Ryuichi_KANAI

🚨🚨🚨 Job alert! Research Fellow in Probability Theory at the Department of Statistical Science at UCL. ⏰Closing date: 12-Jan-2026 More information: 👇 ucl.ac.uk/work-at-ucl/se…

Kanai, R., Hernández, N., Gopinathan, D. and Guillas, S., 2025. History matching for functional data and its application to tsunami warnings in the Indian Ocean. arXiv preprint arXiv:2509.04342.

Also, we analyze hypothetical tsunami scenarios in the Indian Ocean, using JAGURS for tsunami simulations. This new method enables highly accurate tsunami forecasting. This is a study in the field of functional data analysis.

Earlier this month, I released one of the projects from my PhD to arXiv. It extends the method known as history matching to make it applicable to functional data. Our work also features an application to tsunami case studies. doi.org/10.48550/arXiv…

🎉🎉🎉Congratulations to Ryuichi Kanai on passing his PhD viva today! Well done Ryuichi!

Receiving this award has been a great pleasure for me. It's an honor to have been able to contribute to the Department of Statistical Science. I hope students from various backgrounds will continue to support each other in their research and PhD life.

@sakiman7263 Did you mean the High-frequency Active Auroral Research Program?

@Ryuichi_KANAI HAARP ?

Detecting changes in electron density in the ionosphere is not easy. Therefore, by utilizing the difference in arrival times of two different frequencies emitted from GPS satellites, the change in electron density can be estimated.

🎉🎉🎉 Many congratulations to Zhuo (@JasonSun10 ) and Ilina (@ilina_yozova), who have won the MAPS Early Career Researchers Award Research Excellence Award and Teaching Excellence Award, respectively. Very well done, both of you!

As a UCL student, I had many invited talks and contributed talks in Japan (Tokyo, Kyoto, and Sendai), the UK (London, Cambridge, and Bath), and China (Beijing and Changsha). All of them are excellent experiences.

I had a contributed talk about my current project at the University of Warwick. Thank you for giving me this opportunity.

🎉🎉🎉 Congratulations to Alberto Caron on passing his PhD viva! He was supervised by Prof. Ioanna Manolopoulou. @ioanna1182 @AlbiCaron

Kaiyu just gave a great talk about Multilevel BQ at AISTATS 2023 !@stats_UCL @fx_briol @GuillasSerge

I’m back to London. This is my final year.

@Prest0n1337 Thank you very much for taking an interest in our research! If you have any question, feel free to ask me!!

In our research, ionospheric anomalies are captured by GPS satellites. However, we focus on the low-frequency component of the anomaly. If we do not remove the high-frequency component of the anomaly, the response of the ionosphere is very unstable as shown in this movie.

Noise reduction is very important. This movie shows the time series variation of the observed data before noise removal. As you can see here, removal of high-frequency components is extremely important and will require further research.

We initially planned to perform the fitting with an ordinary Gaussian process regression. However, the approach takes too long. Therefore, we decided to use a different method, called the INLA-SPDE approach, significantly reduces the computation time. doi.org/10.1111/j.1467…

@fatihkulahci Thank you very much!

Our research makes it possible to estimate ionospheric anomalies in real time, even in places where measurements do not exist. This has been recognized and our research results have been reported in the U.S. by Yahoo! News US and other media. yahoo.com/news/gps-signa…

In the papers introduced yesterday and the day before yesterday, only data observed by GPS satellites were used. However, this alone is not sufficient, and it is important to estimate values for unobserved locations as well.