Please use this identifier to cite or link to this item:
Title: Monitoring Land Cover Change Using GEE and Multi-sources Remote Sensing Images -A Case Study in Vientiane, Lao PDR
Monitoring Land Cover Change Using GEE and Multi-sources Remote Sensing Images -A Case Study in Vientiane, Lao PDR
Authors: Vorlajuck Khantivong
Burapha University. Faculty of Geoinformatics
Keywords: Urban Expansion; Land Cover Changes; Machine Learning; Random Forest; Classification and Regression Trees; Multi-source Imagery; Google Earth Engine
Issue Date:  9
Publisher: Burapha University
Abstract: Increasing urbanization causes a variety of environmental changes, not only in regional but in the global context. Urban expansion is one part of the city growth process. Urbanization has a significant impact on the environment, which land cover change has created a new style of the development of the city. Land cover changes have become an essential element for the current strategy for natural management, resources, and environmental changes monitoring. Using the integration of geographic information system (GIS) and remote sensing provides us with an effective result in determining the land use/land cover pattern changes as well as providing valuable information needed for planning and researches. Thus, the information such as rate, pattern, and model of urban expansion is required by the city planners to accurately determine city planning direction and policies management for better urbanization.  Since its establishment in 1975 to the Lao people’s Democratic Republic (hereinafter, referred to as Lao PDR or Laos), Laos is attempting to prioritize relations between states, people, land, and the natural environment in a manner that is conducive to achieve state goals in making political security, economic development, and environmental protection and conservation. Therefore, this research will be targeting on the study of the urban expansion and land cover changes in Vientiane. The research focuses on its dynamic pattern changing of urban development for three decades between 1989 and 2019. To quantitatively and assess the rate of the changes in land cover and measure the direction of the urban expansion over the period time in the study area, an innovation and modernization technology must be utilized. A new approach has been introduced for detecting and monitoring urban expansion. The changes in land cover patterns by using the Big Data platform via the Google Earth Engine (GEE) cloud computing. The machine learning algorithms have been recommended for this research, such as Random Forest (RF), Classification and Regression Tree (CART), Continuous Naïve Bayes (CNB), GMO Max Entropy (GMO Maxent), and Minimum Distance (MD). The different classifiers will be used to compare the performance of those machine learning algorithms and determine the adequate performance of the classifiers. Analyzing and monitoring urban areas is an essential factor for town or city planning. Many kinds of research tried to understand and explain the expansion of nature for the increasing and growth direction of the city. In this research, I conducted GEE combining with multiple sources of satellite optical images time-series from three main satellites, Landsat 5 and Landsat 8 from the partnership of the National Aeronautics and Space Administration (NASA) and the U.S. Geological Survey (USGS), and Sentinel-2, Which has been developed and is being operated by European Space Agency (ESA), These three satellites will be used for assessments and feature extraction for urban expansion and land cover changes during the 30-year in Vientiane Capital, Lao PDR. This research is the first attempt, using the GEE platform for detecting and monitoring the urban growth and land cover changes in Laos. It takes advantage of machine learning classifiers by using satellite image analysis for 30 years. The findings of this study include: The GEE can provide a significant role with highly competent and precisely for land cover changes detection and used for city administration and policymaking of the Lao government in the future. GEE platforms are quick and freely provides simple tools for users. Therefore, the combination of machine learning algorithms and high-resolution images (Sentinel-2 satellite image) has a high effect. This study found that CART can achieve overall accuracy of 96.34% and a Kappa coefficient of 0.95. Moreover, RF can achieve overall accuracy of 95.08% and a kappa coefficient of 0.93. At the same time, CNB, GMO Max Entropy, and MD made overall accuracies of 90.59%, 81.49%, and 73.95%, respectively, and the Kappa coefficients are 0.88, 0.74 and 0.66, respectively. After evaluation and classification, it was found that the urban is increased from 9.11 in 1989 to 13.52 in 1994 and increased to 26.70, 43.50, 47.514, 83.11, 105.06 in 1999, 2004, 2009, 2014, 2019 respectively. This is about 0.25% of the total area in 1989 to 2.90% in 2019 with the direction of expansion generated into three main routes, such as to 13th North, 13th South Road, and along Thadeua Road, connected and shared border to Thailand via Mekong River Friendship Bridge No. I.  This research could be beneficial for the Lao government and the policymaker for performing urban development and planning that would accelerate green growth and sustainable infrastructure management. This research will be led to socio-economic, ecological, prosperity development, and for the betterment of city sustainable development plans in the future
Description: Master Degree of Science (M.Sc.)
วิทยาศาสตรมหาบัณฑิต (วท.ม.)
Appears in Collections:Faculty of Geoinformatics

Files in This Item:
File Description SizeFormat 
61910097.pdf8.82 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.