Envair – C. Blanchard and S. Tanenbaum DRI – E. Fujita and D. Campbell
description
Transcript of Envair – C. Blanchard and S. Tanenbaum DRI – E. Fujita and D. Campbell
1
Understanding Relationships Between Changes in Ambient Ozone and Precursor
Concentrations and Changes in VOC and NOx Emissions from 1990 to 2004 in Central California
Envair – C. Blanchard and S. Tanenbaum
DRI – E. Fujita and D. Campbell
Alpine Geophysics – J. Wilkinson
Eight Hypothesized Explanations of Ozone TrendsJuly 1, 2008
2
Today’s Topics
I. Identify hypotheses to explain ozone trends
II. Evaluate hypotheses
III. Discuss relevance to policy questions
IV. Get feedback
3
Key Question
• Substantial and statistically significant downward trends in NOx, NMOC, CO.
• Generally downward trends in peak 8-hour ozone, but majority of sites less than 0.5 ppbv per year, only 7 of 42 statistically significant, and 10 sites have upward trends.
• Why hasn’t peak ozone decreased as much in some areas as in others?
5
1990 – 2004 AQ Trends Summary
• No precursor trends significantly upward*
• NOx sig* down at 22 of 28 sites**
• CO sig* down at 21 of 25 sites**
• NMOC sig* down at 6 of 7 sites***
* p < 0.05** At least 10 years data. One or both metrics: morning (5 – 11 am) or midday (time of peak 8-hour ozone).*** 7 - 10 years data, morning (5 – 8 am).
6“7 x 7 Emissions” = in the 7x7 array of 4 km grid cells around a site
25 sites 7 sites 28 sites
Average Reductions of Emissions and Ambient Concentrations, 1990 - 2004
0
10
20
30
40
50
60
70
80
90
CO NMOC NOx
Pollutant
Red
uct
ion
(p
erce
nt)
Ambient 7 x 7 Emissions Regional Emissions
7
Hypothesis 2
Ozone precursor emissions did not decline enough in some areas experiencing growth and development to produce a significant ozone response in those areas.
Plausible
8
Sub-Region
N orthern S ierra Footh ills
Southern S ierra Footh ills
Sacram ento Valley
N orthern SF Bay A reaEastern SF Bay A reaSouthern SF Bay A reaN orthern San JoaquinC entra l San Joaquin
Southern San Joaquin Boxes show where both NMOC and NOx emissions in the 7x7 arrays of 4 km grid cells (28 km x 28 km) surrounding monitoring sites decreased by less than 20 percent from 1990 to 2004
9
Sites Where Local NMOC and NOx Emissions Decreased by Less Than 20 Percent From 1990 to 2004
-10
-5
0
5
10
15
20
Turlock-S MinaretStreet
Merced-S CoffeeAvenue
Madera-HealthDept #2
Madera-PumpYard
Parlier Visalia-N ChurchStreet
Site
Em
issi
on
s D
ecre
ase
1990
- 2
004
(%)
NMOC Emissions NOx Emissions
“Local” means in the 7x7 array of 4 km grid cells
10If local NMOC emissions unchanged, how much will ozone change?
Regional and Local Ozone-Season NMOC Emissions in 1990 and 2004
0
100
200
300
400
500
600
700
800
San Francisco Bay Sacramento & NSF N & C SJV & SSF Five SJV Sites: Turlockto Visalia
Region
NM
OC
Em
issi
on
s (t
on
s p
er d
ay)
NMOC 1990 NMOC 2004
11Caveat: ambient NMOC decreases > inventory decreases
NMOC Decrease (1994-2004)
0
20
40
60
80
100
120
140
160
SacramentoDel Paso
Folsom Fresno-FirstStreet
Clovis Parlier Bakersfield-Golden
Arvin
NM
OC
Dec
reas
e (p
erce
nt)
Ambient 3 x 3 Emissions 5 x 5 Emissions 7 x 7 Emissions
12
Hypothesis 3
Background* ozone concentrations increased.
* Background ozone means ozone formed outside the North American boundary layer.
