Local Air Pollution Modelling, Local Air Pollution Modelling, - - PowerPoint PPT Presentation

Local Air Pollution Modelling, Local Air Pollution Modelling, AIM/Air AIM/Air

Takeshi Fujiwara Takeshi Fujiwara Kyoto University Kyoto University 2005.3.10 2005.3.10

Topics Topics

• Calculation of impact of traffic sector on air

Calculation of impact of traffic sector on air quality quality

– – By using traffic volume data measured in traffic By using traffic volume data measured in traffic census, emission from traffic sector was census, emission from traffic sector was disaggregated into line source of road networks. disaggregated into line source of road networks.

• Improvement of AIM/Air: inclusion of land use

Improvement of AIM/Air: inclusion of land use information information

– – Based on land use information, total emission of Based on land use information, total emission of each sector was disaggregated to a set of area each sector was disaggregated to a set of area which is used by the sector. which is used by the sector.

[The 1 [The 1st

st topic]

topic] Calculation of impact of Calculation of impact of Traffic Sector on Air Quality Traffic Sector on Air Quality

Outline Outline

• The method of estimating traffic volume

The method of estimating traffic volume

• n road networks by using limited number
• n road networks by using limited number
• f measurement data are applied.
• f measurement data are applied.
• Diffusion of emission gas and particles

Diffusion of emission gas and particles ( (NOx NOx、

、SOx

SOx、

、PM) from automobiles was

PM) from automobiles was calculated. calculated.

• Line emission source was regarded as a

Line emission source was regarded as a series of point sources. series of point sources.

Back ground Back ground

・ ・Research on emission from automobiles by

Research on emission from automobiles by Dr.

• Dr. Hao

Hao et. al., which had been published

• et. al., which had been published

as a book as a book 「

「城市机动车排放汚染控制 城市机动车排放汚染控制 」 」 ・ ・Hao

Hao’ ’s s group have already calculated group have already calculated emission diffusion from automobiles in emission diffusion from automobiles in Beijing. Beijing.

・ ・Emission condition of automobiles, such as

Emission condition of automobiles, such as emission coefficient for each type of car, emission coefficient for each type of car, was assumed based on MOBLIE 5 model. was assumed based on MOBLIE 5 model.

Flow Diagram of Calculating Concentration Related with Transport Sector

Road and Traffic Data GIS data Meteorology Data Estimation of Traffic Volume Estimated Emission Emission Ratio Calculation of Stability Diffusion Calculation

Section of Traffic Estimation Section of Diffusion Calculation

Traffic Data Required Traffic Data Required

・ ・Road data and traffic volume data measured

Road data and traffic volume data measured

• Road length, width, connection.

Road length, width, connection.

• Measured flow rate of automobile.

Measured flow rate of automobile.

・ ・Traffic volume

Traffic volume

• How many automobiles pass over per hour.

How many automobiles pass over per hour.

・ ・Emission ratio of automobile

Emission ratio of automobile

• NOx

NOx emission ratio of each type of emission ratio of each type of automobiles is used. automobiles is used.

Traffic volume estimation Traffic volume estimation

• Using road network, land use, and

Using road network, land use, and actual traffic volume data, traffic actual traffic volume data, traffic volume of each road was estimated. volume of each road was estimated.

• OD traffic volume estimation method

OD traffic volume estimation method based on link flow was applied. based on link flow was applied.

• To accelerate computation speed

To accelerate computation speed cluster computer system was used. cluster computer system was used.

Traffic Volume Estimation Traffic Volume Estimation

• We adopted a method to approximate unknown traffic

We adopted a method to approximate unknown traffic volume with the small number of traffic volume data. volume with the small number of traffic volume data.

• Traffic volume is estimated based on the following

Traffic volume is estimated based on the following equation, it is so called gravity model. equation, it is so called gravity model. OD traffic volume OD traffic volume

・ ・This model requires information of trip generation (A

This model requires information of trip generation (Ai

i) at

) at

• ne node and trip attraction (
• ne node and trip attraction (B

Bj

j) at the other node.

) at the other node.

・ ・Land use information is used to identify the

Land use information is used to identify the characteristics of the node. characteristics of the node.

