Japan Geoscience Union Meeting 2023

Presentation information

[J] Online Poster

M (Multidisciplinary and Interdisciplinary) » M-IS Intersection

[M-IS10] Mountain Science

Fri. May 26, 2023 3:30 PM - 5:00 PM Online Poster Zoom Room (11) (Online Poster)

convener:Yoshihiko Kariya(Department of Environmental Geography, Senshu University), Akihiko SASAKI(Department of Geography and Environmental Studies, Kokushikan University), Chiyuki Narama(Niigata University, Program of Field Research in the Environmental Sciences), Motoshi Nishimura(Arctic Environmental Research Center, National Institute of Polar Research)

On-site poster schedule(2023/5/26 17:15-18:45)

3:30 PM - 5:00 PM

[MIS10-P08] Prototype development of robust ground observation system using Raspberry Pi

*Sakino Izawa1, Hiroto Nagai1 (1.Waseda Univ., School of Education)

Keywords:Raspberry Pi, Ground observation

In recent years, satellites have been launched more frequently and Artificial Intelligence (AI) technology has made remarkable progresses as well as data processing technologies. On the other hand, ground observation technology for calibration and validation has not shown remarkable development. In this research, therefore, an initial development of a new ground observation system is carried out focusing on mobility and robustness, which are appropriate for an operation in cold regions with snow fall conditions.
We developed a compact ground observation system using a single-board computer, "Raspberry Pi". Through field demonstrations, we assess practical feasibility and versatility of the system. The system is powered by dry batteries to simplify the structure and longer-time operations. Recorded data is transferred to a server of Google Drive from remote locations.
In two in-situ assessments and laboratory experiments we had following findings. The system was operated at the 5th and the 6th stations of Mt. Fuji. Data acquisition and uploading of acquired data to the drive were confirmed. In a refrigerator, normal operation was confirmed in a low-temperature environment about -15℃. Spatial resolution of an attached camera is adequate to recognize 5-mm difference with a distance of 3 m. Power supply from four D-type dry batteries enables 14-hours running and eight batteries does 41-hours running.
These results confirm primary functions and performances of a basic structure of automatic observation system based on Raspberry Pi with a simple optical sensor. By using modern AIs, it will enable automatic recognition and data collection of snow depth and other climatic parameters, sending images to a cloud server. Improvement of housing structure/materials will be needed for operations in snowy environments, deserts, wetlands, tropical rainforests, and any other locations that are currently difficult to observe on the ground. We will discuss versatility of this system focusing of contributions for future earth-observation system.