Reducing Uncertainties in National Smoke Emissions

Reducing Uncertainties in National Smoke Emissions Modeling as Applied in the

BlueSky Framework

Dana Coe Sullivan, Sonoma Technology, Inc. Petaluma, California

Narasimhan Larkin, USDA Forest Service AirFire Team,

Pacific NW Research Station, Seattle, Washington

For presentation at the Public Health and Air Quality Applications Program Review

Newport, Rhode Island

September 19, 2012

STI-909046-5465

2

What Are the BlueSky Systems? •  BlueSky Framework and SmartFire

–  Decision-support systems (DSS): (1) a modeling framework and (2) a fire information processing and reconciliation system

–  Function: facilitate modeling of the air quality and climate impacts of smoke and fire

•  Examples of decisions supported include –  Air quality management (prescribed burning: burn/no-burn) –  Air quality forecasting (public health: air quality alerts) –  Forest management (prescribed burn mid- to long-range planning) –  National emissions inventory (climate change, National Ambient

Air Quality Standards [NAAQS]) –  And others: research, incident command, interdisciplinary

collaboration with other DSS

The BlueSky Smoke Modeling Framework

3

Fire Info

Fuels

Total Consumption

Time Rate

Emissions

Plume Rise

Dispersion / Trajectories

SmartFire ICS-209 Rx Sys Manual Other FCCS

NFDRS Hardy LANDFIRE Ag FINN FLAMBÉ FBP Other

Consume 3 FOFEM FEPS EPM ClearSky (Ag) FINN FLAMBÉ FBP

Rx / WF FEPS FOFEM EPM WRAP Idealized Manual Other

FEPS Literature EPM FOFEM FINN FLAMBÉ Other

Briggs FEPS Other

CalPuff HYSPLIT CMAQ GEMAQ

Meteorological Input

MM5 WRF NAM NARR Other

BlueSky is not a model itself, but a framework that contains many models. It connects these modeling steps seamlessly.

4

Project Objective: Reduce Key Uncertainties and Information Gaps

Fire Size & Location

Available Fuel

Fuel Consumed

Fuel Moisture

Emission Factors

All of the steps in the modeling process are “big knobs” with substantial uncertainty.

Plume Injection Height

Dispersion

Addressed by previous NASA ROSES (2006–2009) Addressed by current project (new or improved modules) Other modeling steps

ROSES: NASA’s Research Opportunities in Space and Earth Sciences program

5

Overview of NASA Earth Science in BlueSky

MIN

X tool

GroundReports

NEWSMARTFIRE

FIRE INFO(SMARTFIRE)

FUELS

CONSUMPTION

EMISSIONS

DISPERSION/CHEMISTRY

FUELS

EMISSIO

NS

FIRE R

AD

IATIVE POW

ER

PLUME RISE

CURRENT BLUESKY &SMARTFIRE (BSF-SF)

FUEL MOISTURE

FLAM

BÉ

PA

THW

AY

GOES-10 and -12:

GOES NOAA-15,-16, and -17: POES

Terra and Aqua:MODIS Fire Detects

NO

AA

Haz

ard

Map

ping

Sys

tem

TRMM: TMI, PR; DMSP: SSM/I; Aqua: AMSR-E; NOAA: AMSU-BRainfall

Suite of GEO Satellites carrying infrared sensorsRainfall D

ead

Fuel

Moi

stur

eTM

PA

-RT

(3B

42R

T)

data

set

Terra/Aqua: MODISDirect Broadcast

Live

Fue

l M

oist

ure

Eas

tfire

Alg

orith

m

Terra: MISRStereo Heights

Unknown: MISP (Future)Stereo Heights

Plum

e H

eigh

tC

onst

rain

tsM

INIX

dat

aset

GOES-11 and 12: GOES

Western Hemisphere Fire Detects

Terra and Aqua: MODIS

Global Fire Detects

Fire Pixel Location &

Sub-Pixel Fire Param

eterizationW

F-AB

BA

Terra/Aqua: MODISFire Radiative Power

GOES-R: ABI (Future)Fire Radiative Power

NPP: VIIRS (Future)Fire Radiative Power

Fire Radiative

Power

MIN

X dataset

NASA/NOAA Earth science results

PROPOSED COMPONENTS AND CONNECTIONS IN RED

Key

GroundReports

NEWSMARTFIRE

FIRE INFO(SMARTFIRE)

FUELS

CONSUMPTION

EMISSIONS

DISPERSION/CHEMISTRY

FUELS

EMISSIO

NS

FIRE R

AD

IATIVE POW

ER

PLUME RISE

CURRENT BLUESKY &SMARTFIRE (BSF-SF)

FUEL MOISTURE

FLAM

BÉ

PA

THW

AY

GOES-10 and -12:

