Scholarlink

<p>Simulating land cover changes using machine learning and Geographic Information System (GIS) techniques has become an important tool for monitoring land dynamics that influence sustainable development goals (SDGs). This study aimed to map, quantify, and predict land use/land cover (LULC) changes in Chaurjahari Municipality, with an emphasis on built-up area expansion. Landsat imagery from 2001, 2010, and 2018 was classified using the maximum likelihood classifier to produce annual LULC maps. Changes were analyzed with the Land Change Modeler (LCM) in ArcMap 10.5.1, supported by explanatory variables from verified sources to model scenarios for 2030 and 2050. Results showed that built-up areas expanded by 4.28%, primarily replacing agricultural land, which declined by 4.03%. Forest, bare land, and water bodies experienced smaller shifts. The Multi-Layer Perceptron (MLP) model demonstrated a calibration accuracy of 73.4% with a skill measure of 0.6813, while the predicted 2018 map showed an accuracy of 78.48%, validating the model for future projections. Built-up areas are projected to grow to 9.02% by 2030 and 12.97% by 2050, while agriculture is expected to shrink further. These findings provide vital insights for the international research community, local governments, and urban planners to support evidence-based planning, sustainable urban growth, and informed policy formulation aligned with global sustainability targets.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning and GIS based simulation of urban expansion for sustainable planning in Chaurjahari of Nepal

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年9月11日

查看全文
  • Milan Budha,
  • Mahendra Singh Thapa,
  • Sandesh Gharti,
  • Badri Bhattarai

摘要

Simulating land cover changes using machine learning and Geographic Information System (GIS) techniques has become an important tool for monitoring land dynamics that influence sustainable development goals (SDGs). This study aimed to map, quantify, and predict land use/land cover (LULC) changes in Chaurjahari Municipality, with an emphasis on built-up area expansion. Landsat imagery from 2001, 2010, and 2018 was classified using the maximum likelihood classifier to produce annual LULC maps. Changes were analyzed with the Land Change Modeler (LCM) in ArcMap 10.5.1, supported by explanatory variables from verified sources to model scenarios for 2030 and 2050. Results showed that built-up areas expanded by 4.28%, primarily replacing agricultural land, which declined by 4.03%. Forest, bare land, and water bodies experienced smaller shifts. The Multi-Layer Perceptron (MLP) model demonstrated a calibration accuracy of 73.4% with a skill measure of 0.6813, while the predicted 2018 map showed an accuracy of 78.48%, validating the model for future projections. Built-up areas are projected to grow to 9.02% by 2030 and 12.97% by 2050, while agriculture is expected to shrink further. These findings provide vital insights for the international research community, local governments, and urban planners to support evidence-based planning, sustainable urban growth, and informed policy formulation aligned with global sustainability targets.