Spatial distribution of the atmospheric mixed-layer depth over Tehran using numerical simulations: Two case studies

The atmospheric boundary layer (ABL) is the lowest part of the atmosphere and it plays an important role in the assessment of air quality, transport process in the lower troposphere and climate change. The thickness of ABL varies in time and space. The Atmospheric boundary layer over an urban area with complex horizontal topography horizontally can be highly non-homogeneous. Daytime ABL, known as mixed layer (ML) can be affected by sloped surfaces and urban roughness therefore it assumes a complex structure, contrary to what happens over a flat terrain. In a non-homogeneous ML, not only turbulent convection is important, but also advection and venting to the free atmosphere is important. Such advections are mainly caused by the sloped surfaces and valley circulations in mountainous areas as that of the north of Tehran. Convection and advection by mountain forcing in the boundary layer can be important for the air pollution problem and even transport processes around cities as for the Tehran urban area.
In this paper the structure of ML over the urban area of Tehran and its surroundings, has been studied using numerical simulations. The simulations have been done with WRF (Weather Research and Forecast) model with four nested, one way grids. The simulations of the finest inner grid with horizontal resolution of 1.33 km and varied vertical grids (typically 70 m near the surface to 600 m in the free atmosphere), with two boundary layer schemes, namely YSU (Yonsei University) and MYJ (Mellor Yamada Janjic) which are non-local and local respectively, are here presented. Radiosonde measurements as well as surface meteorological data at the Mehrabad synoptic station are used to evaluate the performance of the boundary layer schemes. Comparison of surface and boundary layer observations with WRF simulations show that the YSU scheme, which is a nonlocal closure scheme, gives more realistic results. In any case, both schemes gives cooler and more moist boundary layers than observations, as also have been observed in other studies.
The ML height over the area varies and depends on surface parameters as surface elevation, 2m temperature and humidity and surface heat fluxes. Potential temperature advection affects the boundary layer height over slopes, especially in summer season. Cold advection of potential temperature caused by the up-slope stream, decreases instability of air near the surface and thus lowers the ABL height. ML height, derived from the simulated potential temperature and moisture profiles over the city, shows a maximum value near the center of the area and a minimum value near the mountain tops. Results also show that in areas where vertical wind speed and moisture have relatively high values, the potential temperature and water vapor mixing ratio profiles do not show the same values for boundary layer height. Typical ML heights for the summer and winter days are respectively 2400 and 1200 for the central part of the city. Maximum ML height variations over the area for summer and winter times are about 1700 m and 1000 m respectively. The main determining parameter in ML height spatial variations is surface heat flux for winter, while the thermal advection is more important in summer days.