1 Swedish experiences of applying the Quality Objectives for NO2 and PM10 modelling introduction model evaluation for Swedish - street/road stations -

1

Swedish experiences of applying the Quality Objectives for NO2 and PM10 modelling

• introduction

• model evaluation for Swedish - street/road stations- urban background stations

• discussion and conclusions

by L. Gidhagen, G. Omstedt and S. Andersson

Model quality objectives (uncertainty) as described in the AQ directive

2

Modelling

uncertainty

NO2 PM10

Hourly 50 % -Daily average 50 % -Annual average 30 % 50%

p

pp

O

MORPEMRPE

max)max(

LV

MORDEMRDE LVLV

max)max(

The uncertainty of modelling estimation is defined as the maximum deviation between the measured and calculated concentration levels for 90 % of individual monitoring points, without taking into account the timing of the events. The average annual modelling uncertaintyfor NO2 is defined as ±30% and for percentiles ±50%

Fairmode http://fairmode.ew.eea.europa.eu/

Op and Mp are the observed and modelled concentrations at the percentile (p)OLV and MLV are the closed observed and measured concentration to the limit valueconcentration (LV)

Indicators for model quality

3

Gidhagen, L., Johansson, H. and Omstedt, G., 2009: SIMAIR - Evaluation tool for meeting the EU directive on air pollution limits. Atmospheric Environment, 43, 1029-1036, doi:10.1016/j.atmosenv.2008.01.056.Andersson, S. och Omstedt, G., 2009: Validering av SIMAIR mot mätningar av PM10, NO2 och bensen. Utvärdering för svenska tätorter och trafikmiljöer avseende år 2004 och 2005. SMHI Meteorologi, Nr. 137, 125 pp. (In Swedish).

SIMAIR

Model validation in Sweden

~ 30 road/street stations~ 20 urban background stations

Validation performed using both RPE and RDE Swedish EPA recommends the use of RPE for quality check of hourly (NO2) and

daily (PM10, NO2) values.

Swedish EPA supports FAIRMODE recommendation to use, for annual mean values, the RDE calculation when observed value is low compared to limit value. For high observed annual mean values, RPE can be used.

We are left with some confusion, e.g.:- Not clear when to use RDE or RPE- Not clear how the 90% of the station comparisons are selected: - How big can the area be (entire Sweden)? - Should the comparison include only one specific year or can we include comparisons for the same station but for two different years?- Why is max(RPE) or max(RDE) selected, would not median be more useful? Or both?

4

Validation results from 2004 and 2005

5

Model simulations of PM10 for street level show acceptable quality, even for 98-percentiles

Differences MRPE and MRDE?

Validation PM10 from 2004 and 2005

City/ street/yearmeasured modelled RPE RDE measured modelled RPE RDE

Landskrona/Eriksgatan, 2004 23,8 16,6 0,30 0,18 41,1 26,2 0,36 0,35Landskrona/Eriksgatan, 2005 24,2 19 0,21 0,13 43 28,4 0,34 0,37Helsingborg/Malmöleden, 2005 22,4 18 0,20 0,11 38 27,5 0,28 0,23Kristianstad/V,Boulevard, 2004 26,8 17,1 0,36 0,24 54,9 26,7 0,51 0,50Kristianstad/V,Boulevard, 2005 36,5 22,8 0,38 0,34 81 38,5 0,52 0,45Nässjö/Brogatan, 2004 25,8 21,9 0,15 0,10 64,4 38,9 0,40 0,30Norrköping/Kungsgatan, 2004 17,8 25,7 0,44 0,20 37,7 46,4 0,23 0,14Norrköping/Ö,Promenad, 2004 22,9 24,7 0,08 0,05 52,8 43,2 0,18 0,22Norrköping/Ö,Promenad, 2005 27,7 26,9 0,03 0,02 63 58,1 0,08 0,04Norrköping/Söderleden, 2005 19,3 21,2 0,10 0,05 44,9 40,7 0,09 0,09Göteborg/Gårda, 2005 29,6 32 0,08 0,06 55 61,5 0,12 0,07Karlstad/Hamngatan, 2005 22,8 21,6 0,05 0,03 45,8 47,9 0,05 0,01Västerås/Kopparbergsv, 2005 25,3 19,4 0,23 0,15 51,3 33,7 0,34 0,39Västerås/Stora gatan, 2005 27,5 19,1 0,31 0,21 59,9 34,6 0,42 0,41Västerås/Vasagatan, 2005 23,6 14,5 0,39 0,23 43,5 26,7 0,39 0,44Sollentuna/Turebergsl, 2004 18,5 14,5 0,22 0,10 32,7 25,5 0,22 0,31Sollentuna/Turebergsl, 2005 20,2 14 0,31 0,16 37,6 25,7 0,32 0,29

max RPE/ max RDE 0,39 0,24 0,51 0,45median RPE/ median RDE 0,22 0,13 0,32 0,30

Annual mean 90-percentile daily mean

6

Validation PM10 from 2004 and 2005 using RPE and RDE

With RDE the quality is OK, with RPE it is almost OK (although MRPE for daily PM10 is still not defined)

RPE and RDE for PM10

7

There are often a few “poor” stations where measurement errors or bad siting (low representativeness) can be suspected

Could med(RPE) and med(RDE) be a better alternative?

Same station

8

Examples for PM10: How will RPE differ from RDE?

Annual mean:RPE = 1%RDE = 6%

Annual mean:RPE = 30%RDE = 18%

Annual mean:RPE = 44%RDE = 20%

Hornsgatan/Stockholm year 2000 Kungsgatan/Norrköping year 2004

Eriksgatan/Landskrona year 2004

Graphs illustrate RPE and RDE for daily 90-percentiles

9

0 20 40 60 80 100 120

N O 2 [µg m -3] uppm ätt

0

20

40

60

80

100

120

NO

2 [µ

g m

- 3] S

IMA

IR

0 20 40 60 80 100 120

N O 2 [µg m -3] uppm ätt

0

20

40

60

80

100

120

NO

2 [µ

g m

- 3] S

IMA

IR

Beräkningar för gaturum

Årsm edelvärde

98% -il dygnsm edelvärde

98% -il tim m edelvärde

Modell evaluation of NO2 at street/road level

Model simulations of NO2 for street level show acceptable quality, even for hourly values

10

Validation NO2 from 2004 and 2005 using RPE and RDE

City/ street/yearmeasured modelled RPE RDE measured modelled RPE RDE measured modelled RPE RDE

Helsingborg/Drottningg,, 2004 27,9 36,3 0,30 0,21 58,8 80,6 0,38 0,37 80,5 97,6 0,20 0,21Kristianstad/V,Boulevard,, 2004 18,6 22,3 0,20 0,09 35,8 40 0,05 0,12 - - - -Kristianstad/V,Boulevard,, 2005 17,6 20,1 0,14 0,06 46,2 37,8 0,10 0,18 - - - -Göteborg/Gårda, 2004 46,7 34,1 0,27 0,32 84,8 75,6 0,19 0,11 109,4 109,4 0,06 0,00Göteborg/Gårda, 2005 46,6 30,1 0,35 0,41 92 69,9 0,22 0,24 109,4 94,5 0,17 0,14Göteborg/Haga, 2004 42 32,8 0,22 0,23 81,5 66,7 0,16 0,18 104,1 85,2 0,20 0,18Göteborg/Haga, 2005 40,8 29,2 0,28 0,29 86 52,8 0,27 0,39 107,6 66,6 0,34 0,38Göteborg, Mölndal, 2004 28,1 30,3 0,08 0,06 66,2 70,4 0,13 0,06 88,5 101,7 0,14 0,15Uppsala, Kungsgatan 27,2 28,5 0,05 0,03 51 57,1 0,12 69,6 81,7 0,17Uppsala, Kungsgatan2 37,4 37,3 0,00 0,00 64,6 66,9 0,04 88,9 90,4 0,02Malmö, Amiralsgatan 39,4 41,9 0,06 0,06 71,8 72,4 0,01 98 84,1 0,14

RPE max/ RDE max 0,30 0,32 0,38 0,37 0,34 0,38RPE median/ RDE median 0,20 0,09 0,18 0,12 0,19 0,15

Annual mean 98-percentile daily mean 98-percentile hourly mean

Daily and hourly percentiles OK, annual means almost OK

11

Improvements in the model can be demonstrated byimproved RPE

The current quality objectivesare still not fully compliant in urban background annual mean values

BUM new:MRPE annual mean = 0.40MRPE 98-percentile daily mean = 0.48MRPE 90-percentile hourly mean = 0.49

Modell evaluation of NO2 in urban background using RPE

Example: SIMAIR for PM10

12

PM10 annual mean 90-percentil

MRPE 0.39 0.51

MRDE 0.24 0.45

Can we use MRPE and MRDE as uncertainties on the simulated levels?