GOES NOAA-15,-16, and -17: POES

Terra and Aqua:MODIS Fire Detects

NO

AA

Haz

ard

Map

ping

Sys

tem



ead

Fuel

Moi

stur

eTM

PA

-RT

(3B

42R

T)

data

set



ead

Fuel

Moi

stur

eTM

PA

-RT

(3B

42R

T)

data

set

Terra/Aqua: MODISDirect Broadcast

Live

Fue

l M

oist

ure

Eas

tfire

Alg

orith

m

Terra: MISRStereo Heights


Plum

e H

eigh

tC

onst

rain

tsM

INIX

dat

asetTerra: MISR

Stereo Heights


Plum

e H

eigh

tC

onst

rain

tsM

INIX

dat

aset




Global Fire Detects



eterizationW

F-AB

BA




Global Fire Detects



eterizationW

F-AB

BA




Fire Radiative

Power

MIN

X dataset




Fire Radiative

Power

MIN

X dataset

NASA/NOAA Earth science results

PROPOSED COMPONENTS AND CONNECTIONS IN RED

Key

NASA/NOAA earth science results

In red: components and connections currently under development

Unknown: MSPI (Future) Stereo Heights M

INX

tool

TRMM (Tropical Rainfall Measuring Mission) MODIS (Moderate Resolution Imaging Spectroradiometer) MISR (Multi-angle Imaging Spectroradiometer) MINX: MISR Interactive Explorer Follow-on: •  NPP–VIIRS •  GPM

Earth Science Data Sources and Expert Collaborators

•  TRMM Multisatellite Precipitation Analysis (TMPA) –  Dead fuel moisture –  George Huffman (NASA), collaborator

•  EastFIRE live fuel moisture content retrieval data set –  Live fuel moisture –  John Qu and Xianjun Hao (George Mason University),

co-investigators •  MINX

–  Plume injection height –  David Diner (NASA), collaborator

•  Fire Locating and Modeling of Burning Emissions (FLAMBÉ) –  Emission factors –  Ed Hyer (Naval Research Laboratory), collaborator

7

Dead Fuel Moisture

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

50,000

0 20 40 60 80 100

1000-hr Fuel Moisture (%)

CO2

Em

issi

ons

(Kg/

Hec

tare

)

0

50

100

150

200

250

300

350

400

450

PM2

.5 E

mis

sion

s (K

g/H

ecta

re)

CO2 PM2.5 •  Dead fuel moisture content (DFMC) directly influences modeled emissions (i.e., greenhouse gases, fine particulate matter [PM2.5]).

•  BlueSky previously defaulted to static DFMC (absent user-supplied values).

Improvement •  Calculate dynamic DFMC default values

over contiguous United States using daily real-time precipitation data from TMPA real-time data set.

8

Live Fuel Moisture

•  Live fuel moisture content (LFMC) affects fire intensity, rate of spread, and duration, all of which affect pollutant concentrations and transport.

•  BlueSky previously defaulted to the static DFMC for large woody fuels (absent user-supplied LFMC).

Improvement •  Use EastFIRE Lab’s data products to

provide dynamic estimation of LFMC at 1-km spatial resolution.

Example: daily LFMC image for July 4, 2008.

0% >300%

9

Plume Injection Height

•  Plume rise models were designed for industrial sources (e.g., Briggs equation).

•  A high degree of uncertainty arises when applying these models to smoke plumes from fires.

Improvement •  MINX and MODIS data were used to develop a

plume-rise data set (continuous from 2000). That data set was used to develop adjustments to the plume rise models used in BlueSky. Analyses led by Fok-Yan Leung, Washington State University (WSU).

10

Alternative Emissions Models

•  BlueSky lacked fuelbed data outside the United States.

•  Methods to calculate smoke emissions outside the United States will facilitate North American emissions inventories and modeling of transboundary smoke transport.

Image source: http://trendsupdates.com/wp-content/uploads/2009/04/forest-fires-smoke.jpg

Improvements •  Implement parts of the FLAMBÉ program

to estimate emissions globally: •  WF_ABBA (area burned) •  Global fuels database •  FLAMBÉ emissions pathway

•  Incorporate MODIS Fire Radiative Power (FRP) pathway.

Summary Status Assessment

•  Analyses of plume rise and fire radiative power were challenging, but are being wrapped up.

•  Despite delays, project goals are still on track to be met by the end date.

•  Key accomplishment: development of modules. •  New modules developed under current NASA ROSES

project will be integrated with existing BlueSky systems at Technology Readiness Level for Applications (TRL-A) Levels 7–9 by project conclusion:

– Fuel moisture modules (9) ‒ FLAMBÉ emissions model (9) – Plume rise algorithm (9) ‒ FRP emissions (7 or 8)

11

Technology Readiness Level for Applications (TRL-A)

•  BlueSky Framework and SmartFire were at TRL-A Level 9 before project start—i.e., approved, in operational deployment, and in use for decision making (by U.S. Environmental Protection Agency [EPA], USDA Forest Service [USFS], other agencies).

