Relating street level flows to BT Tower level flows: results from the DAPPLE

2004 campaignJ. Barlow1, A. Dobre1, R. Smalley2, S. Arnold1, A. Tomlin2, S.

Belcher1

1Department of Meteorology, University of Reading, UK2Energy and Resources Research Institute, University of

Leeds, UK

• Street canyon flow

Street level flowsStreet canyon , aspect ratio H/W=0.6

Perpendicular flow

• Flow “rectification” means that street level flow pattern is very sensitive to outer flow direction

Oblique flows

• Flow “rectification” means that street level flow pattern is very sensitive to outer flow direction

Oblique flows

• For accurate prediction of street level flow or dispersion Need suitable reference measurement

• Vegetation canopies: use windspeed or friction velocity at canopy top

Choice of reference

U/UH

z/H

1

1

• Urban canopies:Roof-top reference practical BUT local obstructions cause wakes, limited representativity

• Vegetation canopies: use windspeed or friction velocity at canopy top

• Urban canopies: Roof-top reference practical BUT local obstructions cause wakes, limited representativity

TODAY:

Results from DAPPLE 2004 campaign in London, referencing street level flow

Compare roof-top and upper level references

Choice of reference

U/UR

z/H

1

1stable• Higher reference “cleaner”

BUT stability can affect flow Klein and Clark (2007)

Oklahoma City, stable conditions, frequent nocturnal jets

better to use rooftop ref as higher ref in “decoupled” flow

DAPPLE 2004 Field Campaign

Equipment:

• 11 3D ultrasonic anemometers

• Qinetiq Zephir Doppler lidar (3rd June)

Campaign duration:• 19th April to 13th June 2004

Site:

• mean building height 21m (radius of 200m)

• plan area index λP ~ 0.5

• frontal area index λF ~ 0.2 (bearing ~240°)

WCC

LIB BT

WCC ref

WCC

LIB BT

LIB ref

WCC

LIB BT

Heights of measurement

Uninterrupted flow

Flow influenced bybuildings

ZWCC, LIB = 17m

H= 21 m

zBT = 190 m

Z ~ 2-3H

Z = H

Z ~ 9H

Z ~ 0.2H

Evaluating reference sites

Turbulenceintensity

Local flow direction

• BT Tower: circlesSmall, approx. constant• LIB: trianglesPeaks associated with wakes

Windspeed and direction

0

0.5

1

1.5

2

2.5

3

3.5

0 2 4 6 8 10 12 14 16

Ubt

Ulib

/Ub

t

• Windspeed ratio:LIB near neutral limit 0.23

• Direction:LIB better correlation with BT ref

Vector decomposition model

Roof top wind = channelled + perpendicular

In-street wind components:

u1=aur1

u2=bur2

In-street wind direction:

tan u2u1 = b/a tanr

(Dobre et al. (2005), Atmospheric Environment, 39(26), 4647-4657 )

θr

ur1

ur2

θu1

u2

Evaluation using Dobre et al. model

BT Tower: best fit of predicted direction to data

WCC ref

LIB ref

Conclusions

• Mean flow pattern in street is most closely related to upper level reference on BT Tower (z ~ 9H)

Develop BT Tower as centralised reference in London (5 year long ACTUAL project)

• Occasional “decoupling” events when stable overnight or in low wind periods

• Stable layers not common for London (<1% of 6 weeks) therefore upper level reference more representative than for Oklahoma / Klein and Clark 2007

Overnight urban stability depends on regional scale forcing, not just local urban energy balance j.f.barlow@reading.ac

.uk

Stable “decoupling” of turbulence 2nd May ‘04

-0.5

0

0.5

1

29/4/04 0:00 30/4/04 0:00 1/5/04 0:00 2/5/04 0:00 3/5/04 0:00 4/5/04 0:00

time

norm

alis

ed

turb

ulen

t ki

netic

ene

rgy

e/

U^2

0.5

1

1.5

2

2.5

3

3.5

4

ratio

of

pote

ntia

l son

ic t

em

pera

ture

s,

BT

/roo

f

BT e/U^2

Lib e/U^2

ratio temp

Correlation street level TKE with UBT

2 = 0.24 (whole campaign) cf. 0.14 for ULIB

2

Correlation street level TKE with UBT

2 = 0.14 (overnight)

0

5

10

15

20

25

30

29/4/04 0:00 30/4/04 0:00 1/5/04 0:00 2/5/04 0:00 3/5/04 0:00 4/5/04 0:00

time

soni

c te

mpe

ratu

re (

degC

)

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

stan

dard

dev

iatio

n so

nic

T (

deg

C)

BT Ts

Lib Ts

BT sigT

Lib sigT

Bulk Richardson number >0.25, stable conditions

Stable “decoupling” of turbulence 2nd May ‘04

Dobre et al. model for different sites

Site 3 z ~0.3H Site 4 z ~0.3H Site 11 z ~0.1H

WCCLIB BT