γ ij j i ij

t B αA X =

Utilization of Land Use Data Utilization of Land Use Data

・ ・A circle, which radius is 300m, shows a

A circle, which radius is 300m, shows a node of road networks. node of road networks.

・ ・A sector which has the largest area in the

A sector which has the largest area in the circle is identified. circle is identified.

・ ・Initial value of trip generation (A

Initial value of trip generation (Ai

i) and trip

) and trip attraction ( attraction (B Bj

j) between two sectors is set.

) between two sectors is set.

・ ・Traffic volume data measured are used to

Traffic volume data measured are used to predict unmeasured traffic volume. predict unmeasured traffic volume.

Procedure of traffic volume Procedure of traffic volume estimation estimation -

• 1

1-

• 1) Traveling time at the initial condition that

1) Traveling time at the initial condition that each link traffic volume is zero is estimated. each link traffic volume is zero is estimated. 2)A path with minimum traveling time 2)A path with minimum traveling time between OD pairs is searched. between OD pairs is searched. 3)Using DIJKSTRA algorithm, each OD traffic 3)Using DIJKSTRA algorithm, each OD traffic volume is distributed to the path with volume is distributed to the path with minimum traveling time. minimum traveling time.

Procedure of traffic volume Procedure of traffic volume estimation estimation -

• 2

2-

• 4)Update the minimum traveling time of each

4)Update the minimum traveling time of each link path. link path. 5)Update the link traffic volume. 5)Update the link traffic volume. 6)Go back to step 2), until each link traffic 6)Go back to step 2), until each link traffic volume is converged. volume is converged.

( )

( )

5

v/C 2.62 1.0 T T × + ⋅ =

a 1) (m a (m) a

v m 1 v m 1 1 v + ⋅ ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ − =

−

Validation of Traffic Volume Validation of Traffic Volume Estimation Estimation

y = 1.0525x 1000 2000 3000 4000 5000 6000 7000 8000 2000 4000 6000 8000 実測交通量 (台/時) 推定交通量 (台/時) 相関係数0.77 Traffic volume measured (cars/hour) Traffic volume estimated (cars/hour)

Relation coefficient

Diffusion Calculation Diffusion Calculation

• By using

By using ArcGIS ArcGIS, road is divided into , road is divided into fractions by 100m x 100m mesh in fractions by 100m x 100m mesh in

• rder to regard emission source on
• rder to regard emission source on

the road as a series of point sources. the road as a series of point sources.

• Diffusion is computed by AIM

Diffusion is computed by AIM-

• AIR

AIR with cluster computers. with cluster computers.

Diffusion equation Diffusion equation

⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ ⎭ ⎬ ⎫ ⎩ ⎨ ⎧ + − + ⎭ ⎬ ⎫ ⎩ ⎨ ⎧ − − ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ − ⋅ =

2 2 2 2 2 2

2 ) ( exp 2 ) ( exp 2 exp 2 ) , , (

z z y z y p

He z He z y u Q z y x C σ σ σ σ πσ ・

Plume diffusion equation

z

：Height of receptor (m)

p

Q ：Emission at point source

(Particles：kg/s，Gas：m3

N/s)

u ：Wind velocity (m/s)

He ：Effective stack height (m)

) (x

y

σ

： Diffusion parameter of y axis at distance x (-)

) (x

z

σ

： Diffusion parameter of y axis at distance x (-) ： Concentration at receptor (x,y,z)

) , , ( z y x C

Demonstration Demonstration

• Area: the center of Beijing city

Area: the center of Beijing city

• Period: Jan 1

Period: Jan 1st

st, 2000 ~ Jan 14

, 2000 ~ Jan 14th

th, 2000

, 2000

• Time step: 1 hour

Time step: 1 hour

• Emission source: automobiles

Emission source: automobiles

• Traffic pattern: hourly change in a day

Traffic pattern: hourly change in a day

• Meteorology data: ECMWF

Meteorology data: ECMWF

• Model: Plume or Puff model for a line of point

Model: Plume or Puff model for a line of point sources (each point source covers emission from sources (each point source covers emission from automobiles on road within 100m x 100m area) automobiles on road within 100m x 100m area)

