2018 Fiscal Year Research-status Report
Data-driven Seasonal Hydrologic Prediction Using Earth Observing Satellites
| Project/Area Number |
18KK0117
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
金 炯俊 東京大学, 生産技術研究所, 特任准教授 (70635218)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
渡部 哲史 東京大学, 大学院工学系研究科(工学部), 特任講師 (20633845)
内海 信幸 東京大学, 生産技術研究所, 博士研究員 (60594752)
|
|---|
| Project Period (FY) |
2018-10-09 – 2023-03-31
|
| Keywords | data-driven modeling / seasonal prediction / satellite remote sensing |
|---|
| Outline of Annual Research Achievements |
In FY2018, Kick-off meeting was held at Washington DC, USA during AGU conference in December, 2018 to discuss on outline the 5-year plan of the project. Since Prof. Famiglietti moved to University of Saskatchewan, Canada, the international counterpart has been extended to a broader community. The other international senior personal Dr. John Reager (JPL/NASA, USA) could not join in neither AGU and the kick-off meeting, but the PI (Hyungjun Kim) could discuss with him on the 5-year plan of this project 2019 during his visiting JPL in February. It has been agreed Japanese research team will visit JPL/NASA in (tentatively) summer, 2019 for a collaborative application combining numerical climate simulations (to be provided by U-Tokyo) and satellite reconstruction technique (to be provided by JPL/NASA).
Using JSPS invitational program Prof. Famiglietti visited Japan in early March and strengthened international partnership with Japan extending to different region (Hokkaido), different discipline (JAXA), and industry (private sector). Additionally, we agreed to co-organize an international conference at Canada in summer, 2019.
Two related papers have been published at international journals, and one of them has been introduced to public by several Japanese and international media platforms.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
* With move of a international senior personnel (Prof. James Famiglietti) to Canada, University of Saskatchewan became additional counterpart institute, and the team is planning an international conference at Canada in summer FY2020.
* As a synergy with another JSPS program, we invited Prof. James Famiglietti to Japan and further broaden the exchange to other region (i.e., Hokkaido University) and industry (i.e., lectures in JAXA and private company) in Japan.
|
| Strategy for Future Research Activity |
A data-driven inference framework for assessing regional vulnerabilities to flood and drought at lead times from 6 months to the present is proposed. An international team among U-Tokyo, NASA, and University of Saskatchewan is organized, and following objectives will be pursued.
1. Global River Water Height Monitoring:A waveform retracker will be developed and produce standard monitoring product applying tropospheric/ionospheric and crustal vertical motion corrections. Such data process sequence will be developed and tested for Jason missions. 2. Teleconnections between Global Sea Surface Temperature and Regional Hydrologic Cycle:Teleconnection patterns be tween SST and river discharge will be found based on the CLIVAR Drought Working Group study. 3. Regional Flood Potential and Drought Assessment from GRACE:Like a bucket that reaches a full capacity, saturation periods indicate the potential for a region to transition to a flood-prone situation. A method will be developed to provide several months of early warning for the largest events. 4. Data-Driven Prediction Using Ensemble Lorentz Attractor and Recurrent Neural Network:The proposed study will develop an ensemble-based Lorentz Attractor model to empirically describe non-linear dynamics between remote/local forcings and river water height response.
Also, further exchange will be promoted, and, in particular, young scientists will be encouraged to visit counterpart institutes.
|
| Causes of Carryover |
データの圧縮や解析手法の効率化ができたため、今年度は計算用・データ保存用サーバの購入を見送った。今後、データ量や解析に必要な計算機資源の増加に合わせてサーバを購入する。
|
