|
Significance of Asian Tropical Forest Conversion to Imbalances in the Global Carbon Budget
Contents:
David L. Skole, Basic Science and Remote Sensing Initiative, Michigan State University, East Lansing, Michigan William A. Salas, Institute for the Study of Earth, Oceans, and Space, University of New Hampshire, Durham, New Hampshire Chaowalit Silapathong, National Research
Council of Thailand, Bangkok, Thailand
The Contemporary Global Carbon Cycle Carbon is an essential element for life on this planet; approximately half of the mass of plant and animal life is composed of carbon. Carbon exists in several forms and is cycled between several biotic and abiotic reservoirs, including the oceans, terrestrial plant biota, and atmosphere (Figure 7.1). The largest pool of carbon is the lithosphere, but it is relatively inactive. On the other hand, the carbon content of the atmosphere has varied considerably over past several thousand years, and this variation has been closely coupled to changes in global climate. Variations in atmospheric carbon dioxide are related to changing fluxes of carbon between the atmosphere and the biota, oceans, and fossil fuels. Since atmospheric carbon dioxide measurements were first systematically made in the late 1950s (Keeling 1960, Keeling et al.. 1976) the concentration has steadily increased (Bacastow and Keeling 1981, Keeling and Whorf 1994). This increase in atmospheric carbon is attributed to two sources: land cover change, and combustion of fossil fuels (Houghton and Skole 1990, Schimel et al. 1995). Fossil fuels have been injected into the atmosphere at a steadily increasing rate since the beginning of the industrial revolution, around 1860 (Marland and Rotty 1984, Keeling 1994, Marland et al. 1994). Carbon has been released from deforestation and other land cover changes over the last 300 years (Houghton and Skole 1990, Houghton 1994). While the annual increase in carbon content
of the atmosphere and the amount of carbon released from fossil fuel combustion
are relatively well known, the exact magnitude of fluxes between the atmosphere
and the biota, and between the atmosphere and oceans, is uncertain. These
uncertainties confound our ability to construct a balanced global budget
for carbon.
Imbalances in the Global Carbon Cycle Reviews of the global carbon cycle can be found in Houghton and Skole (1990) and Schimel et al.. (1995) and we shall cover here only the more salient issues of current uncertainties in the terrestrial carbon cycle. Atmospheric carbon dioxide has increased from 315 ppmv to 356 ppmv in 1992 (Keeling et al. 1989, Thoning et al. 1994, Keeling and Whorf 1994). With the development of techniques to analyze the concentration of CO2 in bubbles of air trapped in deep ice cores, it is also possible to estimate the historical atmospheric concentration of CO2. Since 1750 there has been a steady increase, from 279 ppmv (Neftel et al. 1985). Human activities are largely responsible for the observed increases. Fossil fuel burning is currently the most important source of carbon dioxide. Yet evidence from ice cores suggests atmospheric concentrations of CO2 began to rise prior to 1850, even before significant inputs from fossil fuel combustion. Approximately 30% of the current net flux of carbon dioxide is biogenic and related to anthropogenic land cover changes. Moreover, the long-term historical release of carbon dioxide from human alterations of land cover has been approximately equal to the release from fossil fuel burning (Houghton and Skole 1990). Net additions to the atmosphere from fossil fuel combustion and biomass oxidation are partly removed by oceanic uptake. However, oceanic uptake from the atmosphere is lower than additions to it, and the atmospheric concentration has risen as a result of this imbalance. Measurements of the atmospheric increase when compared to estimates of the net flux of biogenic carbon, fossil fuel emissions, and ocean uptake cannot be accommodated in a balanced carbon budget. Houghton et al. (1983) and Houghton and Skole (1990) have described this situation using a simple equation of the form: D At = B + F - O (1) where ?At is the net increase in the atmospheric carbon at time, t, B is the net flux from land clearing and biomass burning, F is the net flux from fossil fuel burning, and O is the net oceanic flux. The units are 1015 g C year-1. When average annual values for individual terms from the period 1980 to 1990 are substituted into the equation (Schimel et al. 1995), it is possible to describe the current situation: 3.2 ¹ 1.7 + 5.5 - 2.0 (2) There are important uncertainties associated with each of these terms. Nonetheless, the equation is not balanced. Explanations for this imbalance in the budget center on two possibilities. First, the individual terms of the equation may be in error, or improperly modeled -- particularly B and O. The emission from fossil fuels is well known and certain, as are the atmospheric measurements. However, the oceanic and biogenic terms are not. Second, the equation may not be complete, and another term -- the so-called missing sink -- may be important, and hence: 3.2 = 1.7 + 5.5 - 2.0 - b ???