NDP-055
TROPICAL AFRICA: LAND USE, BIOMASS, AND CARBON ESTIMATES FOR 1980
Contributed by:
Sandra Brown*
Greg Gaston**
Work on this project was initiated while at the Department of Natural
Resources and Environmental Sciences, University of Illinois, Urbana, Illinois,
61801, U.S.A.
*Sandra Brown is currently affiliated with Winrock International, Arlington, VA.
(http://www.winrock.org)
**Greg Gaston is currently affiliated with the Department of Geosciences,
Oregon State University, 244 Wilkinson Hall, Corvalis, OR 97331.
Prepared by T.W. Beaty, and L.M. Olsen.
Carbon Dioxide Information Analysis Center
Environmental Sciences Division
Date Published: June, 1996 (Revised 2001)
Prepared by the Carbon Dioxide Information Analysis Center, OAK RIDGE NATIONAL
LABORATORY, Oak Ridge, Tennessee 37831-6335. Managed by the University of
Tennessee-Battelle, LLC for the U.S. DEPARTMENT OF ENERGY under contract
DE-AC05-00OR22725.
CONTENTS
--------
LIST OF TABLES
ABBREVIATIONS
ABSTRACT
1. BACKGROUND INFORMATION
2. SOURCE AND SCOPE OF THE DATA
3. DATA MANIPULATION AND INTERPOLATION
4. MODELING BIOMASS
5. DESCRIPTION OF DATA PROCESSING ROUTINE
6. REFERENCES
7. HOW TO OBTAIN THE DATA AND DOCUMENTATION
8. LISTING OF FILES PROVIDED
9. DESCRIPTION OF THE DOCUMENTATION FILE
10.DESCRIPTION, FORMAT, AND PARTIAL LISTINGS OF THE DATA FILES
11.SAS AND FORTRAN CODES TO ACCESS THE DATA
APPENDIX A: A BREAKDOWN BY COUNTRY OF THE AMOUNT OF LAND AND BIOMASS IN OPEN AND
CLOSED FOREST IN 1980
APPENDIX B: REPRINT OF PERTINENT LITERATURE
APPENDIX C: RECLASSIFICATION SCHEME USED TO OBTAIN THE FOUR LAND USE
CATEGORIES USED IN THIS NDP
LIST OF TABLES
--------------
1. Files in this numeric data package
2. Assignment of the biomass indices to potential biomass densities based on
forest inventories
3. Emission of CO2, CO, and CH4 that would have occurred
if the measured reduction in dry biomass (from potential to actual in 1980)
were burned as fuelwood.
4. Variable formats for biomax.asc (biomax grid), lakes.asc (lakes grid) and
land.asc (land grid)
5. Variable formats for pd80.asc (pd80 grid) and bio1980.asc (bio1980 grid)
6. Variable formats of t_afr.asc and t_afr.pat
7. Variable formats of africa.asc and africa.patsum1980
ABBREVIATIONS
-------------
CDIAC = Carbon Dioxide Information Analysis Center
FTP = file transfer protocol
NDP = numeric data package
QA = quality assurance
GCM = General Circulation Model
UNEP = United Nations' Environmental Program
FAO = United Nations' Food And Agriculture Organization
GIS = geographic information system
PBD = potential biomass density
ARC/INFO is a registered trademark of the Environmental Systems Research
Institute, Inc., Redlands, CA 92372.
SAS is a registered trademark of the SAS Institute, Inc., Cary, NC 27511-8000.
ABSTRACT
--------
Brown, S, and Greg Gaston. 1996. Tropical Africa: Land Use, Biomass, and
Carbon Estimates for 1980. NDP-055, Carbon Dioxide Information Center, Oak
Ridge National Laboratory, Oak Ridge, Tennessee.
This document describes the contents of a digital database containing
maximum potential aboveground biomass, land use, and estimated biomass and
carbon data for 1980. The biomass data and carbon estimates are associated
with woody vegetation in Tropical Africa. These data were collected to reduce
the uncertainty associated with estimating historical releases of carbon from
land use change. Tropical Africa is defined here as encompassing 22.7 x 10E6
km2 of the earth's land surface and is comprised of countries that are located
in tropical Africa (Angola, Botswana, Burundi, Cameroon, Cape Verde, Central
African Republic, Chad, Congo, Benin, Equatorial Guinea, Ethiopia, Djibouti,
Gabon, Gambia, Ghana, Guinea, Ivory Coast, Kenya, Liberia, Madagascar, Malawi,
Mali, Mauritania, Mozambique, Namibia, Niger, Nigeria, Guinea-Bissau, Zimbabwe
(Rhodesia), Rwanda, Senegal, Sierra Leone, Somalia, Sudan, Tanzania, Togo,
Uganda, Burkina Faso (Upper Volta), Zaire, and Zambia).
The database was developed using the GRID module in the ARC/INFO (TM
geographic information system. Source data were obtained from the Food and
Agriculture Organization (FAO), the U.S. National Geophysical Data Center, and
a limited number of biomass-carbon density case studies. These data were used
to derive the maximum potential and actual (ca. 1980) aboveground biomass
values at regional and country levels. The land-use data provided were derived
from a vegetation map originally produced for the FAO by the International
Institute of Vegetation Mapping, Toulouse, France.
Analyses conducted with this database found that 18% of Tropical Africa was
classified as closed forest and 36% was classified as open forest in 1980.
Forested lands contained over 138E+09 Mg of aboveground live
biomass, which is equivalent to 69E+09 Mg of carbon. Closed forests and open
forests had mean aboveground biomass values of 209 Mg/ha and 67 Mg/ha,
respectively, in 1980. These values are down from maximum potential
aboveground live biomass values of 296 Mg/ha for closed forest and 108 Mg/ha
for open forest.
Keywords: Carbon release, climate change, deforestation, forestry, population.
NDP-055 includes the following data and documentation files:
1. Descriptive text file (ndp055.txt)
2. ARC/INFO projection file (trop_africa.prj) with parameters for the
ARC/INFO grids and coverages included in this NDP.
3. ARC/INFO exported integer grid (biomax.e00) containing maximum potential
biomass density data in Mg/ha.
4. ASCII format of BIOMAX grid (biomax.asc)
5. ARC/INFO exported integer grid (lakes.e00) containing the location of all
nonseasonal lakes.
6. ASCII format of LAKES grid(lakes.asc)
7. ARC/INFO exported integer grid (land.e00) containing land-use information.
8. ASCII format of LAND grid (land.asc)
9. FORTRAN code (intgrid.for) to read and print the ASCII data files
(biomax.asc, lakes.asc, and land.asc)
10. SAS retrieval program (intgrid.sas) to read and print the ASCII data files
biomax.asc, lakes.asc, and land.asc.
11. ARC/INFO exported floating point grid (bio1980.e00) containing the
calculated actual aboveground live biomass in open and closed forests
for 1980 in Mg/ha.
12. ASCII format of BIO1980 grid (bio1980.asc)
13. ARC/INFO exported floating point grid (pd80.e00) containing population
density data for 1980.
14. ASCII format of PD80 grid (pd80.asc)
15. FORTRAN code (realgrid.for) to read and print the ASCII files pd80.asc and
bio1980.asc.
16. SAS retrieval program (realgrid.sas) to read and print the ASCII
datafiles pd80.asc and bio1980.asc.
17. ARC/INFO exported coverage (africa.e00) containing the amount and biomass
in open and closed forest by country.
18. ARC/INFO exported coverage (t_afr.e00) containing population data for 1960,
1970, 1980 and 1990 for the subnational political units used in this NDP.
19. ASCII file (africa.asc) containing the attribute information associated
with africa.e00.
20. FORTRAN code (africa.for) to read and print africa.asc.
21. SAS retreival program (africa.sas) to read and print africa.asc.
22. ASCII file (t_afr.asc) containing the attribute information associated
with t_afr.e00.
23. FORTRAN code (t_afr.for) to read and print t_afr.asc.
24. SAS retrieval program (t_afr.sas) to read and print t_afr.asc.
25. Graphic Map, "Maximum Potential Biomass Density" (africa_biomax.gif)
26. Graphic Map, "Land use" (africa_lc.gif)
27. Graphic Map, "Area of Closed Forests (By Country)" (areaclsdforest.gif)
28. Graphic Map, "Area of Open Forests (By Country)" (areaopnforest.gif)
29. Graphic Map, "Mean Biomass of Closed Forests (By Country)" (csldforestbio.gif)
30. Graphic Map, "Mean Biomass of Open Forests (By Country)" (opnforestbio.gif)
31. Graphic Map, "Percent Forest Cover (By Country)" (pctforest.gif)
32. Graphic Map, "Total Forest Biomass (By Country)" (totforestbio.gif)
33. Graphic Map, "Population Density - 1990" (pop1990.gif)
34. Graphic Map, "Population Density - 1980" (pop1980.gif)
35. Graphic Map, "Population Density - 1970" (pop1970.gif)
36. Graphic Map, "Population Density - 1960" (pop1960.gif)
1. BACKGROUND INFORMATION
-------------------------
This database was developed by Sandra Brown, Louis Iverson, Ariel
Lugo, Anantha Prasad, Greg Gaston, and others between 1990 and 1994.
The primary contract that supported the collection of the data was the U.S.
