 This study proposes a deep learning-based forest fire detection algorithm utilizing 10-minute interval high-temporal resolution Himawari, 8-advanced Himawari imager, with a focus on reducing detection latency and incorporating both temporal and spatial information for improved accuracy. The CNN model outperformed the recent operational satellite-based active fire detection algorithms and demonstrated reliable generality under varied environmental conditions. This article was authored by Yu Jin Kang, Yuna Jang, Jong Ho Im, and others.