2026-03-07T21:11:41Z https://chikyu.repo.nii.ac.jp/oai

oai:chikyu.repo.nii.ac.jp:00004788 2023-06-20T15:19:47Z 419:420

Sentinel-2/MSI データを用いたインド・パンジャーブ州における野焼き火元画素の抽出条件 Selection criteria of agricultural burning pixels using Sentinel-2/MSI data in Punjab, India 于, 琨 ユ, クン 村松, 加奈子 ムラマツ, カナコ YU, Kun MURAMATSU, Kanako Aakash Project of Research Institute for Humanity and Nature 短波長赤外の反射率 植生指標NDVI 季節変化 Reflectance of short wavelength infrared vegetation index NDVI seasonal changes application/pdf High concentrations of PM2.5 in Delhi, India, are a health concern. Main emission sources of PM2.5 are smoke from factories and exhaust gas from cars, but considerable amount of smoke from burning of agricultural residues is significant. Satellite observation is useful to monitor the fires from burning agricultural residues. We examined characteristics and seasonal changes in spectral reflectance and normalized vegetation index (NDVI) at the origin-of-fire by burning of agricultural residues. The reflectance in shortwave infrared (SWIR) of 2190 nm (RSWIR2190) was higher than 0.4, and the ratio between two SWIR (RSWIR2190/ RSWIR1610) was higher than 1.25. That was specific characteristic for the origin-of-fire. Most landcover such as water, vegetation and city area did not have that characteristic, but specific city area which had relatively high temperature such as brickyards had similar characteristics of RSWIR2190 was around 0.4. It was needed to remove hot pixels such as brickyards from candidate pixels of origin-of-fires. To confirm the pixel being farm by detecting vegetation signal, the condition was useful that the number of times with NDVI being higher than 0.85 was greater than 1 in a year. The pixels after agricultural burning showed gray in natural color composite image, and the spectral reflectance pattern was different from that in harvest and post-harvest time. From these results, it is suggested that after burning area could be extracted from satellite image, even if satellite could not observe on time of burning agricultural area. Further study with using ground survey data is needed to identify the agricultural burning pixels and estimate identify accuracy. Aakash Project of Research Institute for Humanity and Nature 2023-03 jpn learning object VoR https://chikyu.repo.nii.ac.jp/records/4788 https://aakash-rihn.org/workingpaper/ 2436-9888 (Online) Aakash Working Paper Aakash Working Paper 2 1 18 https://chikyu.repo.nii.ac.jp/record/4788/files/Aakash_Working_Paper02.pdf application/pdf 12.8 MB 2023-04-14