Department of Energy, Office of Health and Environmental Research, contract
DEFG02-90ER61081. Additional support was obtained from the U.S. Environmental
Protection Agency through its post doctoral fellowship program, which is
administered by the National Research Council.
Human-induced changes in the Earth's land cover may influence ecological and
climatic systems. The potential for these changes to have regional or global
implications makes the task of developing an all-inclusive general circulation
model (GCM) with predictive capabilities essential. However, to accurately
predict the effects of long-term increases in the number and concentration of
greenhouse gases on climate, GCMs require information on the timing and
magnitude of past releases of CO2, methane, nitrous oxide, water
vapor, and a host of other anthropogenic gases. An understanding of all of the
major processes leading to the release (sources) and storage (sinks) of these
gases over time is also essential.
The prediction of regional trends in climate will be problematic until a
well- documented understanding of the past, present, and possible future usage
of fossil fuels and land is reached. Such an understanding is possible only
with the concerted effort of many investigators from a diverse set of
disciplines. The perspectives of ecologists, biologists, historians,
geographers, foresters, mathematicians, physicists, and others are needed if
a multifaceted examination of the causes and consequences of land-cover change
and human population growth on the world's climate is to be conducted.
Reductions in the amount of carbon stored in terrestrial ecosystems are
partially responsible for the increase in the concentration of atmospheric
CO2 that has occurred since 1850 (Houghton 1993). For example, live
vegetation in terrestrial ecosystems is estimated to have contained over 900 Pg
of carbon, with 90% of this being in forests, prior to any intentional land
clearing on Earth (Olson et al. 1985, Dale et al. 1991).
A decline in the amount of carbon sequestered in vegetation has been
associated with the rapid increase of human population that began during the
industrial revolution. This population growth created demands for food and
forest products that were fulfilled through the clearing and conversion of
large tracts of forest to agriculture and other uses. This land-cover change
has direct impacts on the concentration of atmospheric CO2. For example, a 1%
change in the total amount of carbon stored in living vegetation would release
about 21 Pg of carbon to the atmosphere, almost four times the present annual
emission from fossil fuels (Houghton 1993).
2. SOURCE AND SCOPE OF THE DATA
-------------------------------
This numeric data package (NDP-055) describes a land-use and biomass
database for Tropical Africa that contains 1980 land cover, maximum potential
aboveground biomass, and biomass estimates for woody vegetation in 1980. In
addition, the database includes information on human populations in each
nation and sub-national political unit in Tropical Africa for the years 1960,
1970, 1980, and 1990. These data were collected to reduce the uncertainty
associated with the possible magnitude and time course of historical releases
of carbon from land-cover change and may be used by demographers, historians,
geographers, and other researchers interested in the relationship between land
cover change, land degradation, climate, and anthropogenic activities.
For purposes of this NDP, Tropical Africa is defined as encompassing
nearly 22.7 x 1E6 km2 of the world's surface and includes the following
countries: Angola, Botswana, Burundi, Cameroon, Cape Verde, Central African
Republic, Chad, Congo, Benin, Equatorial Guinea, Ethiopia, Djibouti, Gabon,
Gambia, Ghana, Guinea, Ivory Coast, Kenya, Liberia, Madagascar, Malawi, Mali,
Mauritania, Mozambique, Namibia, Niger, Nigeria, Guinea-Bissau, Zimbabwe
(Rhodesia), Rwanda, Senegal, Sierra Leone, Somalia, Sudan, Tanzania, Togo,
Uganda, Burkina Faso (Upper Volta), Zaire, and Zambia. As defined here
Tropical Africa does not include countries in Mediterranean and southern
Africa (i.e., Egypt, Libya, Tunisia, Algeria, Morocco, Western Sahara, South
Africa, Lesotho, and Swaziland).
The major effort in the compilation of this database involved the acquisition
and processing of the source data sets. Most of these data sets were acquired
from the United Nations Environmental Program's (UNEP) Global Resources
Information Database (Jaakkola 1990) or from the United Nation's Food and
Agriculture Organization (FAO 1993). These data, in turn, were input,
georeferenced, and converted into a raster format as needed using an
ARC/INFO (TM) geographic information system (GIS). The spatial data sets used
in the development of this NDP are as follows:
1. National boundaries were derived from the U.S. Defense Mapping Agency's
Digital Chart of the World at a scale of 1:1,000,000. Boundaries for
subnational political units were derived from The New International
Atlas, the Encyclopedia of the Nations, and the FAO's Tropical Forest
Resources Assessment Program 1990 population data (Worldmark Press 1984,
Rand McNally 1980). Boundaries for large water bodies (e.g. Lake Victoria)
were copied from the Digital Chart of the World; moderate-size dlakes
(>20 km in length in any direction) were digitized from the UNESCO
Vegetation Map of Africa, scale 1:5,000,000 (UNESCO/AETFAT/UNSO 1981,
White 1983).
2. Population density data for 1960, 1970, 1980, and 1990 for national and
subnational political units were provided by M. Lorenzini and K.D. Singh,
FAO, Tropical Forest Resources Assessment Program 1990 (FAO 1993).
3. Elevation data were derived from a global digital elevation database
originally developed by the U.S. Defense Mapping Agency (ETOPO5). The
elevation data were reclassified into 15-m increments, and all bathymetry
data were set to zero.
4. Soils data were derived from the FAO Soils Map of the World (FAO 1971-81). The soils data were reclassified into five soil texture classes, five
slope classes, and two soil depth classes and were used in calculating
the maximum potential biomass values included with this database. These
soil data are not included with this NDP.
5. Climatic data were obtained from the FAO agro-meteorology database and two
maps were developed using the spatial coordinates for each station and the
mean monthly precipitation, maximum temperature, minimum temperature,
day-night temperature differential, evapotranspiration rate, and vapor
pressure. The first map is a isopleth map of annual precipitation and the
second is of an integrated climate index that provides a measure of the
growing season length based on five climatic variables (Weck 1970).
6. Vegetation data represent conditions in 1980 (approximate date) from the
FAO Vegetation Map of Africa (Lavenu 1987).
3. DATA MANIPULATION AND INTERPOLATION
--------------------------------------
The methodology described in Iverson et al. (1994) and Brown and Gaston
(1995) (reprinted in Appendix B) was used to compute degradation ratios.
These degradation ratios were applied to a map of maximum potential
aboveground biomass density to develop a map of current (ca. 1980) aboveground
biomass. Since the linear regression models used to estimate the degradation
ratios were calibrated for woody vegetation only, this map was masked using
a land cover map (1980). All areas not in a forested or wooded class were
deleted. This process resulted in a map showing the amount of aboveground
biomass in woody vegetation for 1980. The change (emission) of carbon
resulting from land use change and land degradation between the 1880s and 1980
can be estimated for closed and open forests by comparing the aboveground
biomass map with the maximum potential aboveground biomass density map.
Two of the data sources used in calculating the maximum potential aboveground
biomass values needed additional processing before they could be used in the
biomass estimation model. First, the meteorology data were used to calculate
a climatological index for each station based on a modified version of Weck's
(1970) methodology. Second, these climatological indices (point data) were
interpolated into a thematic map using a thin-plate spline technique (Iverson
et al. 1994). Lastly, the FAO vegetation map used (Lavenu 1987) was
reclassified from over 64 ecosystem types into a map showing closed forest,
open forest, grassland savanna, and other non forested lands. See Appendix C
for a description of the reclassification scheme used.
The source data were then projected into a cylindrical equal-area projection
using a central meridian of 15 degrees east and a standard parallel of
0 degrees. This equal-area versions of the source data were then converted
into grids (cellsize 5km) using the ARC/INFO (TM) POLYGRID command.
4. MODELING BIOMASS
-------------------
The maximum potential biomass density of woody vegetation is assumed to be
the product of a combination of the prevailing climatic and geomorphic
conditions in an area. To estimate the maximum potential biomass for each grid
cell, each input layer (i.e., mean annual precipitation, climate index,
elevation/slope, and soil quality/texture) was recoded and assigned values
ranging from 1 to 25 based on the perceived importance of the layer to the
growth of woody vegetation. A weighted overlay of the four layers was then
performed to create a biomass index with a minimum possible value of 4 and a
maximum value of 100 (in Tropical Africa the values ranged from 40 to 100).
The climate-related layers (mean annual precipitation and climate index)
account for 50% of the possible index value. The soil layer accounts for 25%
and the elevation and slope layer [formed by combining the elevation data
(maximum value 13) with the slope data (maximum value 12)] accounts for the
remaining 25%. Iverson et al. (1994) used the same weighting scheme for
Tropical Asia. This scheme was found to give satisfactory results for Africa
when compared to available literature for the region (Graham et al. 1990,
Martin 1991, Lavenu 1987).
