1995

g8f95dddhhhh.v56        ASCII file containing the location, size, temperature, 
                        ecosystem type, and fire flag for each GOES-8 ABBA 
                        (version 5.6) fire pixel identified in GOES-8 imagery
                        for a specific time (hhhh) and day (ddd - julian day) 
                        during the 1995 fire season (June - October 1995). 

g8f95ddd.v56.gif        This is a gif of the daily diurnal fire product showing
                        fire locations for the 1145, 1445, 1745, and 2045 UTC 
                        time periods.

g8year95.v56.gif        This composite gif file shows the locations of all fires 
                        detected with the GOES-8 ABBA (version 5.6) during the
                        1995 fire season.



1997

g8f97dddhhhh.v56        ASCII file containing the location, size, temperature,
                        ecosystem type, and fire flag for each GOES-8 ABBA
                        (version 5.6) fire pixel identified in GOES-8 imagery
                        for a specific time (hhhh) and day (ddd - julian day)
                        during the 1997 fire season (June - October 1997).

g8f97ddd.v56.gif        This is a gif of the daily diurnal fire product showing
                        fire locations for the 1145, 1445, 1745, and 2045 UTC 
                        time periods.

g8year97.v56.gif        This composite gif file shows the locations of all fires 
                        detected with the GOES-8 ABBA (version 5.6) during the
                        1997 fire season.
	

General Documentation:
GOES-8 Automated Biomass Burning Algorithm (ABBA: Version 5.6) Fire Product

Data Source & Points of Contact:

Points of Contact:
Elaine Mae Prins (at the address listed below)
email:  elaine.prins@ssec.wisc.edu
or
Joleen M. Feltz
email:  joleen.feltz@ssec.wisc.edu

Data Source and Address:
NOAA/NESDIS/ORA Advanced Satellite Products Teams (ASPT)
University of Wisconsin-Madison, Space Science & Engineering Center
Cooperative Institute for Meteorological Satellite Studies (CIMSS)
1225 West Dayton St.
Madison, WI 53706

General Overview:

The 1995 fire season in South America represents the first full season of
GOES-8 diurnal observations of biomass burning in South America.  The
Cooperative Institute for Meteorological Satellite Studies (CIMSS) at the
University of Wisconsin-Madison produced diurnal GOES-8 derived fire
products for the 1995 fire season (June-October 1995) with version 5.5 of
the GOES-8 ABBA as part of the Smoke Clouds and Radiation Experiment in Brazil
(SCAR-B).  The GOES-8 ABBA classification scheme was revised in version 5.6
to better characterize non-processed possible fire pixels for interannual 
(1995, 1996, 1997, 1998, etc.) comparisons and trend analyses.  The major difference
between version 5.5 and version 5.6 is in the categorization of possible fire pixels.
Version 5.6 of the GOES-8 ABBA uses statistically derived thresholds to bin possible
fire pixels into high, medium, and low probability categories.  The output files
for version 5.6 include processed, saturated, cloudy, and high possibility fire pixels. 
Diurnal fire products are produced for 1145, 1445, 1745, and 2045 UTC coinciding
with peak burning hours.  The study region extends from 35 to 75W and from 0 to 40S.

GOES-8 ABBA Algorithm and Product Description:

The GOES ABBA is a automated dynamic multispectral thresholding algorithm
which uses visible, shortwave infrared window (SIRW near 4 microns) and longwave
infrared window (LIRW near 11 microns) bands to locate fire pixels.  The
algorithm is based on the sensitivity of the SIRW band to high temperature sub-
pixel anomalies and is an expanded version of a technique originally developed
by Matson and Dozier for NOAA AVHRR (Matson and Dozier, 1981).  Once the GOES
ABBA locates a fire pixel, it incorporates ancillary data to correct for
water vapor attenuation, surface emissivity and solar reflectivity and
subsequently uses numerical methods to determine estimates of sub-pixel
fire size and mean fire temperature.  It is important to keep in mind that
the fire processes occurring within a fire pixel are complex and can change
rapidly.  The GOES-8 ABBA derived estimates of sub-pixel fire size and
temperature give a snapshot of the average burning conditions occurring
at the satellite observation time.  For detailed information on the GOES
fire algorithm and recent improvements in the GOES-8 ABBA see Prins and Menzel, 
1994; Menzel and Prins, 1996; Prins et al., 1997 a, b; and Prins et al., 1998.

File Description:

The data set consists of ASCII text files and corresponding GIF files.

a.)  ASCII text files

1.  File name & Size:
ASCII text files are named using the following convention:  g8fyydddhhmm.v56
where:  yydddhhmm indicates the year (yy), julian date (ddd), and time in
UTC (hhmm). The file size is variable depending on the number of fires
detected for that time period.

