The Urban Microscale Model Intercomparison Project (UMMIP)

A computational fluid dynamics model intercomparison for building-resolving simulation of the urban atmospheric boundary layer.

Motivation

Over the past decades, urban microscale models have been increasingly used to simulate the urban atmosphere. Today, a multitude of models exist that differ in complexity, employed parameterizations, and degree of parallelization. While simple model configurations (i.e. simple building configurations, steady-state conditions, and no thermodynamics) have been often validated against wind tunnel experiments, the evaluation of more complex configurations is difficult. Wind tunnels are not able to take into account complex processes like radiative transfer, non-stationary boundary conditions, and complex thermodynamic interactions with the urban surface. By the same token, real urban environments are far too complex to be aggregated in data sets generated in field experiments. A validation based on such data sets is thus impossible.

The UMMIP aims at providing a comprehensive data set based on multi-model simulations for well-defined set-ups in urban environments. The complexity of the set-ups is increased in several stages based on existing and newly-developed set-ups suitable to be simulated by most of the participating models. UMMIP aims at showing how much the simulations results differ between individual models and therefore will provide an educated estimation of the reliability of urban microscale models. Uncertainties that do not arise from insufficiently accurate input data (like building shape, wall and roof mounting, distribution of vegetation, etc.) or external forcings but from model numerics and physical parameterizations are to be investigated. Finally, we also strive at providing detailed information about model optimization in terms of degree of parallelization and performance.

The model intercomparison encompasses models based on Reynolds-Averaged Navier-Stokes (RANS) equations and turbulence-resolving Large-Eddy Simulation (LES) models.

Set-ups

The setups are XXX, YYY, and ZZZ and are described in detail here?.

Input Requirements

The input requirements are described in detail here?.

Output Requirements

The output requirements are described in detail here?.