2026-03-14T22:57:40Z https://air.repo.nii.ac.jp/oai

oai:air.repo.nii.ac.jp:00002105 2024-08-25T08:08:00Z 590:629:630

ランドサットTM画像を用いたアマゾン川における水域の推定(英文) Estimation of Water Areas in the Amazon on Landsat-TM Images 西田, 眞 景山, 陽一 肥田, 登 NISHIDA, Makoto KAGEYAMA, Yoichi HIDA, Noboru open access リモートセンシング varzea ファジィ推論 エッジデータ 水位変動 ミクセル 土地被覆クラス remote sensing fuzzy estimation edge data water level changes mixel landcover class This paper presents the result of estimating water areas in the Amazon from the optical remote sensing data by using a new fuzzy estimation method and edge data. Conventionally, multi-temporal image data with different water heights are used to estimate the flooded areas. Optical sensors such as Landsat-TM may be limited by their inability to penetrate cloudcover in the Amazon regions. However, they have been well stocked in comparison with SAR (Synthetic Aperture Radar) image data and are useful to classify landcovers. Thus, this paper proposes a method that uses only one optical image data to estimate the flooded area, called várzea. Distinction between a pure pixel and a mixel (mixed pixel) is computed, and the occupancy of landcover classes in the study area is estimated by a fuzzy estimation method. In addition, estimation maps of the water areas at high and low levels are drawn from the occupancy of landcover classes and the situation of water areas is estimated. It is observed that the estimation results using the occupancy can match the local situation in the study area. As a result of this work, the estimation ratio indicates that water areas around Parintins-city, Brazil, at high level is an increase of about 58 % compared with water areas at low levels. The proposed method is applied to this particular regions, but this approach can also be applied to larger water areas considering varzea areas. Therefore, it is confirmed that the proposed method is very useful to estimate water areas in the Amazon from the optical remote sensing images. 東京地学協会 1998-01-01 eng journal article VoR http://hdl.handle.net/10295/2268 https://air.repo.nii.ac.jp/records/2105 https://doi.org/10.5026/jgeography.107.61 地學雜誌 107 1 61 76 https://air.repo.nii.ac.jp/record/2105/files/nishidam4.pdf application/pdf 16.8 MB 2017-02-16