(3) The recent report of the Intergovernmental Panel on Climate Change (Houghton et al. 1995) suggests several possible mechanisms to explain b ??including increased forest C storage due to climate-induced changes in ecosystem metabolism, fertilization by carbon dioxide or nitrogen, or unaccounted land use changes such as forest regrowth. At any rate, the value of this term is rather large (~ 2) and probably cannot be attributed to errors in any one of the other terms alone. Recent analyses by Tans et al. (1990) and others have suggested a large, unaccounted net sink in undisturbed, predominantly northern, forest ecosystems. The analysis of Tans et al. (1990) utilized a constrained approach, coupling measurements of the north-south gradient of atmospheric [CO2], measurements of oceanic uptake and pCO2, estimates of land surface fluxes from fossil fuels and deforestation, and a general circulation model. The approach provides a geographically-specific assessment of sources and sinks which is highly dependent on the magnitude of the uncertain biogenic tropical source term. In another constrained analysis, Houghton and Skole (1990) compared a model estimate of the historic net flux of carbon from land cover change with estimates of the total biospheric net flux derived from a deconvolution of ice core data and ocean models (Siegenthaler and Oechgher 1987). This analysis, reproduced in Figure 7.2, shows general agreement between the estimates prior to about 1930, after which the net flux from land cover change increases while the total biotic flux declines. This suggests either the land use or deconvolution analyses are wrong, or the difference reflects a net sink in undisturbed ecosystems. If a sink in undisturbed ecosystems exists, and is somewhat evenly distributed geographically, the amount per unit area would be too small to measure in the field. Thus, a combination of constrained analyses of this kind and ecosystem models would be required to further elucidate the nature of the sink. An important consideration in analyses such as Tans et al. (1990) and Houghton and Skole (1990) is knowing the biogenic flux from land cover change, particularly the tropical deforestation component. Improved definition of the imbalance in the global carbon budget will come with better definition of the individual terms of the equation above, with the biogenic term being particularly important. Moreover, the use of constrained analyses, in which additional terms to the equation are inferred as residuals from known terms, will require increased geographical and temporal specificity and confidence in the estimate of biogenic fluxes. Net Fluxes of Terrestrial Carbon in Asia Of the combined global net flux of carbon from fossil fuel combustion and land use change in 1990, approximately 30% is from temperate and tropical Asia (Table 7.1). The global net flux from fossil fuels is ~6.1 x 1015 g C year-1 (Marland et al. 1994), while the global net flux from land cover conversion is ~1.7 x 1015 g C year-1 (Houghton and Hackler 1995). The net flux from fossil fuel combustion in Asia is ~1.5 x 1015 g C year-1, and the net flux from biogenic sources is 0.8 x 1015 g C year-1. The fossil fuel emissions from Asia represent 25% of the global total fossil fuel inputs to the atmosphere, while the biogenic emissions from Asia represent almost half of the global total inputs from land cover change. The contribution of China’s fossil fuel emissions is approximately 43% of the regional total fossil fuel emissions, 10% of the regional biogenic flux, and 32% of the combined fossil fuel and biogenic emissions. When we exclude the Pacific Developed countries, which include Japan, Australia, and New Zealand the fossil fuel emissions from Asia are only 18% of the global total fossil fuel inputs of which China’s contribution is almost 60%. Table 7.1 Terrestrial carbon
emissions from fossil fuel consumption and land cover conversion in Asia,
1990
Units: 1015 g C year -1 Currently this is an important region in
terms of its biogenic contribution to the global totals, probably more
so than in terms of its contribution from fossil fuels. The Southeast Asia
subregion is significant in its contribution of biogenic emissions, contributing
almost half of the global total. However, it should be noted that growth
rates of fossil fuel emissions since 1970 have been significant in this
area of the world, and future emissions of fossil fuel carbon will likely
be very high in rapidly developing China and the so-called "Asian Tiger"
countries of Southeast Asia. The growth rates in fossil fuel emissions
in the last 20 years for China and Southeast Asia have exceeded the overall
global rate of increase several fold (213% and 240% vs. 49%).
