*Takemasa Miyoshi1, Shun Ohishi1, Hirofumi Tomita1, Shigenori Otsuka1, James Taylor1, Jianyu Liang1, Rakesh Teja Konduru1, Masaki Satoh2, Shuhei Matsugishi2, Shunji Kotsuki3, Atsushi Okazaki3, Takumi Honda4, Kozo OKAMOTO5, Yasutaka Ikuta5, Koji Terasaki5, Hisashi Yashiro6, Kaya Kanemaru7, Akira Yamazaki8
(1.RIKEN, 2.The University of Tokyo, 3.Chiba University, 4.Hokaido University, 5.Meteorological Research Institute, 6.National Institute for Environmental Studies, 7.National Institute of Information and Communications Technology, 8.Japan Agency for Marine-Earth Science and Technology)
Keywords:OSSE, numerical experimental platform, data assimilation, satellite data
We propose to develop a unique platform with advanced data assimilation technology, realizing numerical simulation experiments for the advanced evaluation of future satellite missions in their design phase. This will provide a scientific, objective approach to an effective design of future satellite missions to maximize the benefit to practical applications such as hydrometeorological prediction and resource control of fisheries. The authors have been leading data assimilation research to explore the best mix of computer model simulations and real-world data for weather forecasting, taking advantage of new-generation technologies such as the supercomputers K and Fugaku, as well as Himawari-8 and -9 satellites. The authors’ previous research revealed that every-10-minute Himawari-8 observations could lead to a large improvement in typhoon and heavy rainfall predictions, though every-30-minute observation was insufficient. We would expect that frequent observations would also be beneficial for ocean and land surface prediction. Therefore, we aim to develop a platform for numerical experimentation to evaluate potential future satellite missions such as quasi-zenith orbiters and small cube-sat constellations for frequent observation over Japan and its surrounding area. This platform will be useful for designing future satellite missions with high cost-benefit performance.