Estimation of daily CO, NOx, Black Carbon, and Organic Carbon emission from biomass burning using AVHRR data

Gregory R. Carmichael (gcarmich@engineering.uiowa.edu)

During TRACE-P intensive period, we estimated emissions from various sources in support of chemical transport modeling and flight planning. To provide an estimation of temporal variability, we included intermittent sources like forest fires as an on-line source, i.e., we used remote-sensed (RS) data to "spot" fires and then calculate the emissions from that. We produced daily CO, NOx, Black Carbon(BC), and Organic Carbon(OC) emissions from fire count AVHRR data provided from the World Fire Web(WFW),  and TOMS AI data.(TOMS Aerosol homepage, http://toms.gsfc.nasa.gov/aerosols/aerosols.html)

Emissions from biomass burning were estimated only for field combustion – burning of forest, savanna, and agricultural residues. We called the burning of forest and savanna as out-field combustion, and burning of agricultural residues as in-field combustion.

 The methodology we used to estimate daily emission amount from field combustion consisted of :

 1)      Estimation of total emission amount from field combustion - For in-field combustion amount, we used our emission database which  came from Argonne National Laboratory (Dr. David Streets)  For out-field and background emission, we used the result of Galanter et al.(2000) research.

2)      Finding spatial allocation factors - For only the regions which fires actively occurred from March to April (Indian sub-continent and Southeast Asia), daily fire emission was analyzed. The data from WFW originally from NOAA AVHRR, was used for spatial and temporal (daily) allocation factor.  TOMS-Aerosol Index from NASA was used to reduce both cloud interference and deficiency of satellite coverage.

3)      Estimation of emission factors – Emission factors for each species were estimated by dividing total emission amount by total fire count. We used a 3-year average(1999, 2000, and 2001), AI adjusted fire count data which was derived from previous step 2) for 2-month period.

4)      Estimation of daily emission  - Daily emission of each species were derived by multiplying emission factors by AI-adjusted fire count of each day.

During the intensive field activity, we produced original fire count image(12kb/day), TOMS AI image(20kb/day), non-adjusted fire count image(75kb/day), adjusted fire count image with landcover(75kb/day),  emission time series image(12kb/day), fire count time series image by landcover(12kb/day) and daily emission data of each species(150kb/day/species) to the internet website URL, (http://www.cgrer.uiowa.edu/ACESS/EMISSION_DATA/ED_index.htm). They were usually downloaded around 6 am in field time and about 10minite required for processing data.