**3. Results**

The following Section 3.1 illustrates the results of the spatial analysis of the sea level rise, while Section 3.2 displays the results of the material stock analysis. A machine-readable Supplementary Data File with the values used to plot Figure 2, Figure 3, and Figure 8 is provided.

**Figure 2.** Number of inundated buildings in urban and rural areas of Fiji by province. Scenario 1 refers to projections to 2050 (+ 0.22 m), scenario 2 refers to projections to 2100 (+ 0.63 m).

**Figure 3.** Percentage of inundated buildings compared to the total buildings currently standing. The inner circle represents the percentage for 2050 (+ 0.22 m); the outer circle represents the percentage for 2100 (+ 0.63 m).

#### *3.1. Results of the GIS-DIAM*

Figure 2 displays the number of inundated buildings per province in both scenarios, separated in rural and urban areas, while Figure 3 indicates the percentage of all existing buildings that will be lost in scenarios 1 and 2. The results illustrate that buildings along coastlines will be a ffected and that every province will su ffer losses, albeit to di fferent extents. In total, 7472 buildings will be inundated by 2050 and 10,304 by 2100. This is equivalent to 4.5% of the overall number of currently existing buildings in scenario 1 and 6.2% in scenario 2. Until 2050, the average number of inundated buildings per year will be 241. After that, the rate decreases 57 buildings per year on average.

The rural proportion accounts for about 80% in both scenarios. Nevertheless, Fiji's major urban areas Suva, Lautoka, Lami, Labasa, Nasinu and the suburban region in Nadi will be heavily a ffected. Of all the inundated buildings in Serua, Cakaudrove, Nadroga Navosa, Namosi, Ra, and Tailevu, over 90% is located in rural areas. Conversely, the provinces of Naitasiri and Macuata are expected to experience the majority of their inundated buildings in urban areas.

The spatial distribution of the inundated buildings in 2050 is plotted in Figure 4 (Vanua Levu) and Figure 5 (Viti Levu). The total number of inundated buildings is the highest in Ba (1737 buildings), Rewa (1450 buildings), and Tailevu (1127 buildings). Yet, the percentage of buildings inundated in comparison to the total number of buildings is the highest in Serua (20%), with one out of five buildings flooded. It is remarkable that less than 2% of the buildings in Naitsairi and in Namosi will be directly a ffected.

**Figure 4.** Distribution of permanently inundated buildings in Vanua Levu in 2050. Each province is colored according to the percentage of inundated buildings compared to the total existing ones.

**Figure 5.** Distribution of permanently inundated buildings in Viti Levu in 2050. Each province is colored according to the percentage of inundated buildings compared to the total existing ones.

Areas with many buildings being inundated in a relatively small spatial frame can be identified in the town centers of Lami and Labasa, as well as in the peri-urban areas of Nadi, Nasinu, and Lautoka, including industrial and touristic locations. Additionally, the south eastern part of Viti Levu, which is broadly and relatively densely settled, is expected to face large inundations.

The spatial distribution of the submerged buildings for 2100 is plotted in Figures 6 and 7. There is a relatively low difference in regard to the amount of buildings predicted to be inundated in 2050 and 2100: compared to a SLR difference of 182% between the two, the number of inundated buildings increases by only 38%. Nevertheless, the rise of inundated buildings in urban areas increases by 58%, with Ba (146%) and Naitasiri (90%) being affected the most. Naitasiri, and furthermore, Bua and Nadroga Navosa also show a relatively high rise regarding the additional inundation of rural buildings in 2100.

The province with the highest number of inundated buildings is still Ba (2500 buildings), followed by Rewa (2009 buildings) and Tailevu (1460 buildings). Serua remains the most affected province, with 23% of its currently standing buildings expected to be permanently underwater by 2100.

**Figure 6.** Distribution of permanently inundated buildings in Vanua Levu in 2100. Each province is colored according to the percentage of inundated buildings compared to the total existing ones.

**Figure 7.** Distribution of permanently inundated buildings in Viti Levu in 2100. Each province is colored according to the percentage of inundated buildings compared to the total existing ones.

#### *3.2. Results of the MSA*

Figure 8 plots the lost material stock for the 2050 and 2100 scenarios (an equivalent table can be retrieved in the Supplementary Materials §4, and in the supporting data file for a machine-readable format). Our simulation predicts that, by 2050, 816 gigagrams (Gg) of concrete, 52 Gg of timber and 32 Gg of steel is stocked in buildings that are likely to be inundated. On average, every year will see 26 Gg of concrete, 1.7 Gg of timber, and 1 Gg of steel rendered unusable because of SLR.

**Figure 8.** Amount of construction materials lost in gigagrams (Gg) due to sea level rise in Fiji. S1 refers to the 2050 scenario (+ 0.22 m); S2 refers to the 2100 scenario (+ 0.63 m). The blue colors indicate materials located in rural areas, while the red colors represent urban areas.

In both scenarios, concrete accounts for about 90% of the total material lost by mass, followed by timber and steel, with 6% and 4% respectively. Steel will be mainly lost in urban areas while timber and concrete mainly occur in rural areas, due to the prevalence of this construction type in more urbanized areas. In total, most material will be lost in the provinces of Ba, Tailevu, Serua, Nadroga, Navosa, and Rewa.

The differences between the 2050 scenario and the 2100 scenario are proportional to the number of buildings that are additionally inundated in scenario 2. While steel rises by 62%, concrete rises by only 26% and timber by 34%. Those values are similar to the increases that can be seen when observing the changes of buildings being inundated in rural areas (33%) and urban areas (58%). The amount of lost material grows more than 100% in the urban areas of Naitasiri and Ba.