[The 2 [The 2nd

nd topic]

topic] Improvement of AIM/Air: inclusion Improvement of AIM/Air: inclusion

• f land use information
• f land use information

Purpose of this study Purpose of this study

• Improvement of accuracy in emission

Improvement of accuracy in emission database database

• Spatial distribution of emission by using

Spatial distribution of emission by using land information land information

• Evaluating affection of each sector on air

Evaluating affection of each sector on air quality quality

• System integration of emission estimation

System integration of emission estimation and diffusion calculation in AIM/Air and diffusion calculation in AIM/Air

Diffusion calculation procedure Diffusion calculation procedure

• f AIM/Air
• f AIM/Air

Disaggregation to county level Disaggregation to 30” x 30” mesh Generation of land use data digitized with ArcInfo Rearrangement of emission based on land use

Energy end use database Population distribution Land use map Vector map Satellite image

Diffusion calculation of air pollutants By using AIM/Air air quality model

Meteorology data （ECMWF）

Emission estimation by end use model Confirmation of location of LPS

Large point source Area source

Estimation of emission Estimation of emission

• Large point source

Large point source：

：Power, cement, iron & steel, nonferrous

Power, cement, iron & steel, nonferrous material sectors and other sectors with large emission. material sectors and other sectors with large emission.

• Area source

Area source：

：Transport, commercial and public sectors and

Transport, commercial and public sectors and

• ther sectors with small emission.
• ther sectors with small emission.

発電 セメント 鉄鋼 非鉄金属 その他 運輸 商業 家庭（都市部） 家庭（農村部） 部門

北京市 河北省 福建省 四川省 内蒙古自治区 上海市 山西省 江西省 貴州省 広西チワン族自治区 天津市 遼寧省 山東省 雲南省 寧夏回族自治区 重慶市 吉林省 河南省 陝西省 チベット自治区 黒竜江省 湖北省 甘粛省 新疆ウイグル自治区 江蘇省 湖南省 青海省 浙江省 広東省 安徽省 海南省 エンドユースモデル地域区分（31地域）

4.24E+08 0.0

Disaggregation Disaggregation of Area Source Emission

• f Area Source Emission

to County Level to County Level

• Emission of 31 provinces was disaggregated to

Emission of 31 provinces was disaggregated to more detailed level, 2347 counties. more detailed level, 2347 counties.

Distribution map of SO2 emission from power sector（kg/yr）

Sector Power Cement Iron & Steel Nonferros Transport Commerce Public Other Index to disaggregate from province level to county level GDP of secondary industries GDP of tertiary industries Population

9.12E+05 0.0

Disaggregation Disaggregation to 30 to 30” ”x30 x30” ” Mesh Mesh

• Emission from county level was disaggregated to

Emission from county level was disaggregated to 30 30” ” x 30 x 30” ” mesh by using population mesh by using population distribution. distribution.

人口分布データ 人口分布データ Emission from county Emission from county

Distribution map of SO2 emission from power sector（kg/yr） Population distribution Population distribution

Land Use Map of Urban Area Land Use Map of Urban Area

• A digital map of land use was made by overlaying land use

A digital map of land use was made by overlaying land use map scanned, location information map scanned, location information ( (VMAP VMAP) ), and satellite , and satellite image. image.

Digitized land use information Digitized land use information

公共建築 商業中心区 学校&病院 居住 工業&倉庫

Public building area Commercial area School & Hospital area Residential area Industrial area & Warehouse

1.00E+05 0.0 4.35E+04 0.0

• Emission data of each 30

Emission data of each 30” ” x 30 x 30” ” cells is cells is rearranged to actual one by using sector rearranged to actual one by using sector’ ’s area s area and population and population

Emission distribution by using population density（kg/yr）

Rearrangement by Using Rearrangement by Using Land Use Information Land Use Information

9.33E+04 0.0

Population density Residential area Redisaggregated one by using information of residential area（kg/yr）

Characteristics of AIM/AIR Air Characteristics of AIM/AIR Air quality model quality model

• Diffusion equations: Puff model is applied at the

Diffusion equations: Puff model is applied at the time the wind speed is small (< 1m/s), time the wind speed is small (< 1m/s),

• therwise plume model is applied.
• therwise plume model is applied.
• Accelerated algorithm in diffusion calculation is

Accelerated algorithm in diffusion calculation is adopted. adopted.

