121 RSL history within South and Southeast Asia from the Mid to Late Holocene, it is possible to indirectly indicate critical areas at risk to flooding. The comparison of different sea-level studies presented in the database or with newly studied sample sets, dated to similar ages and investigated in study sites in close proximity to one another, can help not only to detect where study results support each other, but also identify where there are disagreements between datasets and elucidate possible causes for this deviation. In a study conducted within the framework of this thesis, (Bender et al., in proofs), new field data from different islands within the Spermonde Archipelago were compared to a previous study by Mann et al. (2016) from the same study area. Mann et al. identified index points on an uninhabited island that indicated an RSL slightly higher than MSL, while they further showed deviating RSL results for the densely populated island of Barrang Lompo located nearby. Bender et al. (in proof stage) further studied surrounding islands and was able to support the deviating result for Barrang Lompo. The RSL indicated on this island is 0.8 m lower than that found on the remaining islands within this study. After indicating different environmental aspects including, anthropogenic activity, erosion patterns and protection features (Williams, 2013; Tahir et al., 2012), it appears that this lower RSL result is triggered by local subsidence of the island. Therefore, a rise in sealevel is not the only risk factor contributing to flooding. Local specialties, like an overload of the area due to a dense population, or environmental changes triggered by anthropogenic activity that might affect the local coastal environment and natural protection of the area, such as beach ridges or coral reef flats, are aspects that might increase the risk to flooding. Those aspects should be further analyzed since contradicting
The RSL height resulting from load model NAWI-A2 (Figures 5.4 and 5.5, solid lines) mainly decreases until 100 ka BP (SVM) or 110 ka BP (LVM) due to the viscoelastic relaxation following the previous deglaciation. Then, a further decrease is also caused by the glaciation of the northern hemisphere and the associated global sea-level fall. After 50 ka BP, the changes of the RSL height are dominated by the radial displace- ment induced by the increase of the ice load near Berkner Island. The minimum of the RSL height is reached at 56 ka BP (SVM, −141.4 m) or 51 ka BP (LVM, −80.9 m). Time of sediment deposition
Abstract — Climate change and the resulting sealevel rise are a serious threat to coastal zone in the world, including estuaries and coastal villages. The consequence of sealevel rise (SLR) on the Welsh coast is still ambiguous, and updating of current flood risk tools is necessary to mitigate the potential damages arising from the future flood events along the Welsh coasts. This study focuses on hydrodynamic impacts of SLR on Welsh coasts, particularly in the Dyfi Estuary at Mid Wales, using an advanced modelling tool – TELEMAC2D. The analysis of flood dynamics is based on the time varying water depth and velocity from the model, so that flood index, flood duration and hazard rate can be more accurately predicted and assessed. In this study a high-resolution TEELAC2D model was setup, calibrated and validated for the Dyfi Estuary with a number of SLR and river discharge scenarios. Results showed that with the commonly accepted SLR quantities, a considerable area within the Dyfi Estuary is to be affected, but the flood duration and hazard rate vary widely within the study area, providing the insightful details of the flood risk in the region for better and sustainable future coastal management.
Figure 1 presents the contribution to sea-level change resulting from the changes in the GIS following the LGM. We first examine the response when excluding the neoglaciation. The ice model (Fleming & Lambeck, 2004) defines the GIS as having reached its current extent by 7.5 ka BP. The signal is dominated by falling sea levels (crustal uplift), particularly pronounced in the southwest, east and north, where the GIS had expanded the most during the last glacial period relative to the present-day (Figure 1a, see also Le Meur & Huybrechts, 2001; Tarasov & Peltier, 2002). The maximum rates are of the order of -5 mm yr −1 , with the associated uncertainties,
Another small low-lying island-state that is worth noting is Singapore. The rise in sealevel poses threats to Singapore as it lies only 15 meters (m) above the mean sealevel, with about 30% of the island at less than 5 m above the mean sealevel, according to Singapore’s National Climate Change Secretariat (NCCS). Navaratnarajah (2015) relates SLR to increases of carbon emissions and temperature and notes that warming can cause Singapore’s sealevel to increase by 9.5 m, leaving 745,000 Singaporeans’ homes submerged. Moreover, Ng and Mendelsohn (2005) examine the impact of SLR in Singapore and explore whether Singapore should defend its coast or allow it to be inundated. Based on their study, the annual cost of protecting Singapore’s coasts is estimated to rise over time as the sealevel rises and will range from $0.3 million–$5.7 million by 2050 to $0.9 million– $16.8 million by 2100. Depending on the SLR simulation, the present value of these costs ranges from $0.17 million to $3.08 million. Clearly, the regions or countries most vulnerable to SLR include atoll countries and small islands states like Kiribati and Singapore. Hence, there is a need for relevant discussions and consensus on how SLR will be dealt with through adaptation or mitigation.
