Spatio-temporal assessment of land cover dynamics in Türkiye: a comprehensive analysis using remote sensing from 1985 to 2023


YILMAZ O. S., GÜLGEN F., BALIK ŞANLI F., ATEŞ A. M., ACAR U., GÜNGÖR R.

Advances in Space Research, 2026 (SCI-Expanded, Scopus) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Basım Tarihi: 2026
  • Doi Numarası: 10.1016/j.asr.2026.05.005
  • Dergi Adı: Advances in Space Research
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Artic & Antarctic Regions, Compendex, INSPEC, MEDLINE
  • Anahtar Kelimeler: GEE, Image classification, Random Forest, Spatio-temporal analysis, Spectral indices, Türkiye land cover change
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

Understanding long-term land cover (LC) dynamics is essential for effective environmental planning and sustainable resource management. Türkiye, characterized by diverse ecosystems, rapid urbanization, significant agricultural activities, extensive forested regions, and critical freshwater resources, offers an important case study for understanding the complex dynamics of landscape transformation. This research quantifies and analyses LC changes across Türkiye using multi-temporal satellite imagery from Landsat-5 (TM) and Landsat-8 (OLI) spanning 1985–2023. Analysis was conducted via the Google Earth Engine (GEE) platform, employing a pixel-based supervised classification with the Random Forest (RF) algorithm and multiple spectral indices including normalized difference vegetation index (NDVI), normalized difference water index (NDWI), normalized difference built-up index (NDBI), built-up index (BUI) alongside visible bands. Five primary LC classes were delineated across all 81 provinces, producing a total of 729 detailed LC maps that provide province-level land cover information for nine different time periods between 1985 and 2023. In the accuracy evaluation, the overall accuracy values ranged between 72.9% and 98.8%, Kappa coefficients ranged between 0.537 and 0.983, and F1-score — representing the harmonic mean of Precision and Recall — ranged between 0.758 and 0.984. The results revealed that forest areas experienced a net increase throughout the study period, despite a decline between 2020 and 2023. Surface water bodies expanded during 1985–1995 and 2005–2015 but declined in the intervening years. Urban settlements have shown consistent growth throughout the study period, exhibiting the highest rate of acceleration among all LC classes. Barren lands, while representing the dominant LC type, showed an overall decrease, with slight increases during 2000–2005 and 2020–2023. Agricultural areas expanded over the study period, although periodic fluctuations were observed. This comprehensive analysis of Türkiye’s LC dynamics, based on 8,937 satellite images, provides essential information for policymakers and environmental managers. The findings underscore the importance of continuous monitoring to develop effective land management strategies that can address climate change adaptation, food security, and ecosystem preservation challenges facing Türkiye in the coming decades.