High-resolution land use and land cover change analysis using GEOBIA and landscape metrics: A case of Istanbul, Turkey

Topaloglu R. H., Aksu G. A., Ghale Y. A. G., Sertel E.

GEOCARTO INTERNATIONAL, cilt.37, sa.25, ss.9071-9097, 2022 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 37 Sayı: 25
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1080/10106049.2021.2012273
  • Dergi Adı: GEOCARTO INTERNATIONAL
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Aqualine, Aquatic Science & Fisheries Abstracts (ASFA), Environment Index, Geobase, INSPEC
  • Sayfa Sayıları: ss.9071-9097
  • Anahtar Kelimeler: Geographic Object-Based Image Analysis (GEOBIA), CORINE, URBAN ATLAS, SPOT, Landscape Metrics, IMAGE SEGMENTATION, SATELLITE IMAGERY, CLASSIFICATION, URBANIZATION, ALGORITHMS, PATTERN, AREA
  • İstanbul Teknik Üniversitesi Adresli: Evet

Özet

Determination of the spatio-temporal distribution of Land use and Land cover (LU/LC) is important to understand the dynamics of urbanization, agricultural abandonment, and industrialization. This study aims to create multi-temporal high-resolution LU/LC maps and analyze thematically extensive LU/LC changes using Geographic Object-Based Image Analysis (GEOBIA) and Landscape Metrics for the selected study region in the Istanbul metropolitan city of Turkey. HR SPOT 6/7 images acquired in 2009, 2013, and 2019 were used as main Earth Observation data to create LU/LC maps. Open-source geospatial data were also integrated into classification to better identify some LU/LC classes to increase total classification accuracy. Overall classification accuracy of 2009, 2013, and 2019 dated LU/LC maps are 87.45%, 88.16%, 90.74% respectively. Principal Component Analysis (PCA) and Pearson correlation were used to selecting the landscape metrics and evaluate the results. PCA resulted in three principal components and the total variance was found as 87.3%.