High Resolution Fire Emission Estimates

1
High Resolution Fire Emission Estimates

• Christine Wiedinmyer
• Atmospheric Chemistry Division/
• National Center for Atmospheric Research

2
Fire Emissions Development

• Timing/location
• Area Burned
• Biomass (fuel) burned
• Emission Factors
• Chemical
• physical

3
Many efforts to develop fire emissions
inventories models

• Global, regional, local
• Various resolutions
• Various species
• Driven by different data (remote
  sensing/reports/both)
• I try here to summarize some of the major efforts
  that of which I am aware

4
Local/Regional

• U.S. NEI
• 2002 inventory, detailed, developed by Air
  Resources, Inc. (Golden)
• Emissions are included in the national EI, but
  unclear how emissions are spatially or temporally
  resolved.
• STI/USFS
• The USFS AirFire Team and Sonoma Technology, Inc.
  have developed the SMARTFIRE fire information
  system, which, in conjunction with the BlueSky
  Framework (http//blueskyframework.org), has been
  used by the EPA to develop wildland fire
  emissions for the National Emissions Inventory
  for recent years
• Combines HMS data with local reports
• Working with WRAP FETS
• Local Agencies
• E.g., CARB Emissions for California
• David Lavoue, Canada
• Detailed Canadian fire emissions
• Working on expanding to N. America including
  CONUS and AK

5
Continental/North Central America

• Wiedinmyer/NCAR model
• Daily, 1km2 resolution
• North and Central America 2001-present
• Global 2007-present
• Particles/ trace gases, working on speciation
  profiles for modelers
• USFS Emissions (Wei Min Hao/Shawn Urbanski)
• NASA NACP Project
• 1 km2 resolution
• CONUS / Canada/ North America
• The project currently has completed the mapping
  of burned areas at a 1 km 1 km resolution from
  2002 to 2008 and has produced the daily emission
  rates in 2007

6
Continental/Global Efforts

• Global Fire Emissions Database
• Guido van der Werf and Jim Randerson
• 8-day averages, 1 degree resolution
• Based on MODIS fire counts (climate model
  resolution)
• 1997-2008
• NASA Effort (Arlindo da Silva)
• QFED
• near-real time daily biomass burning emissions to
  support GEOS-5 aerosol forecasting
• Up to 0.25 degree resolution

7
NOAA/NESDIS

• WF_ABBA
• Satellite detections from geostationary
  satellites
• Global coverage coming soon (B. Pierce, NOAA)
• Emissions modeling by Shobha Kondragunta and
  Xiaoyang Zhang
• Operational for forecasting applications
• Biomass burning emissions are produced once a day
  from GOES-E. The output parameters include
  hourly emissions in PM2.5, CO, CO2, CH4, N2O,
  NH3, NOX, SO2, TNMHC, separately.
• The emission data are available at
  http//satepsanone.nesdis.noaa.gov/pub/EPA/GBBEP
• The product from GOES-W is available at
  http//satepsanone.nesdis.noaa.gov/pub/EPA/GBBEP_W
  /
• A product is running internally every day to
  create emissions from MODIS, GOES and AVHRR. It
  provides hourly PM2.5, hourly burned area, and
  hourly FRP. The data cover CONUS originally and
  entire NA currently.
• Hourly Global biomass burning emissions (PM2.5)
  is produced using fire radiative power from
  GOES-W and GOES-E (GOES 10, GOES 11, and GOES
  12), MET09 and MTS01.

8
Issues

• Political boundary issues
• E.g., fuel loadings, reported information
• Area and fuel burned estimation processes
• Reports versus remotely sensed fire detections
• Fire detections versus Fire Radiative Power to
  drive emissions
• Emission factors
• Plume Rise

