Modelling 4D climate variables using functional data analysis and the D-STEM software
Francesco Finazzi (University of Bergamo)
Friday 5th October, 2018 15:00-16:00 Maths 311B
4D spatio-temporal data naturally arise in atmospheric physics, meteorology and climatology, where physical variables are predicted or observed within the spherical‐shell atmosphere and over time. In this talk we discuss a space-time functional data approach for modelling data coming from a radiosonde global network. The data set consists of misaligned vertical profiles of essential climate variables collected through weather balloons while the modelling is based on a multivariate space-time model with latent components. The D-STEM software is adopted for model fitting and kriging.