Urban Canyon Microclimate

Comprehensive 2D analysis of urban microclimate aerodynamics, covering canonical configurations that model how wind interacts with buildings at the neighbourhood scale. The analysis reproduces the three flow regimes of Oke 1988 as a function of building aspect ratio H/W: isolated roughness (H/W < 0.3), wake interference (0.3-0.65), and skimming flow (> 0.65). Two top-down orientations now cover wind travelling in both directions through a street network.

1. Side-View Canyon (Cross-Section)

Buildings of equal height H form a row of street canyons of width W. The flow approaches from the left, separates over each upstream building, and either reattaches between the buildings or skims over the top depending on H/W. The first three cases use two buildings; the H/W = 0.6 case uses three buildings to emulate a denser downtown street network.

1.1 Oke 1988 Flow Regimes

H/W	Buildings	Regime	Description
0.3	2	Isolated roughness	Buildings act as independent obstacles; wakes do not interact. A recirculation zone behind each building reattaches before the next.
0.5	2	Wake interference	Wakes interact moderately. The downstream building sits within the wake of the upstream one, creating an elongated recirculation.
0.6	3	Wake interference (dense)	Denser downtown network; wakes interact across three buildings, stronger canyon recirculation.
0.8	2	Skimming flow	A single stable recirculation vortex fills the canyon. Flow skims over the rooftops with minimal vertical exchange.

---

2. Top-Down Plan View: Vertical Buildings

Three tall rectangular building footprints (long in the cross-flow direction) in a wider channel (NX=900, NY=400) with realistic street spacing (canyon = 2 × building width). Wind travels left to right around the buildings, showing street-level wind patterns with high-speed flow in the gaps and low-speed recirculation in the wakes.

---

3. Top-Down Plan View: Horizontal Buildings (Funnel Flow)

Three long rectangular buildings (long in the flow direction, short across it) stacked across the channel so wind is forced through the gaps between them. This is the complementary orientation: wind is funneled along the pedestrian through orifice-like gaps, accelerating in the passages -- the same physics as the orifice-plate internal flow, but at the street scale. Together with Section 2, both wind directions through a city network are now covered.

---

4. Building Downwash

A tall building upstream and a low-rise building downstream. The tall building deflects the high-speed wind downward, creating a downwash effect that amplifies pedestrian-level winds. This configuration is relevant to the Hunt 1984 criterion for pedestrian comfort in urban developments.

---

Key Findings

• Regime transitions match Oke 1988: H/W = 0.3 shows isolated roughness (wakes don't interact), H/W = 0.5 shows wake interference (elongated recirculation), H/W = 0.8 shows skimming flow (stable vortex fills the canyon).
• Dense downtown (H/W = 0.6, 3 bldgs): Three-building configuration produces stronger canyon recirculation than the 2-building case at the same H/W, demonstrating that building density amplifies the wake interference effect.
• Top-down orientation matters: Vertical buildings (long across flow) create strong lateral deflection with high-speed gaps, while horizontal buildings (long along flow) funnel wind through orifice-like passages -- both are realistic street-network scenarios.
• Downwash amplification: The tall-to-low building height ratio produces a measurable downwash effect that would increase pedestrian-level wind speeds in real urban developments.

LBM Analysis

All urban simulations run at Re = 100 on 900x400 grids with the D2Q9 MRT lattice. At this laminar Re, the flow captures the correct qualitative regime transitions predicted by Oke 1988, which are Reynolds-number independent in the mean. The rectangular building geometries are encoded directly in the obstacle mask, and the Bouzidi interpolated bounce-back provides second-order accurate boundary conditions on the building surfaces.

The side-view canyon series systematically varies H/W from 0.3 to 0.8, reproducing the full range of Oke regimes. The H/W = 0.6 case uses three buildings (NY * 2/5 = 160 cells tall, 40% of domain height) to model a denser urban fabric. The two top-down configurations complement each other: vertical buildings (long in the cross-flow direction) produce lateral deflection and gap acceleration, while horizontal buildings (long in the flow direction) create orifice-like funnel flow through street passages. Together they cover both principal wind directions through a city network.

Discussion

Urban microclimate CFD at LBM-relevant Re values (100-200) is laminar, while real urban flows are turbulent (Re > 10⁶). However, the qualitative flow regimes predicted by Oke 1988 are Reynolds-number independent in the mean; the aspect ratio H/W determines the recirculation pattern regardless of Re magnitude. These laminar simulations capture the correct regime transitions.

Real-world applications: pedestrian wind comfort assessment, pollutant dispersion in street canyons, natural ventilation potential for buildings, and urban heat island mitigation. The two top-down orientations together cover wind approaching a street network from both principal directions. Future work: extend to 3D with the D3Q19 lattice for corner-flow effects around finite-length buildings.