*5.1. Conclusions*

This study presents an estimation of the lost construction material stock due to the permanent rise of sea levels caused by man-made climate change. The results are designed to provide governments, research institutions, non-governmen<sup>t</sup> organizations and residents a statistical foundation for sustainable long-term planning. In particular, they can be used as baseline data for processes of spatial planning, especially to identify highly a ffected areas and to plan potential resettlements.

This methodology can be easily applied to other regions and can be fairly simply automated through the use of ArcGIS ModelBuilder or similar applications. This would provide vital information to local and national policymakers for deciding on the best course of actions to adapt to climate change and prepare for the consequences of SLR.

The present study could be expanded in various directions. The GIS-DIAM showed that infrastructure, especially the coastal highways on Viti Levu, are threatened by inundation. While this will not generate demolition flows, as roads are rarely removed [67], new roads will need to be constructed. To plan future material consumption even more su fficiently, future viability plans and associated material requirements should be considered. Additionally, more precise material indicators which include additional materials such as glass, aluminum, or copper, should be calculated to increase the robustness of the results.

Relocations, waste flows, loss of agricultural land, and future resource demands could be placed in a more encompassing economic model to quantify the overall costs associated to SLR. If possible, integrating the current digital terrain model with a digital elevation model which includes buildings could allow for the calculation of the exact height of constructions. Moreover, a highly detailed LiDAR map should be used to analyze in more detail the inundation pattern. This would allow for the possibility of simulating water barriers, allowing for the study of di fferent solutions in relation to their cost and e ffectiveness. For better predictions, future research should consider increasing the epochs in future time where inundation maps are generated for. By doing so, trends can be better analyzed and the study's predictions could be compared to what has actually happened in the future. The present study's results could be additionally used on Fiji to create maps with secondary resources including an approximate date on which they would become available.

#### *5.2. Limitations and Assumptions*

The main limitation of this study comes from the limited data available in relation to building typologies and construction techniques, which is a common problem when conducting research in developing regions [68]. Material intensities di ffer for each building, while construction typologies are not provided in o fficial maps. The spatial distribution of the typologies, as well as material intensities, were thus assigned on a provincial level rather than being specifically assigned to each building. Additionally, o fficial maps do not report the number of floors for each building, and digital elevation maps only report the terrain height instead of the actual ground elevation. Moreover, over 7000 buildings that were visible from aerial photos were not included in the vectorized digital maps, and had to be manually added.

Additional limitations occur from the DEM and the predicted SLR in regard of a certain point of time. The choice of the years the study was conducted for is to give an overview on two periods in time, one relevant on short-term planning and one relevant on long-term planning. The results stay relevant even if SLR predictions change and therefore provide a good orientation. The accuracy of the DEM is unknown, which unfortunately limits the accuracy of the results. However, this DEM was the best accessible DEM for Fiji during the time this research was conducted (June 2019). Further, this study is meant to provide an overview on the situation and identify the most critical areas rather than presenting highly detailed results on MS. Manual observation was used as a comparative method to verify whether the federal data are more suitable than the SRTM data or the ALOS data. Using ArcMap, it was proven that the governmental data are more consistent and furthermore, less restricted by the limitation of displaying surface elevation heights rather than ground elevation heights. This is particularly due to the fact that the governmental data incorporate the limitation of showing surface elevation more precisely than the SRTM data and the ALOS data. Research was performed to identify whether lost material stock in buildings inundated due to climate change induced SLR is an issue that should be considered in further climate change adaptation research and policies.

Further limitations can be found regarding the implications of actions to adapt to climate change. The construction of water barriers can significantly change the predicted inundated areas, rendering buildings that were predicted to be underwater still usable. As it is not possible to forecast if and when a water barrier will be constructed, this study disregards this option and presents results as if no water barriers will be erected.

In addition, results account only for buildings which will be actually inundated. Buildings which are not inundated might be included in relocation processes, especially in those cases where the majority of the buildings in a village will be flooded. This decision has to be considered from case to case and is therefore not possible to be included in this study.

Storms, coastal erosion, and the salination of agricultural land and water storages might cause additional relocations. Models on the permeation of seawater in agricultural land and aquifers are required to include these aspects for a more accurate prediction.

This study neglects approximately 36,000 people (about 4% of the national population) that are not living on the two main islands. These primarily rural residents could cause further challenges because relocations can eventually not happen on smaller islands and therefore increase the density of the larger islands.

In spite of its limitations, this study was conducted to provide an overview on the situation and its significance for climate change adaptation. The limitations do not influence the goodness of results in a way that would result in significant di fferences.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2071-1050/12/3/834/s1, Table S1: Typical number of floors in buildings by province, including references, Table S2: Distribution of construction typologies per province for scenario 1 (2050), Table S3: Distribution of construction typologies per province for scenario 2 (2100), Table S4: Material intensities of cement block masonry walls, construction typology based on [13], measurements based on [14] densities according to [15] (concrete), [16] (steel), [14] (timber). Note that GFA = Gross Floor Area, Table S5: Material intensities of corrugated iron walls and steel frame buildings, construction typology based on [13], measurements based on [14], densities according to [15] (concrete), [16] (steel), [14] (timber). Note that GFA = Gross Floor Area, Table S6: Material intensities of timber frame buildings, construction typology based on [13], measurements based on [14], densities according to [15] (concrete), [16] (steel), [14] (timber). Note that GFA = Gross Floor Area, Table S7: Results of the material stock analysis. Unit: Gg (note: 1 Gg = 109 g = 103 t = 1 kt).

**Author Contributions:** Conceptualization, S.M., A.M., S.W., H.T. and L.S.; Data curation, A.M.; Formal analysis, S.M. and A.M.; Funding acquisition, S.W. and L.S.; Investigation, S.M.; Methodology, S.M., A.M. and S.W.; Project administration, A.M. and S.W.; Resources, S.M.; Software, S.M.; Supervision, H.T. and L.S.; Validation, A.M., S.W., H.T. and L.S.; Visualization, S.M., A.M. and S.W.; Writing—original draft, S.M., A.M. and S.W.; Writing—review & editing, S.M., A.M., S.W., H.T. and L.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** The authors acknowledge the financial support by the DFG (German Research Foundation) in the framework of the Excellence Initiative, Darmstadt Graduate School of Excellence Energy Science and Engineering (GSC 1070) as well as by the Open Access Publishing Fund of Technische Universität Darmstadt.

**Acknowledgments:** The authors are grateful to the staff at Geospatial Division of the Ministry of Lands & Mineral Resources Fiji, in particular M. Hicks, V. K. Raqona and N. N. Kumar for their kind assistance in providing data.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Maps throughout this paper were created using ArcGIS ® software by Esri. ArcGIS ® and ArcMap ™ are the intellectual property of Esri and are used herein under license. Copyright © Esri. All rights reserved. For more information about Esri ® software, please visit www.esri.com.
