# Natural Variability and Vertical Land Motion Contributions in the Mediterranean Sea-Level Records over the Last Two Centuries and Projections for 2100

^{1}

^{2}

^{3}

^{4}

^{*}

## Abstract

**:**

^{−1}(locally up to 8.45 ± 1.69 mm a

^{−1}) is driving the local sea-level rise acceleration.

## 1. Introduction

^{−1}and 3.2 mm a

^{−1}for the 20th century and the last two decades, respectively [5,6,7]. The Mediterranean Sea has been rising at the mean rate of ~1.8 mm a

^{−1}in the last two-three centuries [8,9,10,11,12]. At global scales, the sea-level change is described as the sum of eustatic, glacio-hydro-isostatic, and land-hydrology components [13], which are independently modeled to provide future sea-level predictions. Local sea-level change can significantly differ from mean global sea-level rise because local factors become relevant and their contribution adds up to the global components mentioned above [14]. Oceanographic and climatic effects, glacial isostatic adjustment (GIA), and continuous vertical land motion, especially coastal subsidence due to tectonics (volcanic included) or anthropic origins, are some of the factors that mainly affect sea-level changes at local scales [15]. Therefore, depending on the geographical location they cause spatial variability in the sea-level change pattern. While local effects have little influence at global scale, since they are smoothed out by averaging operations, they become critical in adaptation planning and risk management, where localized assessments are needed.

## 2. Materials and Methods: Data Set and Natural Variability

_{j}(t), whose functional shape is not fixed a priori, but obtained from the data, and a residue r

_{n}(t) providing the nonlinear trend [40]:

_{j}(t), if it has exactly one zero crossing between two consecutive local extrema and its mean is zero. The first condition ensures a global narrowband requirement, namely periods which are too different are not mixed together into an IMF; the second requirement ensures getting oscillations about the zero level. Both these conditions ensure that L

_{j}(t) is an oscillating function with time-dependent amplitude and phase, namely it can be expressed as L

_{j}(t) = B

_{j}(t)cos[ϕ

_{j}(t)]. If none of these conditions are fulfilled, then step (iii) is repeated by using h(t) as input data. This procedure is iterated until the resulting signal fulfills the IMF properties and it is repeated to find the next IMF by using L(t) − L

_{j}(t) as new input data. When no more modes can be extracted, the difference between the original L(t) and the sum of all IMFs results in the residue r

_{n}(t). Technical details on the procedure for IMF extraction are reported in [41]. The EMD allows calculating a meaningful instantaneous phase for each IMF by using the Hilbert transform

_{j}* represents the complex conjugate of L(t) as ϕ

_{j}(t) = arctan (L

_{j}*/L

_{j}). Finally, the instantaneous frequency, ω

_{j}(t), is given by dϕ

_{j}(t)/dt. For each L

_{j}(t) we can define a characteristic timescale τ

_{j}, as the time average <2π/ω

_{j}(t)>, that represents an average period of the IMF and provides an estimate of the timescale characterizing the EMD mode. We remark that τ

_{j}is not to be intended as the Fourier one. It just gives an indication of the timescale characterizing the EMD mode for which it is computed, although many modes with different average periods may contribute to the variability of the actual signal at a particular timescale. The IMF statistical significance with respect to white noise can be evaluated through a test based on the comparison between IMF amplitude from the real signal and from a white noise series [46] at different confidence levels. Since IMFs are a local, complete and orthogonal set, the EMD is useful to filter raw signal through partial sums in Equation (1). Moreover, since a low number of modes are produced, this property can be exploited to build up low order theoretical models to describe the dynamics of the investigated systems.

## 3. Results: The EMD Analysis and a Generalized Model for Long-Term Sea-Level Variations

_{j}are shown in Table 1) and significant with respect to a white noise at the 90th significance level. The long-term variations of the measured sea level, L

_{LT}, can be defined as:

_{LT}, for the analyzed tide gauges, is shown in Figure 3. All records show a long-term variability in a well-defined range of timescales of about 20–30 and, for the longer records, >45 years, in the typical ranges where internal modes of natural variability are detected [47]. Observational datasets and simulations of the typical scales of Atlantic internal modes of natural variability show variations at decadal timescales that are linked to the ocean-atmosphere coupling [47,48,49,50]. In particular, the 20–30 year variability has been related to the Atlantic Meridional Overturning Circulation (MOC), while the >45 year fluctuations are plausibly due to salinity and matter exchange processes between the Atlantic and Arctic Ocean causing the Atlantic Multidecadal Oscillation (AMO) pattern. Both patterns are quasi-persistent and present up to 8000 years ago [51,52]. The similarity of timescales we found in the tide gauge data with those of the Atlantic modes of variability suggests a connection between climate variability and sea-level variations in the Mediterranean Sea as also indicated by [53,54]. The combined analysis of global oceanic circulation, climate models and observations, indeed, indicates that the main physical mechanism driving the connection between sea level in the Mediterranean Sea and internal modes of natural variability is the mass exchange through the Strait of Gibraltar [53]. This mechanism generates a coherent and uniform signal in the whole Mediterranean at decadal time scales [54,55]. The detected sea-level variations at 20–30 years can be thus related to the combined effect of the North Atlantic Oscillations (NAO) and AMO [26]. An additional contribution could come from the Inter-decadal Pacific Oscillation (IPO), causing variations of the global surface temperature at these timescales, which affects also the climate dynamics in the Atlantic Ocean [56,57]. Likewise, major volcanic eruptions, affecting the global temperature by acting on the radiative forcing [57,58], could also contribute to the SLNV. Fluctuations at longer timescales can be due to both atmospheric forcing [59] and exchange processes [47].

_{n}(t) has not been added up), for these records. For the four stations the variability is very similar with maxima and minima almost in phase. Small differences can be due to border effects for shorter series (e.g., Rovinij) and/or presence of data gaps affecting the IMF calculation or natural reasons such as small local differences in the signal variability. We note that the Venice station shows a different behavior with a flatter signal between 1945 and 1980 due to temporary anthropogenic subsidence induced by groundwater [28].

