**4. Discussion**

It is unknown to which extent SLR induced by man-made climate change will influence future resource demand and waste flows. However, this aspect needs to be considered in regard to adaptation actions, especially in countries characterized by large coastal regions with limited economic possibilities and resource availability. One first step in trying to fill this knowledge gap is to estimate the material stock that will be lost due to the permanent inundation of buildings. This study is a first attempt at quantifying the amount of materials rendered unusable by SLR. By explicitly quantifying the number of buildings and amount of materials that will be under water, policymakers have a valid dataset which can contribute to planning adaptation actions to climate change.

The results show that the main material that will be lost is concrete, a structural material which is notorious for its high carbon emissions and energy requirements [51,52]. This is bad news for the environment but it is also an opportunity for transitioning towards more sustainable building solutions. Fijian policy makers are in time for planning relocations, incentivize the use of alternative structural materials, and consider appropriate actions for the large amount of concrete that will need to be demolished. Concrete can be recycled into new concrete [53–55], or downcycled for the formation of road beddings and railway ballasts [56–58]. Timber can be carefully disassembled and used in new constructions [59–61], or it can be used for generating energy [62–64]. Steel should be brought to smelting facilities and remanufactured into new steel products [65]. The fraction of materials that cannot be kept within the economy should be opportunely disposed in landfills, whose number, location, and size should be discussed with local communities and waste managers. The discussed waste treatment would require the establishment of new recycling and incineration facilities. Therefore, reusing and recycling demolition materials should be considered and will likely bring benefits to the environment in the long term, but will add to the costs of relocation in the short term (e.g., operation costs, facility costs, labor costs).

By 2050, the total expected demolition flows due to SLR are 900 Gg, averaging 30 Gg per year. Considering that in 2017 and 2018, the amount of construction materials in new buildings was on average 1170 Gg per year [36,37], SLR demolition flows equate to 2.6% of the total yearly material requirements for new constructions. Such a small fraction has the potential for being absorbed into new constructions without the need for dumping materials into landfills—albeit this does not consider demolition flows coming from normal activities not related to SLR. Fijian policymakers should consider facilitating the inclusion of construction and demolition waste into new buildings by legislating opportune regulations.

While concrete and timber are mainly lost in rural areas, urban areas show a large amount of steel loss. Nevertheless, adaptation policies might prevent urban areas from being flooded (e.g., constructing dams), which would thus protect existing buildings and prevent their demolition.

In both scenarios, rural buildings account for more than 80% of the inundated buildings. This is indicative of the fact that largest settlements are located inland, while coastlines are characterized by scattered buildings. The retreat of coastlines will push people further inland, likely increasing urbanization rates and shifting material demands toward reinforced concrete, as this is the preferred structural material used in cities. Future research shall include the influence climate change has on migration patterns in Fiji.

The relocation of Vunidogoloa cost US\$15,625 per house [66]. If applied to the number of inundated rural buildings in this study, this would tally to US\$96.13 million for scenario 1 and US\$128.42 million for scenario 2. Yet, this simple estimation does not take into account the full impact of SLR, as aside of relocating, people will lose agricultural land. Considering that over 80% of inundated buildings are in rural areas, the people a ffected by SLR will have either to purchase or be assigned new land, or find di fferent occupations in cities. A thorough analysis of the economic implications of the impacts of SLR on the Fijian economy should be addressed by future research.

The overall di fference between scenario 1 and scenario 2 (+ 28%) is relatively low in comparison to the di fference of SLR between the two (+ 182%). These findings shed new light on the results of Gravelle and Mimura [13]. While any increase in SLR will results in a proportional reduction of land, the same does not hold for buildings, as losses are dependent on their spatial distribution. This shows that there is not necessarily a relation between the total area inundated and the number of buildings flooded, and that for an accurate assessment of the techno-economic impacts of SLR on human settlements specific studies are needed.