Potential biomass density (PBD) indices for forests, woodlands, and woody
savannas were calibrated by assigning aboveground biomass density estimates
(obtained from the literature) to the range of index values. This task posed a
challenge because few measurements of mature forest biomass have been taken in
this region (Brown and Lugo 1982). The highest reported biomass densities
were about 500 Mg/ha (Huttel and Bernhard-Reversat 1975, Huttel 1975) for
tropical moist forests in western Africa. Western Africa is also associated
with the highest range of PBD indices (>96). A PBD index value of 51
was assumed to represent the lower limit for tree or shrub savannas and was
assigned a biomass density of 15 Mg/ha based on data in Lamotte (1975),
Lamotte and Bourliere (1983), and Menaut et al. (1991). It was then assumed
that biomass density exhibited a logistic-shaped function between the PBD
index and biomass. Biomass values (in Mg/ha) were assigned to the PBD values
as shown in Table 2. The lower index value (i.e., ~40) and upper index value
(i.e., 100) for Tropical Africa were equated with 1Mg/ha and with 500 Mg/ha,
respectively. Between these two bounds the potential biomass index was divided
into twelve ranges. Intermediate values of biomass density were obtained or
estimated from Greenland and Kowal (1960), Pierlot (1966), Feson et al. (1974),
and Brown et al. (1994).
The remainder of this methodology is designed for woody vegetation only.
Thus, the next step in determining the actual biomass in forests for 1980 was
to remove non-woody vegetation associations from the map of maximum potential
biomass density that is distributed with this database. This was done by
taking the reclassified FAO vegetation map for 1980 (Lavenu 1987) [i.e., with
four classes: closed forest, open forest (woodland/wooded savanna), grassland
savanna, and other] and setting those grid cells with closed forest or open
forest to a value of 1. All other cells were set to NODATA. This temporary
map was multiplied by the potential biomass density map. This process set or
masked most grid cells that were not classified as woody vegetation (as of 1980)
to NODATA. The resulting maximum potential biomass density map, therefore, is
of forested areas only. The only non-wooded areas remaining in this map are
grid cells located within lakes.
To estimate the actual biomass remaining in live woody vegetation in 1980,
researchers developed degradation ratios based on population data and applied
them to the maximum potential biomass density map for forested lands. The
degradation ratios were based on the assumption that the degree to which
forests are reduced from their potential biomass is a function of the
population density and forest type or ecofloristic zone. For example, high
population densities will result in low degradation ratios (i.e., highly
degraded forests), and drier regions will be degraded more severely than
moist regions because of the ability of moist forests to replace biomass faster
(Brown and Lugo 1982).
Other socioeconomic factors such as the ratio of rural to urban population,
the importance of subsistence versus export-oriented agriculture, and the state
of industrialization affects the rate of biomass reduction. In developed
countries a population density and a stage of industrial development
have been reached at which no further biomass loss occurs and other factors
come into play, which may actually lead to increases in biomass (Flint
and Richards 1994).
In most of Tropical Africa this threshold has not been reached, as
exemplified by the fact that 90% of the population in Africa still uses wood
for cooking. The average amount of wood used for cooking per family is
equivalent to about 1.5 tons of fuel oil per year. It has been estimated that
the annual addition to the fuelwood supply through the growth of biomass was
less than consumption in 1980 and that the half-life of tree stocks (the time
it would take for stocks to decline by 50%) averaged about 25 years in 1980.
Furthermore, tree stocks are declining at accelerating rates because
consumption is increasing exponentially with population growth (Anderson and
Fishwick 1984).
Degradation equations have been estimated for two forest types (closed
forest and woodland/wooded savanna) by comparing the potential biomass
density in 1980 with biomass densities obtained from forest inventories for
the same location. This process produced unitless factors that indicate the
reduction from potential biomass to actual biomass as a function of population
density. Eight reliable forest inventories for Tropical Africa were used to
develop the degradation equations (Clement and Nouvellet 1978, Development
and Resources Corporation 1967, Republique de Cote d'Ivoire 1975, United
Nations 1972, FAO 1989). Due to the limited number of suitable forest
inventories of tropical Africa, 27 additional inventories from tropical South
and Southeast Asia were also included in the analysis (Iverson et al. 1994).
The data for the closed forest zone of Tropical Africa were combined with
the moist forest data from South and Southeast Asia, while data for the
woodland/wooded savanna zone were combined with the dry forest data from South
and Southeast Asia. A least-squares regression was conducted for each zone,
and the following degradation equations were produced.
The equations, sample sizes, and correlation results for closed forest are:
PD > 12: DR = 0.847 - 0.091 ln (PD), (n = 25, r2 = 0.72), (1a)
PD <= 12: DR = 1 - 0.032 PD, (1b)
where DR is the degradation ratio and PD is the population density in people
per km2.
The equations, sample sizes, and correlation results for open forest
(woodland/wooded savanna) are:
PD > 7: DR = 0.866 - 0.118 ln (PD), (n = 10, r2 = 0.54), (2a)
PD <= 7: DR = 1 - 0.050 PD, (2b)
where DR is the degradation ratio and PD is the population density in people
per km2.
Equation 1a and 2a for each forest type is based on the range of available
data, and the minimum PD for which it is applicable. For PDs less than these
minimums, simple linear equations (Eq. 1b and 2b) were developed between the DR
at zero PD (assumed to be one), and the value of DR from equation 1a and 2a at
the minimum PD. A linear function of PD was then calculated for each land
cover class between these DR values.
Use of the natural logarithm in equations 1a and 2a allows negative DR
values to occur for PDs greater than 11,022 for closed forest and 1,539 for
the open forest (woodland/wooded savanna) land cover classes. This was not a
problem in most cases, however, since forest lands tend to have lower
population densities. For this reason, calculated DR values less than 0.1
were reset to 0.1.
The actual-biomass-density map for 1980 is the product of the
potential-biomass- density map for woody vegetation, the population density map
for 1980 stratified by land cover type (closed forest and open forest), and
the calculated degradation ratios. The map of actual biomass densities for
1980 was calculated on a cell-by-cell basis. All cells within a political
unit were assigned the same population density value. The forest strata,
however, may vary within a political unit causing the degradation equation
used to vary. Thus, the map of actual biomass density for woody vegetation
in 1980 reflects the predicted maximum potential biomass, the forest type
stratification, and the population density in 1980.
Due to the coarse scale of the input data used in generating the potential
biomass density map, some large water bodies within Africa were not
represented. Boundaries for large water bodies (e.g., Lake Victoria) and
moderate-sized lakes (>20 km in length in any direction) were digitized and
entered into a lake coverage (map). Following a process similar to that used
for the land cover data, those grid cells within a lake (i.e., > 50%) were
given a value of NODATA, and the remainder were set to one. This coverage was
multiplied by the grid cells in the map of actual biomass density for 1980
to set all "wet" grid cells to NODATA.
This final map of actual live aboveground biomass density only contains
lands that were in closed forest (moist forest and seasonal forest) or open
forest (woodland/wooded savanna) in 1980 (per Lavenu 1987). The final map
contains 4,025,275 km2 of closed forest land with a mean biomass
density of 208.7 Mg/ha and 8,156,575 km2 of open forest land (woodland/ wooded
savanna) with a mean biomass density of 67.1 Mg/ha. As expected, the minimum
biomass densities were located in lowland dry zones to the north and south,
maximum values were located in the humid lowlands in southwest and south
central Africa (i.e., Congo, Gabon, Zaire), and highland areas had slightly
less biomass than their lowland counterparts. The presence of these known
patterns increased our confidence in this methodology and its ability to
generate reasonable estimates of biomass density for Tropical Africa.
The primary product of this GIS methodology is the actual biomass density
information for woody vegetation in 1980. Actual biomass density in year X
may be determined (as described for 1980) using a land-cover map and population
data for year X (e.g., 1990). The results may then be compared to determine
the change in biomass (carbon) of woody vegetation in the intervening years.
The biomass for 1980 may also be compared to the maximum potential aboveground
biomass for the region. The difference between the mean maximum potential
biomass density for closed forest lands and the mean actual biomass density
in 1980 is 86.9 Mg/ha (40.5 Mg/ha for open forest). On the basis of the
average biomass to carbon conversion factor of 1 kg of wood to 0.5 kg of
carbon (Smith 1991), and the assumption that the removed wood was eventually
burned, the amount of CO2, CO, CH4, and other gases emitted from this per ha
reduction in biomass may be estimated for closed and open forests. Table 3
shows the estimated emission of CO2, CO, and CH4 that would have occurred if
86.9 (40.5) Mg of dry biomass were burned using traditional West African
stoves, based on emission factors calculated by Brocard et al. (1995). (If
forest fire based emission factors were used the emission of CO2 and CO would
increase while the amount of CH4 and carbon in the "other" category
would decrease.)
Africa had an estimated total population of 100 million people in 1650 (density
of 0.03 people per km2). Two hundred years later (1850) the population
was still around 100 million (Udo 1982). Thus, it may be assumed that prior to
the colonial period, which began in the mid to late 1800s, the amount of land
in forests had remained relatively stable for over two hundred years. Since
the land in the closed forest and open forest (woodland/wooded savanna)
land-cover classes in 1980 were located in the interior of larger forests in
1880, it can be assumed that they were at or near their maximum potential
biomass density at the beginning of the colonial period. If this was the case,
the emission of CO2 due to the degradation of these forested lands (i.e.,
reduction of biomass from the maximum potential to the "actual" biomass in
1980) would total 27.21E+12 kg/C. If this CO2 had been emitted at a
constant rate since 1880, 272.1E+06 Mg C per year would have been released
into the atmosphere. To put this figure into perspective, Houghton and Hackler
(1995) calculated that the regional net flux of carbon in Tropical Africa from
all land-use changes was 238.69E+06 Mg C in 1980 and had increased to
341.5E+06 Mg C by 1990.
