MAB, after the large increase in 2009-2010. In the SEUSA, in particular from Miami to Charleston, the trends were near 20 mm/yr. The only previously observed period of a similarly high rate of rise in sea level occurred in the mid-1940s (Fig. 4a). Altimeter data from the Archiving, Validation and Interpretation of Satellite Oceanographic data (AVISO) products display the same overall increase of sea-level rise offshore of Florida since 2010 as tide gauge data (data not shown). Hence, both tide gauge and altimetry data have indicated a definitive acceleration in sea-level rise after 2011.
The main finding of this study is that sea-level rise has accelerated recently in the SEUSA from rates of 1-2 mm/yr to rates of 18-20 mm/yr. These findings extend previous studies that documented sea-level rise accelerations in the MAB over the past several decades, with no corresponding acceleration south of Cape Hatteras in records through 2012 (e.g. Sallenger et al., 2012; Ezer et al., 2013). An anomalous high rate of sea-level rise was also recorded in the MAB during 2009-2010 with no corresponding increase observed in the SEUSA (Ezer 2015; Goodard 2015). The recent increase of sea level in the SEUSA seems to be rare, with a similar rate of rise only observed from 1947 to 1948, according to the Probability Density Function of the rates observed in the last 90+ years (Fig. 4).
Possible reasons for the recent acceleration in sea-level rise in the MAB have been attributed to different phenomena, including wind stress and ocean circulation (e.g. Ezer et al., 2013; Woodworth et al., 2014; Thompson and Mitchum, 2014). Some of these causes can be from natural variability in atmospheric and ocean processes over the Atlantic Ocean (North Atlantic Oscillation (NAO) and the Atlantic Multi-decadal Oscillation (AMO)), and other processes (not mutually exclusive) associated to global changes that could be influenced by humans.
The NAO has a periodicity of 2-15 years and can produce sea level variations through the inverse barometer effect (e.g. Olafsdottir et al., 2013; Piecuch and Ponte, 2015). Spectral amplitudes of EOF Modes (Fig. 3) indicated periodicities that fall within those expected from these inter-decadal and multi-decadal oscillations. In fact, there are significant coherences, albeit only with values of ~0.5 and at only 90% confidence, between Mode 3 and the NAO (Fig. 3f). Significant coherences occurred only at periods of ~10 years with lags (not shown) close to 180º, which indicated opposite changes between Mode 3 water level and the NAO index, likely from the inverse barometer effect. There was no significant coherence (at 90% confidence) with EOF Mode 1 and the NAO, although there was weak coherence with Mode 2 at periods of 2 and 3 years. Therefore, linkages between observed variations of sea level in the eastern United States and the NAO is tenuous.
The AMO has periods between 60 and 80 years (Schlesinger and Ramankutty, 1994) and shows anomalously warm sea surface waters (positive phase) in the North Atlantic between 1925 and 1965, and after 2000. There was significant broad-band coherence (95% confidence with values >0.6) between the AMO index and Modes 2 and 3 at periods >20 years (Fig. 3f). The lag for Mode 2 was ~90º (not shown) indicating that its response to the AMO index was greatest at the transition between negative and positive phases. The lag for Mode 3 was around 50º (not shown). There was no coherence between the AMO index and EOF Mode 1. Relationships between AMO and sea level oscillations in the eastern United States also seem inconclusive.
A frequently mentioned attribution to the recent acceleration of sea-level rise in the eastern United States is the slowdown of the Atlantic meridional overturning circulation (AMOC) (e.g., Sallenger et al., 2012; Ezer, 2015; Rhamstorf et al., 2015). The overturning circulation is related to the strength of North Atlantic Deep Water formation. Such overturning circulation is linked to the North Atlantic Subtropical Gyre through the Gulf Stream. Therefore, any perturbations to the AMOC would affect the subtropical gyre and the Gulf Stream. Introduction of freshwater to the surface ocean at high latitudes would increase the static stability of North Atlantic waters and slow down the overturning circulation. There are several studies pointing out the decline of the overturning circulation (e.g. Smeed et al., 2014; Boulton et al, 2014). There is also evidence for reduced Gulf Stream transport (Ezer et al., 2013, Ezer, 2015) and Florida Current transport (Park and Sweet, 2015) by at least 1.5 Sv after 2005, linking it to a decline in the AMOC. Further evidence of drastic changes in the North Atlantic circulation system is the anomalously large amounts of sargassum running aground on the beaches of the entire Caribbean Sea from late 2014 to the summer of 2015. It is possible that these strandings are related, at least partially, to the relaxation of the North Atlantic Subtropical Gyre because of declines in Gulf Stream and Atlantic Meridional Circulation transport, though this linkage is speculative and requires further scrutiny.
