Previous works

PREVENIR

My main work at RIKEN was the development of data assimilation and numerical weather prediction system for urban areas of Argentina utilizing weather radars, under the project PREVENIR, a 5-year collaborative project between Japan and Argentina.
See our project website for details.

Radar data assimilation with SCALE-LETKF

I was the main developer of the SCALE-LETKF, a regional data assimilation and numerical weather prediction system. I have worked on several research topics using the SCALE-LETKF, especially regarding the assimilation of the Phased Array Weather Radar (PAWR) data with very high spatial and temporal resolution utilizing the supercomputer Fugaku.

Post-processing and visualization of ensemble forecasts

I worked on an explorative study on the post-processing and visualization of ensemble forecasts for in a convective scale. This study is conducted under the KAKENHI Grant-in-aid project (https://kaken.nii.ac.jp/ja/grant/KAKENHI-PROJECT-21K13996/).

SCALE-LETKF real-time precipitation forecast experiment

A large-scale real-time precipitation forecast experiment was conducted in the summer of 2021 using the system based on the SCALE-LETKF and PAWR.
See RIKEN news for details.
Our group was nominated as a finalist for the 2023 Gordon Bell Prize for Climate Modelling with this experiment.

Model bias correction using data assimilation and machine learning

Using a simple Lorenz96 system, this study demonstrated the potential advantage of using machine learning in correcting systematic model biases in data assimilation.

Dynamics of the Asian monsoon anticyclone

The dynamical variability of the Asian monsoon anticyclone (Tibetan High / South Asian high) was studied using a simplified two-dimensional dynamical model. The variability of this planetary-scale anticyclone has essential roles not only in regional climate but also in stratospheric chemistry, as it forms a fragile boundary between stratospheric and tropospheric air.

Gravity wave parameterization

A new gravity wave parameterization scheme including three-dimensional propagation is developed. Atmospheric gravity waves have significant contribution to the momentum budget in the middle atmosphere, yet they are still hard to sufficiently resolve in climate models. Improvement of parametrization is necessary to simulate various important dynamical features of the middle atmosphere such as QBO, SSW, and seasonal evolution of the ozone hole.