_{j}is the time average of the instantaneous frequency of IMFs in S. This represents the most elementary approximation to reproduce the long-term variability inferred by the EMD through modes with time-dependent amplitude and phase. The free parameters in Equation (4) (rate r, intercept c, A

_{j}and Φ

_{j}) are calculated by fitting L

_{M}(t) to L

_{LT}(t). To avoid possible end effects due to the procedure of IMF extraction and the computation of derivative at borders, we cut three years of data at both boundaries of L

_{LT}(t) before performing the fit. In addition, since data from Trieste, Genoa, Venice, and Bakar show gaps in the initial part of their records, for these stations the fit starts from year 1901, 1931, 1914, and 1949, respectively, when recordings begin to be continuous for many decades. Two sources of uncertainty were taken into account. Firstly, the uncertainty due to the use of constant frequencies, instead of time varying ones, is evaluated by building 1000 L

_{M}(t) realizations whose Ω

_{j}are randomly extracted from a Gaussian distribution with mean < ω

_{j}(t) > and standard deviation σ [ω

_{j}(t)]. Secondly, effects of possible undetected modes of climate variability, with periods of the order and/or longer than the record lengths, are taken into account as an increase of the estimated uncertainty. Indeed, since these very long period modes are not included in the model, the accuracy for estimating the sea level is reduced. This additional uncertainty is accounted for by adding to each realization L

_{M}(t) a normally distributed zero average white noise. Since we expect that modes with periods of the order and/or longer than the record lengths are settled on the EMD nonlinear residue, the standard deviation of the white noise signal was taken as the standard deviation of the EMD residue from which we subtracted a linear trend.

_{M}(t) to L

_{LT}(t). Fit parameters and adjusted R-square statistics are shown in Table 2. R-square values close to 1 indicate that a good fit is achieved for the analyzed records. Rates r obtained by a fit through Equation (4) are compatible or slightly higher than rates estimated by a linear fit to the raw data for the same time intervals shown in Figure 2 and varying between 0.35 ± 0.38 mm a

^{−1}(Split M.) and 2.43 ± 0.23 mm a

^{−1}(Venice P.S.). Only for Split Gradska Luka, the linear rate from Equation (4) is higher than the linear fit. This is possibly due to a steeper sea-level rise observed in L

_{LT}at this station after 2006. In the raw data, this feature is hidden by short-term fluctuations, thus resulting in a lower rate when a linear fit is computed on the original time series. The origin of this steepening has been proposed to be the coincidence of AMO-NAO phase opposition and warm AMO phase [60]. The same feature is also present in L

_{LT}for the Trieste and Rovinj data. However, for Trieste the linear rate from Equation (4) does not show significant differences with respect to the rate of the linear fit likely due to the longer duration of recordings for which the fitting procedure is very robust and not affected by variations in the last part of the data.

## 4. Discussion: The Vertical Land Motion and Relative Sea Levels by 2050 and 2100

^{−1}at Marseille, Genova, and Split and it is about 1.5 mm a

^{−1}at Dubrovnik. Sudden temporary tectonic episodes, which can cause permanent signatures during the tidal recordings at these stations, have not occurred (www.psmsl.org).

^{−1}. Part of the detected subsidence is due to the continuing global glacio-hydroisostatic signal, estimated in this region to be between −0.12 and −0.21 mm a

^{−1}[15], and the remaining part can be related to natural and anthropogenic effects [64,65]. We can reasonably assume that the VLM rate will remain unchanged until 2100 AD, in the absence of additional episodes of land movement that may result from eventual earthquakes with significant magnitude in the area [66] (http://www.isc.ac.uk/). The remarkable subsidence in the Venice lagoon results in a steeper sea-level rate with respect to those provided by IPCC estimations in the area [29,67].

_{2}emission, as the Earth nowadays and possibly in future years. Studies on the effects of the long-term variability on regional sea levels from different climatic models indicate that the internal variability contribution is already relevant when 100 year integrations are considered [68,69,70]. Moreover, the estimated time of emergence, namely the time at which the CO

_{2}-forced sea level signal starts to dominate the internal variability, is close to or longer than a century in most of the Oceans. Other studies [71,72] indicate that the interplay of natural and forced variability in sea-level changes is expected to persist in time throughout the 21st century. In addition, possible variations of the modes of natural variability, between 2016 and 2100, were taken into account in our analysis by using modes with several frequencies for the uncertainty calculation.

_{j}> 15 years (τ

_{9}= 33.4 ± 5.4 years and τ

_{10}= 58.1 ± 7.6 years) were found. Then, sea levels for the next 50 years (1957–2007) were evaluated through the model, by using the linear trend from the fit to Equation (4) and were compared to the real complete data set.

- The normalized mean squared error (NMSE) between the 15 years smoothed and the model curves, after 1957, is 4.4 × 10
^{−6}(if a model has a very low NMSE then it reproduces the data well); - The maximum and minimum discrepancy between the 15 years smoothed data and the model-derived sea levels are found in 1965 and 1997, respectively. Sea-level values are 6976 and 6977 mm for smoothed data, 6974 (6956, 6992) and 7000 (6968, 7032) mm for model data (values in parentheses represents the upper/lower 90% confidence interval).