Implausible
13
Increasing Sites(north to south)
No. HighlandsPittsburgConcordTurlockMerced
Fresno SkyparkClovis
Fresno DrummondParlierOildaleArvin
Maricopa
D ecreasing
Increasing
Average M axim um 8-hour O zone
Increasing Sites(north to south)
No. HighlandsPittsburgConcordTurlockMerced
Fresno SkyparkClovis
Fresno DrummondParlierOildaleArvin
Maricopa
D ecreasing
Increasing
Average M axim um 8-hour O zone
Ozone increases occurred at scattered sites in separate subregions
Mean peak 8-hour ozone in 1990-94 and in 2000-04
14
0
10
20
30
40
50M
ean
Ozo
ne (
ppb
v)
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005Year
Crater LakePoint ReyesLava Beds
Crater Lake = -395.536 + .219 * Year; R^2 = .046Point Reyes = -1209.936 + .619 * Year; R^2 = .368Lava Beds = -1011.904 + .528 * Year; R^2 = .438
Possible upward trend in background ozone of ~0.5 ppbv per year- if true, may make overall progress more difficult
National Park Service Passive Ozone Monitoring
16Four meteorological types identified by PCA and K-means clustering
Number of Days in Each Meteorological Type, By Month
0
50
100
150
200
250
300
350
400
May June July August Sep Oct
Month
Nu
mb
er o
f D
ays
Type 1 Type 2 Type 3 Type 4
17
Frequency of Days In Each Meteorological Type (May - October)
0
20
40
60
80
100
120
140
160
180
200
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Nu
mb
er o
f D
ays
Type 1 Type 2 Type 3 Type 4
Year-to-year variations occurred in frequency of met types ….
18
Central San Joaquin High-Ozone Days By Meteorological Type
0
10
20
30
40
50
60
70
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Nu
mb
er o
f D
ays
Type 1 Type 2 Type 3 Type 4
…. And the distributions of high-ozone days varied ….
19…. But downward ozone trends did not depend upon met type ….
Number of Sites With Downward Ozone Trends, 1994-2004, by Meteorological Type
0
10
20
30
40
50
60
70
Type 1 Type 2 Type 3 Type 4 Total Sites With 5+Years Data (All
Trends)
Meteorological Type
Nu
mb
er o
f S
ites
20
Number of Sites With Upward Trends
0
2
4
6
8
10
12
Annual Mean Annual Mean Annual Mean Annual Mean Annual Mean Annual
4th-Highest Peak 8-Hour Peak 8-Hour Peak 8-Hour Peak 8-Hour Peak 8-Hour
Peak 8-Hour on Top 60 Days Met Type 1 Met Type 2 Met Type 3 Met Type 4
Trend Test
Nu
mb
er o
f S
ites
…. And upward ozone trends were consistent across met types
21
Hypothesis 5
Decreases in VOC emissions and reactivity slowed the rate of ozone formation but did not reduce the ultimate amount of ozone formed. Therefore, reductions of peak ozone occurred in urban cores and at near-downwind sites, but not at far downwind locations.
Plausible
22
Measurements show ~30 – 120% decreases in ambient NMOC concentrations
Sum of Species, Reactivity-Weighted Sum of Species, and Total NMOC Decreases (1994-2004)
0
20
40
60
80
100
120
140
160
180
200
Sacramento DelPaso
Folsom Fresno-FirstStreet
Clovis Parlier Bakersfield-Golden
Arvin
Dec
reas
e (p
erce
nt)
Sum of Species Reactivity-Weighted Sum of Species NMOC
23
Average reactivity* shows some declines(but mass decrease is larger and more important)
*kOH weighted by mean annual concentrations (units are ppbC-1 sec-1)
24
0
10000
20000
30000
40000
50000
60000
70000
0 20 40 60 80 100Distance (km)
NMOC 1997-2004NMOC 1990-1996
Fremont
San Jose- Piedmont
San Jose-4th
Los Gatos
San Martin
Gilroy
Southern San Francisco Bay Area
NM
OC
Em
issi
on
s (t
on
s/yr
)
0
10000
20000
30000
40000
50000
60000
70000
0 20 40 60 80 100Distance (km)
NMOC 1997-2004NMOC 1990-1996
Fremont