5. LIMITATIONS AND RESTRICTIONS OF THE DATA
-------------------------------------------
The following paragraphs list the limitations and caveats that should be
considered when using this database. Failure to consider these limitations
could result in erroneous interpretations of the data.
1. All data coverages (maps) in this database have been resampled into 5 x 5 km
grid cells. This cell size and the small cartographic scale of most of the
input data used preclude the use of these data for site-specific studies.
2. The soil texture, slope, and elevation data used in the development of the
map of maximum potential aboveground biomass densities were based on
small-scale data sources (1:1,000,000 to 1:5,000,000).
3. The climate data were extrapolated into areas with little or no data.
Consequently, effects of local topography on precipitation and temperature
patterns are not considered in the map of maximum potential aboveground
biomass.
4. The population data used to calculate the degradation ratios are based on
national internal political subdivisions as of 1980. The population data
provided with this database (i.e., for 1960, 1970, 1980, and 1990) are
keyed to these 1980 divisions.
5. The limited number of forest inventories in Tropical Africa makes
verification of the final map of actual biomass density for woody
vegetation in 1980 difficult. However, average biomass densities were
obtained for Guinea, Gambia, Mozambique, and Burkina Faso from national
level forest inventories and compared to those calculated here. The
values for Guinea, Gambia, Mozambique, and Burkina Faso were within 3%,
19%, 3%, and 38%, respectively, of the inventory values (Brown and Gaston
1995).
6. DATA CHECKS PERFORMED BY CDIAC
---------------------------------
An important part of the data packaging process at the Carbon Dioxide
Information Analysis Center (CDIAC) is the quality assurance (QA) of the data
before its distribution. The QA process is an important component of the
value-added concept of ensuring accurate, usable information. The complete QA
of a data set can be a time- consuming process, since data received by CDIAC
are rarely in condition for immediate distribution, regardless of source. The
following summarizes the QA checks performed on the various data groups
presented in this document.
1. The map of maximum potential aboveground biomass density is derived from an
index calculated from six climatic and geomorphic variables. The index values
were associated with the known upper and lower limits of biomass density in
Africa, and the range was divided into 12 density groupings (See Table 2).
The maximum potential aboveground biomass density map (5x5 km grid cells) was
correlated with a 1x1 degree map of carbon in live vegetation (Olson et al.
1983). The analysis obtained an r2 value of 0.4, n = 1195611,
and alpha = 0.0001. Because of cell size differences, a higher correlation
value should not be expected; however, the significance of the correlation
(>99%) implies that the two independently derived data sources are measures of
the same variable.
2. The map of maximum potential aboveground biomass (5x5 km grid cells) and
the Olson et al. (1983) 1 x 1 degree map of carbon in live vegetation were
printed and visually compared to ensure that the location of low and high
carbon/biomass densities matched.
3. The land-use map for 1980 derived from Lavenu (1987) was printed and
compared to a similar map prepared by the United Nations (White 1983).
4. On the basis of QA steps 2 and 3, it was determined that neither the land
cover nor potential-biomass-density map had accounted for internal water bodies
in Tropical Africa. A new map was created for use with this database that
contains all permanent (versus seasonal) lakes larger than 20 km in length in
any direction.
5. The population data for 1960, 1970, 1980, and 1990, were aggregated by
country and compared to secondary data sources (e.g., WorldMark Press 1984, Udo
1982, and Goode 1992).
6. The final actual-biomass-density map for 1980 was aggregated on a
cell-by-cell basis to obtain the mean biomass density by country (see
Appendix A). These mean values were compared with national level forest
inventories available for Guinea, Gambia, Mozambique, and Burkina Faso to
ensure that the calculated values were within accepted ranges for tropical
regions of the world.
5. DESCRIPTION OF DATA PROCESSING ROUTINE
-----------------------------------------
The data distributed with this NDP were initially received as ARC/INFO GRIDs
and the base map used was derived from the 1:1,000,000- scale Digital Chart of
the World developed by the U.S. Defense Mapping Agency with additional
information digitized from a 1:5,000,000 scale land use map of Africa
(White 1983). The data contained in this NDP are available as flat ASCII
files and as ARC/INFO export files (*.e00).
The ARC/INFO GRID data is provided in two formats. This data was exported
using both the ARC/INFO EXPORT command and the ARC/INFO GRIDASCII command.
The ARC/INFO coverages were exported using the ARC/INFO EXPORT command and the
attribute information is provided in ASCII format. The biomass density values
are expressed in Mg per ha (i.e., 2 Mg of dry biomass = 1 Mg/C or 1 metric ton
of carbon).
The degradation ratio equations used in calculating BIO1980.ASC are described
below. For a detailed description of these equations see Part 1, Section 8.2,
of the documentation. (The documentation for NDP-055 is available free-of-charge
from the Carbon Dioxide Information Analysis Center, Oak Ridge, Tennessee).
Degradation equations were estimated for two ecofloristic zones [closed
forest and open forest (woodland/wooded savanna)] by comparing the potential
biomass density with current biomass and population densities obtained for the
same location. This process produced unitless factors that indicate the
reduction from potential biomass to "current" biomass as a function of
population density. A least-squares regression was then conducted for each
zone and the following degradation equations were produced. The equations,
sample size (n), and correlation coefficient for closed forest are:
PD > 12: DR = 0.847 0.091 ln (PD), (n = 25, r^2 = 0.72), (1a)
PD <= 12: DR = 1 - 0.032 PD . (1b)
The equations, sample size, and correlation coefficient for open forest
(woodland/wooded savanna) are:
PD > 7: DR = 0.866 0.118 ln (PD), (n = 10, r^2 = 0.54), (2a)
PD <= 7: DR = 1 - 0.050 PD . (2b)
Where DR is the degradation ratio and PD is the population density in people
per km2.
6. REFERENCES
-------------
Anderson, D. and R. Fishwick. 1984. Fuelwood consumption and Deforestation
in African Countries. World Bank Staff Working Papers No. 704. The World
Bank, Washington, D.C.
Brocard, D., C. Lacaux, J.P. Lacaux, G. Kouadio, V. Yoboue, M. Assa Achy,
B. Ahoua, and M. Koffi. 1995. Emissions from the combustion of biofuels in
the African tropics. In Biomass Burning and Global Change, J.S. Levine (ed.).
MIT Press, Cambridge, MA.
Brown, S., Gaston G., and R. Daniels. 1996. Tropical Africa: Land Use,
Biomass, and Carbon Estimates for 1980. ORNL/CDIAC-92, NDP-055.
Carbon Dioxide Information Analysis Center, Oak Ridge National
Laboratory, Oak Ridge, Tennessee.
Brown, S. and G. Gaston. 1995. Use of forest inventories and geographic
information systems to estimate biomass density of tropical forests:
Application to Tropical Africa. Environmental Monitoring 38: 157-168.
Brown, S. and A.E. Lugo. 1982. The storage and production of organic matter
in tropical forests and their role in the global carbon cycle. Biotropica 14:
161-187.
Brown, S., L.R. Iverson, A. Prasad, and D. Liu. 1993. Geographical
distributions of carbon in biomass and soils of tropical Asian forests.
Geocarto International 4: 45-59.
Clement, J. and Y. Nouvellet. 1978. Inventaire Forestier dans la Zone
d'Edenzork. Republique du Gabon, Soceite Gabonaise de cellulose, Centre
Technique Forestier Tropical, Nogent-sur-Marne, France.
Dale, V.H., R.A. Houghton, and C.A.S. Hall. 1991. Estimating the effects of
land-use change on global atmospheric CO2 concentrations. Canadian Journal of
Forest Research 21: 87-90.
Development and Resources Corporation. 1967. Forestry Resources of the
Southwest Region. Government of the Republic of the Ivory Coast, New York.
Food and Agriculture Organization (FAO). 1993. Forest Resources Assessment
1990: Tropical Countries. Forestry Paper 112, FAO, Rome, Italy.
FAO. 1989. Studies on the Volume and Yield of Tropical Forest Stands: 1.
Dry forest Formations. Forestry Paper 51/1, FAO, Rome, Italy.
FAO. 1971-81. Soils Map of the World 1:5,000,000. FAO/UNESCO, Rome,
Italy.
Flint, E.P. and J.F. Richards. 1994. Trends in carbon content of vegetation in
South and Southeast Asia associated with changes in land use, pp. 201-300. In
Effects of Land-Use Change on Atmospheric CO2 Concentrations: South and
Southeast Asia as a Case Study, V.H. Dale (ed.). Springer-Verlag Inc., New York.
Freson, R., G. Goffinet, and F. Malaisse. 1974. Ecological effects of the
regressive succession mhulu-miombo-savannah in Upper-Shaba (Zaire), pp.
365-371.In Proceedings 1st International Congress of Ecology. The Hauge,
The Netherlands.
Goode, J.P. 1992. Goode's World Atlas. 18th Edition, Rand McNally,
Chicago, Illinois.
Graham, R.L., R.D. Perlack, A. Prasad, J.W. Ranney, and D.B. Waddle. 1990.
Greenhouse Gas Emissions in Sub-Saharan Africa. ORNL-6640. Oak Ridge
National Laboratory, Oak Ridge, Tennessee.
Greenland, D.J. and J.M.L. Kowal. 1960. Nutrient content of the moist tropical
forest of Ghana. Plant and Soil 12: 154-173.
Houghton, R.A. and J.L. Hackler. 1995. Continental Scale Estimates of the
Biotic Carbon Flux from Land Cover Change: 1850 to 1980. ORNL/CDIAC-79,
NDP-050. Carbon Dioxide Information Analysis Center, Oak Ridge National
Laboratory, Oak Ridge, Tennessee.
Houghton, R.A. 1993. Changes in terrestrial carbon over the last 135 years, pp.
139-157. In The Global Carbon Cycle, M. Heimann (ed.). NATO ASI Series, Vol.
I 15. Springer-Verlag Inc., New York.
Houghton, T.A., W.H. Schlesinger, S. Brown, and J.F. Richards. 1985. Carbon
dioxide exchange between the atmosphere and terrestrial ecosystems, pp.
113-140. In Atmospheric Carbon Dioxide and the Global Carbon Cycle,
J.R. Trabalka (ed.). DOE/ER-0239. U.S. Department of Energy, Office of
Energy Research, Washington, D.C.
Huttel, C.H. 1975. Root distribution and biomass in three Ivory Coast rain
forest plots, pp. 123-130. In Tropical Ecological Systems, F.B. Golley
and E. Medina (eds). Springer Verlag Inc., New York.
Huttel, C.H., and F. Bernhard-Reversat. 1975. Recherches sur l'ecosysteme de
la foret subequatoriale de base Cote D'Ivoire. Cycle de las matiere
oranique. Terre et Vie 29: 203-228.
Iverson, L.R., S. Brown, A. Prasad, H. Mitasova, A.J.R. Gillespie, and A.E.
Lugo. 1994. Use of GIS for estimating potential and actual forest biomass
for continental South and Southeast Asia, pp. 67-115. In Effects of Land-
Use Change on Atmospheric CO2 Concentrations: South and Southeast
Asia as a Case Study, V.H. Dale (ed.). Springer-Verlag Inc., New York.
Jaakkola, S. 1990. Managing data for the monitoring of tropical forest cover:
The Global Resources Information Database approach. Photogrammetric
Engineering and Remote Sensing 56: 1355-1358.
Lamotte, M. 1975. The structure and function of a tropical savanna ecosystem,
pp. 179-222. In Tropical Ecosystems: Trends in Terrestrial and Aquatic
Research, F.B. Golley and E. Medina (eds). Springer-Verlag Inc., New York.
Lamotte, M. and F. Bourliere. 1983. Energy flow and nutrient cycling in
tropical savannas, pp. 583-603. In Ecosystems of the World 13:Tropical
Savannas, F. Bourliere (ed.). Elsevier, Amsterdam, The Netherlands.
Lavenu, F. 1987. Vegetation Map of Africa 1:5,000,000. Institue de la
Carte Internationale de la Vegetation, Universitie Paul Sabatier, Toulouse,
France. Distributed by FAO Forest Resources Assessment Project-1990, Rome,
Italy.
Martin, C. 1991. The Rainforests of West Africa. Birkhauser Verlag,
Basel, Switzerland.
Menaut, J.C., L. Abbadie, F. Lavenu, P. Loudjami, and A. Podaire. 1991.
Biomass burning in west African Savannas, pp. 133-142. In Global Biomass
Burning: Atmospheric, Climatic and Biospheric Implications, J.S.
Levine (ed). MIT Press, Cambridge, Massachusetts.
Olson, J.S., J.A. Watts, and L.J. Allison. 1983. Carbon in Live Vegetation
of Major World Ecosystems. TR004, U.S. Department of Energy, Office of
Energy Research, Washington, D.C.
Olson, J.S., R.M. Garrles, R.A. Berner, T.V. Armentano, M.I. Dyer, and D.H.
Yaalon. 1985. The natural carbon cycle, pp. 175-214. In Atmospheric
Carbon Dioxide and the Global Carbon Cycle, J.R. Trabalka (ed.).
DOE/ER-0239. U.S. Department of Energy, Office of Energy Research,
Washington, D.C.
Pierlot, R. 1966. Structure et composition de forets denses d'Afrique,
especialement celles du Kivu. Academie Royale des Sciences d'Outre-mer,
Classes des Sciences naturelles et Medicales N.S.-XVI-4, Bruxelles, Belgium.
Rand McNally Co. 1980. The new International Atlas. Rand McNally,
Chicago, Illinois.
Republique de Cote d'Ivoire. 1975. Analyse et Commentaires des Resultats
de l'Inventaire Forestier de la Region Nord-Ouest: Resultats de la Region du
Nord-Ouest. Inventaire Forestier National, Direction des Inventaire et
Amenagement, Abidjan, Ivory Coast.
Smith, K.R. 1991. Biomass cook stoves in global perspectives: Energy, health,
and global warming. In The Earth as System Central. EWC/ESMAP/UNDP
Report 3, The World Bank, Washington, D.C.
Udo, R.K. 1982. The Human Geography of Tropical Africa. Heinemann
Educational Books, London, United Kingdom.
United Nations. 1972. Forestry Inventory of the Gola Forest Reserves:
Report to the Government of Sierra Leone. FAO, Rome, Italy.
UNESCO/AETFAT/UNSO. 1981. Vegetation Map of Africa 1:5,000,000. Oxford
University Press, Hammond and Kell Ltd., Mitcham, Surrey, United Kingdom.
Weck, J. 1970. An improved CVP-index for the delimitation of the potential
productivity zones of forest lands of India. Indian Forester 96: 565-572.
White, F. 1983. The vegetation of Africa: A descriptive memoir to accompany
the Unesco/AETFAT/UNSO vegetation map of Africa. UNESCO, United Nations,
New York.
Worldmark Press, Ltd. 1984. Encyclopedia of the Nations, vol. 2
(Africa). John Wiley and Sons Inc., New York.
7. HOW TO OBTAIN THE DATA AND DOCUMENTATION
-------------------------------------------
These data may be used with a vector or raster geographic information system
(GIS) or non-GIS database systems. This database (NDP-055) is available free
of charge from CDIAC. The files are available, via the Internet, from CDIAC's
World-Wide-Web site (http://cdiac.esd.ornl.gov), or from CDIAC's anonymous
file transfer protocol (FTP) area (cdiac.esd.ornl.gov) as follows:
1. FTP to cdiac.esd.ornl.gov (128.219.24.36).
2. Enter "ftp" as the user id.
3. Enter your electronic mail address as the password (e.g., fred@zulu.org).
4. Change to the directory "pub/ndp055" (i.e., use the command "cd
pub/ndp055").
5. Set ftp to get ASCII files by using the ftp "ascii" command.
6. Retrieve the ASCII database documentation file by using the ftp "get
ndp055.txt" command.
7. Retrieve the ASCII data files by using the ftp "mget *.asc" command.
8. Retrieve the ASCII FORTRAN files by using the ftp "mget *.for" command.
9. Retrieve the ASCII SAS (TM) files by using the ftp "mget *.sas"
command.
10. Set ftp to get BINARY files by using the ftp "binary" command.
11. Retrieve the binary ARC/INFO (TM) export files by using the ftp
"mget *.e00" command.
12. Exit the system by using the ftp "quit" command.
For non-Internet data acquisitions (e.g., floppy diskette or CD-ROM) or for
additional information, contact:
Carbon Dioxide Information Analysis Center
Oak Ridge National Laboratory
P.O. Box 2008
Oak Ridge, Tennessee 37831-6335, U.S.A.
Telephone: 1-865-574-3645
Telefax: 1-865-574-2232
E-mail: cdiac@ornl.gov
8. LISTING OF FILES PROVIDED
----------------------------
This database consists of 36 files: two *.txt files seven *.e00 GIS data
files, one *prj projection file, seven *.asc data files, four FORTRAN input
statement files, and four SAS input statement files, and 13 *.gif graphics
files. See Table 1 (table.txt) for a complete listing and description of
these files.
9. DESCRIPTION OF THE DOCUMENTATION FILE
----------------------------------------
The ndp055.txt (File 1) file is an ASCII text file (this document).
10. DESCRIPTION, FORMAT, AND PARTIAL LISTINGS OF THE DATA FILES
---------------------------------------------------------------
Table 1 lists and describes the files included in this numeric data package.
FLAT ASCII DATA FILES:
Eight flat ASCII data files are provided with this data package. These files
contain the same information as the ARC INFO grids and coverages included in
this NDP. The ASCII files were either created directly from the ARC/INFO grids
of the same name using the GRIDASCII command or are ASCII versions of ARC/INFO
attribute tables associated with polygon coverages included in this NDP. SAS
and FORTRAN programs designed to read and print the contents of each ASCII
data file are also provided. The information within these files will allow the
user to ascertain population densities, the amount of land remaining in forests
in 1980, the maximum potential aboveground biomass densities within Africa, and
the biomass in aboveground woody vegetation in Tropical Africa in 1980. What
follows is a description of the contents and formats (e.g., variable names,
widths) of the eight flat ASCII data files distributed with this data package.
The flat ASCII files BIOMAX.ASC (File 4), LAKES.ASC (File 6), and LAND.ASC
(File 8) are formatted as shown in the following FORTRAN code and may
be read using the input/output programs INTGRID.FOR (File 9) or INTGRID.SAS
(File 10).
