Modelling Climate Migration in Southeast Asia due to Sea-Level Rise

Southeast Asia (SEA) is among the regions that are most at risk of experiencing climate change-induced human migration. It is home to more than 675 million people, many of whom reside in lower-middle income countries, with pre-existing social vulnerabilities and poverty, largely in low-lying coastal cities. SEA regularly experiences natural hazards, including floods and typhoons, whose impacts are expected to be magnified by climate change, and compounded by newer hazards, such as sea-level rise (SLR). Some level of climate migration is inevitable in this region, but with appropriate management, the scale of migration can be reduced and can serve as an effective adaptation strategy.

To better understand the scale of SLR-induced migration, as well as to identify the areas and population groups most at risk, this study employs a radiation model, which has previously been shown to be successful in predicting internal migration flows. In this study, the radiation model is used to predict the migration flows in SEA due to SLR, and to identify hotspots of sending and receiving regions of migrants. Specifically, future migration flows in Indonesia and Philippines are modelled from 2050 to 2100, looking at baseline non-climate migration, as well as SLR-induced climate migration, for a range of Representative Concentration Pathways (RCPs) and population growth projections. Preliminary results show a ~25% increase in migrant outflows under SLR scenarios relative to baselines, and a ~50% increase in inter-provincial migration rates. This indicates that SLR-driven migration will be significant for these countries and needs to be managed in a way that does not exacerbate existing vulnerabilities. This entails implementing in-situ adaptation measures in locations where it is still feasible, and facilitating relocation where in-situ adaptation would not be feasible. The radiation model is an effective way to predict migration flows at sub-national and national scales, and can be applied to other countries in SEA, given that the data inputs for the model are readily available.