San Jose- Piedmont
San Jose-4th
Los Gatos
San Martin
Gilroy
Southern San Francisco Bay Area
NM
OC
Em
issi
on
s (t
on
s/yr
)
40
50
60
70
80
90
100
110
Ozo
ne (
ppb
v)0 20 40 60 80 100
Distance (km)
Peak 8-hour Ozone 1997-2004
Peak 8-hour Ozone 1990-1996
Southern San Francisco Bay Area
Fremont
San Jose-Piedmont
San Jose-4th
Los Gatos
San Martin
Gilroy
40
50
60
70
80
90
100
110
Ozo
ne (
ppb
v)0 20 40 60 80 100
Distance (km)
Peak 8-hour Ozone 1997-2004
Peak 8-hour Ozone 1990-1996
Southern San Francisco Bay Area
Fremont
San Jose-Piedmont
San Jose-4th
Los Gatos
San Martin
Gilroy
In S Bay area, ozone declined at both urban and downwind sites
25
40
50
60
70
80
90
100
110
Ozo
ne (
ppb
v)
0 20 40 60 80 100Distance (km)
Peak 8-hour Ozone 1997-2004Peak 8-hour Ozone 1990-1996
Sacramento and the Northern Sierra Foothills
Sac-T St
Sac-Del Paso
N. Highlands
Roseville
Elk Grove
Folsom
Auburn
Placerville
Grass Valley
40
50
60
70
80
90
100
110
Ozo
ne (
ppb
v)
0 20 40 60 80 100Distance (km)
Peak 8-hour Ozone 1997-2004Peak 8-hour Ozone 1990-1996
Sacramento and the Northern Sierra Foothills
Sac-T St
Sac-Del Paso
N. Highlands
Roseville
Elk Grove
Folsom
Auburn
Placerville
Grass Valley
0
10000
20000
30000
40000
50000
60000
70000
0 20 40 60 80 100Distance (km)
NMOC 1997 -2004NMOC 1990 -1996
Sac-T St
Sac-Del Paso
N. Highlands
Roseville
Elk Grove
Folsom
Auburn
PlacervilleGrass Valley
Sacramento and the Northern Sierra Foothills
NM
OC
Em
issi
on
s (t
on
s/yr
)
0
10000
20000
30000
40000
50000
60000
70000
0 20 40 60 80 100Distance (km)
NMOC 1997 -2004NMOC 1990 -1996
Sac-T St
Sac-Del Paso
N. Highlands
Roseville
Elk Grove
Folsom
Auburn
PlacervilleGrass Valley
Sacramento and the Northern Sierra Foothills
NM
OC
Em
issi
on
s (t
on
s/yr
)
In Sacramento, ozone declines varied among sites.
26
10000
20000
30000
40000
50000
60000
70000
0 20 40 60 80 100 120 140Distance (km)
NMOC 1997 -2004NMOC 1990 -1996
NM
OC
Em
issi
ons
(ton
s/yr
)
Northern San Joaquin and Southern Sierra Foothills
Stockton-Hazelton
Stockton-Mariposa
Modesto
San Andreas Jackson
Turlock
Sonora
Merced
Yosemite-Turtleback
10000
20000
30000
40000
50000
60000
70000
0 20 40 60 80 100 120 140Distance (km)
NMOC 1997 -2004NMOC 1990 -1996
NM
OC
Em
issi
ons
(ton
s/yr
)
Northern San Joaquin and Southern Sierra Foothills
Stockton-Hazelton
Stockton-Mariposa
Modesto
San Andreas Jackson
Turlock
Sonora
Merced
Yosemite-Turtleback
40
50
60
70
80
90
100
110
Ozo
ne
(pp
bv)
0 20 40 60 80 100 120 140Distance (km)
Peak 8-hour Ozone 1997-2004Peak 8-hour Ozone 1990-1996
Northern San Joaquin and Southern Sierra Foothills
Stockton-Hazelton
Stockton-Mariposa
Modesto
San Andreas
Jackson
Turlock
Sonora
Merced
Yosemite-Turtleback
40
50
60
70
80
90
100
110
Ozo
ne
(pp
bv)
0 20 40 60 80 100 120 140Distance (km)
Peak 8-hour Ozone 1997-2004Peak 8-hour Ozone 1990-1996
40
50
60
70
80
90
100
110
Ozo
ne
(pp
bv)
0 20 40 60 80 100 120 140Distance (km)
Peak 8-hour Ozone 1997-2004Peak 8-hour Ozone 1990-1996
Northern San Joaquin and Southern Sierra Foothills
Stockton-Hazelton
Stockton-Mariposa
Modesto
San Andreas
Jackson
Turlock
Sonora
Merced
Yosemite-Turtleback
Ozone
NMOC Emissions
NMOC emissions did not decline in some parts of the N SJV and SSF. This makes it difficult to assess the argument that NMOC-control was not effective at the far downwind sites.
27
Hypothesis 6
Decreases in NOx emissions and concentrations resulted in less titration of ozone. Therefore, ozone concentrations increased in NOx-rich areas, but peak ozone declined in NOx-limited areas.
