LBA-ECO LC-35 Active Fire Masks Derived from ETM+ Data (Draft)


Summary:

Active fire detection produced derived from Landsat 7 Enhanced Thematic Mapper Plus (ETM+) data for various locations in Brazilian Amazonia during 2001-2003. Data set contains three separate file types: 1- Fire masks containing two values: 0 - no active fire detected, and 1 - active fire detected. 2- Segmented fire masks: contiguous active fire clusters are assigned a specific identification number 3 - Comma separated ASCII files containing information on active fire coordinates and cluster identification numbers

Data Citation:

Cite this data set as follows:

Schroeder, W., E. Prins, L. Giglio, I. Csiszar, C. Schmidt, J. Morisette, and D. Morton. 2008. LBA-ECO LC-35 Active Fire Masks Derived from ETM+ Data. Data set. Available on-line [http://www.daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A.

Implementation of the LBA Data and Publication Policy by Data Users:

The LBA Data and Publication Policy [http://daac.ornl.gov/LBA/lba_data_policy.html] is in effect for a period of five (5) years from the date of archiving and should be followed by data users who have obtained LBA data sets from the ORNL DAAC. Users who download LBA data in the five years after data have been archived must contact the investigators who collected the data, per provisions 6 and 7 in the Policy.

This data set was archived in September of 2008. Users who download the data between September 2008 and August 2013 must comply with the LBA Data and Publication Policy.

Data users should use the Investigator contact information in this document to communicate with the data provider. Alternatively, the LBA Web Site [http://lba.inpa.gov.br/lba/] in Brazil will have current contact information.

Data users should use the Data Set Citation and other applicable references provided in this document to acknowledge use of the data.

Table of Contents:

1. Data Set Overview:

Project: LBA (Large-Scale Biosphere-Atmosphere Experiment in the Amazon)

Activity: LBA-ECO

LBA Science Component: Land Use and Land Cover

Team ID: LC-35 (Csiszar / Longo / Setzer)

The investigators were Csiszar, Ivan Andras; Longo, Karla Maria; Brunner, Jason; Freitas, Saulo Ribeiro de; Morisette, Jeffrey Thomas; Prins, Elaine; Schmidt, Christopher C.; Schroeder, Wilfrid and Setzer, Alberto W . You may contact Schroeder, W. (schroeder@hermes.geog.umd.edu)

Active fire detection product derived from Landsat 7 Enhanced Thematic Mapper Plus (ETM+) data over Brazilian Amazonia during 2001-2003. The spatial resolution of the product is either 28.5 or 30 meters, depending on the data provider for the images used. The time of revisit of Landsat7 ETM+ is 16 days, however, the image selection contains only a limited number of locations imaged 16 days apart. For the majority of the situations, images were selected based on data quality and availability as well as on the occurrence of vegetation fires as indicated by other data sets (e.g., 4 km resolution active fire detection product derived from the Geostationary Operational Environmental Satellite).

Archive Team Reference Section (Temporary):

2. Data Characteristics:

Three main file types are contained in the data set. The following file naming convention was used: L7_PPPRRR_YYYYMMDD_fire_T*. L7 = Landsat7 data PPP = Landsat7 path number RRR = Landsat7 row number YYYY = acquisition year MM= acquisition month DD = acquisition day T* = type (mask, clusters, ASCII file (txt)) File types are described as: 1 - Active fire mask composed of two values: 0 - no fire activity; 1- active fire. This is an image file, data type is byte (8 bits). Each data file has an accompanying header file (file extension *.hdr) including essential information to open and project the data correctly. The header file conforms with ENVI's file format (ITT Corporation) and includes the following information: description = file origin and date of production samples = number of samples lines = number of lines bands = number of bands header offset = this information is not relevant (=0) file type = ENVI definition of file type data type = 1: byte (8 bits); 12 : unsigned integer (16 bits) interleave = raster data storage method. BSQ = Band sequential format (each line of the data is followed immediately by the next line byte order = intel map info = {projection type, reference image pixel in x, reference image pixel in y, coordinate value in x of reference pixel, coordinate value in y of reference pixel, pixel size in x, pixel size in y, UMT zone, UTM zone reference (South), datum, units=Meters, image rotation in degrees} wavelenght units = not applicable band name = informs the name given to band (e.g., Fire Mask or Fire Clusters) 2 - Segmented fire mask. Value of pixels without fire activity receive is zero. For all other pixels showing fire activity, a distinct cluster identification number is assigned to each individual group of contiguous pixels. This is an image file,date type is unsigned integer (16 bits). Each file has a accompanying header file similar to the one described for the fire masks. 3 - Comma separated ASCII file. These files contain the fire detection information derived from the fire masks. They include three collumns describing the latitude, longitude and cluster identification number of each active fire pixel in the corresponding image. An auxiliary file (summary_details.txt) is provided. The file contains 8 comma separated fields, namely: path = Landsat7 path number row = Landsat7 row number year = acquisition year month = acquisition month day = acquisition day hour = acquisition hour (UTC) min = acquisition minute (UTC) active_fire_pixels = number of active fire pixels detected for that image

Site boundaries: (All latitude and longitude given in degrees and fractions)
Site (Region) Westernmost Longitude Easternmost Longitude Northernmost Latitude Southernmost Latitude Geodetic Datum
Brazilian Amazonia (Brazilian Amazonia) -68-435-18World Geodetic System, 1984 (WGS-84)

3. Data Application and Derivation:

This product can be used to identify areas of active fires at 30m nominal resolution. Active fire pixels are classified as such using a contextual approach based on top of atmosphere reflectance values derived from ETM+ bands 4 and 7. The classification method is described in detail in the work of Schroeder et al. [2008] (title: Validation of GOES and MODIS active fire detection products using ASTER and ETM+ data)

4. Quality Assessment:

Visual quality assessment of the product was performed for each individual image. Commission errors could not be found. Omission errors may occur in areas of ambiguous fire activity (e.g., smoldering fires)

5. Data Acquisition Materials and Methods:

The data set is composed of active fire masks derived from Landsat ETM+ data at 30 m nominal resolution. Product is generated from bands 4 and 7 using a contextual fire detection scheme (details given in Schroeder et al. [2008])

6. Data Access:

Data Set Users Guide: LC35_Landsat7_Fire_Masks.html

LME file: /data/bluangel_harvest/Preparation_for_Archive/LC35_Landsat7_Fire_Masks.xml

This data is available through the Oak Ridge National Laboratory (ORNL) Distributed Active Archive Center (DAAC).

Data Archive Center:

Contact for Data Center Access Information:

E-mail: uso@daac.ornl.gov
Telephone: +1 (865) 241-3952
FAX: +1 (865) 574-4665

7. References:

Schroeder, W., Prins, E., Giglio, L., Csiszar, I., Schmidt, C., Morisette, J., and Morton, D. (2008). Validation of GOES and MODIS active fire detection products using ASTER and ETM+ data. Remote Sensing of Environment, 112, 2711-2726.

Related Publications