Potential Land-Use Functional Conflict Maps and Near-Term Built-Up Expansion: A Prospective Assessment on Hainan Island, China

Suitability-overlap maps of potential functional conflict are widely used to identify potential land-use trade-offs, yet their ability to identify subsequent conversion is rarely tested. We examined whether potential functional conflict (C), measured by suitability overlap, adds information about near-term conversion to built-up land (used here as the umbrella term for construction land in the Chinese land-use context) beyond the spatial continuity of existing development. Six CNLUCC land-use datasets for Hainan Island from 2000 to 2025 were analyzed as five 5-year windows on a 1 km grid. Baseline C and development pressure were evaluated against subsequent conversion of cropland and forest/grassland using event coverage, PR-AUC, temporal validation, geographically buffered spatial validation, and a 2020–2025 out-of-time evaluation. Conversion of both source types peaked in 2010–2015. The 20% of eligible grid-period observations with the highest C scores captured 14.91% of cropland events and 5.43% of forest/grassland events, compared with 53.05% and 58.22% for development pressure. Adding C to the development-pressure model produced gains that differed by source type and validation design; gains in the 2020–2025 evaluation were small and uncertain. All candidate models performed markedly worse under buffered spatial validation than under temporal validation. Sensitivity analyses showed that the contribution of C depends on analysis scale and model specification. Suitability overlap and development continuity therefore identify different planning signals: potential functional trade-offs and spatial exposure to near-term conversion.