Biogenic Fluxes of Carbon from Land Cover Change in Asian and Other Equatorial Tropical Forests Human activities have transformed the natural landscape over the past 300 years in a variety of ways and to varying degrees (Turner et al. 1990, Turner and Meyer 1994). Quantitatively, the most important type of change in Asia and elsewhere has been the direct conversion of natural land cover to agriculture (Richards 1984, 1990). Large areas of Asia and other parts of the world are undergoing extensive land transformation as a result of population growth and economic development. Such changes in land cover influence the net exchange of carbon between the land and atmosphere (Detwiler and Hall 1988, Houghton et al. 1987, Houghton et al. 1985, Houghton et al. 1983, Moore et al. 1981, Bolin 1983). Although large releases of carbon are historically attributed to land cover conversion in the temperate zone, tropical deforestation appears to be the major biogenic source of carbon today. The long term cumulative release of carbon from land cover change between 1850 and 1990 is ~122 x 1015 g C (Schimel et al. 1995, Houghton 1994). The current annual net flux of carbon from land cover conversion in the tropics is between 0.4 and 2.5 x 1015 g C year-1 (Houghton et al. 1985, Houghton et al. 1987). Ten countries comprise two-thirds of the net release: Brazil, Columbia, Indonesia, Ivory Coast, Laos, Malaysia, Mexico, Peru, Thailand, and Zaire. The long term historical net flux of carbon due to land conversion has been documented by analyses based on direct estimates of land use data and ecosystem response characteristics (Houghton and Skole 1990). These studies suggest that while some land transformations increase the amount of carbon stored on land, the overall historical trend has been one of increasing forest conversion to agriculture. This has reduced the amount of carbon stored on land and resulted in a net flux of carbon to the atmosphere. However, as we point out below, most of these analyses have focused on the long term, inter-decadal historical trend, rather than fine temporal analysis of particular years. The amount of carbon held in terrestrial ecosystems is changed as a result of (1) direct human effects of land use, such as deforestation, and (2) indirect human effects on ecosystems, such as increased concentrations of atmospheric CO2 or climate change, which influence changes in ecosystem metabolism. At the present time, the net flux of carbon between the biota and atmosphere due to land use change (deforestation) can be calculated with more confidence than changes in carbon storage associated with large- scale changes in ecosystem metabolism. This is because of the large difference in biomass between forests and the agricultural systems which replace them, and because deforestation and reforestation can be readily documented and quantified, particularly if remote sensing data are utilized. The current net flux of carbon between the biota and the atmosphere due to land use change is uncertain. Three factors contribute to the range of uncertainty: (1) rates of deforestation, (2) the fate of deforested land (i.e. the amount of secondary forest regrowth and re-clearing), and (3) the stock of biomass and soil organic matter and their response to disturbance, including anthropogenic reductions of carbon stocks within forests due to thinning or degradation. Models developed with improved geographic and temporal data on deforestation rates, better parameterization of the dynamic nature of deforestation and reforestation, and improved data on above- and below-ground carbon response characteristics are needed. In this chapter we wish to distinguish between changes in carbon storage due to forest growth and changes in carbon storage due to forest regrowth and secondary succession. The former has been postulated as a component of the so-called missing sink, resulting from changes in ecosystem metabolism in undisturbed forests. The latter is related to land use changes in areas where significant abandonment of agriculture is taking place. One of the central tenants of this paper is that this latter term has been poorly documented, and that in fact this could be a significant factor in the global carbon budget. Moreover, asynchronies in clearings in one place and time with abandonment and regrowth in other places and times could result in dampened net biogenic sources during certain periods of time. Most estimates of land cover changes are net changes; there have been few studies which document both deforestation and regrowth simultaneously and virtually no studies document both processes on an annual basis. Should the relative magnitude of these two processes change significantly from one year to the next, there exists a potential for large temporary variations in the magnitude of the biogenic source term, and possibly a significantly dampened source term, even while long term trends suggest otherwise. Interannual Variations in Terrestrial Carbon Fluxes. Tans et al. (1990) used a global circulation model to redistribute atmospheric CO2 based on surface fluxes of CO2 from fossil fuel consumption, tropical deforestation, and oceanic uptake. Their analysis predicted that the atmospheric [CO2] should exhibit a strong north-south gradient of 5.7 to 7.3 ppm. This estimate was significantly higher that the observed meridianal gradient of 3ppm based on the NOAA flask measurements (Tans et al. 1990). They discussed that the source of this discrepancy could either be their atmospheric transport model or an unaccounted extra-tropical terrestrial sink in the Northern Hemisphere. Analysis of 85Kr and CFC tracer data indicates that inaccuracies in their transport model could