## 5. Conclusions

## Author Contributions

## Funding

## Acknowledgments

## Conflicts of Interest

## References

- Vermeer, M.; Rahmstorf, S. Global sea level linked to global temperature. Proc. Natl. Acad. Sci. USA
**2009**, 106, 21527–21532. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Church, J.A.; White, N.J. Sea-Level Rise from the Late 19th to the Early 21st Century. Surv. Geophys.
**2011**, 32, 585–602. [Google Scholar] [CrossRef] [Green Version] - Kemp, A.C.; Horton, B.P.; Donnelly, J.P.; Mann, M.E.; Vermeer, M.; Rahmstorf, S. Climate related sea-level variations over the past two millennia. Proc. Natl. Acad. Sci. USA
**2011**, 108, 11017–11022. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Lambeck, K.; Woodroffe, C.D.; Antonioli, F.; Anzidei, M.; Gehrels, W.R.; Laborel, J.; Wright, A.J. Paleoenvironmental Records, Geophysical Modeling, and Reconstruction of Sea-Level Trends and Variability on Centennial and Longer Timescales. In Understanding Sea-Level Rise and Variability; Blackwell: Hoboken, NJ, USA, 2010; pp. 61–121. [Google Scholar]
- Meyssignac, B.; Cazenave, A. Sea level: A review of present-day and recent-past changes and variability. J. Geodyn.
**2012**, 58, 96–109. [Google Scholar] [CrossRef] - Mitchum, G.T.; Nerem, R.S.; Merrifield, M.A.; Gehrels, W.R. Modern sea level changes estimates. In Understanding Sea Level Rise and Variability; Church, J.A., Woodworth, P.L., Aarup, T., Wilson, W.S., Eds.; Wiley-Blackwell Publishing: London, UK, 2010. [Google Scholar]
- Jevrejeva, S.; Moore, J.; Grinsted, A. Sea level projections to AD2500 with a new generation of climate change scenarios. Glob. Planet. Chang.
**2012**, 80, 14–20. [Google Scholar] [CrossRef] - Wöppelmann, G.; Marcos, M. Coastal sea level rise in southern Europe and the nonclimate contribution of vertical land motion. J. Geophys. Res. Space Phys.
**2012**, 117. [Google Scholar] [CrossRef] [Green Version] - Anzidei, M.; Lambeck, K.; Antonioli, F.; Furlani, S.; Mastronuzzi, G.; Serpelloni, E.; Vannucci, G. Coastal structure, sea-level changes and vertical motion of the land in the Mediterranean. Geol. Soc. Lond. Spéc. Publ.
**2014**, 388, 453–479. [Google Scholar] [CrossRef] - Tsimplis, M.N.; Calafat, F.M.; Marcos, M.; Jorda, G.; Gomis, D.; Fenoglio-Marc, L.; Struglia, M.V.; Josey, S.A.; Chambers, D.; Fenoglio-Marc, L. The effect of the NAO on sea level and on mass changes in the Mediterranean Sea. J. Geophys. Res. Oceans
**2013**, 118, 944–952. [Google Scholar] [CrossRef] [Green Version] - Fenoglio-Marc, L.; Rietbroek, R.; Grayek, S.; Becker, M.; Kusche, J.; Stanev, E. Water mass variation in the Mediterranean and Black Seas. J. Geodyn.
**2012**, 59, 168–182. [Google Scholar] [CrossRef] - Fenoglio-Marc, L.; Mariotti, A.; Sannino, G.; Meyssignac, B.; Carillo, A.; Struglia, M.; Rixen, M. Decadal variability of net water flux at the Mediterranean Sea Gibraltar Strait. Glob. Planet. Chang.
**2013**, 100, 1–10. [Google Scholar] [CrossRef] [Green Version] - Lambeck, K.; Purcell, A. Sea-level change in the Mediterranean Sea since the LGM: Model predictions for tectonically stable areas. Quat. Sci. Rev.
**2005**, 24, 1969–1988. [Google Scholar] [CrossRef] - Kopp, R.E.; Horton, R.M.; Little, C.M.; Mitrovica, J.X.; Oppenheimer, M.; Rasmussen, D.J.; Strauss, B.H.; Tebaldi, C. Probabilistic 21st and 22nd century sea-level projections at a global network of tide-gauge sites. Earth’s Future
**2014**, 2, 383–406. [Google Scholar] [CrossRef] - Lambeck, K.; Antonioli, F.; Anzidei, M.; Ferranti, L.; Leoni, G.; Scicchitano, G.; Silenzi, S. Sea level change along the Italian coast during the Holocene and projections for the future. Quat. Int.
**2011**, 232, 250–257. [Google Scholar] [CrossRef] - Douglas, B.C. Global sea rise: A redetermination. Surv. Geophys.
**1997**, 18, 270–292. [Google Scholar] [CrossRef] - Knutti, R.; Stocker, T.F. Influence of the Thermohaline Circulation on Projected Sea Level Rise. J. Clim.
**2000**, 13, 1997–2001. [Google Scholar] [CrossRef] - Rahmstorf, S. A Semi-Empirical Approach to Projecting Future Sea-Level Rise. Science
**2007**, 315, 368–370. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Rahmstorf, S.; Cazenave, A.; Church, J.A.; Hansen, J.E.; Keeling, R.F.; Parker, D.E.; Somerville, R.C.J. Recent Climate Observations Compared to Projections. Science
**2007**, 316, 709. [Google Scholar] [CrossRef] [PubMed] - Milne, G.A.; Gehrels, W.R.; Hughes, C.W.; Tamisiea, M.E. Identifying the causes of sea-level change. Nat. Geosci.
**2009**, 2, 471–478. [Google Scholar] [CrossRef] [Green Version] - Tsimplis, M.; Marcos, M.; Colin, J.; Somot, S.; Pascual, A.; Shaw, A. Sea level variability in the Mediterranean Sea during the 1990s on the basis of two 2d and one 3d model. J. Mar. Syst.
**2009**, 78, 109–123. [Google Scholar] [CrossRef] - Tsimplis, M.; Spada, G.; Marcos, M.; Flemming, N. Multi-decadal sea level trends and land movements in the Mediterranean Sea with estimates of factors perturbing tide gauge data and cumulative uncertainties. Glob. Planet. Chang.
**2011**, 76, 63–76. [Google Scholar] [CrossRef] - Church, J.A.; Aarup, T.; Woodworth, P.L.; Wilson, W.S.; Nicholls, R.J.; Rayner, R.; Lambeck, K.; Mitchum, G.T.; Steffen, K.; Cazenave, A.; et al. Sea-Level Rise and Variability: Synthesis and Outlook for the Future. In Understanding Sea-Level Rise and Variability, 1st ed.; Church, A.J., Woodworth, P.L., Aarup, T., Wilson, S., Eds.; Blackwell Publishing Ltd.: Hoboken, NJ, USA, 2010. [Google Scholar]
- Cazenave, A.; Remy, F. Sea level and climate: Measurements and causes of changes. Wiley Interdiscip. Rev. Clim. Chang.
