Land Use/Land Cover Change, Artificial Surfaces, and Land Surface Temperature in a Border City: A GIS-Based Assessment for Planning in Edirne, Türkiye

Medium-sized border city contexts remain underrepresented in land use/land cover (LULC)–land surface temperature (LST) research despite distinctive governance con-straints and riparian eco-connectivity. Multi-temporal Landsat imagery (2000–2018) was processed using Google Earth Engine (GEE) to generate LULC and LST dynamics maps in Edirne (north-western Türkiye and structured by the Meriç–Tunca–Arda corridor system). We implemented a reproducible GEE workflow integrating Coordination of Information on the Environment (CORINE) Land Cover (CLC)-based LULC change assessment with Landsat-derived summer LST to diagnose urban heat island (UHI)–relevant thermal patterns. LULC mapping was robust (overall accuracy: 92.3–93.9%; κ = 0.85–0.88) and trends were quantified via linear regression. Agricultural land declined from 79,114 ha (93.4%) to 72,815 ha (85.9%) (net ≈−6300 ha; −326.2 ha yr−1; R2 ≈ 0.80), while built-up area more than doubled from 1086 ha (1.28%) to 2516 ha (2.97%) (+85.0 ha yr−1; R2 ≈ 0.78). Natural/semi-natural cover increased by ≈ +4800 ha (R2 ≈ 0.73), yet the area-weighted mean LST increased from 29.66 °C to 30.91 °C (Δ = +1.26 °C). Decomposition indicates modest cooling from composition change (−0.14 °C) but dominant within-class temperature changes consistent with climatic forcing (+1.40 °C). Thermal hotspots align with artificial surfaces and transport corridors, while riparian blue–green infrastructure (BGI) provides cooling; future work should link observed transitions to independent driver evidence.