URBAN EMISSIONS AND PROJECTIONS Rafael Borge 1 , Julio Lumbreras 1 , - - PowerPoint PPT Presentation

▶
urban emissions and projections
SMART_READER_LITE
LIVE PREVIEW

URBAN EMISSIONS AND PROJECTIONS Rafael Borge 1 , Julio Lumbreras 1 , - - PowerPoint PPT Presentation

Dec 15, 2023 273 likes •499 views

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. Parallel session FAIRMODE 2nd June 2010 2nd June 2010 URBAN EMISSIONS AND PROJECTIONS Rafael Borge 1 , Julio Lumbreras 1 ,

slide-1
SLIDE 1

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L.

2nd June 2010 Parallel session FAIRMODE

URBAN EMISSIONS AND PROJECTIONS

2nd June 2010

Rafael Borge1, Julio Lumbreras1, David de la Paz1, M. Encarnación Rodríguez1 Panagiota Dilara2, and Leonor Tarrason3

1 Laboratory of Environmental Modelling. Technical University of Madrid (UPM) 2 European Commission, Joint Research Centre, Ispra, Italy 3 Norwegian Institute for Air Research. Kjeller, Norway

rborge@etsii.upm.es ; jlumbreras@etsii.upm.es

slide-2
SLIDE 2

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L.

1. Introduction 2. Methodology OUTLINE 3. Results 4. Conclusions

slide-3
SLIDE 3

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 3

  • 1. Introduction

FAIRMODE flowchart as agreed on 2nd plenary meeting (Nov. 2009)

SG(2) + SG(1) Combination of monitoring and modelling (data SG (3) Emission inventories and

Benchmarking SG (4)

Protocols and

modelling (data assimilation) SG(5) Contribution of natural sources and Source apportionment inventories and scenarios

Protocols and Tools for benchmarking of AQ models Urban Agglomeration

slide-4
SLIDE 4

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 4

SG(3) on urban emissions and projections

  • Background document on the emission needs at local scale
  • Needs for guidance on emission compilation at urban level

– Consistency with national inventories – Top down vs bottom up approches – Use of GIS tools

  • Urban emission compilation is a key issue at European level
  • Both guidance and relevant exchange fora are needed
  • Next step:

‒ Proposal for a framework for the development of emission inventories at local scale

  • Links to TFEIP/EIONET, NIAM, GEIA, JRC-EDGAR
slide-5
SLIDE 5

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 5

  • Uncertainties for Air Quality Models (AQMs)
  • Meteorology

Air Quality Modelling in the AQD

assessment of ambient air quality planning and mitigation strategies assessment of the contribution of natural sources, road dust and sea salt short-term forecast for threshold exceedances

  • Meteorology
  • Modelling system
  • Boundary and Initial conditions
  • Emission input
  • Uncertainties from emission inputs → emission inventories:
  • Emission data accuracy
  • Temporal disaggregation
  • Spatial resolution and emission allocation
  • Chemical speciation and mass distribution

Consistent emission estimates across the scales, inventory

  • harmonization. Criteria for

local scale EI development

slide-6
SLIDE 6

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 6

1) Emission data accuracy (Cho et al., 2009) 2) Temporal disaggregation (Wang et al., 2010, Kühlwein et al., 2002) 3) Spatial resolution and emission allocation (Mensink et al., 2008, Cheng et al., 2008, Pisoni et al., 2010)

slide-7
SLIDE 7

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 7

  • 2. Methodology
  • To analyse two approaches for different scale emission inventory

compilation for an inland city and surroundings : − National calculation using country statistics and some regional data with spatial disaggregation afterwards − Regional calculation using regional data

  • Compare AQM results (whole year, 1-h resolution) with monitoring data

Relate these differences with emission compilation methods for the dominant source in the grid cell

  • Select and analyse a

number

  • f

representative stations where the alternative inventories produce important discrepancies in AQM results Understand reasons for discrepancies, get an idea about emission accuracy, and identify options for multi-scale emission inventory harmonization

slide-8
SLIDE 8

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 8

  • AQM domain including AQ monitoring stations
  • Same BC and individual profiles for temporal and chemical speciation.

Differences in model performance due to:

  • Emission data accuracy (total figures and sectoral figures)
  • Emission allocation (source apportionment at grid cell level)
slide-9
SLIDE 9

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 9

  • Emission inventory aggregated comparison (INV1 – INV2)

SNAP Difference SO2 NOX NMVOC CH4 CO NH3 PM2,5 PM10 1 Absolute

  • 2622
  • 212

23

  • 438
  • 179
  • 33
  • 81

Relative

  • 100%
  • 42%

750%

  • 41%
  • 42%
  • 58%
  • 77%

2 Absolute

  • 1729

16269

  • 1001
  • 1154
  • 15749
  • 811
  • 882

Relative

  • 37%

302%

  • 59%
  • 68%
  • 78%
  • 83%
  • 83%

3 Absolute

  • 1565
  • 9824
  • 1265

209

  • 3072
  • 221

74 Relative

  • 27%
  • 46%
  • 57%

34%

  • 41%
  • 52%

13% 4 Absolute 8 11

  • 1653

566 12 24 Relative 6% 6%

  • 44%
  • 6%
  • 6%

6% 5 Absolute

  • 2156

27 1 Relative

  • 58%

0%

  • 112%

112% Absolute

  • 2636
  • 78
  • 3. Results

6 Absolute

  • 2636
  • 78

Relative

  • 4%
  • 81%
  • 7

Absolute 2605 26601 2573 2067 8124 77 397

  • 187

Relative 963% 53% 17% 267% 9% 11% 10%

  • 4%

8 Absolute 416 4576 287 7 3533

  • 468
  • 468

Relative 67% 53% 32% 22% 75%

  • 18%
  • 71%
  • 71%

9 Absolute 501 310

  • 3142
  • 45044

187 765 5 6 Relative 1.8E05% 119%

  • 60%
  • 51%

234% 78% 80% 78% 10 Absolute 6

  • 110

190

  • 647
  • 1209
  • 3782

565 4317 Relative 41%

  • 40%

13%

  • 7%
  • 87%
  • 69%

1139% 1285% 11 Absolute 2

  • 665

8518

  • 1134

344

  • 299

Relative 117%

  • 97%

53%

  • 96%

117%

  • 98%
  • TOT

Absolute

  • 2376

36956

  • 261
  • 46105
  • 7455
  • 3317
  • 553

2803 Relative

  • 17%

42% 0%

  • 42%
  • 5%
  • 44%
  • 9%

36%

slide-10
SLIDE 10

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 10

  • Emission allocation (Gridded total NOx emissions according to the

inventories considered)

NOX emissions (t year-1)

INV1 INV2

  • Largest discrepancies related to road transport and domestic/commcial/institut. heating
  • Some differences in industry-related combustion processes and off-road mobile sources
  • Different spatial allocation patterns
slide-11
SLIDE 11

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 11

  • Emission allocation: a) Source apportionment at grid cell level: SNAP 02

SNAP 02 contribution to NOX emissions at grid cell level (%)

CAM INV1 INV2

  • Differences:
  • statistical basis used for activity rates estimation
  • population as spatial surrogate (uniform emission distribution across a given

municipal urban area vs. CORINE land cover population density)

slide-12
SLIDE 12

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 12

  • Emission allocation: b) Source apportionment at grid cell level: SNAP 07

INV1 INV2

SNAP 07 contribution to NOX emissions at grid cell level (%)

  • Differences:
  • discrepancies regarding driving patterns and road classification
  • differences in mileage estimation per vehicle (daily average intensities and

road length vs. prescribed total mileage values depending on vehicle type)

  • road maps considered
slide-13
SLIDE 13

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 13

  • AQM results:

a) NO2 annual mean

  • AQM results:

b) NO2 99.8th 1-h percentile

slide-14
SLIDE 14

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 14

  • AQM results: c) NO2 Mean Bias (ppb) at station level
  • 30
  • 25
  • 20
  • 15
  • 10
  • 5

5 10 15 20 25 30 Mean bias (ppb) CAM NAT INV1 INV2

a

Monitoring stations: A – traffic

  • 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 23 24 26 28 33 34 35 37 39 49 52 Station ID

  • 30
  • 25
  • 20
  • 15
  • 10
  • 5

5 10 15 20 25 30 22 25 29 30 40 41 43 44 45 46 50 53 55 Station ID Mean bias (ppb) CAM NAT INV1 INV2

b

  • 30
  • 25
  • 20
  • 15
  • 10
  • 5

5 10 15 20 25 30 27 31 36 54 Station ID Mean bias (ppb) CAM NAT INV1 INV2

c

A – traffic B – urban background C – industrial

slide-15
SLIDE 15

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L.

E-21 NAT % % % % % % % % % % % E-21 CAM 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% SNAP-1 SNAP-2 SNAP-3 SNAP-4 SNAP-5 SNAP-6 SNAP-7 SNAP-8 SNAP-9 SNAP-10

Station A – INV1 Station A – INV2

15

  • Station A (traffic)

% CO NOX VOC NH3 SO2 PM10 PM2_5 0% CO NOX VOC NH3 SO2 PM10 PM2_5 SNAP-11

  • INV1 more than double NOx emissions in the corresponding grid cell
  • SNAP 07 (road traffic) is the predominant source (consistent with station label)
  • INV2 considers a significant contribution from other sources
  • NO2 underestimated with INV2 and overestimated with INV1 similarly
  • Absolute mean errors (ME) and the correlation coefficient are similar

2418 t/y 1093 t/y

SNAP 07 emissions largely overestimated in INV1 (excessive contribution of heavy duty vehicles in highway driving patter), although activity ratios are more specific. Inaccurate secondary EF

slide-16
SLIDE 16

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L.

E-43 NAT 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% SNAP-1 SNAP-2 SNAP-3 SNAP-4 SNAP-5 SNAP-6 SNAP-7 SNAP-8 SNAP-9 SNAP-10 E-43 CAM 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% SNAP-1 SNAP-2 SNAP-3 SNAP-4 SNAP-5 SNAP-6 SNAP-7 SNAP-8 SNAP-9 SNAP-10

Station B – INV1 Station B – INV2

16

  • Station B (urban background)

0% CO NOX VOC NH3 SO2 PM10 PM2_5 SNAP-10 SNAP-11 0% CO NOX VOC NH3 SO2 PM10 PM2_5 SNAP-10 SNAP-11

  • INV1 more than double NOx emissions in the corresponding grid cell
  • Source apportionment resulting in this grid cell is more balanced for INV2
  • Non-LPS allocated using covers (INV2) and area-to-point algorithm (INV1)
  • NO2 slightly underestimated with INV2 and overestimated with INV1
  • Better statistics for INV2

797 t/y 352 t/y

SNAP 07 emissions overestimated in INV1 (dominating source in urban background) Not enough information to support and area-to-point allocation strategy (spatial surrogates provide a more reasonable picture)

slide-17
SLIDE 17

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L.

E-27 NAT 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% SNAP-1 SNAP-2 SNAP-3 SNAP-4 SNAP-5 SNAP-6 SNAP-7 SNAP-8 SNAP-9 SNAP-10 E-27 CAM 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% SNAP-1 SNAP-2 SNAP-3 SNAP-4 SNAP-5 SNAP-6 SNAP-7 SNAP-8 SNAP-9 SNAP-10

Station C – INV1 Station C – INV2

17

  • Station C (industrial)

0% 10% CO NOX VOC NH3 SO2 PM10 PM2_5 SNAP-10 SNAP-11 0% 10% CO NOX VOC NH3 SO2 PM10 PM2_5 SNAP-10 SNAP-11

  • INV1 more than triple NOx emissions in the corresponding grid cell
  • Road traffic emissions are in relatively good agreement
  • INV2 considers larger industrial emissions
  • NO2 overestimated with INV1 (MB = 14.8 ppb, ME = 20.3 ppb)
  • NO2 less underestimated with INV2 (MB = -3.6 ppb, ME = 12.6 ppb)

3616 t/y 672 t/y

Apparently, an excessive emission allocation from industry in INV1 in general terms

slide-18
SLIDE 18

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 18

  • Station C (industrial)

E-27 10 20 30 40 50 60 70 80 90 100 ppbV (NO2) NAT CAM Observed

Station C INV2 INV1

  • r coefficients similar
  • Important seasonal differences
  • Better agreement with observations during most of the year for INV2, except for

particular periods concentrated in August-November No significant differences in temporal patterns in those periods ⇒ misrepresentations of the chemical split of NOx and VOCs for particular industrial activities (most likely) high NO2 levels due to non-local contributions

10

1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 162 169 176 183 190 197 204 211 218 225 232 239 246 253 260 267 274 281 288 295 302 309 316 323 330 337 344 351 358 365

Day

slide-19
SLIDE 19

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 19

  • 4. Conclusions
  • There

is an increasing demand for high-resolution, fine-scale emission inventories for air quality modelling activities

  • It was agreed within FAIRMODE that this need is the most relevant

emission-related issue for the application of the AQD

  • A reliable air quality model may be useful to discriminate the
  • A reliable air quality model may be useful to discriminate the

uncertainty of emission inventories

  • AQ monitoring sites should be carefully selected to guarantee the

correctness and representativeness

  • f

the

  • bservational

data considering the spatial and temporal resolution of the model

  • It is essential that the methodology used at different scales is known

and transparent for all the inventories involved

slide-20
SLIDE 20

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 20

  • Emissions from the road traffic are the key issue in an urban-scale

inventory. Traffic flow measurements and accurate fleet characterization are crucial to get a reasonable estimate of traffic

  • emissions. However, energy balances, computation methods and

underlying hypotheses are, at least, equally important

  • A previous analysis of main statistics used to derived activity rates at

different scales is needed.

  • The bottom-up approach is preferred when there is information

enough to support a very detailed emission estimation, but a top- down approach in combination with an updated high-resolution land use/population cover may provide a more accurate picture of general emission distribution pattern.

  • If basic reference statistics are properly harmonised, both approaches

should lead to quite similar results, being the differences due to the use of more specific information available only at finer scales

slide-21
SLIDE 21

Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L.

THANK YOU FOR YOUR ATTENTION!

Laboratory of Environmental Modelling. Technical University of Madrid (UPM)

THANK YOU FOR YOUR ATTENTION!

rborge@etsii.upm.es ; jlumbreras@etsii.upm.es