**2011**, 2, 647–662. [Google Scholar] [CrossRef] - Zhang, X.; Church, J.A. Sea level trends, interannual and decadal variability in the Pacific Ocean. Geophys. Res. Lett.
**2012**, 39, 21701. [Google Scholar] [CrossRef] - Galassi, G.; Spada, G. Sea-level rise in the Mediterranean Sea by 2050: Roles of terrestrial ice melt, steric effects and glacial isostatic adjustment. Glob. Planet. Chang.
**2014**, 123, 55–66. [Google Scholar] [CrossRef] - Han, W.; Stammer, D.; Meehl, G.A.; Hu, A.; Sienz, F.; Zhang, L. Multi-Decadal Trend and Decadal Variability of the Regional Sea Level over the Indian Ocean since the 1960s: Roles of Climate Modes and External Forcing. Climate
**2018**, 6, 51. [Google Scholar] [CrossRef] - Brambati, A.; Carbognin, L.; Quaia, T.; Teatini, P.; Tosi, L. The Lagoon of Venice: Geological setting, evolution and land subsidence. Episodes
**2003**, 26, 264–268. [Google Scholar] - Carbognin, L.; Teatini, P.; Tosi, L. Eustacy and land subsidence in the Venice Lagoon at the beginning of the new millennium. J. Mar. Syst.
**2004**, 51, 345–353. [Google Scholar] [CrossRef] - De Conto, R.M.; Pollard, D. Contribution of Antarctica to past and future sea-level rise. Nature
**2016**, 531, 591–597. [Google Scholar] [CrossRef] - Syvitski, J.P.M.; Kettner, A.J.; Overeem, I.; Hutton, E.W.H.; Hannon, M.T.; Brakenridge, G.R.; Day, J.; Vörösmarty, C.; Saito, Y.; Giosan, L.; et al. Sinking deltas due to human activities. Nat. Geosci.
**2009**, 2, 681–686. [Google Scholar] [CrossRef] - Antonioli, F.; Anzidei, M.; Amorosi, A.; Presti, V.L.; Mastronuzzi, G.; Deiana, G.; De Falco, G.; Fontana, A.; Fontolan, G.; Lisco, S.; et al. Sea-level rise and potential drowning of the Italian coastal plains: Flooding risk scenarios for 2100. Quat. Sci. Rev.
**2017**, 158, 29–43. [Google Scholar] [CrossRef] [Green Version] - Bosman, A.; Carluccio, R.; Casalbore, D.; Caracciolo, F.D.; Esposito, A.; Nicolosi, I.; Pietrantonio, G.; Vecchio, A.; Carmisciano, C.; Chiappini, M.; et al. Flooding scenarios due to land subsidence and sea-level rise: A case study for Lipari Island (Italy). Terra Nova
**2017**, 29, 44–51. [Google Scholar] - Marsico, A.; Lisco, S.; Presti, V.L.; Antonioli, F.; Amorosi, A.; Anzidei, M.; Deiana, G.; De Falco, G.; Fontana, A.; Fontolan, G.; et al. Flooding scenario for four Italian coastal plains using three relative sea level rise models. J. Maps
**2017**, 13, 961–967. [Google Scholar] [CrossRef] [Green Version] - Douglas, B.C. Global sea level rise. J. Geophys. Res.
**1991**, 96, 6981–6992. [Google Scholar] [CrossRef] [Green Version] - Douglas, B.C. Global sea level rise acceleration. J. Geophys. Res.
**1992**, 97, 12699–12796. [Google Scholar] [CrossRef] - Visser, H.; Dangendorf, S.; Petersen, A.C. A review of trend models applied to sea level data with reference to the “acceleration-deceleration debate”. J. Geophys. Res. Oceans
**2015**, 120, 3873–3895. [Google Scholar] [CrossRef] - Huang, N.E.; Shen, Z.; Long, S.R.; Wu, M.C.; Shih, E.H.; Zheng, Q.; Tung, C.C.; Liu, H.H. The empirical mode decomposition and the Hilbert spectrum for non-stationary time series analysis. Proc. R. Soc. Lond.
**1998**, 454, 903–995. [Google Scholar] [CrossRef] - Vecchio, A.; Capparelli, V.; Carbone, V. The complex dynamics of the seasonal component of USA’s surface temperature. Atmos. Chem. Phys.
**2010**, 10, 9657–9665. [Google Scholar] [CrossRef] - Vecchio, A.; Anzidei, M.; Capparelli, V.; Carbone, V.; Guerra, I. Has the Mediterranean Sea felt the March 11th, 2011, Mw 9.0 Tohoku-Oki earthquake? Europhys. Lett.
**2012**, 98, 59001. [Google Scholar] [CrossRef] - Vecchio, A.; Anzidei, M.; Carbone, V. New insights on the tsunami recording of the May, 21, 2003, Mw 6.9 Boumerdès earthquake from tidal data analysis. J. Geodyn.
**2014**, 79, 39–49. [Google Scholar] [CrossRef] - Breaker, L.C.; Ruzmaikin, A. The 154-year record of sea level at San Francisco: Extracting the long-term trend, recent changes, and other tidbits. Clim. Dyn.
**2011**, 36, 545–559. [Google Scholar] [CrossRef] - Ezer, T.; Corlett, W.B. Is sea level rise accelerating in the Chesapeake Bay? A demonstration of a novel new approach for analyzing sea level data. Geophys. Res. Lett.
**2012**, 39, L19605. [Google Scholar] [CrossRef] - Ezer, T.; Atkinson, L.P.; Corlett, W.B.; Blanco, J.L. Gulf Stream’s induced sea level rise and variability along the U.S. mid-Atlantic coast. J. Geophys. Res. Ocean
**2013**, 118, 685–697. [Google Scholar] [CrossRef] - Chambers, D.P. Evaluation of empirical mode decomposition for quantifying multi-decadal variations and acceleration in sea level records. Nonlinear Proc. Geophys.
**2015**, 22, 157–166. [Google Scholar] [CrossRef] [Green Version] - Wu, Z.; Huang, N.E. A study of the characteristics of white noise using the empirical mode decomposition method. Proc. R. Soc. Lond. A
**2004**, 460, 1597–1611. [Google Scholar] [CrossRef] - Frankcombe, L.M.; Von der Heydt, A.; Dijkstra, H.A. North Atlantic multidecadal climate variability: An Investigation of dominant time scales and processes. J. Climatol.
**2010**, 23, 3626–3638. [Google Scholar] [CrossRef] - Marshall, J.; Johnson, H.; Goodman, J. A study of the interaction of the North Atlantic Oscillation with the ocean circulation. J. Clim.
**2001**, 14, 1399–1421. [Google Scholar] [CrossRef] - Bellucci, A.; Gualdi, S.; Scoccimarro, E.; Navarra, A. NAO-ocean circulation interactions in a coupled general circulation model. Clim. Dyn.
**2008**, 31, 759–777. [Google Scholar] [CrossRef] - Fan, M.; Schneider, E.K. Observed decadal North Atlantic tripole SST variability. Part I: Weather noise forcing and coupled response. J. Atmos. Sci.
**2012**, 69, 35–50. [Google Scholar] - Knudsen, M.F.; Seidenkrantz, M.-S.; Jacobsen, B.H.; Kuijpers, A. Tracking the Atlantic Multidecadal Oscillation through the last 8,000 years. Nat. Commun.
**2011**, 2, 178. [Google Scholar] [CrossRef] - Hubeny, J.B.; King, J.W.; Santos, A. Subdecadal to multidecadal cycles of Late Holocene North Atlantic climate variability preserved by estuarine fossil pigments. Geology
**2006**, 34, 569–572. [Google Scholar] [CrossRef] [Green Version] - Calafat, F.M.; Jordà, G.; Marcos, M.; Gomis, D. Comparison of Mediterranean sea level variability as given by three baroclinic models. J. Geophys. Res.