3.4 Cost determination
Once the kilometres of roads inundated are determined for each grid cell, the determination of costs for replacing these roads is performed. The costs for replacement are based on individual determinations in each grid cell for each road type. In the case of SLR, current costing analysis is carried out using the cost to build a new kilometre of road of each type, since sealevel inundation projects a total loss of road inventory. Other adaptation options are not explored with this analysis. The totals for each cell are then combined at a provincial level to provide a total road impact cost. Table 2 provides a detailed illustration depicting costing and summation of provincial damage. As illustrated, an analysis of the impact of one metre of SLR on the existing roadstock inventory of Province 1 and Province 2 is shown. The roadstock inventory is broken down by type and allocated to the corresponding CRU grid cell. The SLR projected in that grid is applied to the roadstock inventory and damages are assessed. The far right columns show the total kilometres and total cost impact of the projected SLR: 21.15 km and US$3 million, respectively, for Province 1 and 100 km and US$12.4 million for Province 2. These are examples only; full datasets can be seen in Appendix A.
Overall, the impact of the SLR is more pronounced in Pulau Duyong compared to Sandakan Town. This is probably because Sandakan Town is built up on a raised platform level that is relatively safe from the projected SLR for the year 2060. On the other hand, majority of Pulau Duyong which comprise of a low lying area and covered by mangrove forests that grow below the Mean High Water (MHW) mark. It is anticipated that there will be a reduction of about 53% of the existing 1,800 hectares of mangrove forest (about 958 hectares) due to erosion and inundation caused by the rise of 0.5 m sealevel in the year 2060. The loss is estimated to be around RM8.8 million.
Changes in tidal current velocities induced by sea-level rise are mostly compensated by the considered topographic changes of the Wadden Sea. The results demonstrate the significance of sea- level rise induced topographic changes in the Wadden Sea for estimating local effects of sea-level rise on tidal dynamics.
In the past decades, there has been much eort on estimating the inuence of internal climatic variability phenomena (e.g. ENSO, PDO, etc.) on regional sealevel. In the European region the dominating atmospheric mode is NAO. Recent studies have demonstrated the impact of the NAO on sealevel in the North Sea, and the inuence has been noted to be variable in time ( Jevrejeva et al. , 2005 ; Yan et al. , 2004 ; Tsimplis and Shaw , 2008 ; Dangendorf et al. , 2012 ). Woolf et al. ( 2003 ) have shown that a linear relationship between the winter sealevel anomalies and the NAO Index can be used to explain most of the variability in the North Sea, the Mediterranean and the eastern parts of the North Atlantic. Wakelin et al. ( 2003 ) found high correlations between sea levels along the western European coast and the NAO index in winter, especially for the German Bight, where wind stress accounted for over 90% of the observed sealevel variability during winter over the period from 1871 to 2008 ( Dangendorf et al. , 2013 ). Wakelin et al. ( 2003 ) estimated the sealevel changes per unit NAO over the northwest European continental shelf by considering results from a two-dimensional, depth averaged tide and surge model, and they found that the NAO sensitivities from observations were, in general, higher than those estimates from the model results. This discrepancy might in part be caused by other contributions such as temperature and rainfall which are also aected by the NAO, but were not included in their two-dimensional model. The latter was also conrmed by Tsimplis et al. ( 2005 ) in an overview about recent ndings to the NAO inuence on sealevel over the Northern European Shelf. In a companion paper
