Reliability of Predicting Air Quality from Transportation Projects Gilad Kozokaro Department of Geography and Environment. Bar-Ilan University

Air Pollution in urban areas During the last decades, air-pollution emitted by transportation in urban areas has become a major problem, many researches considering it to be the main cause for morbidity and mortality rate increase in urban areas.

Air pollution in urban area caused by:

•

Significant increase in the amount of vehicles that travel in the city.

•

The character of air pollutants emission and the dispersion by transportation compared to industries pollution.

Air Pollution in urban areas There are two main strategies of decreasing air pollution from transportation

Decreasing the air pollutants emission by

attempts to change land uses and

developing alternative technology and

transportation use patterns in order to

fuels, in order to design more effective

significantly decrease air pollution in

and less polluting vehicles

urban areas

Environmental impact assessments Researches have demonstrated that the amount and character of mileage as well as the patterns of land uses for transportation affect the demand for travelling, and are the main causes for air pollutants emission increase over the years.

In order to decrease air pollution levels, policy makers must have a long run planning and predicting horizon, taking into account high levels of uncertainty.

The decision makers rely on environmental impact assessments designed to assess all expected environmental impacts of implementing a specific plan.

Purposes and hypothesis Our Purposes was: •

Examine the reliability of air pollution forecasts in environmental impact assessments in transportation projects.

•

Examining the reliability of air pollution dispersion using a combination of a Lagrangian model and a regional meteorological model, compared to Gaussian models commonly used.

Our hypothesis was: 1. That a difference might be found between the air pollution levels predicted by the assessments Compared to the actual measured levels. 2. That the use of a complex dispersion equations model taking into account detailed meteorological data would yield results closer to the actual data, compared to the currently used models.

Tools • Data from three main highways in Israel built for the last years (roads no. 471, 431 and 6) and data from three roads that are near the main roads and have monitoring stations that work in 2010 (roads no. 4, 412 and 574). •

CALINE4, CAL3QHC and CAL3QHCR

are Gaussian models to predict carbon

monoxide (or other inert pollutant) concentration from motor vehicles at roadway. •

CALPUFF is Lagrangian model that can predict range of pollutant from different sources and from different time and space scale.

•

Weather Research and Forecasting (WRF) model is mesoscale numerical weather prediction system serve both atmospheric research and operational forecasting.

• Quantum GIS (Qgis) is Open Source Geographic Information System.

The study stages 1. Re-running the model using the data used for the original assessments. 2. Running the CAL3QHCR model for the three roads under study, using data updated to 2010, in the following categories: emission quotients, traffic volume, and meteorological data. 3. Running the CALPUFF for the three roads under study, using data updated to 2010, in the following categories: emission quotients, traffic volume, and meteorological data. 4. Running both CAL3QHCR and CALPUFF models, using 2010 data, for the three roads located near the roads under study, where air quality monitoring stations are located.

Results Stage 1 Road No’

Original CAL3QHC

Re-running CAL3QHC

Different (%)

IOA

471

1327.3

1350.1

101.7%

0.95

431

808.4

970.8

120.1%

0.73

6

356.3

384.1

107.8%

0.87

It mean that the different results later on are due to a data update and the use or another type of model.

Results Stage 2 Road No’

Original Data year2010 using CAL3QHC

Update Data year 2010 using CAL3QHC

Different (%)

471

1327.3

842.7

63.5%

431

808.4

1111.4

137.5%

6

356.3

247.8

69.6%

We found that the different results are related to a data update. 1.

Emission quotients was low in 2010 then in the Origin runs in all 3 roads.

2.

Traffic volume use low in road 471 and 6 and Equal in road 471.

3.

Meteorological data in all roads didn’t take the all stability class that are present in the area.

Results Stage no’ 3 Road No’

Original assessment to 2010 using CAL3QHC

Update Data to 2010 using CALPUFF

Different (%)

471

1327.3

1949.5

144.4%

431

808.4

3696.0

380.7%

6

356.3

374.9

97.6%

We found that the different results are related to a data update and the model. 1.

Emission quotients was low in 2010 then in the Origin runs in all 3 roads.

2.

Traffic volume use low in road 471 and 6 and Equal in road 471.

3.

Meteorological data in all roads didn’t take the all stability class that are present in the area. The CALPUFF model give high prediction of concentration

Results Stage no’ 4 Road No’

Road No’

Max Actual Measured

Max CAL3QHCR

Max CALPUFF

471

4

1579.6

596.0

1016.9

431

412

1951.9

750.9

995.2

6

574

562.7

122.1

201.5

Statistical parameter

CAL3QHCR

CALPUFF

IOA

0.59

0.54

FAC2

4.56

1.94

FB

1.17

0.50

Results Stage no’ 4

Part of The Conclusions……. 1.

The original assessments under assessed the traffic volumes.

2.

The original assessments failed to present all actual meteorological conditions, such as wind speeds and atmospheric stability levels. Inherent limitations in CALINE4, CAL3QHC, and CAL3QHCR models equation make them to failed to present the pollution in low speed that occur in realty.

3.

CALPUFF & WRF representation all the wind speed and Weather Conditions in winter which may predicted to High pollution level that actual occur in realty.

4.

Running CALPUFF model, combined with WRF weather model give a relatively high level of accuracy, when compared actually measured.

Thank you for your attention