Post on 31-Aug-2021
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
Experimental Campaign in a Heavily Trafficked
Roundabout in Madrid for the Assessment of Air Quality
Monitoring Station Representativeness in Terms of
Population Exposure to NO2
R. Borge, C. Quaassdorff, D. de la Paz, A. Narros, J. Pérez, J.M. de
Andrés, R. Viteri, M. Paredes
Laboratory of Environmental Modelling
Technical University of Madrid (UPM)
rborge@etsii.upm.es
Session A4: Healthy-Polis Workshop on Climate Change and Urban Health – II
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
1. Introduction
2. Methodology
2.1. NO2 concentration
2.2. Pedestrian simulation
2.3. Exposure assessment
3. Results and discussion
4. Conclusions
OUTLINE
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
1. INTRODUCTION
Summary of the AQ Directive’s limit values, target values, long-term objectives, information
and alert threshold for the protection of human health. Source: EEA (2014)
•Poor urban air quality is one of the main environmental concerns worldwide
• According to WHO (WHO, 2014) outdoor air pollution caused 3.7 million premature
deaths in 2012, most of them in urban areas where both, emission sources and
population concentrate
•As a consequence,
increasingly stringent air
quality standards for the
protection of human health
are being put into force: e.g.
Directive 2008/50/EC on
ambient air quality and
cleaner air for Europe
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
Annual mean NO2 (2012). Red and dark dots correspond to
exceedances of the annual limit value (40 µg/m3). Source: EEA (2014)
•Many urban areas in Europe are struggling to meet these air quality standards, particularly
for NO2
Percentage of the urban population in the EU-28 exposed to air
pollutant concentrations above EU and WHO reference levels (2010-
2012). Source: EEA (2014)
•Exceedances imply excessive
population exposure
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•However, exceedances are determined from routinely measurements of urban air quality
monitoring stations
•It is unclear what is the
temporal and spatial
representativeness of such
monitoring stations
•It is difficult to assess the actual
population exposure
To what extent information from air quality
monitoring stations is representative for
population exposure and therefore the
assessment of AQ limit values compliance?
Urban
backgroundTraffic
Industrial
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•This study presents a methodology to assess the representativeness of air quality
monitoring stations in terms of population exposure to ambient air pollution
•It uses a particular location in Madrid (Spain) as a case study:
• Madrid city (Spain): 3.2 million
inhabitants in the city, more
than 5 million people in the
metropolitan area
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
Observed NO2 values (corresponding to the annual
and hourly NO2 limit values defined in the European
AQ Directive) in the Madrid air quality monitoring
network for the years 2010-2013
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
Traffic stations
Urban background stations
Suburban background stations0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
Traffic stations
Urban background stations
Suburban background stations0
50
100
150
200
250
300
350
400
0 10 20 30 40 50 60 70 80
Annual mean NO2 concentration (µg/m 3)
19th
hig
hes
t h
ou
rly
NO
2 c
on
cen
trat
ion
(µg
/m3)
Traffic stations
Urban background stations
Suburban background stations
2010
2012
2011
2013
•Measures and policies are being implemented in the city and a positive trend of AQ levels
is observed
NO2 annual mean (Annual LV)
NO2 annual 1h 99.8th percentile (1-h LV)
Expected (modelled) NO2 ambient
concentration in the Madrid metropolitan area
as a result of Madrid’s Air Quality Plan
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•However issues in specific traffic-related hot-spots such as Fernandez Ladreda (FL)
square, remain
FL air quality
monitoring station
NO
2A
nnualm
ean (
2014)
Traffic stations in the Madrid City Council Air Quality Network
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
2. METHODOLOGY
•206 passive samplers (diffusion
tubes) were deployed in the area
•21 days (9 Feb. – 2 Mar. 2015)
•Unexpensive method useful to
understand spatial (not temporal)
gradients in urban areas with high
resolution
•Airbore pollutants (e.g. NO2) are
diffused into the tube and captured
by an absorbent (e.g.
triethanolamine –TEA-)
NO2 concentration
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
10
•Three tubes in the FL air quality
monitoring station to compute
relative error (10.1%) and correction
factor (0.91)
•The resultant specie is assess in a
spectrophotometer and related to know-
concentration patterns to derive concentration
results
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•A concentration map
was created by
interpolating individual
tube results through a
spline with barriers
algorithm (minimum
curvature method)
NO2
(µg/m3)
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Hourly NO2 concentration values
were taken from the FL air quality
monitoring station
•An average daily variation pattern
was computed
0
20
40
60
80
100
120
140
160
180
1
14
27
40
53
66
79
92
10
5
11
8
13
1
14
4
15
7
17
0
18
3
19
6
20
9
22
2
23
5
24
8
26
1
27
4
28
7
30
0
31
3
32
6
33
9
35
2
36
5
37
8
39
1
40
4
41
7
43
0
44
3
45
6
46
9
48
2
49
5
50
8
52
1
Ob
serv
ed N
O2
(µg
/m3)
Day
M T W T F S S M T W T F S S M T W T F S S
0
10
20
30
40
50
60
70
80
90
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Ob
serv
ed N
O2
(µg
/m3) –a
ve
rag
e-
Hour
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Spatial information from the passive tubes was combined with the temporal information
from the air quality monitoring station to produce 24 hourly concentration maps
•Results were interpolated to a 5 meter resolution grid to cross the results with those from
the pedestrian simulation
This assumes that temporal concentration variations recorded by the air quality monitoring station are representative all over the
domain (this hypothesis is deemed reasonable since concentration values in the area are directly related to road traffic that
follows a similar temporal pattern everywhere in the domain, but needs to be tested)
NO2
(µg/m3)
NO2
(µg/m3)
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Both traffic and pedestrian fluxes were simulated with a micro-scale modelling system
• The PTV VISSIM 6.00-19 microscale traffic flow model was selected to generate realistic
traffic data while the VISWALK module was used for pedestrian simulation
•15 representative scenarios (1 hour length) were simulated
Pedestrian simulation
Avera
ge tra
ffic
in
tensity
Week days Weekends
Avera
ge tra
ffic
in
tensity
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•An intensive field campaign was made to compile all the information needed to feed the
traffic model:
Detailed network (lanes):
19 links, 22 connectors Movements
EMT buses Traffic lights
location and
phases
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•The additional information provided for the pedestrian simulation includes:
- Definition pedestrian areas (sidewalks), conflict areas (crosswalks) and obstacles
- Routes and number of pedestrians in each of them
- Bus lines, stops and frequency. Pedestrians boarding and exiting the bus.