9
Inter-comparison of methods
Area Burned
CO Emission
J. Al-Saadi, 2008
10
SPECIFICS
11
CARB Emissions

• The California Air Resources Board uses a version
  of the federal First Order Fire Effects Model
  (FOFEM) customized to California and implemented
  in ArcGIS, to generate geospatially and
  temporally explicit inventories of criteria
  pollutant emissions from wild and prescribed
  fires in the state. The model uses GIS-based fire
  perimeters from a statewide fire geodatabase
  maintained by the Fire and Resource Assessment
  Program (FRAP) of the California Dept. of
  Forestry and Fire Protection (CAL FIRE).  Fire
  perimeter GIS data are developed from field GPS
  data or from digitized aerial imagery.  Rasters
  of thousand-hour fuel moistures (NFDR-TH) are
  also used by the model.  NFDR-TH rasters are from
  the USDA-FS WFAS.  For photochemical modeling,
  modeled emissions are apportioned by hour of the
  day and vertically into the boundary layer using
  a scheme developed by Dave Sandberg (USDA-FS PNW
  Research Station) for the Fire Emissions Joint
  Forum of the Western Regional Air Partnership
  (FEJF-WRAP).  The fire emission inventories are
  fed into regional photochemical models used to
  simulate ozone and particulate matter statewide. 
  CARB plans to also use the model to develop the
  statewide forest and rangeland greenhouse gas
  emission inventory.  Updates to the model are
  also underway via a contract with Dr. Peng Gong
  of UC Berkeley.

12
STI/USFS

• The USFS AirFire Team and Sonoma Technology, Inc.
  have developed the SMARTFIRE fire information
  system, which, in conjunction with the BlueSky
  Framework (http//blueskyframework.org), has been
  used by the EPA to develop wildland fire
  emissions for the National Emissions Inventory
  for recent years. SMARTFIRE currently uses data
  from both the NOAA Hazard Mapping System (HMS)
  fire detects and the Incident Command System
  (ICS) 209 ground based reports. BlueSky can use a
  number of different fuel loading maps and
  consumption models, and various model pathways
  have been used, but the current standard is the
  FCCS fuel loading maps combined with the CONSUME3
  consumption model. Significant upgrades to both
  SMARTFIRE and BlueSky are ongoing in
  collaboration with the WRAP, RSAC, CIRA, NRL, and
  UCAR. Near-term work for SMARTFIRE is being done
  to enable the inclusion of additional datasets
  that can be reconciled together to avoid double
  (or multiple) counting of the same fire. Datasets
  identified for inclusion first are the prescribed
  fire data from the WRAP's FETS system in
  conjunction with the WRAP and AirSciences, and
  the MBTS burn scar data from the USFS RSAC. A
  pilot study to examine how best to combine these
  datasets has been planned as collaborative
  project between the USFS, WRAP, STI, AirSciences,
  and CIRA. BlueSky improvements underway include
  the addition modules for satellite emissions
  calculation systems - specifically the system
  developed by Weidinmyer at UCAR and the FLAMBE
  system developed by the NRL Monterey, and other
  satellite data being incorporated for fuel
  moisture, plume rise, and other framework pieces.
  Comparison studies between various emissions
  calculation pathways (including the 12 currently
  in BlueSky and ones being added above) are
  currently being done under the Smoke and
  Emissions Model Intercomparison Project (SEMIP,
  http//semip.org), a JFSP project that has
  identified the 2008 NEI as a specific test case
  for comparison. Comparison results are currently
  available online at http//data.semip.org.

13
WRAP

• The WRAP's Fire Emissions Tracking System (FETS)
  is a web-enabled database for planned and
  unplanned fire events in the Western U.S.  It is
  intended as a planning tool for daily smoke
  management coordination, and retrospective
  analyses such as emission inventories and
  regional haze air quality planning tasks. 
  Current data status by fire type and smoke
  management program for activity reporting and
  subsequent emissions estimation are detailed in
  the attached FETS Data Status document, which is
  updated daily.    Current FETS development work
  through 2010 will support routine daily FETS
  operations and add reporting of fire activity
  data for missing planned fire source categories
  (Prescribed Wildland, Agricultural, and Rangeland
  fire types) for Western smoke management programs
  not currently reporting to FETS.  A proposal is
  in development for the EPA Emissions Inventory
  Group to expand FETS-style reporting for other
  state and smoke management programs across the
  U.S., to assist in review, quality assurance, and
  completeness of fire emissions reported in the
  EPA National Emissions Inventory system. 
  Development of a top-down, first principles
  estimate of Residential Wood Combustion emissions
  across the West is under consideration to
  supplement the other biomass burning categories
  already tracked in FETS.  WRAP-Air Sciences will
  be providing FETS applications and analyzing data
  as an integral part of a pilot study to examine
  how best to combine datasets from USFS-STI
  SmartFire/BlueSky Framework and USFS RSAC with
  FETS data, through a collaborative project
  between the USFS, WRAP, STI, AirSciences, EPA,
  and NPS.  A key feature of the collaborative
  pilot project is to try to identify optimum
  combinations of data available and processing
  effort needed to produce high-quality datasets in
  3 time frames 1) short-term forecasting, 2)
  medium-term emissions inventory
  reporting/tracking, and 3) longer-term
  retrospective air quality planning datasets.   

