Urbanization pattern using Machine Learning on Google Earth Engine

Tool Used: Google Earth Engine, ArcGIS Pro
Skills Applied: Random Forest Classification, LULC Change Detection, Remote Sensing, Land Surface Temperature Analysis
Dataset: Landsat Imagery (2000–2020)
Project Type: Independent Research (MS Thesis Work)

Project Overview:
This project analyzes Land Use Land Cover (LULC) change across nine South Asian cities over a 20-year period using the Random Forest machine learning algorithm on the Google Earth Engine platform. The selected cities represent diverse geographic and climatic contexts across five countries:

Bangladesh: Rajshahi, Chattogram
India: Ahmedabad, Bhubaneswar, Kanpur
Pakistan: Gwadar, Quetta
Nepal: Kathmandu
Afghanistan: Kabul
Satellite imagery was used to classify LULC into categories such as urban, vegetation, water, and bare land. The results revealed a substantial increase in built-up areas, often replacing natural land covers. Additionally, the study assessed the impact of urbanization on Land Surface Temperature (LST), highlighting spatial correlations between urban growth and thermal intensity. The findings emphasize the environmental trade-offs of rapid urban expansion in South Asia and inform future planning and climate adaptation efforts.