Wind speed shows an equivocal tendency in the last 5-10 years (Fig. 5a). However, both east and north wind velocity components seem to have become weaker since 2011 (Fig. 5b). This apparent wind relaxation could contribute to waning of the Gulf Stream as wind is a major driving force of the subtropical gyre. Reduced wind velocities have also been consistent with decreased significant wave heights (Fig. 5c), most evidently since 2005. Atmospheric pressure has not shown a clear trend in the period of record (Fig. 5d) nor their oscillations seem to be tied to the variations in sea level of Figure 2b. But there is a clear increasing trend in air temperature and water temperature, consistent with global tendencies. It is possible that the increasing sea level can have a contribution from regional ocean warming, but this would need a thorough exploration beyond the scope of this study.
Despite many possible explanations for the regional sea level patterns identified here, there are no clear candidates that explain the magnitude and pattern of sea-level change witnessed along the eastern seaboard of the United States over the past 5 years. The increase in the rate of sea-level rise in the SEUSA may not be linked to a decline in the AMOC given that models predict an effect on coastal sea level in the MAB, but not in south of Cape Hatteras (e.g., Yin and Goddard, 2014). Although the high rates of sea-level rise currently observed in the SEUSA are not without precedent, the only previous such event on record that occurred in the mid 1940s did not occur at a similar phase of the NAO or AMO to that of present. These differences put into doubt whether these decadal-scale oscillations are responsible for the current situation.
Importantly, the high magnitude of sea-level rise in the SEUSA contributes to a higher base sea level upon which extreme events such as storm surges and king tides are superimposed. For example, the tide gauges on the east coast of Florida display >100 mm of sea-level rise over a mere 5 years (old fig 2b new figure). These rates exacerbate efforts to defend coastlines using smooth sea level projections that do not consider the possibility for such variability in the rate of sea-level rise, even if it turns out to be short-lived. Additionally, because seasonal variability determines the amplitude of king tides that regularly cause incursions of seawater into streets of many coastal communities, including—but not limited to—south Florida, it is essential to understand how this observed rate of sea-level rise is distributed throughout the year. Though not shown here, most of the increase in the rate of sea-level rise in the SEUSA since 2011 has been driven by higher sea levels during what is normally the seasonal low (the spring). If the seasonal high in the tidal cycle in the SEUSA (September and October) was somehow similarly affected in the future, the king tides and late season hurricanes will have a much more devastating effect.
Boon, J.D., 2012. Evidence of sea level acceleration at U.S. and Canadian tide stations, Atlantic coast, North America. J. Coast. Res. 28 (6), 1437–1445. http://dx.doi.org/10. 2112/JCOASTRES-D-12-00102.1.
Boulton, C.A., Allison, L.C., Lenton, T.M., 2014, Early warning signals of Atlantic Meridional Overturning Circulation collapse in a fully coupled climate model, Nat. Commun., 5,
Church, J. A. et al., 2013, in Climate Change 2013: The Physical Science Basis, eds Stocker, T. F. et al., Cambridge University Press.
Ezer, T., L. P. Atkinson, W. B. Corlett and J. L. Blanco, 2013, Gulf Stream’s induced sea-level rise and variability along the U.S. mid-Atlantic coast, J. Geophys. Res. Oceans, 118, 685–697, doi:10.1002/jgrc.20091.
Ezer, T. 2015, Detecting changes in the transport of the Gulf Stream and the Atlantic overturning circulation from coastal sea level data: The extreme decline in 2009–2010 and estimated variations for 1935–2012, Global and Planetary Change, 129, 23–36.
Fang, T., Piegl, L., 1992. Algorithm for Delaunay triangulation and convex hull computation using a sparse matrix. Computer Aided Design 24, 425–436.
Fang, T., Piegl, L., 1993. Delaunay triangulation using a uniform grid. IEEE Computer Graphics and Applications, 36–47.
Frankecombe, L. M. and Dijkstra, H. A., 2009, Coherent multidecadal variability in North Atlantic sea level, Geophys. Res. Lett., 36, L15604, doi:10.1029/2009GL039455.
