气候变化领域集成服务门户(CCPortal): Land cover classification in the tropics, solving the problem of cloud covered areas using topographic parameters

CCPortal

DOI	10.1016/j.jag.2018.12.010
	Land cover classification in the tropics, solving the problem of cloud covered areas using topographic parameters
	Shrestha D.P.; Saepuloh A.; van der Meer F.
发表日期	2019
ISSN	15698432
起始页码	84
结束页码	93
卷号	77
英文摘要	Optical remote sensing data has been extensively used since early seventies for mapping and monitoring land cover. But, cloud cover has always been hindering the optimal use of the data, especially in the tropical and temperate regions. Because of cloud cover, researchers are forced to use only cloud free images or generate a composite using data from different dates to fill in the cloud and cloud-shadow areas. It is not a problem if the filling image is from the same season and from the same year. If there is large time gap between the acquisition date of image with the clouds and that of the filling image it will have an effect on classification accuracy. In this paper, we describe an approach in land cover classification of areas covered by cloud and cloud shadows. We use the synergy of optical image classification, processing of SAR data taking advantage of its ability of penetrating clouds, and decision rules using terrain parameters. We hypothesize that surface roughness of different land cover types derived from processing of SAR data and in combination with other topographic parameters (slope and elevation) can be applied to classify areas covered by clouds and cloud-shadows. The method was applied in a case study in Mts. Wayang-Windu and Patuha areas, near Bandung in West Java, Indonesia. Landsat 8 OLI-TIRS, Sentinel-1 A C band and SRTM DEM data of the study area were obtained. Initial classification of the optical data was carried out using training samples and fuzzy classification. Class labelling was done by applying fuzzy c-means algorithm. The Sentinel-1 A C band data was processed to get surface roughness. Decision boundaries for surface roughness, terrain slope and elevation were generated for each land cover type, which were implemented in the classification of the areas covered by cloud and cloud shadows. The result shows that it is possible to solve the problem of cloud cover using terrain parameters. © 2019 Elsevier B.V.
英文关键词	Cloud and cloud shadow; Decision rules; Optical and SAR data; Surface roughness; Terrain slope and elevation
语种	英语
scopus关键词	cloud cover; elevation; image analysis; land classification; land cover; remote sensing; satellite data; slope; surface roughness; synthetic aperture radar; terrain; topographic effect; tropical region
来源期刊	International Journal of Applied Earth Observation and Geoinformation
文献类型	期刊论文
条目标识符	http://gcip.llas.ac.cn/handle/2XKMVOVA/156509
作者单位	University of Twente, Faculty of Geo-information Science and Earth Observation (ITC), P O. Box 217, Enschede, 7500 AE, Netherlands; Bandung Institute of Technology, Faculty of Earth Sciences and Technology (ITB), Jl. Ganesha No. 10, Bandung, West Java, Indonesia
推荐引用方式 GB/T 7714	Shrestha D.P.,Saepuloh A.,van der Meer F.. Land cover classification in the tropics, solving the problem of cloud covered areas using topographic parameters[J],2019,77.
APA	Shrestha D.P.,Saepuloh A.,&van der Meer F..(2019).Land cover classification in the tropics, solving the problem of cloud covered areas using topographic parameters.International Journal of Applied Earth Observation and Geoinformation,77.
MLA	Shrestha D.P.,et al."Land cover classification in the tropics, solving the problem of cloud covered areas using topographic parameters".International Journal of Applied Earth Observation and Geoinformation 77(2019).