Remote sensing and satellite images are important tools for monitoring land cover changes. The MODIS Land Cover product (MCD12Q1) which follows 17 International Geosphere- Biosphere Programme (IGBP) global land cover classifications, is widely used for global-scale land cover monitoring. Its accessibility makes it a practical choice for stakeholders, as it allows easier data processing compared to other complex machine learning techniques that require extensive training and expertise. However, its coarse resolution and classification system do not capture some ecological details and struggles to distinguish small and diverse land cover types, especially wetlands, particularly at a regional or sub-national scale. To address this limitation, we integrated MCD12Q1 data with very high-resolution (VHR) imagery from Google Earth to uncover certain categories within the IGBP permanent wetlands in Myanmar for the year 2018. By using widely and publicly accessible optical satellite data from MODIS and Google Earth, along with minimal GIS and computational expertise, we ensure that our approach remains practical for regional applications. We analyzed 250 random sample points taken from the 500- meter MCD12Q1 raster polygons for the permanent wetland category, reclassified these points by following standard guidelines from various references and visually verified them using VHR Google Earth imagery and finally applied the calibration approach to estimate the actual area of each land cover type found within the IGBP permanent wetlands in Myanmar. Our findings reveal that the majority of areas classified as IGBP permanent wetlands were actually mangroves (54.8%), while others include aquaculture farms (16.4%), mangroves and croplands (4.8%), reservoirs (15.2%), river/lake (4.8%), mudflats (2%) and peatlands (2.4%).
저자
Yadanar Ye Myint [ Division of Forest and Biomaterials Sciences, Graduate School of Agriculture, Kyoto University, Kyoto, Japan ]
Corresponding Author
Joseph Emile Honour Percival [ Division of Forest and Biomaterials Sciences, Graduate School of Agriculture, Kyoto University, Kyoto, Japan/ Spatial Data Analysis and Visualization Lab, University of Hawaii at Hilo, Hilo, Hawai'i, USA/ Department of Geography & Environmental Science, University of Hawai'i at Hilo, Hilo, Hawai'i, USA ]
Kaoru Kitajima [ Division of Forest and Biomaterials Sciences, Graduate School of Agriculture, Kyoto University, Kyoto, Japan ]
