Wednesday, 25 January 2017
4E (Washington State Convention Center )
Handout (487.3 kB)
Academic studies of urban transport and dispersion phenomena usually focus on highly idealized conditions. Countless studies of both simulations and wind-tunnel models involve perfectly two-dimensional street canyons under perfectly steady wind speed and wind direction perpendicular to the street canyon. The Joint Urban 2003 field campaign includes a continuous plume release under nearly ideal conditions with steady 15-minute averaged wind speed and direction, and direction within about 10 degrees of perpendicular to the street canyons. In spite of the nearly ideal conditions, the observed plume behavior for this case is complex and difficult to simulate. The observed plume spreads much more rapidly than one would normally expect, even when simulating the effect of buildings on the plume. I will discuss the subtle details, that are neglected in overly idealized models and simulations but commonly occur in real cities, which contribute to the observed complex plume behavior. Specifically, while the 15-minute average wind direction was very steady, oscillations exist at smaller time scales. These oscillations around a nominally perpendicular wind direction will cause the channeling flow through the street canyons to switch directions and potentially causing complex interactions and enhanced vertical mixing. These oscillations will not be found in the prevailing flow in overly idealized models and simulations. In fact, perfectly perpendicular winds to the street canyons in an urban area are a rare exception to the more general case where an along-canyon component affects the flow in the canyon. Additionally, some of the complex plume behavior observed in this case suggests that infiltration into and subsequent exfiltration out of a large underground parking garage near the source may also significantly affect the initial plume spread.
![](images/flag.png)
![](images/blue_ribbon.png)