**2012**, 117, c02009. [Google Scholar] [CrossRef] - Calafat, F.M.; Chambers, D.P.; Tsimplis, M.N. Mechanisms of decadal sea level variability in the eastern North Atlantic and the Mediterranean Sea. J. Geophys. Res.
**2012**, 117, c09022. [Google Scholar] [CrossRef] - Tsimplis, M.N.; Shaw, A.G.P. The forcing of mean sea level variability around Europe. Glob. Planet. Chang.
**2008**, 63, 196–202. [Google Scholar] [CrossRef] - Kosaka, Y.; Xie, S.-P. Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature
**2013**, 501, 403–407. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Zhang, L. The roles of external forcing and natural variability in global warming hiatuses. Clim. Dyn.
**2016**, 47, 3157–3169. [Google Scholar] [CrossRef] - Kosaka, Y.; Xie, S.-P. The tropical Pacific as a key pacemaker of the variable rates of global warming. Nat. Geosci.
**2016**, 9, 669–673. [Google Scholar] [CrossRef] - Marullo, S.; Artale, V.; Santoleri, R. The SST Multidecadal Variability in the Atlantic–Mediterranean Region and Its Relation to AMO. J. Clim.
**2011**, 24, 4385–4401. [Google Scholar] [CrossRef] - Vigo, M.; Sánchez-Reales, J.; Trottini, M.; Chao, B. Mediterranean Sea level variations: Analysis of the satellite altimetric data, 1992–2008. J. Geodyn.
**2011**, 52, 271–278. [Google Scholar] [CrossRef] - Devoti, R.; D’Agostino, N.; Serpelloni, E.; Pietrantonio, G.; Riguzzi, F.; Avallone, A.; Cavaliere, A.; Cheloni, D.; Cecere, G.; D’Ambrosio, C.; et al. The mediterranean crustal motion map compiled at INGV. Ann. Geophys.
**2017**, 60. [Google Scholar] [CrossRef] - Serpelloni, E.; Faccenna, C.; Spada, G.; Dong, D.; Williams, S.D.P. Vertical GPS ground motion rates in the Euro-Mediterranean region: New evidence of velocity gradients at different spatial scales along the Nubia-Eurasia plate boundary. J. Geophys. Res. Solid Earth
**2013**, 118, 6003–6024. [Google Scholar] [CrossRef] [Green Version] - Altamimi, Z.; Collilieux, X.; Métivier, L. ITRF2008: An improved solution of the international terrestrial reference frame. J. Geod.
**2011**, 85, 457–473. [Google Scholar] [CrossRef] - Antonioli, F.; Anzidei, M.; Lambeck, K.; Auriemma, R.; Gaddi, D.; Furlani, S.; Orrù, P.; Solinas, E.; Gaspari, A.; Karinja, S.; et al. Sea level change during Holocene from Sardinia and northeastern Adriatic (Central Mediterranean sea) from archaeological and geomorphological data. Quat. Sci. Rev.
**2007**, 26, 2463–2524. [Google Scholar] [CrossRef] - Carbognin, L.; Gatto, P.; Mozzi, G.; Gambolati, G.; Ricceri, G. New trend in the subsidence of Venice. In Proceedings of the Anaheim Symposium, Anaheim, CA, USA, 13–17 December 1976. [Google Scholar]
- Guidoboni, E.; Comastri, A. Catalogue of Earthquakes and Tsunami in the Mediterranean Area from the 11th to the 15th Century; INGV-SGA: Rome, Italy, 2007. [Google Scholar]
- Carbognin, L.; Teatini, P.; Tomasin, A.; Tosi, L. Global change and relative sea level rise at Venice: What impact in term of flooding. Clim. Dyn.
**2010**, 35, 1039–1047. [Google Scholar] [CrossRef] - Lyu, K.; Zhang, X.; Church, J.A.; Slangen, A.B.A.; Hu, J. Time of emergence for regional sea-level change. Nat. Clim. Chang.
**2014**, 4, 1006–1010. [Google Scholar] [CrossRef] - Bordbar, M.H.; Martin, T.; Latif, M.; Park, W. Effects of long-term variability on projections of twenty-first century dynamic sea level. Nat. Clim. Chang.
**2015**, 5, 343–347. [Google Scholar] [CrossRef] - Carson, M.; Köhl, A.; Stammer, D. The Impact of Regional Multidecadal and Century-Scale Internal Climate Variability on Sea Level Trends in CMIP5 Models. J. Clim.
**2015**, 28, 853–861. [Google Scholar] [CrossRef] - Kopp, R.E.; Hay, C.C.; Little, C.M.; Mitrovica, J.X. Geographic Variability of Sea-Level Change. Curr. Clim. Chang. Rep.
**2015**, 1, 192–204. [Google Scholar] [CrossRef] [Green Version] - Little, C.M.; Horton, R.M.; Kopp, R.; Oppenheimer, M.; Yip, S. Uncertainty in Twenty-First-Century CMIP5 Sea Level Projections. J. Clim.
**2015**, 28, 838–852. [Google Scholar] [CrossRef] - Church, J.A.; Clark, P.U.; Cazenave, A.; Gregory, J.; Jevrejeva, S.; Levermann, A.; Merrifield, M.; Milne, G.; Nerem, R.S.; Nunn, P.; et al. Sea level change. In Climate Change 2013: The Physical Science Basis; Stocker, T.F., Qin, D.G.-K., Plattner, M., Tignor, S.K., Allen, J., Boschung, A., Nauels, Y., Bex, X.V., Midgley, P.M., Eds.; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2013. [Google Scholar]
- Church, J.A.; Clark, P.U.; Cazenave, A.; Gregory, J.; Jevrejeva, S.; Levermann, A.; Merrifield, M.; Milne, G.; Nerem, R.S.; Nunn, P.; et al. Sea Level Change Supplementary Material. In Climate Change 2013: The Physical Science Basis; Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Stocker, T.F., Qin, D.G.-K., Plattner, M., Tignor, S.K., Allen, J., Boschung, A., Nauels, Y., Bex, X.V., Midgley, P.M., Eds.; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2013. [Google Scholar]
- Peltier, W.R. Global glacial isostasy and the surface of the ice-age earth: The ICE-5G (VM2) model and GRACE. Annu. Rev. Earth Planet. Sci.
**2004**, 32, 111–149. [Google Scholar] [CrossRef] - Lambeck, K.; Smither, C.; Johnston, P. Sea-level change, glacial rebound and mantle viscosity for northern Europe. Geophys. J. Int.
**1998**, 134, 102–144. [Google Scholar] [CrossRef] - Aral, M.M.; Chang, B. Spatial Variation of Sea Level Rise at Atlantic and Mediterranean Coastline of Europe. Water
**2017**, 9, 522. [Google Scholar] [CrossRef] - Chang, B.; Guan, J.; Aral, M.M. A Scientific Discourse: Climate Change and Sea-Level Rise. ASCE J. Hydrol. Eng.
**2015**, 20, A4014003. [Google Scholar] [CrossRef] - Chang, B.; Guan, J.; Aral, M.M. Modeling Spatial Variations of Sea Level Rise and Corresponding Inundation Impacts: A Case Study for Florida, USA. Water Qual. Expo. Health
**2013**, 6, 39–51. [Google Scholar] [CrossRef]