• Concentration every hour is calculated.

Concentration every hour is calculated.

• Emission pattern (time change of emitting

Emission pattern (time change of emitting pollutants) is definable for each sector. pollutants) is definable for each sector.

• Meteorology data are interpolated at the

Meteorology data are interpolated at the receptors where concentration is calculated. receptors where concentration is calculated.

Emission Patterns -1-

• Hourly change

Hourly change

– – Cement, Iron & Steel, Non Cement, Iron & Steel, Non-

• ferrous, Commercial

ferrous, Commercial sectors sectors: Constant emission : Constant emission from 9 to 19 o from 9 to 19 o’ ’clock clock – – Power and Public sectors Power and Public sectors: : Emission pattern was set Emission pattern was set based on actual based on actual consumption of electricity in consumption of electricity in a day a day

• Daily change

Daily change

– – Cement, Iron &steel, Non Cement, Iron &steel, Non-

• ferrous sectors

ferrous sectors: No : No emission on Sat. & Sun. emission on Sat. & Sun. Constant emission on the Constant emission on the

• ther days
• ther days

0.00 0.01 0.02 0.03 0.04 0.05 0.06 6 12 18 24 O'clock

Magnitude of emission

Emission Patterns Emission Patterns -

• 2

2-

• Monthly change

Monthly change

– – Commercial and Public Commercial and Public sectors: sectors: Emission fraction Emission fraction corresponding to each corresponding to each month was decided so month was decided so that heating devices that heating devices may be much used in may be much used in the month which the month which averaged temperature averaged temperature

• f the month is more
• f the month is more

than 10 degree. than 10 degree.

• Special term or day

Special term or day – – Cement, Iron & Steel, Cement, Iron & Steel, Non Non-

• ferrous sections:

ferrous sections: No emission on the No emission on the special term or day special term or day

1 2 3 4 5 6 7 8 9 10 11 12 M

• nth

Emission fractions 0.009 0.013 0.243 0.121 0.051 C

• m

m erci al Publ i c 0.447 1.000 0.071 0.178 0.125 0.066 0.011 0.009 1.000 0.078 0.076 0.052 0.051 0.009 0.009 0.051 0.169 0.051 0.051 0.052 0.009

新年 1-Jan 春節 5-Feb ～7-Feb 労働祭 29-Apr ～5-May 国慶祭 29-Sep ～3-O ct Special term or day

System Integration of System Integration of Emission Projection Emission Projection and Air Quality Model and Air Quality Model

• To integrate AIM/

To integrate AIM/ Enduse Enduse and AIM/Air, and AIM/Air, reconfirmation of LPS data base, reconfirmation of LPS data base, rearrangement of AIM/Air data format, rearrangement of AIM/Air data format, documentation were conducted. documentation were conducted.

• Spatial

Spatial disaggregation disaggregation of emission to sector

• f emission to sector’

’s s area by using land use information was area by using land use information was performed with performed with ArcInfo ArcInfo and Access. and Access.

• 32 clustered computers aided diffusion

32 clustered computers aided diffusion calculation. calculation.

Result of estimated emission by using Result of estimated emission by using end use model, AIM/Local China end use model, AIM/Local China

• Total emission of SO

Total emission of SO2

2 was 18,206Kt

was 18,206Kt、

、 total emission of

total emission of NOx NOx was was 11,033Kt. 11,033Kt.

• Contribution from power sector was the largest in both SO

Contribution from power sector was the largest in both SO2

2 and

and NOx NOx emission emission

• Emission from public sector occupied 15.7% of total emission.

Emission from public sector occupied 15.7% of total emission.

• Emission from traffic sector occupied 12.0% of total emission.

Emission from traffic sector occupied 12.0% of total emission.

Share of SO2 emission by sector Share of NOx emission by sector

Public（Urban area） 1.3% Public（Rural area） 2.6% Commerce 0.9% Nonferrous 2.7% Iron & Steel 7.5% Cement 4.4% Transport 12.0% Power 51.6% Others 17.0%

Power 41.1% Others 23.5% Commerce 5.1% Transport 1.0% Public（Urban area） 5.3% Public（Rural area） 10.4% Cement 5.8% Nonferrous 4.1% Iron & Steel 3.7%

General Result of Air Quality General Result of Air Quality Estimation Estimation

• AIM/Air Air quality model was applied to

estimation of SO2、NOx concentration of Beijing, Shanghai, and Chongqing.

