با همکاری انجمن آبخیزداری ایران

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشجوی دکتری علوم و مهندسی مرتع، دانشکده منابع طبیعی و علوم زمین، دانشگاه شهرکرد، ایران

2 استادیار گروه مهندسی طبیعت، دانشکده منابع طبیعی و علوم زمین، دانشگاه شهرکرد، ایران

3 دانشیار گروه مهندسی طبیعت، دانشکده منابع طبیعی و علوم زمین، دانشگاه شهرکرد، ایران

4 استادیار گروه احیاء مناطق خشک و کوهستانی، دانشکده منابع طبیعی، دانشگاه تهران، ایران

10.22092/ijwmse.2025.371067.2135

چکیده

نقشه‌های کاربری/پوشش زمین نقش کلیدی در مدیریت منابع طبیعی و برنامه‌ریزی سرزمین دارند و پایش دقیق آن‌ها برای مدیریت پایدار ضروری است. در سال‌های اخیر، توسعه سنجش‌ازدور و الگوریتم‌های یادگیری ماشین همراه با سامانه گوگل ارث انجین (GEE)، امکان تحلیل کلان‌داده‌های مکانی–زمانی را فراهم کرده است. در این پژوهش، تغییرات کاربری/پوشش زمین حوزه آبخیز وسیع و ناهمگن کارون ۱ با استفاده از تصاویر لندست ۷ (۲۰۰۲) و لندست ۸ (۲۰۲۴) و به‌کارگیری سه الگوریتم CART، RF و SVM مورد بررسی قرار گرفت. تصاویر ترکیبی ۹ صحنه‌ای پس از اعمال فیلتر میانه تهیه شد و ۱۹۲۰ نقطه آموزشی در قالب ۷ کلاس کاربری/پوشش زمین استخراج گردید. شاخص‌های NDVI، NDBI، NDWI و مدل ارتفاعی (DSM) به‌عنوان داده‌های کمکی وارد مدل شدند. نتایج نشان داد هر سه الگوریتم از دقت بالایی برخوردارند و الگوریتم SVM بالاترین دقت کلی و ضریب کاپا را در هر دو سال (بیش از ۹۲ درصد) به دست آورد. نتایج همچنین حاکی از غالب بودن مراتع (حدود ۴۰ درصد) و جنگل‌ها (حدود ۲۷ درصد) در سطح حوضه و کاهش زمانی این کاربری‌ها، به‌ویژه پهنه‌های آبی، است. یافته‌ها نشان می‌دهد ترکیب الگوریتم‌های یادگیری ماشین با GEE ابزاری کارآمد برای پایش تغییرات محیطی در مقیاس حوضه‌های بزرگ است. با این حال، محدودیت وضوح مکانی داده‌های لندست از چالش‌های اصلی مطالعه محسوب می‌شود و استفاده از تصاویر سنتینل برای پژوهش‌های آتی پیشنهاد می‌شود.

کلیدواژه‌ها

عنوان مقاله [English]

Monitoring and Detection of Land Use/Land Cover Change Using Machine Learning Algorithms on the Google Earth Engine Platform: A Case Study of the Karun 1 Watershed

نویسندگان [English]

  • Sina Nabizadeh 1
  • Ali asghar Naghipour 2
  • Ataollah Ebrahimi 3
  • Hamidreza Keshtkar 4
  • Elham Ghehsareh 3

1 Ph.D. Student, Department of Nature Engineering, Faculty of Natural Resources and Earth Sciences, Shahrekord University, Iran.

2 Assistant Professor, Department of Nature Engineering, Faculty of Natural Resources and Earth Sciences, Shahrekord University, Iran

3 Associate Professor, Department of Nature Engineering, Faculty of Natural Resources and Earth Sciences, Shahrekord University, Iran.

4 Assistant Professor, Department of Arid and Mountainous Regions Reclamation, Faculty of Natural Resources, University of Tehran, Iran.

چکیده [English]

Land use/land cover (LULC) maps play a key role in natural resource management and sustainable land-use planning. Recent advances in remote sensing, machine learning, and cloud-based platforms such as Google Earth Engine (GEE) have enabled efficient large-scale spatiotemporal analyses. In this study, LULC changes in the large and heterogeneous Karun-1 watershed were assessed using Landsat 7 ETM+ (2002) and Landsat 8 OLI (2024) imagery. Cloud-free composite images derived from nine scenes during the peak growing season were generated using a median filter. A total of 1,920 training samples for seven LULC classes were extracted based on field data, aerial images, and Google Earth. Vegetation and auxiliary indices (NDVI, NDBI, NDWI, and DSM) were integrated with spectral bands. Supervised classifications were performed using CART, Random Forest (RF), and Support Vector Machine (SVM) algorithms in GEE. The results indicated that all three algorithms produced highly accurate LULC maps, with SVM achieving the highest overall accuracy and kappa coefficient (above 92% in both years). Rangelands (≈40%) and forests (≈27%) dominated the watershed area, while a declining trend was observed in rangelands, forests, and especially water bodies over time. The findings confirm the effectiveness of integrating machine-learning algorithms with GEE for large-scale environmental monitoring. However, limitations related to the spatial resolution of Landsat imagery remain a challenge. Therefore, the use of higher-resolution data such as Sentinel imagery is recommended for future studies.

کلیدواژه‌ها [English]

  • Remote sensing
  • Support Vector Machine (SVM)
  • Random Forest (RF)
  • Classification and Regression Tree (CART)
  • Auxiliary data