A stochastic storm model with multiple advection directions

• Grant number: 24/20354-6
• Support Opportunities: Scholarships in Brazil - Master
• Start date: August 01, 2025
• End date: February 28, 2027
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Statistics
• Principal Investigator: Guilherme Vieira Nunes Ludwig
• Grantee: Pedro Francisco Godoy Bernardinelli
• Host Institution: Instituto de Matemática, Estatística e Computação Científica (IMECC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 23/02538-0 - Time series, wavelets, high dimensional data and applications, AP.TEM

Abstract

Rainfall models based on Gaussian random fields have difficulties simulating intervals in space and/or time where there is no rain. These models can, however, incorporate advection effects in their covariance functions, emulating the movement of atmospheric masses caused by wind. Stochastic rectangular pulse models (RPM), on the other hand, can describe regions (space) and intervals (time) without rain. The main objective of this project is to construct a hierarchical approach, driven by a latent Gaussian spatio-temporal process with advection, that allows simultaneously for varying intensity as well as bounded intervals / regions of rainfall during a storm.