Development of a microscale model for the thermal environment in complex areas
Along with the increasing fraction of urban population, there is increasing demand for the assessment of thermal conditions for humans within the urban environment. This can be achieved best by calculating thermal indices, that represent the thermal perception and thermal stress of an averaged human being in high spatial and temporal resolution. Thermal indices are based on a number of meteorological and physiological parameters, that need to be provided in the desired spatial and temporal resolution. The required high spatial resolution makes it impossible to measure all of the required data. It therefore has to be calculated by a numerical model using data from a reference station.
The most important meteorological input parameter for the calculation of thermal indices is the mean radiant temperature (Tmrt ). It summarizes the effect of short- and longwave radiation. Tmrt can not be measured, but needs to be calculated by complex radiation modelling taking all effects of the current environment into account. In urban areas with all types of different vertical structures, materials and surfaces Tmrt is highly volatile in space and therefore needs to be calculated in high spatial resolution.
An other very important parameter in urban biometeorology is wind speed. This is because it shows great impact on humans itself, but also because it influences other parameters that are important in urban biometeorlogy. Assessment of wind speed in complex urban environments requires spatially resolved data, as it shows strong spatial variability. This spatially resolved data can only be provided by numerical modelling. A wind model suitable for the application in a small scale model has to meet quite some demands. One of them is that it should be rather fast, to allow for lots of data sets and, thus, for more representative assessment. The model should also be able to cope with time independent input data, as long data series often contain gaps. Therefore, only a diagnostic wind model comes into consideration, as it is time independent.
Currently only few operational diagnostic wind models exist. Most of them are designed for the use in particle dispersion modelling. Examples for models like this are TALdia or QUIC-URB. They are unfortunately not suitable for the integration in a small scale model for technical reason or legal issues. The construction of a new wind model was therefore considered to be the best option.
The diagnostic wind model developed in the course of this thesis is based on the approach and the parametrisations of the ABC model. Following this approach, first an initial wind field is constructed that already must contain the different modifications to the undisturbed wind field caused by the obstacles in the model domain. This initial wind field contains a lot of divergence that is reduced by the multiplication by an Langranian multiplier. The Langranian multiplier has to be determined previously by solving a Poisson equation using the Successive-Over-Relaxation method.
During the construction of the initial wind field modifications have to be calculated for every obstacle. The model calculates four different types of modifications. A stagnation zone windward from the obstacle, a recirculation on the lee-side, as well as a velocity deficit zone adjacent to the recirculation. If a street canyon is detected, a vortex is placed inside it.