JpGU-AGU Joint Meeting 2017

講演情報

[JJ] ポスター発表

セッション記号 M (領域外・複数領域) » M-GI 地球科学一般・情報地球科学

[M-GI30] [JJ] 情報地球惑星科学と大量データ処理

2017年5月22日(月) 15:30 〜 17:00 ポスター会場 (国際展示場 7ホール)

コンビーナ:村田 健史(情報通信研究機構)、大竹 和生(気象庁気象大学校)、野々垣 進(国立研究開発法人 産業技術総合研究所 地質情報研究部門 情報地質研究グループ)、堀之内 武(北海道大学地球環境科学研究院)

[MGI30-P08] Development of database system for cruise information of JAMSTEC vessels and statistical analysis of observation downtime

*山岸 保子1阪口 秀1山室 悠太2 (1.国立研究開発法人海洋研究開発機構 数理科学・先端技術研究分野、2.国立研究開発法人海洋研究開発機構 海洋工学センター)

Japan Agency for Maine-Earth Science and Technology (JAMSTEC) has seven research vessels and controls the research cruise of these vessels. Each year, several tens of research cruises are carried out, and a large amount of marine observation data is acquired. JAMSTEC has not only the observation data but also a large amount of navigation data of past cruises. There were cruises in which many of the scheduled observations could not be implemented due to various causes. For example, in some research cruises carried out in the fall in the surrounding sea of the Japanese Islands, many of observations were canceled by typhoon. Scheduled observations cannot be performed and sufficient observation results cannot be obtained so that progress of the research will be prevented. However, even though there are navigation data, it has not been examined how many observations were canceled in past cruises and what caused the observation downtime. At present, JAMSTEC has to schedule all research cruises for the next fiscal year one year before, which may prevent efficient operation of the research cruise. In this study, we are developing database system for the cruise information operated in the past several tens years to clarify observation downtime for each cruise and what caused the downtime. We are also analyzing statistically the downtime data to describe the relationship between the downtime and various factors of the cruise such as season, sea area, observation equipment, vessel, etc. The analysis results will provide useful information to plan the cruise. Furthermore we will analyze all information of past cruises by machine learning, and we will predict the downtime of the planned cruise and propose better research cruise plan, which will help to obtain the sufficient observation results and to advance the research. Acknowledgments: we are grateful to Mr. Morisaki and Ms. Sada for their supports.