• The estimated and the observed values of year

averaged concentration were compared.

Estimated Observed Estimated Observed Beijing 0.11 0.08 0.08 0.14 Shanghai 0.03 0.04 0.03 0.10 Chongqing 0.11 0.17 0.07 0.06 SO2(mg/m3) NOx(mg/m3)

*) The observed values are reported in China Energy Databook

Result of SO Result of SO2

2 Concentration

Concentration in Beijing in Beijing

• Remarkable seasonal

change, specially high concentration in winter season.

• Concentration on 63 days

was over the 2nd level standard (0.15mg/m3)

• Concentration on 35 days

was over the 3rd level standard (0.25mg/m3)

• Commercial and power

sectors were main contributors to SO2 concentration

0% 20% 40% 60% 80% 100%

Jan. Feb. Mar. Apl. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Contribution to SO2 concentration

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1/1 3/1 5/1 7/1 9/1 11/1 Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Daily averaged SO2 concentration

Comparison of SO Comparison of SO2

2 Concentrations of

Concentrations of Shanghai and Shanghai and Chongqing Chongqing with Beijing with Beijing’ ’s s

Shanghai Chongqing

0% 20% 40% 60% 80% 100%

Jan.

• Feb. Mar.

Apl. May Jun. Jul. Aug. Sep.

• Oct. Nov.

Dec.

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Contribution to SO2 concentration

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 1/1 3/1 5/1 7/1 9/1 11/1

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Daily averaged SO2 concentration (mg/m3)

0% 20% 40% 60% 80% 100%

Jan. Feb. Mar. Apl. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Contribution to SO2 concentration

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 1/1 3/1 5/1 7/1 9/1 11/1

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Daily averaged SO2 concentration (mg/m3)

Result of Result of NOx NOx Concentration in Beijing Concentration in Beijing

• Seasonal change and

high concentration in winter season.

• Concentration on 64 days

was over the 2nd level standard (0.15mg/m3)

• Concentration on 11 days

was over the 3rd level standard (0.25mg/m3)

• Power sectors was a main

contributor to NOx concentration.

0% 20% 40% 60% 80% 100%

Jan. Feb. Mar. Apl. May Jun. Jul.

• Aug. Sep.

Oct. Nov. Dec.

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Contribution to Nox concentration

0.00 0.05 0.10 0.15 0.20 0.25 0.30 1/1 3/1 5/1 7/1 9/1 11/1 Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Daily averaged NOx concentration (mg/m3)

Comparison of Comparison of NOx NOx Concentrations of Concentrations of Shanghai and Shanghai and Chongqing Chongqing with Beijing with Beijing’ ’s s

0% 20% 40% 60% 80% 100%

Jan. Feb. Mar. Apl. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Contribution to NOx concentration

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 1/1 3/1 5/1 7/1 9/1 11/1

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Daily averaged NOx concentration (mg/m3)

Shanghai Chongqing

0% 20% 40% 60% 80% 100%

Jan.

• Feb. Mar.

Apl. May Jun. Jul.

• Aug. Sep. Oct. Nov. Dec.

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Contribution to NOx concentration

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 1/1 3/1 5/1 7/1 9/1 11/1

Public Commerce Transport Others Nonferrous Iron & Steel Cement Power

Daily averaged NOx concentration (mg/m3)

Conclusion

• By estimating emission from road networks,

By estimating emission from road networks, diffusion of air pollutant from transport sector diffusion of air pollutant from transport sector was calculated. was calculated.

• Land use information added more reality to

Land use information added more reality to location of area emission source. location of area emission source.

• Concentration of SO

Concentration of SO2

2、

、NO

NOｘ

ｘ in Beijing, Shanghai,

in Beijing, Shanghai, and and Chongqing Chongqing were estimated and evaluated . were estimated and evaluated .

• This system becomes a powerful tool to assess

This system becomes a powerful tool to assess impact on air pollution briefly. impact on air pollution briefly.