[3I5-OS-27b] OS-27

[3I5-OS-27b-01] (OS invited talk) Data driven methodology advancing agricultural digital application development

〇Hironori Arai¹ (1. the university of osaka)

Keywords:digital-twin, DMD, DDLCN, data-assimilation, MRV

In terms of GHG accounting, Monitoring, Reporting and Verification (MRV) systems with high transparency is essential to formulate Nationally Determined Contributions. The presenters have been working to establish efficient/transparent MRV system in a tropical rice cropping system by synthesizing multipoint & long-term field observation data, satellite remote sensing data, and a local/regional/global data-assimilation system. The presenters are developing a multi-scale observation data integration system based on the data assimilation system supported by machine-learning (ML) technologies. The various scale observations consist of satellite/UAV/ground-IoT sensing are enabling the realistic projection of real-world status in simulations precisely. In this presentation, the presenters will demonstrate the state of the art digital-twin technologies enabling the realistic projection of rice cultivation status in the Vietnamese Mekong Delta. We also present our ML methodology enabling efficient data assimilation with such observation data which plays core role of the mission accomplishment.
Finally, we will discuss the direction of AI technology development to support decision making of internationally diverse stakeholders.