TA698 : Change Detection in Water Bodies of Lakes Using Spatio-temporal Methods and Remote Sensing Data
Thesis > Central Library of Shahrood University > Civil & Architectural Engineering > MSc > 2022
Authors:
Abstarct:
Abstract
One of the most important human needs for survival is water resources. Among environmental changes, water plays a very vital role in the political, social and economic issues of countries, which can be mentioned as one of the most practical sources of water supply available to humans and animals. Investigating the fluctuations of the water level of the lakes in terms of the importance, location and nature of these water bodies has gained special importance in recent years. Since the past times, the identification and monitoring of surface water and other effects of the earth's surface using remote sensing has been very much considered and used. In this study, with two different approaches and using Landsat-8 and Sentinel-2 optical sensors, the surface changes of Lake Urmia, vegetation and soil around it were studied during the years 2013 to 2021. In this research, after performing radiometric and atmospheric data corrections, as well as pen-sharpening the data using the Gram Schmidt and LMVM methods to increase the spatial resolution for Landsat-8 and Sentinel-2 sensors, respectively, indices such as NDWI, MNDWI, AWEI, WI2015 and NDVI were extracted from the images. Then by combining the obtained indices and the spectral bands of the sensors with each other, the classification of images using SVM, NN, ML and MD algorithms and also the integration of the results of the classifiers using the majority voting method to improve the classification results are done.
Finally, the results of the classification were validated by calculating the overall accuracy and Kappa coefficient, and the amount of changes in the water surface of Lake Urmia, the vegetation and the surrounding soil between 2013 and 2021 were calculated.
The results of the research showed that the majority voting method with the highest overall accuracy (99.07%) and Kappa coefficient (0.986) was chosen as the most suitable classification method for both approaches.The water level of Lake Urmia, vegetation and soil around it also had significant changes during the years 2013 to 2021. The amount of changes in lake water level, vegetation and soil in 2021 compared to 2020 decreased by 29%, increased by 16% and increased by 17.5% respectively.
Keywords:
#Keywords: Surface Water #Lake Urmia #Remote Sensing #Supervised Classification #Data Fusion #Change Detection Keeping place: Central Library of Shahrood University
Visitor:
Visitor: