160
Space and Earth Remote Sensing: GIS Application to Environmental Impact Studies
1 INTRODUCTION
Global change is more than climate change or global warming; it also concerns
changes in land use and land use practices over large areas. Vegetation fires are a
key element of land use practices in the tropics. Much of the world's tropical
Savannahs are periodically burned: in Africa alone an area of more than 300
million hectares, 75% of the Savannah, bums annually. Episodic fires in the
tropical rain forests are not uncommon and are thought to be on the increase. The
ecological, environmental and economic effects of biomass burning occur on all
scales, from local to global, and systematic documentation and history of the
phenomenon are not yet available.
While the subject of fire in the tropical environment has been widely studied, there
is now a need to assess the characteristics of fire on unexplored scales. Biomass
burning is a phenomenon of continental proportions, capable of causing large
scale environmental changes (Malingreau and Tucker, 1988). This calls for a
better understanding of burning on regional and continental scales, a better
appreciation of historical trends and, if possible, improved capabilities to chart
future trends and possible impacts.
Earth observations from space by remote sensing satellites provide systematic and
consistent measurements of a series of parameters related to vegetation fires and
their environmental impacts. Fire detection is normally done on the basis of the
visible wavelength and surface brightness temperatures derived from Landsat,
SPOT and NOAA-A VHRR sensors (Langaas, 1992; Gregoire, 1993).
The aim of this work is to explore spatial characteristics of vegetation fire on
regional and continental scales of representation and to quantitatively describe the
evolution of spatial patterns of vegetation fire on regional and continental scales in
Africa.
2
BIOMASS BURNING DOCUMENTATION USING
AVHRR-GAC TIME SERIES
Sub-continental patterns of fire distribution, both in time and space, may vary
from year to year. These variations may be due to climate, interannual variations
in rainfall distribution, or to anthropic factors, such as movements of populations
and land use changes. Data from NOAA-A VHRR allow regular observation of
biomass burning at regional and continental scales.
While the sampling procedure used to produce the A VHRR Global Area Coverage
(GAC) data, 4 kilometers ground resolution, makes it difficult to collect
quantitative information related to fire (Justice et at, 1989), GAC archives are very
useful for getting information on the main regions of fire occurrence and its
spatio-temporal variability, as an index of fire activity (Belward et at, 1993).
Moreover, the availability for the research community of large volumes of these
images, constituting a historical archive covering the last decade from July 1981,
both of 4 kilometers (GAC data) and 1 kilometer (LAC data: Large American
Coverage) ground resolution, can give a temporal perspective to the analysis of
burning patterns (Fig. 1).
Space and Earth Remote Sensing: GIS Application to Environmental Impact Studies
1 INTRODUCTION
Global change is more than climate change or global warming; it also concerns
changes in land use and land use practices over large areas. Vegetation fires are a
key element of land use practices in the tropics. Much of the world's tropical
Savannahs are periodically burned: in Africa alone an area of more than 300
million hectares, 75% of the Savannah, bums annually. Episodic fires in the
tropical rain forests are not uncommon and are thought to be on the increase. The
ecological, environmental and economic effects of biomass burning occur on all
scales, from local to global, and systematic documentation and history of the
phenomenon are not yet available.
While the subject of fire in the tropical environment has been widely studied, there
is now a need to assess the characteristics of fire on unexplored scales. Biomass
burning is a phenomenon of continental proportions, capable of causing large
scale environmental changes (Malingreau and Tucker, 1988). This calls for a
better understanding of burning on regional and continental scales, a better
appreciation of historical trends and, if possible, improved capabilities to chart
future trends and possible impacts.
Earth observations from space by remote sensing satellites provide systematic and
consistent measurements of a series of parameters related to vegetation fires and
their environmental impacts. Fire detection is normally done on the basis of the
visible wavelength and surface brightness temperatures derived from Landsat,
SPOT and NOAA-A VHRR sensors (Langaas, 1992; Gregoire, 1993).
The aim of this work is to explore spatial characteristics of vegetation fire on
regional and continental scales of representation and to quantitatively describe the
evolution of spatial patterns of vegetation fire on regional and continental scales in
Africa.
2
BIOMASS BURNING DOCUMENTATION USING
AVHRR-GAC TIME SERIES
Sub-continental patterns of fire distribution, both in time and space, may vary
from year to year. These variations may be due to climate, interannual variations
in rainfall distribution, or to anthropic factors, such as movements of populations
and land use changes. Data from NOAA-A VHRR allow regular observation of
biomass burning at regional and continental scales.
While the sampling procedure used to produce the A VHRR Global Area Coverage
(GAC) data, 4 kilometers ground resolution, makes it difficult to collect
quantitative information related to fire (Justice et at, 1989), GAC archives are very
useful for getting information on the main regions of fire occurrence and its
spatio-temporal variability, as an index of fire activity (Belward et at, 1993).
Moreover, the availability for the research community of large volumes of these
images, constituting a historical archive covering the last decade from July 1981,
both of 4 kilometers (GAC data) and 1 kilometer (LAC data: Large American
Coverage) ground resolution, can give a temporal perspective to the analysis of
burning patterns (Fig. 1).