Yunsoo Choi, PI Beata Czader, co-PI

08/08/2014

Lamar University/TARC

Impact of uncertainties in NO2 and HONO emission and

chemistry on radicals and ozone in southeast Texas

• NOx emissions has a large uncertainties

• Reaching to a factor of two

• Remote sensing gives some NOx emission constraints (e.g., Choi et al., 2008, 2012)

• Nitrous acid (HONO) is a major source of hydroxyl radical (OH) in the morning time

• Accurate estimation of HONO in air quality modeling is important as it affects predictions of HOx (OH+HO2) and ozone concentrations (e.g., Czader et al., 2012)

• Inaccurate model predictions of NOx concentrations results in mispredictions of HONO concentrations.

Uncertainty of NOx emissions

NO emissions particularly decrease in the urban areas over the southern US

Remote sensing adjusted NOx emission inventory

The large NOx emissions reduction decrease surface NOx concentrations over Houston, which migrate the difference between model and observation (Choi, ACP, 2014)

Emission adjustment impact on NOx over Houston

Circle: CMAQ NOx – Obs NOx

Uncertainty of HONO emissions

Background

University of Houston

1. Nitrous acid (HONO) is an important source of hydroxyl radical (OH), which plays a crucial role in oxidation of volatile organic compounds (VOCs) leading to the formation of ozone.

Czader et al., JGR, 2013

CMAQ results for SHARP, 2009

Background

University of Houston

2. Accurate estimation of HONO in air quality modeling is important as it affects predictions of HOx (OH + HO2) as well as ozone concentrations.

3. Current mobile-emission model, MOVES, estimates HONO emissions based on the HONO/NOx ratio derived from the tunnel studies done in 2001.

HONO sources

Emissions from combustion processes

1. Gas phase formation OH + NO → HONO 2. Heterogeneous formation

2 NO2 + H2O → HONO + HNO3

can be parameterized as NO2 → HONO with reaction rate coefficient

3. Photolytic sources Experiments show enhanced HONO formation when the sunlight is available, suggested uptake coefficient r = 2∙10-5 with dependence on light intensity

3

x

1083NO

HONO

Based on Kurtenbach et al.,2001.

Direct emissions Chemical formation

University of Houston

r = 1 – 5 ∙10-6 r = 1∙10-6 used in CMAQ

Recent HONO measurements in Houston

University of Houston

Rappenglueck et al., JAWMA, 2013 reports HONO measurement in Houston and suggest much higher HONO/NOx emissions ratio than Kurtenbach et al. 2001.

xNO

HONO 0.017 (±0.0009) r2=0.75 based on all data, July 15 – Oct. 15, 2009

0.016 r2=0.88 for Sep. 28, 2009

Highway Junction I-59 South/610 in Houston

Partial view of where the measurements were taken

Goals

University of Houston

1. Evaluate NOx emissions

2. Apply the latest HONO/NOx ratio in estimating emissions of HONO

from mobile sources;

3. Perform air quality simulations with the CMAQ model;

4. Evaluate the effect of changing mobile emissions on HONO

predictions as well as on O3 and HOx mixing ratios.

Methodology

University of Houston

Time period: September 2013, same as DISCOVER AQ in Houston Emissions: 2008 NEI processed with SMOKE v. 3.1 Meteorological parameters: WRF v. 3.5 driven by NAM for AQF and re-simulated with NARR data AQM: CMAQ v. 5.0.1, cb05tucl_ae5_aq chemical mechanism

Evaluation of NOx modeling

University of Houston

HONO mixing ratios depends on accuracy of NOx emissions and NO2 mixing ratios.

Therefore, for evaluation of HONO modeling one should use the most accurate NOx emissions.

Our simulations for September 2013 use 2008 NEI but emission reductions from mobile sources as well as point sources occurred between 2008 and 2013 due to emission restrictions and technological developments resulting in lower mobile emissions.

NOx emissions in our modeling system are overpredicted at many locations and need to be adjusted to 2013 levels before proceeding with HONO research.

National emissions trends:

Emissions reductions since 2008

Emission inventories:

TEXAS Harris County

Sector NEI2008 NEI2011 % decrease NEI2008 NEI2011 % decrease

mobile 600493 474137.2 21 71484.25 51142.79 28

others 665662.7 518628.4 16 17943.44 17272.06 21

NOx 2008 2009 2010 2011 2012 2013

mobile 6,941 6,241 5,734 5,786 5,398 5,010

other 9,968 9,636 9,240 8,789 8,309 8,109

total 16,909 15,877 14,974 14,574 13,707 13,119

mobile % 100 17 28

other % 100 12 19

total 100 14 22

University of Houston

Our modeling case: 30 % reduction in mobile sources NOx and 20 % reduction in point sources

EI shows more reduction than NET

NOx measurements

University of Houston

Map of CAMS and the Moody Tower measurement sites. Color symbols show NOx mean values for September 2013.

• The main source of NOx emission is traffic.

