Impact of Prescribed burning In the Flint Hills...
Transcript of Impact of Prescribed burning In the Flint Hills...
Biological and Agricultural Engineering 1
Impact of Prescribed burning In the Flint Hills Region
Dr. Zifei Liu
May 10th, 2018
2
The
public
K-State,
KDHE
Land
managers
How will smoke
affect me?Why burning
is important?
How to reduce smoke impact?
3
April 16, 2018
4
• The impact of smoke
• The ozone story
• Why do we burn? When is the best time to burn?
• Monitoring smoke using drone
Outline
5
The impact of smoke
6
Source: Kansas Geological Survey
Open-file Report 2016-1
Flint Hills Discovery Center Foundation
Largest remaining intact
tallgrass prairie
7 million acres of rangeland
Approximately 1/3 of the
rangeland are burned each year
21 counties in Kansas and
Oklahoma
7https://firms.modaps.eosdis.nasa.gov/map
Satellite fires, 4/1 to 5/1, 2017
National Aeronautics and Space Administration (NASA)
Moderate Resolution Imaging Spectrometer (MODIS) fire products
Fire Information for Resource Management System (FIRMS)
8
Acres burned
0
1
2
3
4
Acr
es b
urn
ed (
Mil
lion
)Average: 2.1 Million acres
Median: 2.5 Million acres
(Mohler and Goodin, 2012)
Smoke is mainly a product of incomplete combustion
9
• Unburned fuel or fuel
degradation products such
as aromatic hydrocarbon
Fuel
(C/H/O)O2
• Partially oxidized products
such as CO
CO2 and H2O
+ Incomplete
combustion
Complete
combustion
Organic
irritants
Toxic
Form O3 in the presence of sunlight
Complete combustion
70-90% PM2.5
10
Smoke and air quality
Combustion
CO NOx
PM
H2O SVOC
CO2
VOC
11
Air pollutantsObserved concentrations in literature NAAQS
24-hr
standardsAt the fires At downwind communities
PM2.5 148-6865 μg/m3 63-400 μg/m3 35 μg/m3
The Tallgrass monitoring site
12
Biological and Agricultural Engineering
Five PM2.5 source categories at the Tallgrass site
(Results from Unmix receptor modeling)
13
11%
0
2
4
6
8
10
12
14
Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
PM
2.5
(µ
g/m
3)
U5:Secondary organic particles
U4:Primary smoke particles
U3:Crustal/soil
U2:Sulfate/industrial
U1:Nitrate/agricultural
11%
49%
0
2
4
6
8
10
12
14
Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
PM
2.5
(µ
g/m
3)
U5:Secondary organic particles
U4:Primary smoke particles
U3:Crustal/soil
U2:Sulfate/industrial
U1:Nitrate/agricultural
Five PM2.5 source categories at the Tallgrass site
(Results from PMF receptor modeling)
14
41%
0
2
4
6
8
10
12
Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec.
PM
2.5
(µg
/m3)
P5:Traffic/SOA
P4:Smoke
P3:Crustal/soil
P2:Sulfate/industrial
P1:Nitrate/agricultural
15
U4-Primary smoke particles
16
U5-Secondary organic particles
17
0
5
10
15
20
25
30
35
40
45
50
PM
2.5
(µ
g/m
3)
Daily PM2.5 at the Tallgrass site
24hr PM2.5 standard
18
Highest daily PM2.5 in April vs. acres burned(Tallgrass site, 2002-2014)
R² = 0.1744
0
10
20
30
40
50
60
70
80
0 0.5 1 1.5 2 2.5 3 3.5 4
PM2.5
(μg/m3)
Acres burned (Million)
2010
2014 2003
2005
2009
20082011
24hr PM2.5 standard
19
Smoke PM2.5 at the Tallgrass site and the Kansas City site in April (in µg/m3)
0
1
2
3
4
5
6
7N
N
E
E
S
E
S
S
W
W
N
W
FH PMF smoke(FH-P4)
KC PMF smoke(KC-P4)
Tallgrass site
Kansas City site
In April, on average,
smoke contributed
4.4 g/m3 at Tallgrass site, and
1.1 g/m3 at Kansas City site.
20
• Size near the wavelength of
visible light (0.4-0.7µm) can
efficiently scatter light and
reduce visibility.
• Can reach deeper into
human respiration system
Human hair (50-70μm)
PM10
(10μm)
PM2.5
(2.5μm)
21
OC ~60-70%
EC ~2-20%
Inorganic ash
~10-30%
Composition of smoke PM
Smoke PM
could be more
damaging to
human health
than normal
urban particles.
