首页|MAPPING SINGAPORE'S LAND USE BASED ON IPCC GUIDELINES: A FRAMEWORK-BASED TRAINING GUIDE FOR EFFECTIVE CLASSIFICATION
MAPPING SINGAPORE'S LAND USE BASED ON IPCC GUIDELINES: A FRAMEWORK-BASED TRAINING GUIDE FOR EFFECTIVE CLASSIFICATION
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
A comprehensive framework-based training guide designed to enhance the accuracy and efficiency of land use classification in Singapore is presented。 The guide is created to align with the Intergovernmental Panel on Climate Change (IPCC)'s land use categories。 This guide primarily utilises the capabilities of Google Earth Engine (GEE) to access extensive geospatial data and perform supervised classification on the land cover。 The analysis is based on 10-m spectral bands of Sentinel-2 Harmonized collection， with cloud removal techniques applied and clipped to the boundary of Singapore。 Training data for several machine learning classification algorithms is defined from two scenes of Sentinel-2 acquired in 2019 and 2020， by visually examining and labelling areas in these images to create corresponding feature collections。 Results of a preliminary accuracy assessment show that Random Forest provides a higher overall accuracy of 80% to 83%。 The validation accuracy was derived by calculating the confusion matrix of the trained model and then using it to compute the overall accuracy of the classification via Google Earth Engine。 In conclusion， this guide aims to enhance classification accuracy of Singapore's land use for measurement， reporting， and verification (MRV) of greenhouse gas (GHG) emissions from Singapore's Land Use， Land-Use Change， and Forestry (LULUCF) sector。
Land Use ClassificationLULUCFIPCC
Jacylin Jie Yi Tham、Kim Hwa Lim、Soo Chin Liew
展开 >
Centre for Remote Imaging, Sensing and Processing (CRISP), National University of Singapore S17 Level 2, Lower Kent Ridge Road, Singapore 119260
NETL
