Graduate Students Seminar

Speaker 1: Dinesh Ekanayaka

Title

Enhancing Numerical Modeling for the Viscous-Plastic Formulation of Sea Ice

Abstract

This study investigates the dynamics of viscous plastic (VP) sea ice and the numerical challenges in its modeling. Comparing with an existing one-dimensional model as a prototype, which employs backward Euler (BE) for temporal discretization, Newton for the nonlinear solver, and central difference (CD) for spatial discretization for the momentum equation, the researchers found that higher-order methods, such as the Weighted Essentially Non-Oscillatory (WENO) method, provided better results. The WENO method efficiently captured sharp gradients and maintained stability, outperforming traditional methods. It effectively handled discontinuities in ice thickness and concentration without causing significant oscillations or numerical blow-ups, thus demonstrating its superiority in maintaining accurate and stable solutions for sea ice dynamics.

Speaker 2: Mouhamed Oloude

Title

Modeling Malaria Incidence with a Spatio-Temporal Approach

Abstract

Tn Sub-saharan Africa, managing malaria incidence and prevalence represent a major public health challenge. The disease is primarily transmitted through the bites of female Anopheles mosquitoes, which breed in water-concentrated areas, such as reservoirs, wetlands, dams, etc. This talk addresses the study of the effect of a mega hydropower dam on malaria incidence. Poisson and hazards models will be used to explore the impact of distance to dam and seasonal factors on transmission rates. Using simulations and data collected from villages located at varying distances from a dam in South Ethiopia, malaria cases are modeled over multiple periods and years to evaluate incidence rates and associated survival times. The results show that the two regression models provide similar estimates of seasonal effects and interesting counterintuitive estimates of the effects of distance to the dam.