Analyzing and investigating the dynamics of Land Use and Land Cover (LULC)) transformation using satellite data: A case study of Adana, Türkiye

The main aim of this research is to analyze and investigate the dynamics of Land Use and Land Cover (LULC) Transformation in Adana, Turkiye on a yearly basis from 2018 to 2022 using Machine Learning (ML) in Google Earth Engine (GEE) and GIS software tools such as ArcGIS Pro for mapping and presentation of results. Sentinel 2A Image collections were used for classification due to their high resolution (10m), wide swath and multi-spectral imaging capabilities. Distinctive and separate supervised classification was done on image collections for 2018, 2019, 2020, 2021 and 2022 using JavaScript in GEE. Seven (7) useful classes namely Water, Trees, Flooded Vegetation, Crops, Built Area, Bare Ground and Rangeland with reference to the ESRI Global Land Cover Map were analyzed and mapped. To achieve high accuracy results, various Supervised Machine Learning Algorithms were tested for the best solution, combined with visual inspection. The Overall Accuracy and Kappa Statistics for the main region on a yearly basis was 85% and 86% for 2018, 86% and 81% for 2019, 90% and 85% for 2020, 85% and 83% for 2021 and lastly 89% and 83% for 2022 respectively. However, despite various methods of obtaining accurate results for a large area, discrepancies and misclassifications in the classes with complex pixels and spatial information were observed. To address this problem, a distinctive subset covering key features and geography from the selected region and surrounding region was used to show the trend in changes and obtain better accurate results related to the difference in the classes over time. The Overall Accuracy and Kappa Statistics of the subset was 91% and 88% for 2018, 88% and 85% for 2019, 90% and 84% for 2020, 89% and 86% for 2021 and finally 85% and 80% for 2022 respectively. The main drivers for the changes in LULC areas were attributed to the expansion of Built-up Areas, Growth in Crop Production, Deforestation and Afforestation and Climate Change related issues.