• The highest values occur in heavy traffic areas such as downtown Houston and along highways

• High NOx values are also in the industrial areas east of downtown

Low NOx sites – benefits of NOx emissions reduction

University of Houston

Industrial

industrial

Low NOx sites – NOx emissions reduction lead to NOx underprediction

University of Houston

BAD CASE

Mid-range NOx sites affected by traffic emissions

University of Houston

High NOx sites - benefits of NOx emissions reduction

University of Houston

High NOx sites - benefits of NOx emissions reduction

University of Houston

Comparison of NO for MT for Sep. 2013

University of Houston

Statistics NO

Number of points 716

Mean Observed 2.54

Sim. Reg. 5.52

Sim. Red. NOx 3.05

Max. value Observed 49.97

Sim. Reg. 98.80

Sim. Red. NOx 59.41

Correlation

coefficient

Sim. Reg. 0.55

Sim. Red. NOx 0.54

Mean Bias Sim. Reg. 2.98

Sim. Red. NOx 0.51

Absolute

Mean Error

Sim. Reg. 4.48

Sim. Red. NOx 2.56

Index of

agreement

Sim. Reg. 0.59

Sim. Red. NOx 0.71

Reduction of NOx gives better prediction of NO at the Moody Tower site

Comparison of NO2 for MT for Sep. 2013

University of Houston

Statistics NO2

Number of points 703

Mean Observed 7.35

Sim. Reg. 14.97

Sim. Red. NOx 11.51

Max. value Observed 34.24

Sim. Reg. 67.16

Sim. Red. NOx 62.09

Correlation

coefficient

Sim. Reg. 0.65

Sim. Red. NOx 0.64

Mean Bias Sim. Reg. 7.62

Sim. Red. NOx 4.17

Absolute

Mean Error

Sim. Reg. 8.48

Sim. Red. NOx 5.95

Index of

agreement

Sim. Reg. 0.63

Sim. Red. NOx 0.71

Reduction of NOx gives better prediction of NO2 at the Moody Tower site

Comparison of O3 for MT for Sep. 2013

University of Houston

Statistics O3

Number of points 712

Mean Observed 30.96

Sim. Reg. 31.61

Sim. Red. NOx 34.15

Max. value Observed 98.39

Sim. Reg. 75.65

Sim. Red. NOx 80.37

Correlation

coefficient

Sim. Reg. 0.73

Sim. Red. NOx 0.74

Mean Bias Sim. Reg. 0.64

Sim. Red. NOx 3.18

Absolute

Mean Error

Sim. Reg. 9.88

Sim. Red. NOx 9.97

Index of

agreement

Sim. Reg. 0.84

Sim. Red. NOx 0.83

Minimal impact on ozone from reduced NOx case

Increasing HONO emissions

University of Houston

SMOKE speciation profiles for mobile sources:

NOX → NO2 → 9.2% → 8.4% NOX → NO → 90% → 90% NOX → HONO → 0.8% → 1.6%

Kurtenbach et al.,2001. CMAQ

xNO

HONO0.017

Rappenglueck et al.,2013

0.008 0.003-0.008

0.016

Increasing HONO emissions

University of Houston

Statistics HONO

Number of points 200

Mean Observed 0.69

Sim. Red. NOx 0.30

Sim. H 0.41

Max. value Observed 3.15

Sim. Red. NOx 2.62

Sim. H 2.93

Correlation

coefficient

Sim. Red. NOx 0.58

Sim. H 0.57

Mean Bias Sim. Red. NOx -0.39

Sim. H -0.28

Absolute

Mean Error

Sim. Red. NOx 0.46

Sim. H 0.43

Index of

agreement

Sim. Red. NOx 0.63

Sim. H 0.70

The base case simulation (reduced NOx) resulted in too low HONO mixing ratios Increasing HONO emissions resulted in higher HONO mixing ratios, especially at during morning peak values

Effects of changing HONO on OH

University of Houston

Photolysis of HONO is a source of NO and OH. Increased HONO emissions resulted in up to 6% increased OH during mid-morning. An increase in OH occurs along highways corresponding to increased HONO mobile emissions.

12 LT 12 LT

Effects of changing HONO on ozone

University of Houston

Increasing HONO emissions resulted in increased ozone prediction of up to 1 ppb during morning time. Changes in the afternoon ozone, at the time of ozone peak, are not significant.

11 LT 15 LT

Conclusions

University of Houston

Reductions in NOx improved NOx predictions especially at regional sites for which model used to overpredict NOx as well as sites affected by heavy traffic and in the industrial areas.

Increasing HONO emissions from mobile source according to recent estimates resulted in increased HONO mixing ratios that is closer to measured values.

Increased HONO emissions impacted OH during morning and midday hours, with up to 6% increase. Changes in ozone are not significant, up to 1 ppb.

Following work: Humidity dependence of HONO formation

Humidity dependence of HONO formation

Stutz et al., 2004 show that NO2 → HONO conversion is

humidity depended.

They recommended that RH should be considered in the

parameterization of HONO formation in air pollution

models

How to implement

For CMAQ implementation we will scale the reaction uptake

coefficient (r) with humidity

e.g., r scaled by a factor (humidity)/30

Products from this project

University of Houston

Choi, Y., The impact of satellite-adjusted NOx emissions on simulated NOx and O3 discrepancies in the urban and outflow areas of the Pacific and Lower Middle US, 2014, Atmospheric Chemistry and Physics, 14, 675-690

Czader, B., Choi, Y., and Li, X., Impact of updated traffic emissions on HONO mixing ratios simulate for urban site in Houston, Texas, 2014, acp-2014-549

Enhanced capability of UH air quality forecasting system (http://spock.geosc.uh.edu)

UH air quality forecasting system

University of Houston