Goddard, P.B., J. Yin, S.M. Griffies, and S. Zhang, 2015, An extreme event of sea-level rise along the Northeast coast of North America in 2009–2010, Nature Communications, DOI: 10.1038/ncomms7346.
Haigh, I.D.,Wahl, T., Rohling, E.J., Price, R.M., Pattiaratchi, C.B., Calafat, F.M., Dangendorf, S., 2014. Timescales for detecting a significant acceleration in sea-level rise. Nat. Commun. http://dx.doi.org/10.1038/ncomms4635.
Knight, J. R., Allan, R. J., Folland, C. K., Vellinga, M. & Mann, M. E. 2005, A signature of persistent natural thermohaline circulation cycles in observed climate. Geophysical Research Letters 32.
Lambeck, K., Rouby, H., Purcell, A., Sun, Y. & Sambridge, M. 2014, Sea level and global ice volumes from the Last Glacial Maximum to the Holocene. Proceedings of the National Academy of Sciences 111, 15296-15303.
McCarthy, G., E. Frejka-Williams, W.E. Johns, M.O. Baringer, C.S. Meinen, H.L. Bryden, D. Rayner, A. Duchez, C. Roberts, S.A. Cunningham, 2012. Observed interannual variability of the Atlantic meridional overturning circulation at 26.5°N, Geophys. Res. Lett. Doi:10.1029/2012GL052933.
Olafsdottir, K.B., A. Geirsdottir, G.H. Miller and D.J. Larsen, 2013, Evolution of NAO and AMO strength and cyclicity derived from a 3-ka varve-thickness record from Iceland, Quat. Sci. Rev., 69(1), 142-154.
Park, J. and Dusek, G., 2013, ENSO components of the Atlantic multidecadal oscillation and their relation to North Atlantic interannual coastal sea level anomalies, Ocean Sci., 9, 535–543, doi:10.5194/os-9-535-2013.
Park, J. & Sweet, W., 2015, Accelerated sea-level rise and Florida Current transport. Ocean Science Discussions 12, 551-572.
Piecuch, C.G., R.M. Ponte, 2015, Inverted barometer contributions to recent sea level changes along the northeast coast of North America, Geophysical Research Letters, 42(14), 5918–5925.
Rhamstorf, S., J.E. Box, G. Feulner, M.E. Mann, A. Robinson, S. Rutherford and E.J. Schaffernicht, 2015, Exceptional twentieth-century slowdown in Atlantic Ocean overturning circulation, Nature Climate Change, DOI: 10.1038/NCLIMATE2554.
Rohling, E. J. et al. 2009, Antarctic temperature and global sea level closely coupled over the past five glacial cycles. Nature Geosci 2, 500-504.
Sallenger Jr, A.H., K.S. Doran, and P.A. Howd, 2012, Hotspot of accelerated sea-level rise on the Atlantic coast of North America, Nature Climate Change, DOI: 10.1038/NCLIMATE1597.
Schlesinger, M. E., and N. Ramankutty (1994), An oscillation in the global climate system of period 65–70 years, Nature, 367, 723–726.
Smeed, D. A. et al., 2014, Observed decline of the Atlantic Meridional Overturning Circulation 2004 to 2012. Ocean Sci. 10, 29–38.
Thompson, P. and Mitchum, G., 2014, Coherent sea level variability on the North Atlantic western boundary. J. Geophys. Res. 119, 5676–5689.
Thomson, R.E. and W. Emery, 2014, Data Analysis Methods in Physical Oceanography, 3rd Edition, Elsevier, 728 pp.
Trenberth, K. E., and D. J. Shea (2006), Atlantic hurricanes and natural variability in 2005, Geophys. Res. Lett. 33, L12704, doi:10.1029/2006GL026894.
Woodworth, P. L., M. A. M. Maqueda, V.M. Roussenov, R. G. Williams, and C. W. Hughes (2014), Mean sea-level variability along the northeast American Atlantic coast and the roles of the wind and the overturning circulation, J. Geophys. Res. Oceans, 119, 8916–8935, doi:10.1002/2014JC010520.
Figure 1. a) Eastern United States showing the station locations for the long period (in white letters and black stars) and the short period (in red numbers and yellow stars). Numbers and letters correspond to station names given in other figures. b) one-year filtered sea level at 3 stations in Southeastern Florida). LABEL MAD & SEUSA5>