PSL welcomes CIRES research scientist Joseph Bellier, who joined the Attribution and Predictability Assessments Team in December. Joseph is primarily working with Michael Scheuerer and Tom Hamill to develop statistical methods for post-processing and downscaling precipitation ensemble forecasts coming from ensemble prediction systems like GEFS. Most of these methods have the detrimental effect of breaking the spatio-temporal structure of the raw forecasts, although this coherence is crucial for hydrological modeling. Thus, the main question on which Joseph is currently working is how to reconstruct the adequate spatio-temporal structure, with the objective of delivering seamless probabilistic forecasts over areas such as the contiguous United States.

Joseph graduated with a master degree in hydrology and water sciences at University Grenoble Alpes (France), prior to completing his PhD in 2018 on probabilistic hydrological forecasting.

Outside of work, Joseph can usually be found on the local trails, either running, mountain biking or backcountry skiing.

Posted: February 19, 2019