22
Observed VOC/SVOC in literature NIOSH or
OSHA 8-hr
exposure
limitsAt the fires At downwind communities
Acrolein 0.018-0.071ppm 0.009 ppm 0.1 ppm
Formaldehyde 0.03-0.47 ppm 0.02-0.047 ppm 0.016 ppm
Isocyanic acid - 600 ppb -
PAHs - - 200 μg/m3
• BaP 0.10-0.16 μg/m3 0.007 μg/m3 -
• Acenaphthene 0.57-1.53 μg/m3 0.83-0.89 μg/m3 -
• Naphthalene 0-3.27 μg/m3 0-3.53 μg/m3 -
• Phenanthrene 0.38 μg/m3 - -
23
Air pollutantsObserved mixing ratios in literature NAAQS
8-hr
standardsAt the fires At downwind communities
CO 1-140 ppm 1-6 ppm 9 ppm
24
Air pollutantsObserved mixing ratios in literature NAAQS
8-hr
standardsAt the fires At downwind communities
O3 - Up by 50 ppb 70 ppb
• Strong oxidant that damages cells lining the respiratory system
• Cough
• Sore or scratchy throat
• Pain with deep breath, or chest pain
• Fatigue
• Aggravation of lung diseases
Formation of O3
25
(VOCs NOx)
O3
Existing
pollutants in airSmoke
(VOCs, NOx)
Evolution of O3 standards
26
8-hour 1-hour
1979 120ppb
1997 80ppb
2008 75ppb
2015 70ppb
55
60
65
70
75
80
02-04
Average
03-05
Average
04-06
Average
05-07
Average
06-08
Average
08-10
Average
09-11
Average
10-12
Average
11-13
Average
12-14
Average
Kansas City JFK Heritage park Leavanworth
Wichita HD Peck Sedgwick
Topeka KNI Cedar Bluff
Kansas 8hr O3 design values (ppb)
27(Data from KDHE, 2010 and 2015. 5-Year Ambient Air Monitoring Network Assessment )
New standard
Old standard
Consequences of nonattainment
28
• Enhanced emissions
inventory
• Photochemical
modeling
• Economic development
curtailed
State Implementation
Plan (SIP)Transportation
conformityContingency
measures
• Potential for loss of
highway funds and
restrictions on how
highway funds can
be spent.
• Expanded burning
restrictions
29
The O3 story
30
CO, VOC, Aerosol,
Carbon cycle
O3
Climate
Temperature
Humidity
Dynamics
RF
(UV+IR)L
igh
tenin
g
Precip
itation
Tem
peratu
re
CO2
CH4
BC
RF
N2O
RF
Biomass burning
NOx,
Nitrogen cycle
Interaction of fire, atmospheric composition and climate in the earth system
(RF: radiative forcing)
Nine O3 monitoring sites
31
32
0
10
20
30
40
50
60
70
80
90
1/1/2002 1/1/2003 1/1/2004 1/1/2005 1/1/2006 1/1/2007 1/1/2008 1/1/2009 1/1/2010 1/1/2011 1/1/2012 1/1/2013
O3
(pp
b)
8hr O3 standard
Daily max 8hr O3 at the Konza Praire site (2002-2013)
33
O3(d) = c1+c2 sin2𝜋 𝑑+283
365
+c3 [O3 on the previous day]
+c4 [Air temperature]
+c5 [Solar radiation]
-c6 [Relative humidity]
-c7 [Wind speed]
O3 regression models
R2=0.66 to 0.76
34
8hr O3 standard
Daily max 8hr O3 at the Konza Praire site (2002-2013)
35
8hr O3 standard
Daily max 8hr O3 at the Wichita Health Department site (2001-2016)
36
Multi-year average O3 at different sites
(non-rainy days in April )
Measured O3 (ppb)-----------------------------------------------------------------------------------
Modeled O3 (ppb)
53.2------------------
50.1
48.9------------------
47.2
43.9------------------
43.550.5------------------
48.147.7
------------------
47.0
46.2------------------
45.2
47.6------------------
46.4
49.6------------------
48.3
Highest model R2 at
Cedar Bluff, lowest
burning interference,
R2=0.76
48.3------------------
47.6
49.2------------------
47.7
