新任メンバー紹介 - ジョナサン（ジョン）・デロ 博士
公開日 2019年05月16日
自己紹介, Introduction
ジョナサン（ジョン）・デロ 博士, 学振外国人特別研究員
Dr. Jonathan Derot, JSPS Postdoctoral Research Fellow (Short term)
I have completed my BSc and MSc in France at the Mediterranean Institute of Oceanography (Marseille). As part of the validation of my research for Master's degree, I participated in a sea field (GEOTRACES West Atlantic leg 3) for one month, aboard the RSS James Cook. A few months later, I started my thesis at LOG (Laboratory of Oceanology and Geosciences) on: "Use of MAREL Carnot automatics biogeochemical data for understanding the mechanisms of extremes in water quality in Boulogne-sur-Mer (France)". The main objective of this thesis is the characterization of high frequency dynamics in coastal areas and in particular their extremes (i.e. the pollution events), through the study of long-term biogeochemical time series registered by automated systems.
Following my PhD, I have performed short-term postdoc (8 months) at the Japanese University of Tottori, with the professor Hiroshi Yajima. Our fundamental objective was to improve manage of water quality in Japanese lakes and reservoirs. This franco-japanese collaboration, allowed us to highlight that was possible to use a machine learning model and more precisely a Random Forest (RF) model, to forecast the primary production in these different types of aquatic ecosystems. Later, I worked two years at IRSTEA (French National Research Institute of Science and Technology for Environment and Agriculture) of Bordeaux. This research work was exclusively linked to the European Water Framework Directive (WFD). I used a RF model and data mining methods with a morpho-functional phytoplankton traits database; in order to improve the French Phytoplankton Index for Lakes (IPLAC) with new diagnostic metrics.
Now, I joined the Estuary Research Center of the Shimane University from 8 Mai 2019 for one year as a JSPS-funded postdoc (short-term program). My research works are focused on the development of a hybrid hydro-ecological and machine learning model for water quality management, mainly at the HAB (Harmful Algal Bloom) level.
