Abstract
Urban land grading can support sustainable urban planning by revealing spatial differences in commercial, transport, public service and environmental conditions. We developed a reproducible 100 m × 100 m grid-based framework for Haikou’s main urban area, China, using OpenStreetMap and planning-unit boundaries. We organized 14 secondary indicators into four dimensions representing commercial prosperity and centrality, transport conditions, public services and infrastructure, and environmental amenities and ecological constraints. We applied Criteria Importance Through Intercriteria Correlation (CRITIC) to derive local weights within dimensions and initial weights across dimensions. Final dimension weights minimized deviations from these initial weights under the standard’s range and ordering constraints. Across 71,292 evaluation cells, composite scores exhibited a pronounced right-skewed distribution. Grades 1–3, 4–5 and 6–7 accounted for 22.36%, 40.47% and 37.17% of the study area, respectively. A spatial gradient extended from Central’s contiguous high-grade core through transitional Changliu to predominantly middle- and low-grade Jiangdong hinterlands. High-grade areas generally combined commercial, transport, service and environmental advantages rather than a single transport or waterfront advantage. External assessment used 612 common 1 km cells with GDP and LandScan population data. Spearman’s ρ was 0.604 for GDP and 0.580 for population. The corresponding cluster-adjusted rank correlations were 0.424 and 0.405. These associations support external consistency, not parcel-level price validation. The framework provides relative land rankings and reveals spatial differentiation where market observations are incomplete. These outputs can support sustainable urban planning by informing the joint consideration of service provision and environmental conditions. They complement statutory boundary delineation and parcel-level land valuation.