DETERMINING REGIONAL CARBON EMISSIONS UNDER VARIABLE FIRE REGIMES

1
DETERMINING REGIONAL CARBON EMISSIONS UNDER
VARIABLE FIRE REGIMES IN CENTRAL SIBERIA D.J.
McRae1, S.G. Conard2, G.A. Ivanova3, S.P. Baker4,
A.I. Sukhinin3 , W.M. Hao4, and T.W. Blake1
AGU Fall Meeting 11 December 2007 Paper Number
GC23A-0999
RUSSIAN FIRE BEAR (Fire Effects in the Boreal
Eurasia Region) PROJECT A major goal of this
project is to provide fire data and models
required to estimate carbon emissions from
wildfires in central Siberia. Research is being
carried out in Scots pine (Pinus sylvestris) and
larch (Larix sp.) forest types, which together
make up 58 of the forest area in the Asian
portion of Russia. The overall objective of this
research is to develop validated estimates of
fire areas, fire severity, emissions, and the
impact of fire on carbon balance for key forest
types of central Siberia. This project is
building on our past research efforts in Scots
pine forests (2000-2004), while initiating
similar research in larch forests by

• FACTS
• The Russian boreal forest contains up to 20 of
  the global terrestrial carbon storage.
• Wildfires affect from 10-15 million ha annually
  in typical burning years.
• Fires are projected to increase in both frequency
  and severity across Siberia under climate change.
  
• Fire behavior and impacts are highly variable
  across this landscape. Information on fire
  severity and its effects on factors such as
  emissions, carbon storage, and ecosystem recovery
  are scarce.
• Carbon emissions are roughly 45-50 of the fuel
  consumed during a fire, which can be modeled from
  fire data.

• Quantifying and modeling effects of different
  severity fires on Scots pine sites to estimate
  effects of fire severity on carbon cycle, direct
  fire emissions, and forest dynamics.
• Developing models relating fire behavior, fire
  weather, and fuel condition at the time of
  burning to carbon emissions, energy release, and
  other ecosystem impacts.
• Evaluating the potential for estimating emissions
  directly from satellite infrared channels through
  relationships between the fires energy release
  and its emissions.

Results to date are

• 20 experimental fires have been conducted in
  central Siberia.
• Fire behavior (e.g., fire spread) models have
  been created.
• Carbon emission (i.e., from fuel consumption
  data) models have been created.
• Emissions sampling has quantified the composition
  of carbon and aerosols in the smoke.

Low-intensity surface fire burning in a Siberian
Scots pine forest. Eighty percent of fires in a
typical year burn as surface fires, which in many
cases will not cause any damage to the mature
trees.
APPROACH FOR ESTIMATING EMISSIONS

• PROBLEMS
• Changes in boreal fire regimes can be expected to
  lead to large changes in patterns of burn
  severity, with attendant effects on emissions per
  unit burned area and on postfire vegetation
  recovery. Developing accurate regional to
  continental estimates of carbon emissions from
  wildfires in Siberia requires the collection of
  data and creation of models that will enable us
  to accurately quantify not only the areas that
  are burned annually, but the emissions per unit
  of burned area for fires of widely varying
  characteristics.
• Estimates of fire emissions must account for the
  heterogeneous nature of fire behavior. This is
  due to the constant changes in daily burning
  conditions (e.g., drought), topography, and
  wildland fuel types.
• While remote sensing has been proven in
  estimating burn area, the direct estimate of
  carbon emissions from satellite images may not be
  possible as fuel consumption cannot be
  determined.

Carbon emissions 2.233 (0.554FWI) R2 0.82
15 June 2003
2. Creation of carbon emission models from
experimental fire results which documented fuel
consumption (50 carbon content).
1. Determination of burned area through
remote-sensing analysis. Red polygons indicate
areas that have burned between June 6-16, 2003.
Gray polygons indicate areas that have already
burned in the current fire season.
The ranges of emission factors (EF) for the major
carbon emission products for these fires were
correlated with the FWI for each fire. Higher EF
values for CO2 and modified combustion efficiency
(MCE) occurred with increasing FWI values. At the
same time, EF values for CO and CH4 were observed
to decrease.
Arrows show areas of saturation that
differentiate between crown fires and surface
fires. While satellite images can determine
specific fire types, they cannot determine actual
fuel consumption which is imperative in
estimating carbon emissions.
Satellite images provide an excellent means of
obtaining burned area estimates. Polygon outlines
represent burn areas with red shading showing
locations where active fires are burning.

• TAKE-HOME MESSAGE
• Remote-sensing, by itself, cannot be used to
  determine emissions.
• Carbon emission models have been created for
  Russian Scots pine forest types.
• A Russian fire danger database (1953-2007) has
  been created allowing us to use our model for
  estimating emissions for any year, provided we
  have fire size and location information.

QUESTION How can accurate estimates of carbon
emissions from biomass burning be made to better
understand the contribution of Russian fires to
greenhouse gases?
For more information contact 1 Canadian Forest
Service, 1219 Queen St. E., Sault Ste. Marie, ON.
P6A 2E5 Contact Douglas J. McRae, Telephone
705-541-5539 dmcrae_at_nrcan.gc.ca Tom W.
Blake, Telephone [phone]
tblake_at_nrcan.gc.ca 2 USDA Forest Service, RPC-4,
1601 N. Kent St., Arlington, VA. 22209 Contact
Susan G. Conard, Telephone [phone]
sconard_at_fs.fed.us 3 V.N. Sukachev Institute of
Forest Research, Akademgorodok, Krasnoyarsk
660036 Russia. Contact Galina A. Ivanova,
Telephone [phone] green_at_akadem.ru
Anatoly I. Sukhinin, Telephone
7-3912-49-4462 boss_at_ksc.krasn.ru 4 USDA Forest
Service, Fire Sciences Laboratory, 5775 US
Highway 10 West, Missoula, MT. 59808 Contact
Wei Min Hao, Telephone [phone]
whao_at_fs.fed.us Steve P. Baker, Telephone
406-329-4860 sbaker03_at_fs.fed.us
3. Determination of the Canadian Fire Weather
Index (FWI) at target sites in central Siberia
coinciding with fires shown in Figure 1. Note how
the location of fire hot spots, depicted in
yellow from Figure 1, correlate well with the
high FWI values found in this region. Estimated
ranges in carbon emissions for fires burning at
the different FWI in the figure are based on data
from Figure 2.
NASAs Land Cover Land Use Change Program