**Figure 1.**Location of the nine analyzed tidal stations: Marseille (5.35°–43.32°, 1885–2012), Trieste (13.76°–45.57°, 1875–2012), Genova (9.92°–44.41°, 1884–1997), Venice Punta Salute (12.32°–45.39°, 1909–2000), Bakar (14.5°–45.30°, 1930–2011), Dubrovnik (19.13°–42.65°, 1956−2009), Rovinj (13.64°–45.06°, 1955–2011), Split RT Marjana (16.39°–43.51°, 1952–2011), Split Gradska Luka (16.44°–43.51°, 1954–2011). Coordinates of the station and time span of recordings are indicated in parentheses.

**Figure 2.**Time series of the monthly mean tidal records from stations in Figure 1 (data from www.psmsl.org). Red lines show linear regressions on the data (see the text for details).

**Figure 3.**Function L

_{LT}(t) (black line), obtained by summing up IMFs at timescales >15 years and the residue, and the best-fit function L

_{M}(t) (red line) described by Equation (4).

**Figure 4.**EMD reconstruction by using IMFs with timescale >15 years (the residue r

_{n}(t) has not been added up) for the four closest stations in the North Adriatic Sea.

**Figure 5.**Map of the absolute (i.e., IGb08 frame) vertical velocities for the continuous GPS (cGPS) stations used in this work. Black dots refer to the stations ROGS (Bakar) and RIGS (Rovinj) for which, due to the limited available data, calculated vertical land motion (VLM) are not reliable.

**Figure 6.**Sea levels in the period 1888–2007 at Marseille as estimated by deriving the model from the first 70 years (1888–1957) of tidal data. Fit parameters are: r = 1.57 ± 0.04; c = 3882 ± 75; A = 13.8 ± 1.0, 9.8 ± 1.01; Φ = 1.22 ± 0.06, 1.89 ± 0.09, R

^{2}= 0.91. The model (black line) with 90% confidence interval (enclosed by black dashed lines), sea-level data at 1 year of resolution (red dashed line) and smoothed by a 15-year sliding window (red line), are shown. The vertical bar marks the boundary between sea level data used for the model (left) and estimated sea levels (right).