Sub-domain definition 300x300m Pedestrian areas Roads and crosswalks
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Every possible route along the square was defined as a Pedestrian Static Route
Decission using the collected data as input or output of the areas from the simulation.
•The social force approach generates the shortest path for the pedestrians to connect the
defined beginning and end of each route trough the pedestrian areas.
•Pedestrian speed distribution
in the range [2.6 – 5.8] km/h
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Pedestrian counts were made in all the 8 sampling areas defined to provide
representative inputs for all the 15 scenarios defined
•Data (counts) from each scenario were post
processed to assure consistency and flux
continuity
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Pedestrian counts included quantification of bus passengers getting on and off the
buses
•This observations were combined with the information regarding bus routes, stops and
frequencies provided by the Municipal Transport Company (EMT)
•Also metro stations were included in the simulation as input/output areas
PT Stops
PT Route
PT Frequency
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•The location of each individual
pedestrian was computed with 2
second resolution
•Pedestrian locations were
interpolated to a 5 meter resolution
grid and integrated throughout 1
hour time for each scenario
•From the result of the 15 representative scenarios an average day of pedestrian fluxes
was derived
0
100000
200000
300000
400000
500000
600000
700000
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Pe
de
str
ians (
pers
on·s
)
Hour
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•NO2 concentration values and pedestrians figures were multiplied for every grid cell (5x5
meters) within the pedestrian simulation sub-domain (300 x 300 m) to compute total
exposure every hour of the day
Exposure assessment
•Total exposure in the area can be computed by
aggregating individual grid cell results for every
hour or for the whole day (total exposure –TE–)
•In addition, total theoretical exposure –TTE –
was computed as the product of the total number
of pedestrians (persons·s) and the concentration
recorded in the air quality monitoring station
The TTE/TE ratio is propose as an index to assess the representativeness of the
air quality monitoring station to assess population exposure
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
3. RESULTS AND DISCUSSION
•Total exposure
Daily total
exposure
(person·s·µg/m3)
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Maximum exposure in the roundabout area, especially in the crosswalks of the main
streets
•High exposure figures also in bus stops and gathering areas
•Strong variation throughout the day: traffic emissions (and therefore concentration
values) and pedestrian temporal patterns rather similar maximum exposure during
the morning peak hour
0
10000000
20000000
30000000
40000000
50000000
60000000
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21 H22 H23 H24
To
tal e
xp
osu
re (
pers
on·s
·µg
/m3)
Hour
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•Although general spatial patterns are similar throughout the day, changes on traffic
fluxes, pedestrian routes and public transportation schedules also have an effect on
exposure distribution
Total exposure
(person·s·µg/m3)
4-5 AM
Total exposure
(person·s·µg/m3)
8-9 AM
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•This implies that it is difficult to provide exposure-representative information for a single
air quality monitoring station
•In this particular case, the TTE/TE ratio was below 1 (0.78 as an average), indicating that
recorded NO2 levels in FL monitoring station would underestimate population exposure
in that particular 300x300 m domain
0.70
0.80
0.90
1.00
1.10
1.20
1.30
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
TT
E/T
E in
de
x
Hour
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
•The TTE/TE ratio could be used to compute the concentration that the monitoring station
should measure for a value of 1
•If the spatial concentration pattern is known this allows the definition of a exposure-
representative location for the air quality monitoring station (area in blue in the examples)
4-5 AM 8-9 AM
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
4. CONCLUSIONS
•The methodology presented constitutes a preliminary case study for the assessment of
the representativeness of air quality monitoring stations from the population exposure
point of view
•The methods used to simulate pedestrians and to obtain concentration maps can only be
applied at microscale and may be used to support studies of monitoring station micro-
siting
•However, if pedestrians location and concentration values are provided by other means,
the methodology based on the TTE/TE index assessment may be applied to larger areas
and contribute to the definition monitoring points that may be representative of
population exposure and thus, assessment of the compliance of health-related ambient
air quality legal standards
•Meaningful criteria to define the extend of these areas, a critical factor for this kind of
analysis, should also be defined in the future
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
• The TECNAIRE-CM research project was funded by the Madrid Greater Region (S2013/MAE-2972)
• Acknowledgments:
• The Madrid City Council provided the traffic
cameras and partially funded this study
• The micro-scale traffic modelling was possible
thanks to the collaboration of the national traffic
authority (DGT), the Municipal Transport Company
(EMT, S.A.), Madrid Calle 30, S.A.
• VISSIM and VISWALK ware licensed by PTV Group
www.tecnaire-cm.org
2nd Healthy Polis Workshop (during Kunshan Forum)
30 May 2015; Kunshan (China)
Thank you for your attention!rborge@etsii.upm.es
Session A4: Healthy-Polis Workshop on Climate Change and Urban Health – II