Average O3 model residuals in April (non-rainy days) (Likely due to burning)
37
Konza Prairie site 3.1 ppb
Topeka site 2.4 ppb
Three Wichita sites 0.7, 1.3, 1.7 ppb
Three Kansas City sites 0.4, 1.2, 1.5 ppb
Cedar Bluff site 0.7 ppb
Biological and Agricultural Engineering
2001-2016
38
Average of days with
O3>70ppb in April
(47 days in total)April average
Daily Max 8hr O3 77±5 ppb 43.9-53.2 ppb
O3 on the previous day 60±11 ppb -
Daily maximum air temperature 24.5±4.5 ºC 20.7±5.5 ºC
Tmax-Tmin 16.6±5.3 ºC 12.3±5.0 ºC
Solar radiation 738±279 Langley 607±304 Langley
Relative humidity 54±10 % 67±14 %
Wind speed 3.4±1.8 m/s 4.1±2.0 m/s
O3 model residuals 21±9 ppb -
Likely due to burning
39
Acres burned vs. highest 8hr O3 in April
30
40
50
60
70
80
90
100
110
0
1
2
3
4
O3
(ppb)
Acres
burned
(Million)
40
Acres burned vs. highest 8hr O3 in April
y = 9.4711x + 58.589
R² = 0.6283
50
60
70
80
90
100
110
0 0.5 1 1.5 2 2.5 3 3.5 4
O3
(ppb)
Acres burned (Million)
2013
2012
2007
2002
2004
2006
2001
2015 2010
2014 2011
2016
2008
2003
2009
2005
2017
For every one million increase of burn acres, the highest 8-hour
O3 mixing ratios increased around 9 ppb.
41
Acres burned vs. # of days with 8hr O3 >70ppb in April
R² = 0.4323
0
1
2
3
4
5
6
7
8
0 1 2 3 4
# of days with
O3 >70ppb
Acres burned (Million)
2013 2012
2007
2002 2015
2001, 2004
2006
2005
2010 2008
2011
2014
2009
2016
2017
When the acres burned were larger than or equal to 1.9 million,
O3>70 ppb occurred in April at least at one of the ten monitoring sites.
42
Fire count vs. highest 8hr O3 in April
R² = 0.7067
50
60
70
80
90
100
110
0 1000 2000 3000 4000 5000 6000
O3
(ppb)
Fire count
2016
2013
2002
2007
2012
2001 2003
2004
2005
2006
2008
2009
2010
2011
2014
2015
2017
30
50
70
90
110
0
1
2
3
4
O3
(ppb)
Acres
burned
(Million)
43
2/15-5/15, 2002 2/15-5/15, 2003 2/15-5/15, 2004
2/15-5/15, 2005 2/15-5/15, 2006 2/15-5/15, 2007
Biological and Agricultural Engineering 44
2/15-5/15, 2008 2/15-5/15, 2009 2/15-5/15, 2010
2/15-5/15, 2011 2/15-5/15, 2012 2/15-5/15, 2013
30
50
70
90
110
0
1
2
3
4
O3
(ppb)
Acres
burned
(Million)
Biological and Agricultural Engineering 45
April 10, 2015 April 3-4, 2016 April 13, 2016
Sedgwick O3: 78ppb
Wichita O3: 103ppb Topeka O3: 64ppbChanute O3: 77ppb
2/15-5/15, 2014 2/15-5/15, 2015 2/15-5/15, 2016
30
50
70
90
110
0
1
2
3
4
O3
(ppb)
Acres
burned
(Million)
46
Daily fire count vs. highest 8hr O3 in April
R² = 0.2968
20
30
40
50
60
70
80
90
100
110
0 375 750 1125
O3
(ppb)
Daily fire count
• When daily fire count >750, O3>70 ppb occurred in all the 7 days.
• When daily fire count is between 375 to 750, O3>70 ppb occurred in 15 out of 27
days (56%).
• When daily fire count is between 1 to 375, O3>70 ppb occurred in 27 out of 347
days (8%).
• When daily fire count is 0, O3>70 ppb did not occur.
47
y = 1480.2x
R² = 0.7472
0
1500
3000
4500
6000
0 1 2 3 4
Fire
count
Acres burned (Million)
• When daily burned acres >0.5M, O3>70 ppb is most likely to occur.
• When daily burned acres is between 0.25M to 0.5M, the chance of
O3>70 ppb is 56%.
• When daily burned acres <0.25M, the chance of O3>70 ppb is 8%.
• When there is no burning, O3>70 ppb is not likely to occur.