**Figure 7.**Sea-level height up to 2100 for the analyzed stations, as obtained by the model of Equation (4) that includes VLM and AR5 RCP2.6 (blue line) and RCP8.5 (red line) rate. Color bands represent the 90% confidence interval, obtained by including uncertainties from the model, VLM, and AR5 trend. The black line represents sea levels when the constant trend obtained from fits is used. Since cGPS data are not reliable, the contribution of VLM is not included for Bakar and Rovinj.

**Table 1.**Characteristic periods (years) of the significant intrinsic mode functions (IMFs) with a characteristic timescale >15 years, obtained by the Empirical Mode Decomposition (EMD) for the analyzed data set. Errors are estimated as the standard deviation.

Tide Gauge Station | n | j = 7 | j = 8 | j = 9 | j = 10 | j = 11 | j = 12 | j = 13 |
---|---|---|---|---|---|---|---|---|

Marseille | 15 | 29.8 ± 14.5 | 36.5 ± 7.4 | 43.1 ± 15.6 | 63.1 ± 21.2 | |||

Trieste | 15 | 19.8 ± 4.4 | 27.0 ± 3.6 | 37.6 ± 8.8 | 52.2 ± 12.7 | 76.3 ± 45.9 | ||

Genova | 10 | 15.3 ± 2.5 | 32.2 ± 2.1 | |||||

Venice P.S. | 12 | 16.0 ± 2.5 | 21.0 ± 2.1 | 44.3 ± 5.0 | ||||

Bakar | 12 | 16.9 ± 2.8 | 29.4 ± 7.7 | 58.0 ± 4.9 | ||||

Dubrovnik | 11 | 17.0 ± 2.5 | 49.4 ± 19.2 | |||||

Rovinj | 11 | 16.2 ± 5.6 | 36.6 ± 8.3 | |||||

Split M. | 13 | 18.6 ± 3.4 | 21.7 ± 4.5 | 33.7 ± 25.9 | ||||

Split G. | 11 | 17.8 ± 3.7 | 49.6 ± 16.4 |

**Table 2.**Fit parameters and adjusted R-square values for the analyzed data set. Uncertainties define the 90% confidence limits.

Tide Gauge Station | Linear Fit (mm a ^{−1}) | r (mm a ^{−1}) | c (mm) | A (mm) | Φ | R^{2} |
---|---|---|---|---|---|---|

Marseille | 1.22 ± 0.10 | 1.28 ± 0.01 | 4436 ± 12 | 10.1 ± 0.3 | 4.03 ± 0.30 | 0.9948 |

11.1 ± 0.3 | −0.84 ± 0.03 | |||||

16.1 ± 0.5 | 1.84 ± 0.04 | |||||

18.7 ± 0.5 | −0.42 ± 1.54 | |||||

Trieste | 1.22 ± 0.15 | 1.59 ± 0.05 | 3874 ± 86 | 10.8 ± 0.4 | 2.54 ± 2.11 | 0.9844 |

13.2 ± 0.8 | 2.34 ± 0.05 | |||||

10.3 ± 1.7 | 4.87 ± 0.73 | |||||

19.1 ± 2.9 | −3.04 ± 0.13 | |||||

21.9 ± 2.0 | 2.52 ± 0.05 | |||||

Genova | 1.07 ± 0.24 | 1.20 ± 0.03 | 4618 ± 54 | 7.8 ± 0.7 | −1.93 ± 0.08 | 0.9254 |

12.1 ± 0.6 | 2.58 ± 0.24 | |||||

Venice P.S. | 2.43 ± 0.23 | 2.78 ± 0.04 | 1615 ± 74 | 13.1 ± 1.1 | −0.61 ± 0.09 | 0.9579 |

18.8 ± 1.3 | −2.03 ± 0.07 | |||||

23.8 ± 1.3 | 2.10 ± 0.05 | |||||

Bakar | 0.9 ± 0.37 | 0.88 ± 0.15 | 5323 ± 306 | 21.1 ± 1.3 | 3.36 ± 0.89 | 0.9515 |

21.9 ± 2.2 | −1.63 ± 0.11 | |||||

12.1 ± 0.7 | −0.66 ± 2.36 | |||||

Dubrovnik | 0.90 ± 0.44 | 1.24 ± 0.05 | 4520 ± 104 | 27.4 ± 0.6 | 2.41 ± 0.02 | 0.9662 |

20.6 ± 0.8 | −1.74 ± 0.04 | |||||

Rovinj | 0.38 ± 0.43 | 0.97 ± 0.10 | 5144 ± 190 | 26.5 ± 0.9 | −0.45 ± 0.03 | 0.9014 |

20.3 ± 1.2 | 1.76 ± 0.09 | |||||

Split M. | 0.35 ± 0.38 | 0.17 ± 0.02 | 6671 ± 29 | 18.8 ± 0.3 | 2.62 ± 0.02 | 0.9888 |

26.8 ± 0.4 | −1.84 ± 0.01 | |||||

26.4 ± 0.4 | 1.35 ± 0.01 | |||||

Split G. L. | 0.51 ± 0.40 | 1.87 ± 0.07 | 3360 ± 137 | 17.1 ± 0.7 | 0.68 ± 0.04 | 0.9434 |

35.1 ± 0.1 | 1.48 ± 0.03 |

**Table 3.**VLM measured at the cGPS station closest to the tide gauges used in this work. The cGPS station name, position, rate of VLM, time interval of the time series, mean VLM, and the closest tide gauges to the cGPS stations are indicated. Negative sign in the VLM corresponds to land subsidence.