•  The new modules and algorithms for BlueSky, developed under the current ROSES project, began at various TRL-A Levels: –  Fuel moisture modules Level 3: Proof of Concept –  Plume rise algorithm Level 3: Proof of Concept –  FLAMBÉ emissions model Level 6: Demonstration in Relevant

Environment –  FRP emissions model Level 3: Proof of Concept

•  Currently, new modules are at TRL-A levels 5–7 and are projected to be at levels 7 (prototyped), 8 (completed and qualified), or 9 (approved and deployed) by the project end (September 2013).

12


13

0 1000 2000 3000 4000 5000 6000 7000

0 7 14

23

34

37

47

54

64

70

82

95

115

164

231

255

301

417

548

719

796

921

998

1111

1214

1299

1347

1404

1488

1516

1572

1622

1710

1792

1882

2047

2108

2222

2361

2438

2574

Plum

e he

ight ASL (m

eters)

Eleva5on (meters)

WRAP

MISR

0

2000

4000

6000

8000

10000

12000

0 7 14

23

34

37

47

54

64

70

82

95

115

164

231

255

301

417

548

719

796

921

998

1111

1214

1299

1347

1404

1488

1516

1572

1622

1710

1792

1882

2047

2108

2222

2361

2438

2574

Plum

e he

ight ASL

(meters)

Eleva5on (meters)

FEPS

MISR

Quantifiable improvements to modeling results were made. Sample illustration: Comparison of observed (MISR) and plume heights as previously modeled (WRAP or FEPS).


14

Reg. Path FLAMBE Reg. Path FLAMBE2008 14,814,841 14,473,993 98% 2,597,043 5,937,409 2.32009 17,326,747 17,291,675 100% 1,376,974 5,446,343 4.0

YearArea (acres) PM2.5 (tons)

FLAMBÉ Emissions Pathway National Emissions Inventory (NEI) Emissions Pathway

Sample illustration: Results from new emissions pathways vary from prior methods, but within the range of uncertainty that is typical for the present time (a factor of 5 to 10) and with the ability to cover new regions.

NEI NEI Ratio Ratio (%)

Schedule

15

Key next steps: benchmarking and socioeconomic benefits studies.

Budget Status

16

Institution Budget Obligated Unobligated Costed Uncosted

Sonoma Technology, Inc. 54,937 -‐ 54,937 -‐ 54,937 George Mason University -‐ -‐ -‐ -‐ -‐ Washington State University 35,000 -‐ 35,000 -‐ 35,000 Population Research Systems -‐ -‐ -‐ -‐ -‐USFS 21,348 -‐ 21,348 -‐ 21,348


Sonoma Technology, Inc. 271,812 271,812 -‐ 120,516 151,296 George Mason University -‐ -‐ -‐ -‐ -‐ Washington State University 116,745 116,745 -‐ 36,838 79,907 Population Research Systems 90,000 90,000 -‐ -‐ 90,000USFS 83,740 83,740 -‐ -‐ 83,740


Sonoma Technology, Inc. 115,657 115,657 -‐ 115,657 -‐ George Mason University 43,000 43,000 -‐ 43,000 -‐ Washington State University 102,554 102,554 -‐ 102,554 -‐USFS 102,936 102,936 -‐ 90,000 12,936


Sonoma Technology, Inc. 159,563 159,563 -‐ 159,563 -‐ George Mason University 29,000 29,000 -‐ 29,000 -‐ Washington State University 84,633 84,633 -‐ 84,633 -‐USFS 99,045 99,045 -‐ 99,045 -‐

PY12

PY11

PY10

PY09

Thank You Our Co-Investigators: Fok-Yan Leung (WSU); John Qu and Xianjun Hao

(EastFIRE Lab, George Mason University)

Funding from National Fire Plan, USDA Cooperative State Research, Education, and Extension Service’s (CSREES) National Research Initiative, USFS, Joint Fire Science Program, EPA, U.S. Department of the Interior, and NASA.

Our many collaborators and partners, including Mark Ruminski (National Oceanic and Atmospheric Administration’s Hazard Mapping System); Amber Soja (National Institute of Aerospace); Tom Pace (EPA); Pete Lahm (USFS); Susan O'Neill (Natural Resources Conservation Service); and Tweak Software.

Dana Coe Sullivan (707) 665-9900 [email protected]

Narasimhan (‘Sim’) Larkin (206) 732-7849 [email protected]

17

Documents

Reducing Uncertainties in National Smoke Emissions