1500 fire count
≈ 1 million acres burned
48
Daily fire count vs. highest 8hr O3 in April
R² = 0.2968
20
30
40
50
60
70
80
90
100
110
0 375 750 1125
O3
(ppb)
Daily fire count
• When daily burned acres >0.5M, O3>70 ppb is most likely to occur.
• When daily burned acres is between 0.25M to 0.5M, the chance of
O3>70 ppb is 56%.
• When daily burned acres <0.25M, the chance of O3>70 ppb is 8%.
• When there is no burning, O3>70 ppb is not likely to occur.
49
When daily fire count is between 375 to 750,
or daily burned acres is between 0.25M to 0.5M
Average
(7 days)April average
Daily maximum air
temperature22.5±5.0 ºC 20.7±5.5 ºC
Solar radiation 638±218 Langley 607±304 Langley
Relative humidity 51±6 % 67±14 %
Wind speed 2.0±0.7 m/s 4.1±2.0 m/s
50
When daily fire count >750,
or daily burned acres >0.5M
Average of days with
O3>70ppb
(15 days)
Average of days with
O3<70ppb
(12 days)
April average
Daily maximum
air temperature24.4±5.4 ºC 19.2±4.1 ºC 20.7±5.5 ºC
Solar radiation 697±244 Langley 596±98 Langley 607±304 Langley
Relative humidity 54±10 % 54±12 % 67±14 %
Wind speed 2.4±1.1 m/s 2.9±1.2 m/s 4.1±2.0 m/s
51
When daily fire count is between 1 and 375,
or daily burned acres < 0.25M
Average of days with
O3>70ppb
(27 days)
Average of days with
O3<70ppb
(320 days)
April average
Daily maximum
air temperature24.8±3.8 ºC 20.4±6.5 ºC 20.7±5.5 ºC
Solar radiation 705±268 Langley 610±283 Langley 607±304 Langley
Relative humidity 58±11 % 62±13 % 67±14 %
Wind speed 4.0±1.9 m/s 4.2±2.0 m/s 4.1±2.0 m/s
50 52 54 56 58 60
52
Modeling O3 (ppb) in April
Fire count 82±166
O3 on the previous day 55±11 ppb
Daily maximum air
temperature20.7±5.5 ºC
Solar radiation 607±304 Langley
Wind speed 4.1±2.0 m/s
Relative humidity 67±14 %
R2=0.55
April average
53
When seasonal burned acres is <1.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)Fire
count
2012 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)Fire
count
2013 Fire count Ozone
54
When seasonal burned acres is <1.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)Fire
count
2002 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)Fire
count
2007 Fire count Ozone
55
When seasonal burned acres is between 1.5 and 2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2001 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2004 Fire count Ozone
56
When seasonal burned acres is between 1.5 and 2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2006 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2015 Fire count Ozone
57
When seasonal burned acres is between 1.5 and 2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2010 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2017 Fire count Ozone
58
When seasonal burned acres >2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2003 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2005 Fire count Ozone
59
When seasonal burned acres >2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2008 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2009 Fire count Ozone
60
When seasonal burned acres >2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2011 Fire count Ozone
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2014 Fire count Ozone
61
When seasonal burned acres >2.5 million
50
60
70
80
90
100
0
100
200
300
400
500
O3
(ppb)
Fire
count
2016 Fire count Ozone
62
Highest 8hr O3 in April vs. acres burned at the three sites around Wichita
Wichita Health
Department site
(2001-2017)
Sedgwick site
(2009-2017)
Peck site
(2001-2017)
y = 7.468x + 54.12
R² = 0.5596
50
70
90
110
0 1 2 3 4
O3
(ppb)
Acres burned (Million)
y = 8.0796x + 56.15
R² = 0.3571
50
70
90
110
0 1 2 3 4
O3
(ppb)
Acres burned (Million)
y = 5.6835x + 55.387
R² = 0.2912
50
70
90
110
0 1 2 3 4
O3
(ppb)
Acres burned (Million)
63
Highest 8hr O3 in April vs. acres burned at the three sites around Kansas City
Kansas City
JFK site
(2001-2017)
Leavenworth site
(2004-2017)
Heritage Park site
(2004-2017)
y = 1.211x + 60.766
R² = 0.032
50
70
90
110
0 1 2 3 4
O3
(ppb)
Acres burned (Million)
y = 2.4717x + 60.959
R² = 0.1023
50
70
90
110
0 1 2 3 4
O3
(ppb)
Acres burned (Million)
y = 1.7192x + 62.581
R² = 0.0734
50
70
90
110
0 1 2 3 4
O3
(ppb)
Acres burned (Million)
64
Acres
burned
(Million)
Before the SMP was implemented After the SMP was implemented
Year
# of days
with 8hr
O3>70 ppb
Highest
8hr O3
Year
# of days
with 8hr
O3>70 ppb
Highest
8hr O3
<1.92002
2007
0
1
64
73
2012
2013
0
0
64
62
1.9-2.0
2001
2004
2006
2
2
4
76
74
78
2015 1 77
2.4-2.5 2010 3 822014
2017
2
0
85
70
2.7-2.8
2003
2008
2011
2
1
4
76
74
84
2016 1 103
3.2-3.52005
2009
3
3
77
95
O3 levels in April before and after the Flint Hills Smoke Management Plan (SMP)
(based on measurements from the nine regulatory monitoring sites in Kansas)
65
Why do we burn?When is the best time to burn?