GNSS Station | Lon | Lat | VLM (mm a ^{−1}) | Time Interval | Average VLM (mm a ^{−1}) | Tide Gauge Location |
---|---|---|---|---|---|---|

PRIE | 5.3727 | 43.2768 | −0.65 ± 0.55 | 2007.6397–2018.0890 | −0.61 ± 0.43 | Marseille |

MARS | 5.3538 | 43.2788 | −0.58 ± 0.31 | 1998.5465–2018.0917 | ||

TRIE | 13.7635 | 45.7098 | −0.20 ± 0.36 | 2003.1054–2018.0917 | −0.08 ± 0.49 | Trieste |

TRI1 | 13.7878 | 45.6606 | 0.04 ± 0.61 | 2007.7821–2016.8948 | ||

GENV | 8.8809 | 44.4152 | 0.74 ± 0.90 | 2008.4631–2014.7602 | −0.12 ± 1.54 | Genova |

GENO | 8.9211 | 44.4194 | −0.25 ± 0.30 | 1998.5575–2018.0917 | ||

GENA | 8.9482 | 44.3976 | −0.51 ± 4.30 | 2016.6270–2017.9273 | ||

GENU | 8.9593 | 44.4027 | −0.50 ± 0.68 | 2009.9465–2017.9986 | ||

DUB2 | 18.1103 | 42.6502 | −1.13 ± 0.96 | 2011.9739–2018.0917 | −1.56 ± 0.72 | Dubrovnik |

DUBR | 18.1104 | 42.6500 | −1.99 ± 0.48 | 2000.7226–2012.7363 | ||

SPLT | 16.4385 | 43.5066 | 0.56 ± 0.77 | 2005.0013–2012.2527 | 0.56 ± 0.77 | Split |

CGIA | 12.2655 | 45.2065 | −2.91 ± 0.83 | 2010.9630–2018.0917 | −3.30 ± 0.85 | Venice |

SFEL | 12.2913 | 45.2300 | −4.57 ± 0.67 | 2001.5465–2011.1657 | ||

VEAR | 12.3578 | 45.4379 | −2.17 ± 1.35 | 2006.1602–2010.7164 | ||

VEN1 | 12.3541 | 45.4306 | −1.46 ± 0.65 | 2009.8068–2018.0917 | ||

VENE | 12.3320 | 45.4370 | −0.89 ± 0.82 | 2001.0863–2007.5630 | ||

VE01 | 12.3339 | 45.4375 | −2.22 ± 1.56 | 2007.8506–2011.1821 | ||

MSTR | 12.2386 | 45.4904 | −2.50 ± 0.79 | 2007.8890–2014.6315 | ||

TREP | 12.4547 | 45.4677 | −8.45 ± 1.69 | 2004.1871–2008.0724 | ||

CAVA | 12.5827 | 45.4794 | −2.71 ± 0.58 | 2001.5438–2011.1657 | ||

TGPO | 12.2283 | 45.0031 | −4.89 ± 0.53 | 2007.3164–2018.0917 | ||

PTO1 | 12.3341 | 44.9515 | −4.26 ± 0.59 | 2008.4631–2017.9301 | ||

GARI | 12.2494 | 44.6769 | −3.25 ± 0.66 | 2009.3958–2018.0917 | ||

RAVE | 12.1919 | 44.4053 | −4.23 ± 0.58 | 2008.0259–2017.9986 | ||

BEVA | 13.0694 | 45.6719 | −1.69 ± 0.57 | 2008.0368–2018.0917 |

**Table 4.**Sea levels for 2050 and 2100, with respect to 2005, at the tide gauge for the two scenarios RCP2.6 and RCP8.5 including both the VLM and sea level natural variability (SLNV) contribution. Uncertainties define the 90% confidence limits. Time span used for the modeling and duration of raw records (in parenthesis) are also indicated. Since cGPS data are not reliable, the contribution of VLM is not included for Bakar and Rovinj.

Tide Gauge Station (Duration) | Sea Level (mm) RCP2.6 | Sea Level (mm) RCP8.5 | ||
---|---|---|---|---|

2050 | 2100 | 2050 | 2100 | |

Marseille | 182 ± 79 | 364 ± 167 | 208 ± 79 | 602 ± 240 |

1888–2009 | ||||

(128 years) | ||||

Trieste | 142 ± 82 | 336 ± 197 | 150 ± 86 | 523 ± 237 |

1901–2009 | ||||

(138 years) | ||||

Genova | 163 ± 150 | 337 ± 306 | 193 ± 156 | 581 ± 347 |

1931–1992 | ||||

(92 years) | ||||

Venice P.S. | 283 ± 103 | 603 ± 217 | 311 ± 114 | 818 ± 258 |

1914–1997 | ||||

(92 years) | ||||

Bakar | 166 ± 69 | 259 ± 165 | 182 ± 70 | 475 ± 203 |

1949–2008 | ||||

(82 years) | ||||

Dubrovnik | 225 ± 91 | 445 ± 200 | 246 ± 95 | 681 ± 246 |

1959–2006 | ||||

(54 years) | ||||

Rovinj | 149 ± 65 | 295 ± 164 | 177 ± 80 | 510 ± 216 |

1958–2008 | ||||

(57 years) | ||||

Split M. | 191 ± 106 | 322 ± 213 | 220 ± 112 | 567 ± 249 |

1955–2008 | ||||

(60 years) | ||||

Split G.L. | 174 ± 106 | 376 ± 240 | 204 ± 112 | 621 ± 273 |

1957–2008 | ||||

(58 years) |

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Vecchio, A.; Anzidei, M.; Serpelloni, E.; Florindo, F.
Natural Variability and Vertical Land Motion Contributions in the Mediterranean Sea-Level Records over the Last Two Centuries and Projections for 2100. *Water* **2019**, *11*, 1480.
https://doi.org/10.3390/w11071480

**AMA Style**

Vecchio A, Anzidei M, Serpelloni E, Florindo F.
Natural Variability and Vertical Land Motion Contributions in the Mediterranean Sea-Level Records over the Last Two Centuries and Projections for 2100. *Water*. 2019; 11(7):1480.
https://doi.org/10.3390/w11071480

**Chicago/Turabian Style**

Vecchio, Antonio, Marco Anzidei, Enrico Serpelloni, and Fabio Florindo.
2019. "Natural Variability and Vertical Land Motion Contributions in the Mediterranean Sea-Level Records over the Last Two Centuries and Projections for 2100" *Water* 11, no. 7: 1480.
https://doi.org/10.3390/w11071480