S-Space College of Natural Sciences (자연과학대학) Dept. of Earth and Environmental Sciences (지구환경과학부) Theses (Ph.D. / Sc.D._지구환경과학부)
Implementation and modification of an urban canopy parameterization in MM5/URBANIZED
도시캐노피 밑 난류특성매개변수화를 고려한 중규모 모델개발 및 이를 이용한 도시특성 모의
- 자연과학대학 대기과학과
- Issue Date
- 서울대학교 대학원
- Urban Canopy Parameterization; modification of eddy diffusivity; momentum and heat flux parameterization; modification of roughness length scale; UHI; transient events; high ozone concentration in nighttime
- 학위논문 (박사)-- 서울대학교 대학원 : 대기과학과, 2014. 2. 윤순창.
- This study is carried out to simulate the flow and turbulence in urban area for air pollution modeling using a modified urbanized mesoscale model. The Urban Canopy Parameterization (UCP) of Dupont et al. (2004) implemented in MM5v3.7 mesoscale meteorological modeling system (MM5-UCP-Basic) is evaluated against the non-urbanized model (MM5-NoUCP) using measurements taken in urban Phoenix during two field studies. In general, MM5-UCP-Basic improved the predictions of typical meteorological parameters. Nevertheless, significant discrepancies still exist between observations and the predictions of MM5-UCP-Basic, and new parameterizations and land use classes are introduced to improve the model performance.
The parameterization of anthropogenic heat flux from buildings and roadways is also included. The land use classes in the improved model (MM5-UCP-MOD) represent roadways and rivers, in addition to five classes of buildings identified in MM5-UCP-Basic.
New parameterizations considered the appropriate roughness length, velocity decay during evening transition, and heat and momentum diffusivities for the nocturnal period so that account for different heat and momentum transfer rates under stable atmospheric conditions. Five nested grid domains are used for simulations, with the highest resolution (1km) implemented into the MM5-NoUCP, MM5-UCP-Basic and MM5-UCP-MOD.
Improved parameterizations were validated by detailed flow and turbulence measurements which were conducted as the Phoenix SUNRISE field experiment in 2001 (Lee et al., 2003
Doran et al., 2003) and TRANSFLEX in 2006 (Fernando et al., 2013).
This research was to refine the urban land use classification in MM5-UCP-Basic and modify the turbulence parameterizations to better represent surface fluxes and urban effect such as UHI and LLJs. The features of this modified model, MM5-UCP-MOD, are presented in this research together with its validation and comparisons with MM5-UCP-Basic and original MM5v3.7 (referred to as MM5-noUCP).
According to sensitivity tests for parameterization improvement, the parameterization changing the turbulence length scale in TKE is the most significant to develop the high performance of momentum flux in urban simulation.
In general, substantial improvements in the prediction of wind speed, temperature (especially during the nighttime) and momentum flux as well as a smaller improvement in the heat flux are noted, so that is pointing to possible further enhancement onto model performance by including the improved physics.
By the way, the MM5-UCP-MOD is useful to understand the UHI and urban meteorology so as to simulate and predict the nocturnal air pollution in the city, especially. The temperature field and heat flux including effect of UHI in urban center are better simulated by MM5-UCP-MOD with urban data for Phoenix than by the standard version of MM5 (MM5-NoUCP).
Besides, transient events of end of high mountainous area, which enclosed the downtown, are also accomplished to simulate by MM5-UCP-MOD. The transient events are typically developed by coupled effect of topographical condition of Phoenix and thermally driven flow in neighborhood scale. And it seems to be analyzed by drag force approach inside roughness sub-layer.
The budgets of turbulent kinetic energy near the boundary from output of MM5-UCP-MOD are able to understand the turbulent energy transform near the top of building canopy.
Since three-dimensional observations are not enough to verify the simulation and the use of urban and vegetation canopy morphology database with land use type are too simplified, validation of new parameterization is not enough.
Nevertheless, the formation of LLJ owing to UHI and transient events with nocturnal downslope flow in transient time can be explained using MM5-UCP-MOD. And also, this effect produced the nocturnal high ozone concentrations and can be only simulated by MM5-UCP-MOD. Although the limitation of comparison with measurement for ozone and PM10, the analysis of the model outputs emphasized that results from MM5-UCP-MOD and CMAQ are useful to understand and predict the urban meteorology and air pollution.