Hi

I’m research scientist in Meteorology. Since December 2025, I’m SMASH fellow at the University of Ljubljana, developing data assimilation (DA) methods for coupled Earth system models (ESM). Precise representation of ESM components and their interactions are required for accurate weather forecasting on medium to long timescales. Coupled ESMs can predict extreme weather events but rely on background-error covariance models for DA. Traditional methods struggle to capture nonlinear and multiscale error structures, limiting forecast precision. We addresses these limitations by introducing a neural network-based covariance model. My research leverages variational autoencoders (VAEs) to transform innovations, prior errors, and observation errors into a nearly Gaussian latent space, enhancing assimilation performance. Unlike rigid, semi-empirical transformations, neural networks adapt to complex error structures, improving state estimates and forecasts. The model will be trained on reanalysis datasets to ensure robustness across diverse meteorological conditions. My project aims to advance weather forecasting, benefiting sectors like agriculture, and energy.

In my PhD (2020-2024), I studied the assimilation of cloud-affected satellite visible and infrared observations to improve convective-scale numerical weather prediction (i.e. small-scale storms).

Selected talks:

• 25 October 2024: International Symposium on Data Assimilation (Kobe, Japan), see the schedule.
• 7 June 2024: Ambiguity and nonlinearity in the assimilation of visible and infrared observations. ISDA seminar series (online), see the recording.

See my latest publications. See my CV, PhD thesis, MSc thesis, and BSc thesis.

Disclaimer

Co-funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.

Last edit: 5 December 2025