We were investigating the sensitivity of the surface climate with the Earth System Model of Intermediate Complexity (EMIC): Planet Simulator to changes of boundary (topography and sea land distribution) condition. We have demonstrated that changes induced only by slightly changing boundary conditions could have significant signals in the net precipitation and 2 m temperature fields. Spatial pattern of P − E changes certainly need deeper consideration and analysis, but this is out of scope in this work. However, important conclusions can be drawn from the analysis presented above. That is that cooling due to an idealized mountain uplift in the Mediterranean catchment area would cause shift towards dry hydrologic regimes,, while warming causes shift towards wet conditions. These changes are connected with global cli- mate change, such as shift in the global hydrological cycle, and they are due to land and sea redistribution on the glacial time scale. Although the correlation pattern between T and P − E may vary above the globe depending on the sea-land distribution, analysis of absolute fluxes and temperature (Tables 2.4 and 2.5) indicates that T and P − E are positively correlated in the Mediterranean catchment area. This result together with the available geological observation provides an useful consideration when developing reconstruction of hydrological conditions in the Black Sea catchment area during the glacial cycle. In chapter 4 we will analyze available hy- drological data for the past 20 ka. Experiments resembling glacial conditions (MASK, POMA, PELT) also imply less precipitation and evaporation, therefore, an additional sensitivity study is performed in order to investigate how big changes of P and E in the Black Sea catchment area would be needed in order to cause model changes as big as observed sealevel minimum during the past glacial cycle. This analysis is described in chapter 5. In chapter 6, we will syn- thesize results from chapters 2 and 4 together with available proxies to reconstruct hydrological conditions in the Black Sea catchment area during the past 120 ka. Changes of circulation pat- terns could also influence T and P − E patterns, therefore, we are also motivated to investigate changes of teleconnection patterns due to changes of orography (chapter 3).
The return water level assessment is not only uncertain regarding the heterogeneous assessment procedures or the limited water level information but also with respect to possible future projections related to climate change. Recent analyses highlight that global MSL rose by 3.2 mm/year from 1971 to 2010. As consequence from an increased ocean warming and the increased loss of mass from glaciers and ice sheets, future rates of sealevel rise (SLR) are expected to very likely exceed those observed during 1971 to 2010 (IPPC 2013). Until now, most coastal protection strategies assumed that changes in ex- treme water levels during the 21 st century will be dominated by changes in MSL and de- sign water levels were raised by exactly the same amount of the projected SLR (S MITH et al. 2010). These results are limited to the assumption of a similar long-term behaviour between mean and extreme water levels. For the German Bight, however, M UDERSBACH
As stated before, barystatic sealevel can be calculated from monthly GRACE gravity fields in two fairly independent ways: First, a global mean estimate of barystatic sealevel can be obtained from spatially integrating GRACE-based ocean bottom pressure anomalies over a certain spatial domain (i.e., the direct approach). This domain cannot be the whole ocean, since coastal regions are affected by spatial leakage, so that buffer zones extending several hundreds of kilometers off the continents are typically left out. Sec- ond, an observed mass distribution at the continents can be used to solve the SLE which provides a spatially variable mass-induced sealevel anomaly available in all oceanic locations including the coastal seas (i.e., the SLE approach). Any solution of the SLE allows one to calculate regional averages over a certain spatial domain that can be readily compared with the results from the direct approach. The misfit between both methods is indicative of the bias in continental mass caused by spatial leakage across the coast.
