Under clear and calm weather conditions thermal wind systems strongly influence air quality and heat load in urban areas. The city of Munich (Germany) is located in the northern foreland of the European Alps. Due to the undulating and only slightly inclined terrain in Munich, local slope and valley wind systems are found to be rather weak, while on the regional scale a diurnal reversing thermal circulation system is occurring regularly, extending from the Alps to the Danube Valley, i.e. about 50 km north of Munich. The influence of this so-called ‘Alpine pumping’ on urban climate in Munich is currently investigated in a cooperative project of the City of Munich and the German National Meteorological Service (Deutscher Wetterdienst).

Regional climate simulations for a 20 year period using the COSMO-CLM model are carried out to identify days with Alpine pumping and to determine the mean diurnal variation of the direction, intensity, and extension of the regional thermal circulation. In a further step local scale temporal observations and high resolution simulations with the urban climate model MUKLIMO_3 are used to quantify the impact of Alpine pumping on ventilation and thermal conditions in Munich.

In evaluation of unsteady heat and pollutant diffusion, or wind resistance design of buildings and structures in cities, large eddy simulation (LES) is one of the best appropriate numerical techniques for prediction of the wind flow and transport phenomena. Especially it is important to capture the near-ground flow field within the urban canopy layers. Some researchers carried out LES for wind flows over actual urban districts.

However, in order to reproduce the flow field in the urban canopy, not only the local effect by surrounding detailed geometry but also the long-ranged effect based on the developing atmospheric boundary layer should be considered. In particular, the urban geometry located in the windward area is an influential factor determining the development process and the turbulent characteristics of atmospheric boundary layer.

Therefore, this study carries out LES in the large region over 10km and analyzes the statistics of turbulent flow in the development process of atmospheric boundary layer. Moreover, the effect of turbulence characteristics in large region on the flow field in the urban canopy layer is examined.

First, we consider the generation technique of inflow turbulence for connecting with LES in the large urban region. In the process of generating inflow turbulence the driver region is set in order to add high-frequency fluctuation to filtered flow field with low-frequency which is usually obtained by the other type of simulation such as unsteady RANS or the meteorological model. Here, we decompose the obtained unsteady flow field by filtering technique and the fluctuating part of the unsteady flow is used for recovering the fluctuations in the high frequency region. As a result of the spectrum analysis for the obtained time series data, it is confirmed that the wide range of distribution of power spectrum density can be obtained recovering the high frequency fluctuation.

Next, one example using above technique is shown. In order to reproduce the complicated flow field in the urban canopy layer, GIS data of building geometry and topography in the large region over several 10 km is applied to construction of the urban model. Mesh grids, which reproduce detailed geometry such as roughened actual building shape or vegetation aspect in actual urban area, are generated from the collected GIS data. Then, LES computation in the 10km region is performed by using the generated inflow turbulence for inflow boundary condition. In this LES, several sampling points are set along the streamwise direction in order to check the change of vertical profiles of mean flow and turbulent intensity.

In the present study, we consider Tokyo area as numerical model which is located on the coastline. We set the measurement points along 2 straight lines from the sea, which pass over skyscraper and low-rise residential areas. We should note that the boundary layer thickness becomes thin on the sea.

As a result of the simulation, the atmospheric boundary layer develops gradually in inland of the urban districts where the middle rise buildings are packed densely. The thickness reaches to up to 600m in the sky scraper area, which is several km away from the coast of Tokyo bay. On the other hand, the analysis of wind profile on the way passing low-rise residential area shows the smaller development compared with the wind profile on the other way. Moreover, the vertical profile of turbulent intensity at each point shows that some effects of the oncoming turbulent flow generated by the windward isolated building remain even in the point a few km away from the windward building. Also it is recognized that the turbulent fluctuation characteristics in the urban canopy layer are formed due to the surrounding circumstances. It can be said that the characteristics of urban canopy flows are determined by various effects of the approaching turbulent boundary layer itself, some disturbance generated by buildings at relatively distant location and the wake flow of the surrounding obstacles.

Understanding the interaction between large-scale atmospheric and oceanic circulation patterns and changes in land-cover (LCLU) due to urbanization is a relevant subject in many coastal climates. Recent studies by Lebassi et al. (2009) found that summer average-maximum air temperature in two populated California coastal areas showed cooling trends at low elevation coastal areas open to marine air penetration and warming trends at inland and high elevation coastal areas due to an increase in sea-breeze activity during the period. These results require further analysis of the influence of changes in LCLU and other large-scale signals in these changes in the Los Angeles area through numerical modeling.

The aim of this work is therefore to quantify the thermal response of the land-surface and impacts on sea breeze to answer science questions for this process and to assess the suitability of the regional atmospheric modeling system to be used in long-term modeling studies. A field study was configured to observe the sea-breeze intensity and penetration via 18 surface ground stations along the projected transect of the sea breeze during September 24th 2013, one of the days of the NASA HyspIRI Mission preparatory flight campaign over Southern California. The land surface temperature (LST) map collected at approximately 1200 Local Time (LT) over the area with the MASTER sensor showed land-sea surface temperatures differences in the order of 20 °C for great part of the domain. The observed wind speeds reflected the thermal response to this gradient through an increase in the sea-breeze speed from 2 to 5 m s-1 only at coastal sites at 0900 LT to a 9 m s-1 in coastal and inland locations at 1600 LT. New urban land classes were derived from the broadband albedo of the AVIRIS sensor and then ingested into the Weather Research Forecast (WRF) model to represent these coastal/urban processes. Results from the modeling showed an improvement of modeled surface temperatures and wind speeds using the higher resolution HyspIRI derived products into the model against the default model characteristics. The model also captured the diurnal spatial and temporal patterns of the sea breeze in the region. These new data records and updated modeling are enabling numerical ensembles to respond to science questions related to causes of the coastal cooling process due to land use, global warming and large scale oceanic processes.