Global assessment of human exposure to sea-level rise to 2300

The PROTECT project [1] includes a probabilistic integrated assessment of global population exposure to coastal flood hazard under climate-induced sea-level rise (SLR) over the next three centuries (to 2300). The assessment synthesises present-day datasets on population distribution [2], low lying coastal elevations [3] and extreme tides [4] with probabilistic projection datasets of population [5] and sea level [6] to 2300. For the scenarios considered (SSP1-2.6 & SSP2-4.5) and at a global scale, the median human exposure to coastal flood hazards grows substantially but then peaks in the early 2200s and subsequently slowly declines by 2300, despite continued rise in sea level.

Previous assessments have primarily focussed on shorter timeframes [2], typically to 2100, while it is widely acknowledged that even if temperatures are stabilised, sea levels are almost certain to continue to rise for many centuries [7][8][9]. Stakeholder workshops carried out with practitioners under the umbrella of PROTECT [10] and literature reviews [11][12] highlight the importance of extending sea-level rise information beyond 2100, to support strategic coastal adaptation and management, land-use planning, and critical infrastructure design.

Recent advancements in long term socio-economic modelling [13][5] now provide projections of global population and GDP at country level to 2300. These have already been applied to long-term risk assessments for other climate sectors [13][5][14].

For this assessment, the global coastline was split into ~29,000 segments, each assigned an extreme tide curve (from the COAST-RP dataset [4]) and a hypsometric curve, generated from a global terrain model [3] and present-day population distribution [2]. The hypsometric curves aggregate the total land-area and population at each elevation, including consideration of hydraulic connectivity to the coastline. This gives the land area and population that would be exposed at a given coastal flood level (up to 20mAMSL) for each coastal segment.

When sea-level scenarios [6] (SSP1-2.6 & SSP2-4.5) and socio-economic data [5] are combined, the human exposure and land area exposure to coastal flood hazard under a chosen extreme tide return period (or the annual average based on the event-exposure curve) is calculated.

This approach facilitates efficient computations, sampling across probabilistic data, and providing robust statistics at a high spatial resolution compared to traditional methods. The outputs at each coastal segment can be aggregated to sub-national, national, or the global scale.

In this analysis, it is found that the median exposure of people to coastal flood hazards increases fivefold to a peak in the early 2200s and subsequently slowly declines to 2300 in both SSPs, despite the continued rise in sea level. For the 80^th percentile population exposure grows even more (10- to 11-fold) but then stabilises rather than declines. These results reflect the interplay of sea level and demography with fall in global population in the latter half of the assessment period and are contrary to conventional wisdom. This analysis shows that in addition to sea-level rise, it is important to consider demographic trends when considering coastal futures.

Figure 1. Probabilistic annual average global population exposure to coastal flood hazard

References exceed the word limit so not included