14
NASA (Arlindo Da Salva/Peter Colarco)

• We have been developing near-real time daily
  biomass burning emissions in support of our
  GEOS-5 aerosol forecasting system, the so-called
  Quick Fire Emission Dataset (QFED). These
  emissions are based on the MODIS fire products
  from both AQUA and TERRA satellites. The current
  algorithm (QFED v1) uses fire counts with
  constant emission factors calibrated against GFED
  v2.  The new top-down algorithm being evaluated
  (QFED v2) uses MODIS Fire Radiative Power and
  will be attached to a plume rise model. The
  GEOS-5 aerosol forecasting system has a nominal
  resolution of 0.25x1/3 deg, globally. QFED v2
  will match this resolution.    

15
Hao/Urbanski (USFS)

• The fire and emission dataset used in the
  WRF-CHEM model is being developed as an extension
  of Dr. Hao?s ongoing NASA North American Carbon
  Program (NACP) project, Daily, Weekly, Seasonal
  and Interannual Variability of CO2, CO and CH4
  Emissions from Biomass Burning in North America
  and Their Impact on Atmospheric Chemical
  Composition. The final dataset will include the
  daily burned areas (Li et al., 2004), and
  associated emissions of CO2, CO, CH4, volatile
  organic compounds, PM2.5, and black carbon and
  organic carbon of aerosols (Urbanski et al.,
  2008, 2009) at a 1 km 1 km resolution for the
  continental U.S. from 2002 to 2010. It is the
  most comprehensive high-resolution dataset for
  the spatial and temporal distribution of biomass
  burning and associated emissions in the U.S. The
  project currently has completed the mapping of
  burned areas at a 1 km 1 km resolution from
  2002 to 2008 and has produced the daily emission
  rates in 2007.

16
NOAA/NESDIS

• The biomass burning emissions in the North
  America are operational products in NOAA/NESDIS.
  They are publicly accessible via the URL that I
  provided in the slides. The emissions were
  calculated from 2001. The global estimates are
  hourly with a spatial resolution of 4km.

17
NOAA/NESDIS
Operational Products of Biomass Burning Emissions
Derived From Satellite-based Active Fires
Shobha Kondragunta and Xiaoyang Zhang
18
Burned Area Simulated from Active Fires in 2005
Hourly burned areas are produced from active fire
observations using diurnal pattern in near real
time.
19
Fuel Loadings Estimated from MODIS Vegetation
Properties (1KM)

• Forest foliage
• forest branch
• shrub
• grass
• litter
• coarse woody detritus

(ton/ha)
20
Operational product of Biomass Burning Emissions
from GOES-E

• Biomass burning emissions are produced once a
  day. The output parameters include
• hourly emissions in PM2.5, CO, CO2, CH4,
  N2O, NH3, NOX, SO2, TNMHC, separately.
• The emission data are available at
  http//satepsanone.nesdis.noaa.gov/pub/EPA/GBBEP/

21
Operational Biomass Burning Emissions from GOES-W

• Biomass burning emissions derived from GOES-W
  cover west part of North America. The output has
  the same format as GOES-E emissions
• The product is available at http//satepsanone.ne
  sdis.noaa.gov/pub/EPA/GBBEP_W/

22
Emissions from MODISAVHRRGOES
This product is running internally every day. It
provides hourly PM2.5, hourly burned area, and
hourly FRP. The data cover CONUS originally and
entire NA currently.
23
Global PM2.5 Emissions from Fires in September
15, 2009
Hourly Global biomass burning emissions (PM2.5)
is produced using fire radiative power from
GOES-W and GOES-E (GOES 10, GOES 11, and GOES
12), MET09 and MTS01.