From 10:30 - to 12:00 on 1 (Tue) Nov 2022
 

  Lecture room 109 in School of Engineering, Hiroshima University

  # Access > Higashi-Hiroshima Campus:  Place of the lecture room [Google Map]
 

Oceanic environments such as sea surface temperature (SST), current and sea level have a significant impact on climate changes, marine ecosystem, and even human activities. Therefore, scientists have made huge efforts for their prediction, particularly using numerical simulations. Continuous improvements of numerical models make it possible to predict a more realistic ocean with the help of data-assimilation and fine spatio-temporal resolutions. However, on the other hand, the well-developed oceanic model requires high computational power and time, making it hard to be utilized for practical purposes sometimes. To compensate the high computational costs, there is a need to develop novel approaches with efficient computational costs, combined with the numerical model outputs. In that way, artificial neural networks (ANNs) could be one of the solutions because they utilize pre-trained networks that need low computational power by creating multivariate and nonlinear regression equations. In my talk, I will present several frameworks applicable for the prediction or data gap filling of oceanic environments using ANNs. Particularly, the focus will be on the predictions of time-varying features of SST, sea surface current, coastal sea level and surface mixed-layer depth. The network's performances are evaluated by switching input variables based on the dynamics governing oceanic environmental changes to find the optimal ANN models, which demonstrate their strong potential for practical uses.
 

Prof. Hidemi Mutsuda

Department (academic degree program) of Transportation and Environmental Systems
Graduate School of Advanced Science and Engineering
Hiroshima University

Kagamiyama 1-4-1, Higashi-Hiroshima, 739-8527 JAPAN
Telephone: +81-824-24-7778
E-mail: mutsuda[at]hiroshima-u.ac.jp (Please convert [at] to @.)