Plausible
28
Mean Hourly NOx at Merced
0
5
10
15
20
25
30
35
40
45
50
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour
NO
x (p
pb
v)
NOx 1990-94 NOx 1995-99 NOx 2000-04
Mean Hourly Ozone at Merced
0
10
20
30
40
50
60
70
80
90
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour
Ozo
ne
(pp
bv)
O3 1990-94 O3 1995-99 O3 2000-04
Ozone increases at night can be linked to lower NOx ….
… But effects on peak ozone are difficult to interpret
29
Yosemite National Park - Turtleback Dome
0
20
40
60
80
100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Ozo
ne
(pp
bv)
Jerseydale
0
20
40
60
80
100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Ozo
ne
(pp
bv)
Five Mile Learning Center
0
20
40
60
80
100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Ozo
ne
(pp
bv)
Yosemite National Park - Turtleback Dome
0
20
40
60
80
100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Ozo
ne
(pp
bv)
Jerseydale
0
20
40
60
80
100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Ozo
ne
(pp
bv)
Five Mile Learning Center
0
20
40
60
80
100
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Ozo
ne
(pp
bv)
0
10
20
30
40
50
60
70
80
90
Ozo
ne (
ppbv
)
0 4 8 12 16 20 24HOUR
2000_20041995_19991990_1994
Period
Yosemite – Turtleback Dome
0
10
20
30
40
50
60
70
80
90
Ozo
ne (
ppbv
)0 4 8 12 16 20 24
HOUR
2000_20041995_19991990_1994
Jerseydale
Period
Mid-elevation Sierra Nevada sites show peak ozone improvements compared to early 1990s (but less change since the mid-1990s)
Mean Peak 8-Hour Ozone
30
Hypothesis 7
Greater or lesser decreases of VOC emissions compared to NOx emissions resulted in changes in VOC/NOx ratios, which led to changes in the efficiency of ozone production per unit of precursor mass.
Plausible
31
HighVOC/NOx
Low VOC/NOx
Termination Reactionat low VOC/NOx,
removes NOx.
Termination Reactionat high VOC/NOx, removes radicals.
Significance of VOC/NOx Ratios in Ozone Formation
Ozone formation is most efficient at VOC/NOx
(ppbC/ppb) ratios between 10-12.
32
Morning (5 – 8 a.m.) NMOC/NOx ratios declined over time
(Emission ratios average about 30% lower than ambient ratios)
33
NMOC/NOx - Parlier
0
5
10
15
20
25
30
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
NM
OC
/NO
x (p
pb
C/p
pb
v)
5:00 AM 12 Noon 4:00 PM 11:00 PM
Differences between morning and afternoon NMOC/NOx ratios declined
35
Hypothesis 8
The validity of hypotheses 5, 6, and 7 varied during the trend period, with hypothesis 5 (VOC reductions ) having greater relevance early and hypotheses 6 (NOx reductions) and 7 (VOC/NOx ratios) increasing in importance toward the end.
Plausible
36Somewhat larger NMOC declines from early to mid ’90s
Ozone-Season Emission TrendsNineteen Counties in CCOS Domain
0
200
400
600
800
1000
1200
1400
1600
1800
2000
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Em
issi
on
s (t
on
s p
er d
ay)
NMOC NOx NMOC Trend NOx Trend
37
Plausible Hypotheses
H2: Emissions did not decline enough in some areas to affect ozone concentrations
• Applicable in portions of N SJV & C SJV
• NMOC emissions in affected area are ~20 percent of N & C SJV total
38
Plausible Hypotheses (continued)
H5: VOC emission reductions were effective in urban centers and near-downwind areas but not far downwind
• Ambient NMOC reductions are evident,
but spatial patterns of ozone more complex than hypothesis
• Difficult to infer causal link to ozone
39
Plausible Hypotheses (continued)
H6: NOx emission reductions increased ozone in NOx-rich areas but reduced peak ozone at NOx-limited sites
• Higher ozone during non-peak hours is probably related to decreased NO
• Effects of NOx reductions on peak ozone are unclear
40
Plausible Hypotheses (continued)
H7: Changes in VOC/NOx made ozone production more (or less) efficient
• Declines in NMOC/NOx ratios occurred in measurements and the inventory
• Differences between afternoon and morning NMOC/NOx ratios declined - may indicate that air mass ages and aging changed
41
Plausible Hypotheses (continued)
H8: H5, H6, and H7 are all applicable, but differed in importance at different times
• Consistent with data but difficult to demonstrate
42
Preliminary Implications
• Observed ozone trends reflect emission changes and atmospheric chemistry
• Ongoing emission reductions needed – including local reductions in some areas
• NMOC emissions: some areas became more NMOC-sensitive. NMOC mass changes have probably been more important than reductions of reactivity