University of Twente Student Theses
Evaluation of uncertainty associated with projections of climate change-driven coastline variations in Japan
Roover, S.A.W. de (2018) Evaluation of uncertainty associated with projections of climate change-driven coastline variations in Japan.
|
PDF
5MB |
| Abstract: | Global population and wealth are increasingly concentrated in coastal zones. On the other hand, climate change-driven sea level rise (SLR) will result in increased coastal erosion and thus, land loss. To avoid socio-economic losses, projections of coastal retreat have to be made. These projections ideally should fit in contemporary coastal management, in which hazard risk assessment is a key tool; knowledge of exceedance probabilities of coastal recession is therefore important. In Japan, projections of future beach erosion (up to 2100) have already been estimated using the Bruun rule (Udo & Takeda, 2017). However, Bruun rule-derived coastline projections (Bruun, 1962) are difficult to use in coastal hazard risk assessment, because they are deterministic and uncertainty from storms is not included (Cooper & Pilkey, 2004; Ranasinghe & Stive, 2009). The research goal was to quantify uncertainty related to sea level rise and storm definition in the Bruun rule-derived future shoreline positions in Japan, by comparing these projections with the results of the PCR models (Ranasinghe, Callaghan, & Stive, 2012). The comparison was carried out for three sites in Japan with sandy beaches and (almost) no hard structures. Beach evolution time series were unavailable and only a few (dated) profile cross-sections were acquired from literature. Visits to the sites did not resolve these issues. The erosion (Mendoza & Jiménez, 2006) and recession models were adopted from the PCR model for Hazaki beach, Japan (Da Cruz, 2018). Site 1 was considered similar in structure and wave climate and the Hazaki model was applied to site 1 without large differences. Sites 2 and 3 were not similar to Hazaki beach, but the erosion and recession models were still applied. Recalibrating the erosion and recession models was not possible due to insufficient data. For sites 1 and 2, two different storm definitions or detection methods were used. For site 3, one method was used. This resulted in five PCR models. With each model setup, shoreline positions were simulated between 2018 and 2100 over four SLR scenarios. This resulted in twenty data sets of simulated shoreline positions. Empirical cumulative distribution functions (ECDFs) describing coastal recession exceedance probabilities were produced with the PCR model results. Two ECDFs were made per data set based on annual maximum landward shoreline positions (Rmax) and shoreline positions derived from 5-year trend lines (Rtrend). Exceedance probabilities for Bruun estimates were derived with these ECDFs. For 2100 and the most severe SLR scenario, the exceedance probabilities derived with the ECDFs for Rmax for Bruun rule estimates were: 49% and 44% with the two PCR models for site 1; 18% and 77% with the two PCR models for site 2; and 43% with the single PCR model for site 3. The uncertainty due to SLR and storm definition could thus be quantified with the use of PCR produced shoreline ECDFs. The models were significantly more sensitive to the choice of storm detection method than to the choice of SLR scenario. However, no general quantified relation could be evaluated between Bruun rule estimates and their exceedance probabilities derived with the PCR model results. The shape of the curve that describes the temporal change of these exceedance probabilities is closest to an observable general relation. Besides these main conclusions, also other findings were reported: • PCR exceedance probabilities of Bruun rule estimates changed over time due to 1) the temporal increase of the SLR and Bruun rule recession projections and 2) the increase of exceedance probabilities of the same shoreline positions in the PCR model results; • The Hazaki PCR modelling methodology could be generalised to other sites with relative ease. The predictive successfulness of the models was unknown due to the lack of beach evolution data for validation; • Validated dates of erosion at one site could be used to sample storm wave characteristics from wave time series of a nearby beach, given that the nearby beach has a similar wave climate and beach structure. |
| Item Type: | Essay (Master) |
| Faculty: | ET: Engineering Technology |
| Subject: | 56 civil engineering |
| Programme: | Civil Engineering and Management MSc (60026) |
| Link to this item: | https://purl.utwente.nl/essays/76909 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page