Satellite-detection and attribution of rapid increases in tropospheric ozone made in China and its influence on the western United States Folkert Boersma.

Satellite-detection and attribution of rapid increases in tropospheric ozone made in China and its influence on the western United States

Folkert BoersmaNow also at

A short story about ozone made in China

SubmittedWillem W. Verstraeten1,2,3*, K. Folkert Boersma1,2,3, Jason E. Williams1, Jessica L. Neu4, Kevin W. Bowman4 and John R. Worden4

Impact of (SRES-A2) emissions (in ppb)

2030 – 2000

Loss in life expectancy in monthsSource: EU programme CAFE, CAFE Scenario Analysis Report Nr. 2. (Amann et al., 2004)

Szopa et al., 2004

Why do we care about ozone in the troposphere?

I live here

1. Ozone is an air pollutant – also in Europe

Shindell et al., 2009

Why do we care about ozone in the troposphere?

2. Ozone is the 3rd most powerful cause of positive radiative forcing

O3 RF: 0.40 W/m2 CO2 RF: 1.69 W/m2

…but uncertainty in ozone forcing (±0.20 W/m2) higher than for CO2 and CH4

Why do we care about ozone in the troposphere?

3. Ozone drives the oxidative capacity of the atmosphere

O3 + hv O2 + O(1D)O(1D) + H2O 2 OH

OH is the “detergent” of the atmosphere, determining lifetime of CH4, CO, and many other pollutants

Need to understand ozone to improve our understanding of OH and possible trends therein

Model-result: OH very difficult to measure!

How is ozone formed in the troposphere?

Lifetime tropospheric O3: a few weeks

STE

IPCC AR5-reported trends in tropospheric ozone

IPCC conclusions• Ozone increases over Europe and eastern USA have levelled off• Ozone has likely increased over eastern Asia since the 1990s

Ozone increases over western United States

• Compile and analyse large dataset (1984-2008) of ozone measurements over the western United States

• Ozonesondes, lidar, aircraft measurements

Cooper et al., 2008

EU air quality standard

Unclear where that ozone came from

High ozone tends to originate from deep within Asia

LOW OZONE CONDITIONS

HIGH OZONE CONDITIONS Ozone monitoring in the free troposphere since the 1970s is very limited.

What can satellite instruments tell us?

Observes NO2 in VIS

Surface

l1 l2

Scattering by Earth surface and by atmosphere

Satellite instruments: TES and OMI

Observes ozone in thermal IR

SurfaceTo

Il(T0)

blackbodyradiation

absorbing gasT1

elIl(T1)

Broadly sensitive to

free tropospheric ozone

Sensitive to NO2

down to the surface

Vertical sensitivity

464 hPa averaging kernel:Broad sensitivity to free tropospheric ozone

OMI: sensitivity down to surface

Validation and use of TES ozone retrievals

Global map of summertime tropospheric ozone from TES (@464 hPa)

Strong precursor emissions!+ photochemical activity

JJA2005-2010464 hPa

Validation and use of TES ozone retrievals

Global validation with WOUDC ozone sondes for period of 6 years

• Insignificant trend in TES-sonde bias at midlatitudes• TES can be safely used for temporal analysis

Verstraeten et al., AMT, 2013

OMI tropospheric NO2 retrievals

Strong precursor emissions!

Increase in Summertime tropospheric O3 and NO2 observed by TES and OMI over China

2005-2006: May-August 2009-2010: May-August 2010-2005 Difference

• TES: Strong increase in FT ozone over and downwind of China• OMI: Strong increase in BL NO2 pollution over China

TES 464 hPa

Increase of 1.1 ppbv/yr ozone 0.0

15.0

30.0

45.0

60.0

75.0

90.0

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73

Months

Mo

nth

ly o

zo

ne V

MR

at

46

4 h

Pa

-18

-12

-6

0

6

12

18

24

30

36

42

48

Mo

nth

ly a

no

maly

ozo

ne V

MR

at

46

4

hP

a

TES Anomaly TES Lineair (Anomaly TES)

Increase of 1.1 ppbv ozone/yr at 464 hPa

TES 3-9 km partial ozone column 2009-2010

Also increasing but not at a statistically significant pace

Increase in tropospheric O3 observed by TES over western United States

Interpretation of the increases with TM5 CTM

Research Question 1: Can emissions explain the increase in FT O3 over/downwind of China?

3-D global chemistry transport model• 34 vertical layers (surface-0.1hPa), 3˚×2˚• ECMWF meteorological fields (ERA-interim)

Emission inventories: • Anthropogenic inventories: RETRO project + REAS inventory for Asian • Lightning NOx emissions: ~5 Tg N/yr• Biomass burning emissions: GFEDv2• Tropospheric chemistry: 42 species and > 60 reactions (CBM-IV)• Reference: Huijnen et al., 2010.