66
Management objectives
Timing of burns
Fire safety
• Forage production
• Plant species composition
• Soil moisture
• Cattle weight gain
• Relative humidity
• Wind speed
• Wind direction
• Continuous burn window
67(Towne and Craine, 2016)
Does burning have to be restricted to a narrow window in April?
• Forage production (McMurphy and Anderson, 1963; Owensby and
Anderson, 1967); little statistical basis
• Plant composition Burn earlier in the spring increases cool-season
grasses and has little impact on warm-season grasses
• Woody species No scientific data indicated April burning is superior;
frequent burning is most important
• Soil moisture
• Animal performance Only one problematic study (Anderson et al., 1970);
potentially misleading
Much of the data are equivocal
68
How many days are actuallyavailable to conduct prescribed burns?
• 8am to 6pm
• T=1.7 to 43.3ºC
• RH=25 to 80%
• V=1.07 to 4.02 m/s
• No rain
• 3-hour continuous burn
window
Based on Oklahoma Mesonet hourly weather data (Weir, 2011)
Burn day
69
• A 20-year study that looks at the consequences of
burning Flint Hills prairie at different times of the year.
• It finds that burning outside of the current late spring
time frame has no measurable negative consequences
for the prairie and, in fact, may have multiple benefits.
Burning outside the traditional burn season?
(Towne and Craine, 2014)
70
• Although early studies considered any increase in cool-
season grasses as undesirable, that response could
actually be beneficial.
Burning outside the traditional burn season?
(Towne and Craine, 2016)
• A mix of fire seasons, may be necessary to maintain
prairie diversity.
(USDA, 2009)
• Most land management objectives can be achieved
with growing-season burns.
(Weir et al., 2011)
The goal
71
• Keep prescribed burning, but burn in
a manner that minimize adverse
environmental and social effects.
Objectives • To avoid exceedances of the NAAQS.
• To receive an exemption/flag in the
event of an exceedance of the
NAAQS (Exceptional Event).
NAAQS: National Ambient Air Quality Standards
Reduce impact of smokeReduce smoke production
• Frequency of burns
• Managing fuel load and
fuel moistures
• Ignition and burn technique
• Reduce smoldering
• Timing of burns
– Allow for adequate smoke
dispersion
– Minimize exposure of
sensitive populations
– Avoid high O3 day
Two strategies
72
Management objectives
Timing of burns
Fire safety
Smoke dispersion
Sensitive populations
Avoid high O3
Smoke modeling tools
? 73
74
Monitoring smoke using drone
75
• Define sampling packages and standard operating
procedures that can be used on unmanned aircraft
systems (UAS) for the collection of smoke emission and
meteorological data.
• Collect smoke emissions data to develop smoke
emission factors that best represent the Flint Hills fires
and compare to the current FCCS fuelbed#131 to
determine if the current defaults are representative.
• Collect black carbon and thermal image data to ground-
verify the models NASA have created from MODIS and
VHRRS satellite data.
Objectives of the test
76
The sampling packagesDrone: K-State Polytech DJI S1000 multi-rotor (8 rotors) aircraft
Drone#1: Continuous measurements
O3: POM, 2B Tech.;
PM2.5 : MIE pDR-1500, Thermo Sci.
Drone#2: Integrated measurements
Sampling bags for NO/NOx/NO2, CO/CO2, and VOCs
PM: Air-O-Cell impactor
Drone#3: MA200 black carbon
monitor; Thermal sensor
77
Around 180 acres burned for the test
on April 16, 2018
78
79
Preliminary results