READ (5,100) NCOLS
READ (5,100) NROWS
READ (5,110) XLLCORNR
READ (5,110) YLLCORNR
READ (5,100) CSIZE
READ (5,100) NODATA
DO 10 ROWS = 1, NROWS
READ (5,120) (VALUE(ROWS,COLS), COLS=1, NCOLS)
10 CONTINUE
C
100 FORMAT(14X,I5)
110 FORMAT(14X,F15.5)
120 FORMAT(1534(1X,I5))
BIOMAX.ASC contains maximum potential biomass density data in Mg/ha; LAND.ASC
contains land-use information classified into four land-use categories; and
LAKES.ASC contains the location of all non-seasonal lakes >20 km in any
direction in Tropical Africa. Each file of these three files contains header
information containing a set of key words followed by cell (i.e., data) values
in row-major order. The first six lines of BIOMAX.ASC, LAKES.ASC, and LAND.ASC
(header information) are:
ncols 1534
nrows 1502
xllcorner -3616156.5
yllcorner -3644063.04465
cellsize 5000
NODATA_value -9999
The data are in a normal (type 1) cylindrical equal-area projection with
a central meridian of 15 degrees East longitude and a standard parallel of 0
degrees latitude. Data are provided for 1,502 rows and 1,534 columns, with the
first cell being located in the upper-left corner. Each data cell covers a 25
km2 square area, 5,000 m per side. The area covered by the data is bounded by
a box with the following coordinates: lower-left -3616156.500, -3644063.04465;
upper-right 4053843.500, 3865936.955.
The variables in BIOMAX.ASC (File 4), LAKES.ASC (File 6), and LAND.ASC (File 8)
are formatted as listed in Table 4 and are shown in the same order as they
would appear in the files.
The flat ASCII files PD80.ASC (File 14) and BIO1980.ASC (File 12) are formatted
as shown in the following FORTRAN code and may be read using the input/output
programs REALGRID.FOR (File 15) or REALGRID.SAS (File 16).
READ (5,100) NCOLS
READ (5,100) NROWS
READ (5,110) XLLCORNR
READ (5,110) YLLCORNR
READ (5,100) CSIZE
READ (5,120) NODATA
DO 10 ROWS = 1, NROWS
READ (5,130) (VALUE(ROWS,COLS), COLS=1, NCOLS)
10 CONTINUE
C
100 FORMAT(14X,I5)
110 FORMAT(14X,F15.5)
120 FORMAT(14X,F8.2)
130 FORMAT(1534(1X,F8.2))
PD80.ASC contains population-density data for 1980 in people/km2 and
BIO1980.ASC contains the calculated "actual" aboveground live biomass in open
and closed forest for 1980 in Mg/ha. Each of these two files contains header
information with a set of key words followed by cell values in row-major order.
The first six lines of PD80.ASC, and BIO1980.ASC (header information) are :
ncols 1534
nrows 1502
xllcorner -3616156.5
yllcorner -3644063.04465
cellsize 5000
NODATA_value -9999
The data are in a normal (type 1) cylindrical equal-area projection with
a central meridian of 15 degrees East longitude and a standard parallel of 0
degrees latitude. Data are provided for 1,502 rows and 1,534 columns with the
first cell located in the upper-left corner. Each data cell covers a square
area, 5,000 m per side. The area covered by the data is bounded by a box with
the following coordinates: lower-left -3616156.500, -3644063.04465; upper-right
4053843.500, 3865936.955.
The variables in PD80.ASC (File 14) and BIO1980.ASC (File 12) are formatted as
listed in Table 5 and are shown in the same order as they would appear in the
files.
The flat ASCII file T_AFR.ASC (File 22) contains population density data for
1960, 1970, 1980, and 1990 for the subnational political units used in this
database. The data were provided by M. Lorenzini and K. D. Singh (FAO,
Tropical Forest Resources Assessment Program 1990). The file is formatted as
shown in the following FORTRAN code and may be read using the input/output
programs T_AFR.FOR (File 23) or T_AFR.SAS (File 24).
10 READ (5,100,END=999) COUNTRY, COUNTRY_NO, ADMP, ADMP_NM,
1 PD60, PD70, PD80, PD90, LANDAREA
GOTO 10
C
100 FORMAT(A30,I5,I4,A40,4F8.3,I10)
The variables in T_AFR.ASC (File 22) are formatted as listed in Table 6 and are
shown in the same order as they appear in the file.
The first ten data records of t_afr.asc are:
Water 0 0 0 0 0 0 0
Mauritania 136 287 TIRIS_ZEMOUR 0.058 0.077 0.106 0.148 252900
Mali 133 275 TOMBOUCTOU -1 -1 1.215 1.236 408977
Water 0 0 0 0 0 0 0
Niger 158 306 AGADEZ -1 0.111 0.196 0.334 714790
Chad 39 74 BET 0.119 0.138 0.147 0.17 600350
Mauritania 136 283 ADRAR 0.326 0.301 0.253 0.161 215300
Mauritania 136 278 HODH_CHARKI 0.986 0.953 0.877 0.71 182700
Sudan 206 387 NORTH_REGION -1 1.942 2.176 2.339 477405
Sudan 206 388 EAST_REGION -1 4.28 5.952 8.273 334331
The last ten data records of t_afr.asc are:
Mozambique 144 296 INHAMBANE 9.594 12.542 14.517 17.591 68615
Botswana 20 18 KGALAGADI 0.142 0.157 0.217 0.286 106940
Botswana 20 20 KWENENG 1.9 2.027 3.096 4.419 35890
Botswana 20 19 KGATLENG 3.57 4.295 5.445 6.757 7960
Botswana 20 23 NGWAKETSE 2.246 2.812 3.863 5.237 27370
Mozambique 144 298 CID_MAPUTO_MAPUTO_PROV 18.545 34.664 47.177 69.288 26358
Botswana 20 24 BAROLONG_SOUTH_EAST 10.397 21.905 40.573 71.489 2880
Mozambique 144 298 CID_MAPUTO_MAPUTO_PROV 18.545 34.664 47.177 69.288 26358
Zimbabwe 181 337 MATABELELAND_NORTH -1 8.159 11.329 15.806 73537
Zimbabwe 181 339 MIDLANDS -1 13.191 17.577 23.54 58967
The flat ASCII file AFRICA.ASC (File 19) contains the amount of land and
biomass in open and closed forest by country for 1980. The file is formatted
as shown in the following FORTRAN code and may be read by using the
input/output programs AFRICA.FOR (File 20) or AFRICA.SAS (File 21).
10 READ (5,100,END=999) COUNTRY, CNTRY_NO, TOTCELL,
1 TOTAREA, CFAREA, OFAREA, CFBIO, OFBIO,
1 TOTBIO, FOREST
GOTO 10
C
100 FORMAT(A30,I5,I11,I10,2I9,2F12.1,F12.4,F12.1)
The variables in AFRICA.ASC (File 19) are formatted as listed in Table 7 and
are shown in the same order as they appear in the file.
The first ten data records of africa.asc are:
Angola 7 50127 1253175 66750 1063775 109.4 69.7 8144.76 90.2
Botswana 20 23195 579875 0 101800 0 13.1 133.36 17.6
Burundi 29 1021 25525 2575 6900 57.1 40.3 42.51 37.1
Cameroon 32 18529 463225 221750 182500 297.5 120 8787.06 87.3
C African Republic 37 24856 621400 105425 515850 262.9 185.7 12350.96 100
Chad 39 50635 1265875 0 316825 0 38 1203.94 25
Congo 46 13841 346025 247675 375 344.9 36.4 8543.68 71.7
Benin 53 4664 116600 4325 103675 75.6 56.4 617.42 92.6
Equatorial Guinea 61 1072 26800 24725 0 318 0 786.26 92.3
Ethiopia 62 49920 1248000 375450 325400 65.3 35.6 3610.11 56.2
The last ten data records of africa.asc are:
Senegal 195 7916 197900 13500 63975 52.8 25.8 236.34 39.1
Sierra Leone 197 2899 72475 30875 28400 220.6 174.9 1177.82 81.8
Somalia 201 25573 639325 350 165675 11.4 12.4 205.84 26
Sudan 206 99743 2493575 5425 737150 134 63 4716.74 29.8
Tanzania 215 35629 890725 5550 529800 73.5 48.6 2615.62 60.1
Togo 217 2297 57425 725 49575 40.9 71.7 358.42 87.6
Uganda 226 8204 205100 12075 131825 124 101.7 1490.39 70.2
Burkina Faso 233 10957 273925 0 97375 0 34 331.08 35.5
Zaire 250 92433 2310825 1860300 378175 223.9 118.2 46122.15 96.9
Zambia 251 29846 746150 47250 579650 45.9 52.3 3248.45 84
ARC/INFO (Version 7.2.1) EXPORT FILES
Seven ARC/INFO export files (*.e00) are provided with this NDP. These files
contain the same data as the flat ASCII data files described previously. The
ARC/INFO export (*.e00) files were produced using the EXPORT command with the
NONE option in ARC/INFO Version 7.2.1. These files are in an uncompressed,
fixed-block format that may be moved across computer systems. The coverages
are in a normal (type 1) cylindrical equal-area projection with a central
meridian of 15 degrees East longitude and a standard parallel of 0 degrees
latitude. All units are in meters and the spheroid is a sphere with a radius
of 6,370,997 meters (vs. the Clark 1866 spheroid).