2 TM5 simulations:1. Invariable NOx emissions2. OMI-constrained NOx emissions

OMI-constraints on NOx emissions (2005-2010)

• Update a priori emissions to follow OMI-observed trend• Update for all grid cells (no smearing)• Take into account non-linear sensitivities to chemical regime

NO2

OMI,2006

OMI,2006jOMI,2006j Ω

ΩΩβ1EE

New top-down emissions

A priori emissions

Change in year j relative to 2006 in NO2 observed with OMI

Sensitivity of NO2 column to changing emissions

Taking into account chemical regime (β)

Evaluate TM5 response to perturbing NOx emissions (Lamsal et al., 2009):

TM5,2006

TM5,2006

Ω

ΔΩβ

E

ΔE

β=1: linear local response between NOx and NO2 columnβ>1: Rapid negative feedback on NO2 via increased OHβ<1: Accumulation of NO2 in response to emissions

JANUARY

JULY

Increasing emissions by +15%

In winter: β<1 (long lifetime NO2)In summer: β>1 (enhanced loss)

OMI-constrained NOx emissions (2010-2005)

Clean-up of USA, Europe and Japan Strong increases in Asia

OMI-constrained NOx emissions over China (2010-2005)

y = 0.0144x + 3.1972

R 2 = 0.035

y = 0.0142x - 0.5179

R 2 = 0.3057

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

1 13 25 37 49 61 73Months since J anuary 2005

OM

I tr

op

.NO

2 1

0-1

5 m

ole

c/cm

2

Mean China Anomaly Lineair (Mean China) Lineair (Anomaly)

5.4% / yr

y = 0.0144x + 3.1972

R 2 = 0.035

y = 0.0142x - 0.5179

R 2 = 0.3057

y = 0.0268x + 2.4192

R 2 = 0.0849y = 0.0032x + 2.0221

R 2 = 0.0091

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

1 13 25 37 49 61 73Months since January 2005

OM

I tr

op

.NO

2 1

0-1

5 m

ole

c/cm

2

Mean China Anomaly TM5_ ref TM5_ upd

Lineair (Mean China) Lineair (Anomaly) Lineair (TM5_ upd) Lineair (TM5_ ref)

5.4% / yr

TM5 with OMI emissions captures OMI time series

Reproducing the TES-observed O3 increase over China

TM5 with updated OMI emissions better captures TES time series

But substantial increase also in TM5 simulation with constant NOx emissions

Stratosphere-troposphere exchange (STE) over China

Research Question 2: What is the role of stratosphere-troposphere exchange?

Use TM5• Stratospheric ozone in TM5 is constrained by 30-year

(assimilated) 14-satellite dataset of total ozone (van

der A. et al., ACP, 2010)• Net stratosphere-troposphere exchange is governed

by ECMWF meteo + MSR constraints• STE can be tracked in TM5 with tagged tracer ‘O3S’O3 increase due to STE Overall O3 increase

2010-2005 differenceChina

2010-2005 differenceChina

Stratosphere-troposphere exchange (STE)

Deseasonalized changes in tropospheric ozone

Answers to Research Questions:• STE and increasing NOx both explain

the observed increase over China • STE stronger effect

From subtracting TM5 simulations 1 and 2 (same STE)

What is happening with O3 over the western United States?

TES 3-9 km partial ozone column 2009-2010

Also increasing but not at a statistically significant pace

Reduced NOx emissions …

Substantial reductions in NOx emissions over the western USA (-12% between 2005-2010)

Overall O3 increase O3 increase due to STE

…but also strong O3 increase due to STE over the western USA

STE associated with enhanced stratospheric circulation in response to vigorous El Nino & QBO in 2010 relative to 2005

2005-2010 ozone changes over the western US

Deseasonalized changes in tropospheric ozone: increase!

• TES indicates increase over western US• In spite of local decreases in NOx

• STE masks out effects reductions in anthropogenic NOx emissions

How much of the western US ozone increase originated from China?

Compare two 6-year TM5 simulations

2. TM5 with OMI-constrained NOx emissions throughout the world3. As (2) but with constant REAS-2006 NOx emissions over China

Difference between (2) and (3) indicative of the effect of increasing Chinese NOx emissions on western United States

Change in Anthropogenic NOX emissions between 2010 and 2006

ΔO3 due to increasing Chinese NOx emissions

ΔO3 between 2010 and 2005 at ~900 hPa (±1 km)

Western US O3 increases by +0.1 ppbv at ±1 km because of +15% Chinese NOx emission increases

ΔO3 between 2010 and 2005 at ~460 hPa (±6 km)

Western US O3 increases by +0.2 ppbv at ±6 km because of large-scale transport & sustained O3 production driven by decomposition of nitrogen reservoirs in descending air masses

ΔO3 due to increasing Chinese NOx emissions

Deseasonalized timeseries 3-9 km correcting for STE and local effects

FT (3-9 km) O3 increases by 0.01 DU/yr (0.1%/yr)

This import neutralizes one-third of the O3 decreases related to reduction in surface emissions due to stricter air pollution policies.

Discussion

• Combined use of TES and OMI provides a new perspective on changing tropospheric composition in 2005-2010

• TES observes a 10% increase in 6 years in FT O3 over China• Supported by TM5 when OMI-constrained NOx (+2%) and STE

changes (+8%) are accounted for

• TES indicates a 4% increase in FT O3 over western US• Policy-driven NOx reductions alone would have reduced O3 (-1%)

over the western US in the absence of other effects• STE (+5%) and import of ozone made in China (+0.5%) offset

local policy measures

Support for increased STE over China throughout 2005-2010

Wang et al, 2012, ACP

OZONE SONDES OVER BEIJING

--- Sonde: influenced by photochemical production and STE--- Model (CLaMS): only STE