Example: calculated yearly mean PM10 concentration is 25 µg/m3 and calculated 90-percentil is 45 µg/m3 then the uncertainties are:

MRPE:Yearly mean: 25 +/- 9.8 µg/m3 i.e. between 15.3 - 34.8 µg/m3

90-percentile (daily mean): 45 +/- 23.0 µg/m3 i.e. between 22.1 - 68.0 µg/m3

MRDE:Yearly mean: 25 +/- 9.6 µg/m3 i.e. between 15.4 - 34.6 µg/m3

90-percentile (daily mean): 45 +/- 22.5 µg/m3 i.e. between 22.5 - 67.5 µg/m3

Likely those intervals are too large for a general public?

Can MRPE and MRDE be used as uncertainties?

Are the quality objectives (QA) obtainable and relevant?

Answer: Yes, but…

Some confusion on how to calculate MRPE and MRDE which must be eliminated.

The indicator should reflect model uncertainty as much as possible. In our opinion RPE is a better indicator than RDE, especially for Swedish conditions with air quality levels often well below the limited values.

It is unclear if and how these indicators can be used, except for showing compliance of Directive’s “Quality objectives for models”. For describing model uncertainties in a broader sense other and more refined indicators are needed (Delta tool…).

13

Conclusions

Thank you for your attention!

14

Mal

mö

Land

skro

naH

elsi

ngbo

rg 1

He

lsin

gbo

rg 2

Jönk

öpi

ngN

ässj

öN

orrk

öpi

ngG

öte

borg

1G

öte

borg

2M

arie

stad

Kar

lsta

dK

arls

kog

aV

äste

rås

1V

äste

rås

2U

pps

ala

Sto

ckh

olm

1S

tock

holm

2S

tock

ho

lm 3

So

llent

una

Gä

vle

Su

ndsv

all

Öst

ersu

ndU

me

åL

ycks

ele

Ske

lleft

eå

0

10

20

30

40

PM

10 å

rsm

ede

lvär

de [µ

g/m

3]

Lokalt b idrag

Bakgrundshalter

M KN

M iljöm ål 2010

Skåne Sydöst Väst M älardalen S tockholm M itt N orr

PM10 in Swedish cities (2004): Importance of local contribution

15

PM10 in Swedish cities (2004): Local – Urban - Regional

0

10

20

30

40

50P

M1

0 [µ

g m

-3]

års

med

el

Ma

lmö

, D

ala

pla

n

Lan

dsk

ron

a, E

rik

sg.

Kri

stia

ns

tad

, V

. Bo

ule

vard

en

Gö

teb

org

, E

6 vi

d G

ård

a

No

rrk

öp

ing

, Ö.

Pro

men

ade

n

Sto

ckh

olm

, H

orn

sg.

Sto

ckh

olm

, N

orr

lan

ds

g.

Sto

ckh

olm

, S

veav

.

Sto

ckh

olm

, Es

sin

ge

led

en

Up

psa

la, K

un

gs

g.

Um

eå, V

. E

sp

lan

and

en

PM10

Reg

Urb

Gatu

Reg

Urb + Gatu

16

0

10

20

30

40

50

PM

2.5

[µg

m-3

]å

rsm

ed

el

Ma

lmö

, D

ala

pla

n

Kri

sti

an

sta

d,

V.

Bo

ule

va

rde

n

Sto

ck

ho

lm,

Ho

rns

g.

Sto

ck

ho

lm,

No

rrla

nd

sg

.

Sto

ck

ho

lm,

Sv

ea

v.

Sto

ck

ho

lm,

Es

sin

ge

led

en

Up

ps

ala

, K

un

gs

g.

PM2.5

Reg

Urb

Gatu

Reg

Urb + Gatu

PM2.5 in Swedish cities: Local – Urban - Regional