2.  Contents and Format:
Each ASCII text file consists of information for active fires identified
in GOES-8 multispectral imagery by the GOES-8 Automated Biomass Burning
Algorithm (ABBA) at a specific time period.  The first line in each ASCII
text file includes the date and time for the fire product. The second line
contains column description headers.  Each successive line contains
information for a given fire pixel including fire pixel location
(longitude/latitude); estimated fire size (km2) and average fire
temperature (K); fire pixel ecosystem type; and fire pixel flag
(ranging from 0 to 3).  The ecosystem type is based on the Olson World
Ecosystems database (Version 1.4D, Olson, 1992).  The fire flags are
defined as follows:

Fire Flag:
0       The fire pixel was identified and processed by the GOES-8 ABBA.
        The GOES-8 ABBA was able to determine estimates of sub-pixel fire
        characteristics (size and temperature).
1       Although the fire pixel was identified by the GOES-8 ABBA, it was
        saturated and could not be processed for sub-pixel fire characteristics.
2       Although the fire pixel was identified by the GOES-8 ABBA, it could
        not be processed for sub-pixel fire characteristics due to cloud
        contamination.
3       The pixel was identified by the GOES-8 ABBA as a high probability 
        possible fire pixel.

An example of the first few lines of a GOES-8 ABBA fire product ASCII text
file is provided below:

Date:  95150     Time:  1745 UTC^M
Longitude   Latitude   Size(km2)   Temp(K)   Ecosystem   Fire Flag
 -65.68     -10.55      0.0050      987.        32           0
 -55.23     -10.62     -9.0000       -9.        33           1
 -43.22     -10.73      0.0321      645.        29           0
 -48.79     -10.98     -9.0000       -9.        32           2
 -49.76     -12.30     -9.0000       -9.        41           3

b.)     GOES-8 ABBA fire product GIF files

1.  File name & size:
The GIF files are named using the following convention:  g8fyyddd.v56.gif
where:  yyddd indicates the year (yy) and julian date (ddd).  The file size is
variable.
2.  Contents:
The GIF files are graphical depictions of the fire product contained in the
ASCII text files.  Each GIF file displays the diurnal GOES-8 ABBA fire product
for the study region on a given date.  The display consists of four panels
showing the locations of GOES-8 ABBA derived fire pixels on a mercator
projection at four time periods (1145, 1445, 1745 and 2045 UTC).

Associated References:

Matson, M, and J. Dozier, 1981: Identification of subresolution high temperature
  sources using a thermal IR sensor. Photo. Engr. and Rem. Sens., 47, 1311-1318.
Menzel, W.P., and E.M. Prins, 1996:  Monitoring biomass burning with the new
  generation of geostationary satellites.  In Biomass Burning and Global Change,
  edited by J.S. Levine, pp. 56-64, The MIT Press, Cambridge, MA.
Olson, J.S., 1992:  World Ecosystems (WE1.4).  Digital Raster Data on a
  10-minute geographic 1080x2160 grid.  In Global Ecosystems Database, Version
  1.0: Disc A. Boulder, CO: U.S. Dept. of Commerce, National Oceanic and
  Atmospheric Administration, National Geophysical Data Center.
Prins, E.M. and W.P. Menzel, 1994:  Trends in South American biomass burning
  detected with the GOES visible infrared spin scan radiometer atmospheric
  from 1983 to 1991. Jour. Geo. Res., Vol. 99, 16,719-16735.
Prins, E.M., and W.P. Menzel, 1996:  Investigation of biomass burning and
  aerosol loading and transport utilizing geostationary satellite data.
  In Biomass Burning and Global Change, edited by J.S. Levine, pp. 65-72,
  The MIT Press, Cambridge, MA.
Prins, E.M., W.P. Menzel, D.E. Ward, 1997a:  GOES-8 ABBA Diurnal Fire
  Monitoring during SCAR-B. In SCAR-B Proceedings, edited by V.W.J.H. Kirchhoff,
  pp 153-157, Transtec Editorial, Sao Paulo, Brazil.
Prins, E.M., W.P. Menzel, J.M. Feltz, and D.E. Ward, 1997b:  Analysis of GOES-8
  derived fire and smoke products for SCAR-B and the 1995 burning season in
  Brazil. Submitted to the Journal of Geophysical Research - Atmospheres,
  SCAR-B special issue.
Prins, E.M., W.P. Menzel, and J.M. Feltz, 1998:  Characterizing spatial and 
  temporal distributions of biomass burning using multi-spectral geostationary
  satellite data, Ninth Conference on Satellite Meteorology and Oceanography,
  Paris, France, 25-29 May, 1998, pp. TBD.

Requested Form of Acknowledgment:

This version of the GOES-8 ABBA fire product data set has been submitted for
publication but has not yet been published.  Please refer to the Prins et al.,
1997b, 1998 references when citing the data set.

The work was sponsored in part by the National Aeronautics and Space Administration under
contract NAGW-3804.  Ongoing interannual efforts are funded by the National Oceanic
and Atmospheric Administration (NOAA) under NA67EC0100.