The following files (BIOMAX.E00, LAND.E00, PD80.E00, LAKES.E00 and BIO1980.E00)
should be imported with the GRID option and contain 1,502 rows and 1,534
columns, with each data cell covering a square area with 5,000m per side. The
remaining files (T_AFR.E00 and AFRICA.E00) are ARC/INFO polygon coverages and
should be imported using the COVER option. The polygon coverage AFRICA
contains regions. The region subclass within AFRICA is named SUM1980. These
regions were built using countries within Tropical Africa as unique
identifiers. The AFRICA.PAT.SUM1980 INFO file contains summary information by
country for the amount of land and biomass in open and closed forests in 1980.
The area covered by the data is bounded by a box with the following
coordinates: lower-left -3616156.500, -3644063.04465; upper-right 4053843.500,
3865936.955. A listing of the variable names and data formats for these last
two export (E00) files are shown in Tables 6, and 7.
The first ten data records of BIOMAX.E00 are:
EXP 0 /DATA4/OZ1/CDIAC/NDP-055/NEW_E00/BIOMAX.E00
GRD 2
1534 1502 1-0.21474836470000E+10
0.50000000000000E+04 0.50000000000000E+04
-0.36161565000000E+07-0.36440630446500E+07
0.40538435000000E+07 0.38659369553500E+07
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
The last ten data records of BIOMAX.E00 are:
42 90055
75 91074
150 99339
250 81883
350 64032
425 37550
475 17970
500 2264
EOI
EOS
The first ten data records of LAND.E00 are:
EXP 0 /DATA4/OZ1/CDIAC/NDP-055/NEW_E00/LAND.E00
GRD 2
1534 1502 1-0.21474836470000E+10
0.50000000000000E+04 0.50000000000000E+04
-0.36161565000000E+07-0.36440630446500E+07
0.40538435000000E+07 0.38659369553500E+07
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
The last ten data records of LAND.E00 are:
LAND.VAT XX 2 2 8 5
VALUE 4-1 14-1 10-1 50-1 -1 -1-1 1I
COUNT 4-1 54-1 10-1 50-1 -1 -1-1 2-
-9999 1105588
1 175811
2 346001
3 157539
4 519129
EOI
EOS
The first ten data records of PD80.E00 are:
EXP 0 /DATA4/OZ1/CDIAC/NDP-055/NEW_E00/PD80.E00
GRD 2
1534 1502 2-0.34028234663853E+39
0.50000000000000E+04 0.50000000000000E+04
-0.36161565000000E+07-0.36440630446500E+07
0.40538435000000E+07 0.38659369553500E+07
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
The last ten data records of PD80.E00 are:
IFO 2
PD80.STA XX 4 4 32 1
MIN 8-1 14-1 15 3 60-1 -1 -1-1 1-
MAX 8-1 94-1 15 3 60-1 -1 -1-1 2-
MEAN 8-1 174-1 15 3 60-1 -1 -1-1 3-
STDV 8-1 254-1 15 3 60-1 -1 -1-1 4-
-9.99900000000000000E+03 2.19550610351562500E+03-4.78676211358813089E+03 5.00127
655918892469E+03
EOI
EOS
The first ten data records of LAKES.E00 are:
EXP 0 /DATA4/OZ1/CDIAC/NDP-055/NEW_E00/LAKES.E00
GRD 2
1534 1502 1-0.21474836470000E+10
0.50000000000000E+04 0.50000000000000E+04
-0.36161565000000E+07-0.36440630446500E+07
0.40538435000000E+07 0.38659369553500E+07
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
-9999 -9999 -9999 -9999 -9999
The last ten data records of LAKES.E00 are:
STDV 8-1 254-1 15 3 60-1 -1 -1-1 4-
-9.99900000000000000E+03 4.00000000000000000E+00-9.95541616610273741E+03 6.58839
540326123597E+02
LAKES.VAT XX 2 2 8 2
VALUE 4-1 14-1 10-1 50-1 -1 -1-1 1I
COUNT 4-1 54-1 10-1 50-1 -1 -1-1 2-
-9999 2294029
4 10039
EOI
EOS
The first ten data records of BIO1980.E00 are:
EXP 0 /DATA4/OZ1/CDIAC/NDP-055/NEW_E00/BIO1980.E00
GRD 2
1534 1502 2-0.34028234663853E+39
0.50000000000000E+04 0.50000000000000E+04
-0.36161565000000E+07-0.36440630446500E+07
0.40538435000000E+07 0.38659369553500E+07
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04-0.9999000E+04
The last ten data records of BIO1980.E00 are:
IFO 2
BIO1980.STA XX 4 4 32 1
MIN 8-1 14-1 15 3 60-1 -1 -1-1 1-
MAX 8-1 94-1 15 3 60-1 -1 -1-1 2-
MEAN 8-1 174-1 15 3 60-1 -1 -1-1 3-
STDV 8-1 254-1 15 3 60-1 -1 -1-1 4-
-9.99900000000000000E+03 4.66339996337890625E+02-7.86027052510779868E+03 4.13001
816239345681E+03
EOI
EOS
11. SAS AND FORTRAN CODES TO ACCESS THE DATA
--------------------------------------------
BIOMAX.ASC (File 4), LAKES.ASC (File 6), and LAND.ASC (File 8) may be read
using the input/output programs INTGRID.FOR (File 9) or INTGRID.SAS (File 10).
INTGRID.FOR (File 9):
real xllcornr, yllcornr
integer ka
integer ncols, nrows, csize, nodata
dimension ka(2304068)
open(unit=5,file='biomax.asc',status='old',access='sequential')
open(unit=6,file='biomax.out',status='new',access='sequential')
read (5,100) ncols
read (5,100) nrows
read (5,110) xllcornr
read (5,110) yllcornr
read (5,100) csize
read (5,120) nodata
write (6,200) ncols, nrows, xllcornr, yllcornr,
1 csize, nodata
do k=1,1502
read (5,*) (ka(j),j=1,1534)
write(6,*) (ka(j),j=1,1534)
enddo
continue
100 format (14x,i5)
105 format (14x,f10.1)
110 format (14x,f14.5)
120 format (14x,i5)
200 format (1x,i5,/,1x,i5,/,1x,f10.1,/,1x,f15.5,/,1x,i5,/,1x,i5)
999 close(unit=5)
close(unit=6)
stop
end
INTGRID.SAS (File 10):
data in;
infile 'path and file name' firstobs=7 linesize=9203;
input varname@@;
run;
The flat ASCII files PD80.ASC (File 14) and BIO1980.ASC (File 12) may be read
using the input/output programs REALGRID.FOR (File 15) or REALGRID.SAS
(File 16).
REALGRID.FOR (File 15):
real xllcornr, yllcornr, xa
integer ncols, nrows, csize, nodata
dimension xa(2304068)
open(unit=5,file='bio1980.asc',status='old',access='sequential')
open(unit=6,file='bio1980.out',status='new',access='sequential')
read (5,100) ncols
read (5,100) nrows
read (5,110) xllcornr
read (5,110) yllcornr
read (5,100) csize
read (5,120) nodata
write (6,200) ncols, nrows, xllcornr, yllcornr,
1 csize, nodata
do x=1,1502
read (5,*) (xa(j),j=1,1534)
write(6,*) (xa(j),j=1,1534)
enddo
continue
100 format (14x,i5)
105 format (14x,f10.1)
110 format (14x,f14.5)
120 format (14x,i5)
200 format (1x,i5,/,1x,i5,/,1x,f10.1,/,1x,f15.5,/,1x,i5,/,1x,i5)
999 close(unit=5)
close(unit=6)
stop
end
REALGRID.SAS (File 16):
data in;
infile 'path and file name' firstobs=7 linesize=13805;
input varname@@;
run;
The flat ASCII file T_AFR.ASC (File 22) may be read using the input/output
programs T_AFR.FOR (File 23) or T_AFR.SAS (File 24).
T_AFR.FOR (File 23):
character cntry*30, aname*40
integer cntryn, admp, larea
real pd60, pd70, pd80, pd90
open(unit=5,file='t_afr.asc',status='old',access='sequential')
open(unit=6,file='t_afr.out',status='new',access='sequential')
5 read (5,100, end=999) cntry, cntryn, admp, aname, pd60, pd70,
1 pd80, pd90, larea
10 write (6,100) cntry, cntryn, admp, aname, pd60, pd70,
1 pd80, pd90, larea
goto 5
100 format (a30,1x,i5,1x,i4,1x,a39,1x,f8.3,1x,f8.3,
1 1x,f8.3,1x,f8.3,1x,i8)
999 close(unit=5)
close(unit=6)
stop
end
T_AFR.SAS (File 24):
data in;
infile 't_afr.asc';
input country $ 1-30 cntry_no 32-36 admp 38-41
admp_nm 43-82 pd60 84-91 pd70 93-100
pd80 102-109 pd90 111-118 landarea 120-129;
proc print;
run;
The flat ASCII file AFRICA.ASC (File 19) may be read by using the input/output
programs AFRICA.FOR (File 20) or AFRICA.SAS (File 21).
