Modélisation de la pollution par le trafic à Bruxelles

Transcription

1 Modélisation de la pollution par le trafic à Bruxelles Priscilla Declerck, Thierri de Vos, Billie Heene, Olivier Brasseur Laboratoire qualité de l air - Bruxelles Environnement

2 Introduction 2

3 The Brussels-Capital Region: A dense urban Region consisting of 19 municipalities Approximately 1.1 million inhabitants Population has grown fast over the last decade, and is still growing (+ 1.2%) Economic heart of Belgium more than 37. commuters enter the city every day Public transport 15% Walk 4% Bicycle 2% Extremely dense transport network Car 47% Traffic related air pollution Challenges concerning mobility Train 32%

4 CANyon Street model for Black Carbon (CANS BC ) based on the approach of the Operational Street Pollution Model (OSPM) 7

5 CANS BC : no chemical module Input data: Meterological data: Wind a height of 3m (Molenbeek) Wind velocity Urban background: BC conc. from the urban background station of Uccle Traffic emissions Depending on vehicle countings Temporal resolution of 15 Vehicle speeds were kept constant 8

6 Direct validation: the Crown Street 9

7 1

8 Avenue canyon Veh/day= 15. H = 12m W = 19 m 11

9 AIMS Monthly evolution: variable meteorological conditions Half-hourly evolution: variable traffic emissions Sensitivity to the wind direction: capability of estimating the local contribution to BC concentrations Specifications Two years simulation period: 7/211 6/213 Observed BC concentrations: fixed BC measurement station in the Crown Street Simulated BC concentrations: CANS BC

10 7/211 9/211 11/211 1/212 3/212 5/212 7/212 9/212 11/212 1/213 3/213 5/213 BC concentration (µg/m³) Monthly evolution 6 5 Observation Simulation (OSPM-BC) CANS

11 BC concentration (µg/m³) Diurnal evolution Observation (41R2) Simulation (OSPM-BC) CANS Hour (UT)

12 Observation Traffic emission Recirculation Background Sensitivity to wind direction

13 Sensitivity to wind direction Observation Traffic emission Recirculation Background

14 Sensitivity to wind direction Observation 27 9 Simulation Traffic emission (OSPM-BC) (CANS-BC) 26 1 Recirculation Background

15 Traffic emission Recirculation Background Observation Sensitivity to wind direction

16 Sensitivity to wind direction Traffic emission 27 9 Recirculation 26 1 Background Observation Very good agreement between CANS BC and observation

17 Sensitivity to wind direction Overestimated BC concentrations overestimated local contribution Traffic emission 27 9 Recirculation 26 1 Background Observation

18 Sensitivity to wind direction Underestimated BC concentrations too weak recirculation and local contribution? Traffic emission 27 9 Recirculation 26 1 Background Observation

19 Indirect validation: the Belliard street 23

20 24

21 Regular canyon Veh/day= 27. H = 21 m W = 19 m 25

22 AIM Validate simulated BC concentrations based on the linear relation between BC and NOx Specifications Simulation period: 11/212 6/213 Observed NOx concentrations: fixed NOx measurement station in the Belliard Street Simulated BC concentrations: CANS BC

23 NOx Relation between BC and NOx y = 24,823x Black Carbon

24 NOx Relation between BC and NOx y = 24,823x 15 CANS BC [BC] [NOx] = 24.8 [BC] Black Carbon

25 NOx concentration (µg/m³) Diurnal evolution Observation (41B8) (41R2) Simulation (OSPM-BC) CANS Hour (UT)

26 Direct validation: the Boulevard Anspach 3

27 31

28 Avenue canyon Veh/day= 14. H = 22m W = 27m 32

29 AIMS Sidewalk: left vs right Specifications Simulation period: 18 days at rush hours (morning and evening) Observed BC concentrations: mobile BC measurement using AE51 aethalometer Simulated BC concentrations: CANS BC

30 Sidewalk: left vs right Spatial variations of BC concentrations along sidewalk Stop/start at traffic lights Consider «averaged sidewalk»

31 Concentration BC (µg/m3) Sidewalk: left vs right All cases Very good agreement x (m) C(x) CANS all BC mean CANS BC measure Observ.

32 Concentration BC (µg/m3) Sidewalk: left vs right Recirculation on the right Overestimated gradient x (m) C(x) CANS R BC mean CANS BC measure Observ.

33 Concentration BC (µg/m3) Sidewalk: left vs right Recirculation on the left Overestimated gradient x (m) C(x) CANS L BC mean CANS BC measure Observ.

34 Concentration BC (µg/m3) Sidewalk: left vs right No recirculation Strongly underestimated gradient x (m) C(x) CANS P BC mean CANS BC measure Observ.

35 Sidewalk: left vs right No recirculation if: Weak wind (< 2 m/s) : left/right gradient correctly simulated Wind direction close to street axis: Angle street / wind > 1 Angle street / wind < 1 Observation CANS BC Observation CANS BC

36 Conclusions 4

37 Validation tests showed that CANS BC : Performs well at annual scale in simulating average monthly and diurnal BC evolutions The model correctly handles: the meteorological conditions the variation in traffic emissions Is suitable to accurately simulate BC concentrations in canyon streets of the BCR if: Accurate estimate of traffic emissions (counting, driving mode, emission factors, ) Vehicle speeds turbulence generated by vehicles Able to simulate the gradient of BC concentration between left and right sidewalks but improvements are needed when the wind direction is close to the street axis. 41

38 Thank you for your attention