17:15 〜 18:30
[ACG36-P16] Wildfire watch using SWIR data from SGLI/GCOM-C
キーワード：森林火災、SGLI、しきさい
The Second-generation Global Imager (SGLI) is an optical sensor onboard the GCOM-C satellite. SGLI has eleven channels for visible to near-infrared wave range, six channels for short-wavelength to thermal infrared wave range, and two polarization channels for visible and near-infrared wavelength. The spatial resolution is switchable between 250 m and 1 km. In terms of watching wildfire, the spatial resolution of 250 m is an advantage of SGLI. Although lack of mid-infrared (MIR) channels is a disadvantage, SGLI has channels in a short-wavelength infrared (SWIR) range where data is sensitive to high temperatures.
We examined the SWIR data for its performance of watching wildfire. We introduced a threshold-based test for detecting hotspots and a single-channel method to estimate fire radiative power (FRP). The single-channel method followed Wooster et al. 2003. Preliminary results show good agreements with those retrieved from Himawari-8 data. Meanwhile, some of the SWIR data were saturated probably by some wildfire flaming in a wide area. That suggests the upper limit of the dynamic range of SWIR channels is too low for estimating FRP; this is likely to be originated from a mismatch between the high spatial resolution and the narrow dynamic range. However, SWIR can capture signs from wildfire behind the clouds and has detected hotspots that Himawari-8 could not, suggesting wildfire data from SGLI is worthy.
We examined the SWIR data for its performance of watching wildfire. We introduced a threshold-based test for detecting hotspots and a single-channel method to estimate fire radiative power (FRP). The single-channel method followed Wooster et al. 2003. Preliminary results show good agreements with those retrieved from Himawari-8 data. Meanwhile, some of the SWIR data were saturated probably by some wildfire flaming in a wide area. That suggests the upper limit of the dynamic range of SWIR channels is too low for estimating FRP; this is likely to be originated from a mismatch between the high spatial resolution and the narrow dynamic range. However, SWIR can capture signs from wildfire behind the clouds and has detected hotspots that Himawari-8 could not, suggesting wildfire data from SGLI is worthy.