Finally, there are a host of dimensions to explore in future work. To begin with, key issues that have been omitted from this analysis include the special treatment of vulnerable areas such as low-lying islands and ports. In addition, this type of disaggregated coastal impact model could be extended to account for localized impacts like erosion, salt water intrusion on water resources, coastal tourism and recreation, and interactions with agricultural production. The treatment of uncertain extreme surge and the resulting damages could benefit from exploring different attitudes towards risk, as well sensitivity analysis around the potential for nonstationarity in the storm surge distribution, should warming increase the likelihood of sealevel extremes as in Grinsted et al (2013). Furthermore, CIAM considers storm surge in isolation, however natural disasters often combine flooding with wind damage (e.g., Superstorm Sandy in 2012). This work has begun to examine the potential for suboptimal outcomes, however there are deeper dimensions related to insurance markets and maladaptation that merit treatment. Lastly, future studies could extend the direct cost estimates presented here, integrating the high resolution of CIAM’s adaptation decisions with a CGE framework to determine the economy-wide welfare effects of sealevel rise.
3.1 Do SLR’s scientists offer enough certainties to decision-takers? Let us consider an airport’s governance particularly aware about SLR. It intends planning future mitigating investments for the long term. For instance, airport may be located on the Mediterranean coast for near seashore airports (Barcelona (1m) or Nice (2-3m) or, still, Roma (Leonardo da Vinci, 3m). Let us assume that the planner prospects for the future forty years (2055). What are the factors he should take into considerations? Can he reasonably rule out the assumption of finding another location for a new airport? An appropriate answer to this question needs having a clear understanding of the SLR consequences. Indeed, the decision-maker faces several unknown data. Mainly, the analysis defines three scenarios. The first one corresponds to IPCC’s view that considers that the SLR does not exceed one meter during the 21 st century. The second one comes from scientists that raise doubts about it. This need considering then the three following points: i) The relevancy of the 2°C assumption, ii) Sea-level rise and past warm periods and, iii) The relevancy of semi-empirical models that deal with SLR.
methods are based on a combination of models and data, which are more useful when trying to estimate hazard extent where there is limited historical information. Methods for shoreline analysis vary in approach and accuracy. Delineating and using shoreline positions from photos to compute shoreline change analysis also require identification and quantification of biases and uncertainties associated with the shoreline position, such as photo resolution, georeferencing error, shoreline position error (tidal changes), and shoreline digitization error because these errors or uncertainties in turn influence the estimation of the shoreline change rate (Genz et al. 2007). These studies on coastal recession in Ghana have also lacked information on the contributions of regional SLR to coastal recession in Ghana, which to a large extent is a result of deficient historical tide gauge data for Ghana. Despite the questionable quality of historical tide gauge data, the National Oceanic and Atmospheric Administration (NOAA 2013), attempted to compute sea-level trends for the Takoradi tide gauge station in Ghana, yielding a rate of 2.16 ± 0.39 mm/y of relative SLR, taking into consideration only monthly tide gauge data covering the years 1929 to 1965, and some selected monthly records for 1992, 2007, 2008, and 2009 that were considered to provide accurate data. There is the need for more information on SLR, how SLR contributed to historical shoreline change in Ghana, and if possible the need for predictions of future shoreline in Ghana. The research purpose is to provide this useful information for coastal adaptation strategies by analyzing historical shoreline change for the entire Ghana coast, and quantify and predict the contribution of SLR to the shoreline change in Ghana.
the northward shift of the ITCZ (20°N during August) is accompanied by the humid “SW” monsoon south of the ITCZ. Northeast directed trade winds prevail during winter times when the ITCZ has migrated to the southward position of 5°N. The trade winds and the overlying Sahara Air Layer are the prevailing wind systems over NW-Africa (Fig. 1a) and drive major dust transport from the Sahara-Sahel zone towards the Atlantic Ocean (Koopmann, 1981; Prospero and Lamb, 2003). Fluvial supply in the study area is restricted to the Senegal River north of Dakar but plays only a minor role in sediment supply (Redois and Debenay, 1999). Therefore, the surface sediments on the Senegalese shelf consist of quartz sands and carbonate shell fragments (McMaster and Lachance, 1969; Barusseau et al., 1988; Redois and Debenay, 1999). These sands have originated from palaeo-dunes that, in turn, have formed around the Last Glacial Maximum (LGM) as a result of increased aridity and wind strength (Sarnthein and Diester-Haass, 1977; Lancaster et al., 2002), and low sealevel providing widely exposed shelf areas. The Senegalese Shelf varies in width between 50 - 100 km (Hagen, 2001) and the average depth of the shelf break south of 15°N is found in about 100 - 150 m present water depths (McMaster and Lachance, 1969).