AFRICA.FOR (File 20):
character cntry*30
integer cntryn, tcell, tarea, carea, oarea
real cbio, obio, tbio, forest
open(unit=5,file='africa.asc',status='old',access='sequential')
open(unit=6,file='africa.out',status='new',access='sequential')
5 read (5,100, end=999) cntry, cntryn, tcell, tarea, carea,
1 oarea, cbio, obio, tbio, forest
10 write (6,100) cntry, cntryn, tcell, tarea, carea,
1 oarea, cbio, obio, tbio, forest
goto 5
100 format (a30,1x,i5,1x,i11,1x,i10,1x,i9,1x,i9,
1 1x,f12.1,1x,f12.1,1x,f12.2,1x,f12.1)
999 close(unit=5)
close(unit=6)
stop
end
AFRICA.SAS (File 21):
data in;
infile 'africa.asc';
input country $ 1-30 cntry_no 32-36 tcell 38-48
tarea 50-60 carea 62-70 oarea 72-80
cbio 82-95 obio 97-110 tbio 112-126
forest 128-41;
proc print;
run;
------------------------------------------------------------------------------
APPENDICES
------------------------------------------------------------------------------
APPENDIX A - BREAKDOWN BY COUNTRY OF THE AMOUNT OF LAND AND BIOMASS IN OPEN
AND CLOSED FOREST IN 1980 (see ndp055appA.pdf)
------------------------------------------------------------------------------
APPENDIX B - REPRINT OF PERTINENT LITERATURE
Use of Forest Inventories and Geographic Information Systems To Estimate
Biomass Density of Tropical Forests: Application to Tropical Africa
S. Brown and G. Gaston
U.S. Environmental Protection Agency,
200 SW 35th St.,
Corvalis, OR 97333, USA
ABSTRACT
--------
One of the most important databases needed for estimating emissions of carbon
dioxide resulting from changes in the cover, use, and management of tropical
forests is the total quantity of biomass per unit area, referred to as biomass
density. Forest inventories have been shown to be valuable sources of data for
estimating biomass density, but inventories for the tropics are few in number
and their quality is poor. This lack of reliable data has been overcome by use
of a promising approach that produces geographically referenced estimates by
modeling in a geographic information system (GIS). This approach has been
used to produce geographically referenced, spatial distributions of potential
and actual (circa 1980) aboveground biomass density of all forest types in
tropical Africa. Potential and actual biomass density estimates ranged from
33 to 412 Mg/ha (10E06 g/ha) and 20 to 299 Mg/ha, respectively, for
very dry lowland to moist lowland forests and from 78 to 197 Mg/ha and 37
to 105 Mg/ha, respectively, for montane-seasonal to montane-moist forests.
Of the 37 countries included in this study, more than half (51%) contained
forest that had less than 60% of their potential biomass. Actual biomass
density for forest vegetation was lowest in Botswana, Niger, Somalia, and
Zimbabwe (about 10 to 15 Mg/ha). Highest estimates for actual biomass
density were found in Congo, Equatorial Guinea, Gabon, and Liberia (305 to
344 Mg/ha). Results from this research effort can contribute to reducing
uncertainty in the inventory of country-level emission by providing consistent
estimates of biomass density at subnational scales that can be used with other
similarly scaled databases on change in land cover and use.
Complete citation:
Brown, S. and G. Gaston. 1995. Use of Forest Inventories and Geographic
Information Systems to Estimate Biomass Density of Tropical Forests:
Applications to Tropical Africa. Environmental Monitoring and Assessment
38:157-168.
Copyright 1995 Kluwer Academic Publishers, Printed in the Netherlands
This abstract is provided with permission of Kluwer Academic Publishers.
Visit them online at http://www.kluweronline.com
-----------------------------------------------------------------------------
APPENDIX C - RECLASSIFICATION SCHEME USED TO OBTAIN THE FOUR LAND USE
CATEGORIES USED IN THIS NUMERIC DATA PACKAGE.
The 64 land cover classes in the original FAO land cover map
(Lavenu 1987) were combined using the reclassification scheme shown here to
obtain the 4 land use categories used in this numeric data package.
Land use FAO class number and land cover description
Category
Closed Forest Dense Forest
1. Dense Forest (1a)
2. Dense Forest-Dense Tree Thicket (1a-1b)
3. Dense Forest-Dense Tree Thicket-Forest Fallow (1a-1b/6)
4. Dense Forest-Miombo Woodland (1a-2a)
5. Dense Forest-Miombo Woodland-Forest Fallow (1a-2a/6)
6. Dense Forest-Discontinuous Tree/Shrub Thicket (1a-2b)
7. Dense Forest-Discontinuous Tree/Shrub Thicket-
Forest Fallow (1a-2b/6)
8. Dense Forest-Discontinuous Tree/Shrub Thicket-
Savanna Fallow (1a-2b/7)
9. Dense Forest-Tree/Woodland Savanna (1a-2c)
10. Dense Forest-Tree/Woodland Savanna-Cropland/
Plantation (1a-2c)
11. Dense Forest-Grass Savanna (1a-3a)
13. Dense Forest-Forest Fallow (1a/6)
14. Dense Forest-Forest Fallow-Miombo Woodland (1a/6-2a)
15. Dense Forest-Forest Fallow-Tree/Woodland Savanna (1a/6-2c)
Dense Tree Thicket
16. Dense Tree Thicket (1b)
17. Dense Tree Thicket-Discontinuous Tree/Shrub Thicket (1b-2b)
Open Forest Woodland
19. Miombo Woodland (2a)
20. Miombo Woodland-Tree/Woodland Savanna (2a-2c)
21. Miombo Woodland-Tree/Woodland Savanna-Forest
Fallow (2a-2c/b)
22. Miombo Woodland-Shrub Savanna/Steppe (2a-2d)
23. Miombo Woodland-Grass Savanna (2a-3a)
24. Miombo Woodland-Grass Steppe (2a-3b)
26. Miombo Woodland-Forest Fallow (2a/6)
Discontinuous Trees
27. Discontinuous Tree/Shrub Thicket (2b)
28. Discontinuous Tree/Shrub Thicket Tree/Woodland
Savanna (2b-2c)
29. Discontinuous Tree/Shrub Thicket Tree/Woodland
Savanna Forest Fallow (2b-2c/6)
30. Discontinuous Tree/Shrub Thicket Tree/Woodland
Savanna Savanna Fallow (2b-2c/7)
31. Discontinuous Tree/Shrub Thicket Shrub
Savanna/Steppe (2b-2d)
32. Discontinuous Tree/Shrub Thicket Shrub
Savanna/Steppe Savanna Fallow (2b-2d/7)
33. Discontinuous Tree/Shrub Thicket Grass Savanna (2b-3a)
34. Discontinuous Tree/Shrub Thicket Grass Steppe (2b-3b)
35. Discontinuous Tree/Shrub Thicket Cropland/Plantation (2b/5)
36. Discontinuous Tree/Shrub Thicket Forest Fallow (2b/6)
Woodland Savanna
37. Tree/Woodland Savanna (2c)
38. Tree/Woodland Savanna Discontinuous Tree/Shrub
Thicket (2c-2b)
39. Tree/Woodland Savanna Shrub Savanna/Steppe (2c-2d)
40. Tree/Woodland Savanna Grass Savanna (2c-3a)
41. Tree/Woodland Savanna Grass Steppe (2c-3b)
42. Tree/Woodland Savanna Cropland/Plantation (2c/5)
43. Tree/Woodland Savanna Forest Fallow (2c/6)
44. Tree/Woodland Savanna Savanna Fallow (2c/7)
Shrub Savanna
45. Shrub Savanna/Steppe (2d)
46. Shrub Savanna/Steppe Grass Savanna (2d-3a)
47. Shrub Savanna/Steppe Grass Steppe (2d-3b)
48. Shrub Savanna/Steppe Pseudodesertic Steppe (2d-4a)
50. Shrub Savanna/Steppe Cropland/Plantation Grass
Savanna (2d/5-3a)
51. Shrub Savanna/Steppe Savanna Fallow (2d/7)
Grassland Grass Savanna
Savanna 52. Grass Savanna (3a)
53. Grass Savanna Shrub Savanna/Steppe (3a-2d)
55. Grass Savanna Savanna Fallow (3a/7)
Grass Steppe
56. Grass Steppe (3b)
57. Grass Steppe Pseudodesertic Steppe (3b-4a)
58. Grass Steppe Cropland/Plantation (3b/5)
58. Grass Steppe Cropland/Plantation (3b/5)
Other Agriculture/Plantation
12. Dense Forest-Cropland/Plantation (1a/5)
18. Dense Tree Thicket-Forest Fallow (1b/6)
25. Miombo Woodland-Cropland/Plantation (2a/5)
49. Shrub Savanna/Steppe-Cropland/Plantation (2a/5)
54. Grass Savanna-Cropland/Plantation (3a/5)
62. Cropland/Plantation (5)
63. Cropland/Plantation-Forest Fallow (5/6)
64. Cropland/Plantation-Savanna Fallow (5/7)
Desert
59. Pseudodesertic Steppe (4a)
60. Desert (4b)
61. Halophytic Soil (4c)
Source map: Lavenu, F. 1987. Vegetation map of Africa 1:5,000,000.
Institute de la Carte Internationale de la Vegetation, Universitie Paul
Sabatier, Toulouse, France. Distributed by FAO Forest Resources Assessment
Project-1990, Rome, Italy.