Filling the gap between CWE and MMM

Abstract: The interaction between the local and the global climate can only be fully understood if the corresponding numerical models are coupled to one another. This research aims at designing a coupling strategy and assessing its performance.

Description: Nowadays, a big gap exists between people doing Computational Wind Engineering (CWE) and people active in the field of Mesoscale Meteorological Modelling (MMM). The first group of people is concerned with the urban scale, i.e. wind flow in streets, squares and courtyards. They use numerical simulation tools based on Computational Fluid Dynamics, in which the urban neighbourhood is modelled in detail, but where the characteristics of the boundary layer are only included in an approximate way. The second group focuses on the regional scale, like for weather forecasting. They use detailed models describing atmospheric stratification, the effect of Coriolis forcing and terrain topology, but include the urban cities by means of a simplified parameterization. Although totally different scales are considered, many applications heavily rely on both scales. For instance, the weather forecast can predict high temperatures, but being seated in the shadow of a single tree can still be comfortable. In other words, local climate predictions depend both on local and on global factors. The opposite also holds. The local production of fine dust in a city causes enhanced cloud formation in the region downstream.

The aim of this project is to develop a multi-scale framework that allows studying exactly those applications where both scales matter. Should different models be applied in different regions? Or should we rely on model nesting? Several one and two-way coupling strategies are evaluated in terms of efficiency and accuracy. Besides spatial issues, also temporal aspects are considered.

Tasks: Literature survey; getting familiar with micro- and mesoscopic numerical models; designing a coupling strategy and proving its efficiency.

Type: Bachelor, Semester or Master thesis

Internal supervisor: Peter Moonen

Bitte wechseln Sie Ihren Browser
Seite zuletzt geändert am 11.10.2010 13:17 Uhr