New Paper
(Lee, Yoo & Park, Artificial Intelligence in Geosciences, 2026)
Downscaling Landsat Land Surface Temperature with SatVu's HotSat-1 Data
SatVu's HotSat-1 data continues to generate meaningful research results — including in Korea.
Launched in 2023, SatVu's first satellite, HotSat-1, collected thermal imagery during its operational period that continues to be used by researchers around the world.
During that period, SatVu proposed making HotSat-1 imagery specially available to Korean researchers for academic use. The data was provided to remote sensing researchers through the Korean Society of Remote Sensing (KSRS), and a number of research projects have made use of it. We're sharing one recently published example below.
Drawing on the HotSat-1 data made available through this process, a joint research team from Seoul National University of Science & Technology and Pusan National University proposed an AI model (SRGAN) that downscales Landsat land surface temperature data from 100m to 30m resolution. (Lee, Yoo & Park, *Artificial Intelligence in Geosciences*, 2026)
Validated across three sites — including Gijang, South Korea — the model outperformed conventional regression-based approaches, reaffirming the academic value of HotSat-1 data.
*Read the paper: https://lnkd.in/e5KCQwuN
Interest in, and research use of, SatVu's thermal imaging technology continues to grow in Korea. As SatVu's official partner in the country, Indyware regularly shares examples like this with the local academic and industry community, and is glad to help bridge data requests like this one when needed.
HotSat-2 is now performing well in orbit. While optical and SAR satellites show that a facility exists, SatVu's thermal imagery shows whether it's operating. HotSat-1 data contributed to monitoring activity at North Korea's Yongbyon nuclear facility; HotSat-2 has since captured operational status at a Cuban refinery, India's Jamnagar refinery, and Australia's Gorgon LNG facility.
----------------------------------------------------------------------------
INDYWARE is the authorized partner of SatVu in Korea.
----------------------------------------------------------------------------
SatVu의 HotSat-1 영상이 한국의 연구자들에게 연구 목적 활용을 위한 특별 제공되었습니다. 한국원격탐사학회를 통해 이를 활용한 연구가 수행한 바 있으며, 이중 최근 논문으로 게재된 사례를 소개합니다.
서울과학기술대학교·부산대학교 공동 연구팀은 이렇게 확보한 HotSat-1 데이터를 바탕으로, Landsat 위성의 지표면 온도(LST) 자료를 100m급에서 30m급으로 정밀화하는 AI 모델(SRGAN)을 제안했습니다.
한국 기장군을 포함한 3개 지역에서 검증된 이 모델은 기존 회귀 기반 방식보다 더 정밀하고 일관된 결과를 보여주며, HotSat-1 데이터의 학술적 활용 가치를 확인하였습니다.
광학·SAR 위성이 시설과 물체가 '있다'는 것을 보여준다면, SatVu의 열화상 영상은 그 시설이 지금 '가동 중인지, 멈춰 있는지'까지 보여줍니다. 실제로 HotSat-1 데이터는 북한 영변 핵시설의 활동 모니터링에 기여한 바 있고, 최근 HotSat-2는 쿠바 정유시설, 인도 잠나가르 정유소, 호주 고르곤 LNG 시설의 실제 가동 상태까지 포착해냈습니다.
*Source: Lee, K., Yoo, C., & Park, S. (2026). Downscaling of Landsat LST with HotSat-1 data and generative adversarial networks. Artificial Intelligence in Geosciences, 7, 100213*