Development of Grid-Based Population Projection Method: A Modified Cohort Component Approach Applied to South Korea

This study develops a method for downscaling population projections from municipalities to a 500 m grid level. The method adapts the cohort component approach, utilizing grid-level age and gender population data as weights to distribute municipal-level projections. This methodology was applied to South Korea, which has recently been experiencing population decline and sharp spatial disparities in population distribution. Using 2018 as the base year, projections are made at 5-year intervals to 2038. Results show significant spatial variation in projected population changes across South Korea, with 66.7% of populated grids expected to experience population decline by 2038. The model's accuracy was evaluated by comparing 2023 projections to actual data, revealing challenges in rapidly developing areas but better performance in stable regions. We further classify grids into five demographic typologies such as extinction risk and functional decline areas, to identify vulnerable locations and support targeted policy responses. These typologies show a contrast between resilient urban centres and shrinking rural peripheries, highlighting the need for differentiated spatial strategies. This grid-based projection method offers a valuable tool for place-based policymaking, urban planning and infrastructure development in the context of demographic change.