Bifurcation of a Summer Convective Thunderstorm over Beijing, Part 2: Analyses of uWRF Simulated Thermodynamic Causes of the Observed Bifurcation

The previous paper discussed the observation and simulation of Beijing urban-area bifurcation of a summer evening convective thunderstorm (TS) precipitation event.  The case was observed during the Study of Urban Impacts on Rainfall and Fog/Haze (SURF) Project IOP on 22 July 2015.  The first paper discussed two uWRF simulations: (a) Urban, which used 2010 land use data and (b) No-Urban, which converted all urban Beijing grid cells into cropland, the dominant surface type surrounding Beijing.  This paper provides additional analyses of the uWRF outputs to provide preliminary insights into the thermodynamic mechanisms leading to the observed storm bifurcation. 

A vertical cross section along the SE to NW wind flow (that existed at 1815 LST, one-hour before the TS outflow reached Beijing) showed a weak 1.2 km deep urban heat island (UHI, defined by the Urban minus No-urban simulated temperature-differences) that had been advected to the (NW) downwind urban edge; the UHI was capped by a cross-over layer (of negative Urban minus No-urban temperatures).  As high pressure existed downwind of the urban area, as opposed to a thermal low pressure, the UHI was expected to have little effects on TS structure.

In the No-urban simulation, the 1.5 km deep (mainly horizontal) SE to NW flow blew past the area normally occupied by Beijing.  This was capped by a reverse-directed (also mainly horizontal) flow from the downwind mountains; these two flows were vertically linked by an upward directed flow (2.5 km deep, with a max speed of >0.2 m/s) over the area normally occupied by Beijing.  In the Urban case, the SE to NW flow below 2.0 km decelerated (as expected) over the rough upwind urban-edge, producing a 2.0 km deep horizontal-convergence (and hence upward-motion) layer above that location, as the flow was bifurcating around the city.  This induced a region of compensating downward motion over the urban center, which completely eliminated the upward motion at that location found in the No-Urban simulation.

At about 1900 LST (45 min later), when the NW to SE directed TS-outflow in the Urban simulation reached the NW edge of the rough urban area, it was slowed and also bifurcated around the city, as was the (still) SE to NW flow at the SE urban edge.  These effects produced upward flow at both lateral edges and strong (>1.0 m/s) downward motion over the urban area.  These up and down motions extended upwards beyond 3 km, and thus were able to impact the thunderstorm dynamics.  Without an underlying rough urban surface in the No-urban simulation, its NW outflow moved faster over the urban center, and a convergence zone (with upward motion of about 1.0 m/s) formed at its forward boundary (over the area of the missing downwind SE urban edge). 

To summarize, the urbanized underlying surface over Beijing urban slowed and bifurcated horizontal winds within the lowest 3 km and thus caused precipitation bifurcation.  Further analyses will focus on the pressure field and on multiple Urban simulations in which different urban forcings are suppressed and/or expanded, e.g., buildings, anthropogenic heat via inclusion of the BEM module, aerosol concentrations, so that the roles of each bifurcation mechanisms can be further evaluated.