Figure 7a displays observational data and sea-level predic- tions generated in this study. It shows that none of GIA mod- els approximates the observations. The difference ranges be- tween around 14 m at 9 ka and 3 m at 2.5 ka. Bohai Bay’s old- est Holocene shoreline (∼ 9.7 ka cal BP) is at −17.2 m (ob- served), at c. −35 m (ANU) or at c. −10 m (ICE-X). The BRAD model predicts this shoreline to be at ∼ −20 m at 10 ka. Our observed shoreline elevation is similar to Sunda Shelf (c. −15 m; Hanebuth et al., 2011) but different to the islands of Tahiti (c. −28 m; Bard et al., 2010) and Barbados (c. −25 m; Peltier and Fairbanks, 2006). There are two ways to interpret this: (i) the age of the lowermost SLIP in core Q7 is overestimated due to old carbon contamination of the dat- ing material, or (ii) the relatively shallow shoreline position in our study area is a deviation from eustasy due to the lev- ering of the broad continental shelf in response to ocean load (e.g. Milne and Mitrovica, 2008). The similarity to the Sunda Shelf and the absence of contamination elsewhere in the sed- iment cores does indeed suggest that the broad-shelf effect (East China Sea shelf; Fig. 1) causes the shallow shoreline position. More SLIP data are needed to provide unequivocal evidence for it.
ABSTRACT: Sealevel change is one of the main factors which cause major impacts along the global coastlines and it vary widely in the past few decades due to global warming. Global sealevel change is usually caused by melting of land-based ice and thermal expansion, as water warms. As the global warm- ing based sealevel rise (SLR) is alarming in ocean waters, it is increasingly important to assess the effect of the same in all the coastal processes. . In this study, the effects of SLR on the tidal hydrodynamics along the Gulf of Khambhat, India are investigated. Selected major diurnal and semi-diurnal constituents, M2, S2, N2, K1, O1 and P1, have been taken up for a detailed investigation through finite-element based numerical modelling, TELEMAC-2D. The numerical model results of the existing conditions are com- pared with available literature data and found to be in agreement. Tidal propagation is predicted for the above mentioned six tidal constituents with and without sealevel rise conditions. Three sealevel rise sce- narios (0.1m, 0.5m and 1m) had been adopted for the study. The amplitudes and velocities of individual constituents, extracted along a specific stretch of gulf were compared with that of no SLR condition, to obtain the percentage variations.
For all other stations, similar improvement rates could be observed. Similarly, postfit residuals decreased for the range observations, while there are still some periodic signals visible in the azimuth residuals. The latter needs more investigations in order to optimize the estimation procedure. From the variance-covariance matrices of the 3D position solutions, error ellipses can be computed and analyzed in a local horizontal coordinate system (North-East-Height). By this the height errors in each direction can be extracted, which is important to quantify the resulting uncertainties of absolute sealevel heights. Ideally, one would expect mostly circular error ellipses, meaning that in all directions the same accuracy is available. In case the ellipse is more oblate, the accuracies vary with respect to the direction, which might again result from phase center variations of the ECR. In addition, these results need to be analyzed more deeply once a longer data span is available.
Figure 2: Maximum flood current velocity in the reference case (Ref) (a), changes due to sea-level rise (SLR - Ref) (b), changes due to sea-level rise and topographic changes ([SLR & TC] – Ref) (c) and residual changes due to topographic changes ([SLR & TC] – SLR) (d). The same for maximum ebb current velocity (e-h) and the ratio of maximum flood current velocity to maximum ebb current velocity (i-l). Depth contour lines are displayed for 0, 3, 10 and 30 m below NHN.