account for differences of only 10% and not the two fold differences measured. Therefore, they assumed that the "missing sink" was a terrestrial sink of 2.0 to 3.4 x 1015 g C year-1 in the Northern Hemisphere temperate zone. Siegenthaler and Sarmiento (1993) updated the Tans et al. oceanic uptake estimates by accounting for differences in ocean skin temperature and bulk water temperature, horizontal transport of carbon from the terrestrial biosphere to the ocean via rivers, and the north-south transport of carbon monoxide and its subsequent uptake in the southern oceans. They conclude that there is a terrestrial sink of about 1.8± 1.3 1015 g C year-1, or that the tropical source term of 1.6± 1.0 x 1015 g C year-1 is too high, or a combination of both. Quay et al. (1992) estimated the net flux of CO2 in the ocean and biosphere using isotopic measurements of d 13C (essentially the ratio of 13C/12C) in the atmosphere and dissolved inorganic carbon from ocean surface waters and atmospheric [CO2]. d 13C varies depending on the source of the carbon and changes depending on the uptake mechanism due to isotopic fractionation, therefore changes in this ratio provide an indication of the uptake mechanisms (Quay et al. 1992, Ciais et al. 1995a). Because terrestrial photosynthesis discriminates against the heavier isotope ( 13C), atmospheric uptake can be partitioned between the ocean and the biosphere (Quay et al. 1992, Sarmiento 1993, Houghton et al. 1995, Ciais et al. 1995a ). Quay et al. (1992) used this technique to estimate that from 1970 to 1990 the average annual oceanic uptake was 2.1 x 1015 g C year-1, while the average annual terrestrial uptake was negligible (0.1 x 1015 g C year-1). More recently Ciais et al. (1995a, 1995b) applied a hybrid approach using a two-dimensional atmospheric transport model (cf. Tans et al. 1990) with [CO2] and d 13C measurements from NOAA’s Climate Monitoring and Diagnostics Laboratory (CMDL) global flask network (cf. Quay et al. 1992) to estimate CO2 partitioning as a function of latitude and time from 1990 to 1993. These results suggest that the northern temperate zone was a large terrestrial sink (3.5 x 1015 g C year-1) in 1992 and 1993, whereas the northern tropics (from Equator to 30° N) are a large source (2.0± 1.3 x 1015 g C year-1). This estimate for the northern tropics is considerably higher than the estimate of 0.9 x 1015 g C year-1 by Houghton et al. (1987). Ciais et al. (1995) attributed this difference to either a problem with their model for intrahemispheric transport between the tropical and temperate latitudes or that previous deforestation estimates, particularly for tropical regions such as Southeast Asia have been underestimated. These results also suggest that the southern tropics were a small terrestrial sink in 1992 and 1993. This is surprising because this latitudinal band contains the Brazilian Amazon and Indonesia. Estimates of CO2 emission from tropical land use rank these two countries as the top two in terms of total emissions due to deforestation (Houghton et al. 1987). The weak terrestrial sink in the southern tropical zone suggests that either the forests have a positive net ecosystem production (due to fertilization of undisturbed forest and/or secondary growth formation) or previous estimates of deforestation have been overestimated (Ciais et al. 1995a). In another approach Keeling and Shertz (1992) and Keeling et al. (1996) relied on new methods for measuring atmospheric [O2]/[N2]. Photosynthesis converts CO2 and water into organic carbon and molecular oxygen (O2). Combustion and respiration pathways work in the opposite direction with the oxidation of organic compounds producing CO2. Thus, atmospheric [O2] is a function of photosynthesis, respiration and combustion and varies inversely with atmospheric [CO2]. However, changes in atmospheric [O2] are difficult to measure because these changes are very small compared with the overall atmospheric concentrations. Nonetheless, changes in [O2] can still be estimated by changes in the ratio of O2/ N2 because [N2] is relatively stable. Since O2 is not chemically active in the ocean due to its low solubility, changes in atmospheric [O2] can then be related to fluxes of CO2 from the biosphere. The net terrestrial exchange of CO2 can be used to constrain the oceanic CO2 uptake based on atmospheric measures of CO2. Keeling and Shertz (1992) used this technique to estimate an annual oceanic uptake of 3.0± 2.0 x 1015 g C year-1 and a small annual net terrestrial sink between 1989 and 1991. Keeling et al. (1996) used O2/ N2 data from 1991-1994 to measure hemispheric gradients in CO2 uptake. They estimated that, overall, on an annual basis from 1991 to 1994 the terrestrial biosphere was a net sink of 2.0± 0.9 x 1015 g C year-1, tropical land biota was not a strong source or sink, and the oceans were a net sink of 1.7± 0.9 x 1015 g C year-1. Using longer term trends in O2/ N2 they estimated that the measured slow down in the rate of increase in atmospheric CO2 was due to either increase CO2 uptake in the temperate land biota or the removal of a net tropical source of CO2 due to a decrease of deforestation rates or increase in the area in forest regrowth. These analyses suggest that the tropics are neither a strong source nor a strong sink (Keeling et al. 1996). Ciais et al. (1995) estimate that the northern tropics are a source, but the southern tropics are a small sink. They propose two explanations for the sink-to-neutral magnitude: (1) an increase in tropical Net Ecosystem Production to offset deforestation sources, and (2) evidence for a combination of reduced rates of deforestation and increased regrowth of previously cleared land "which underscore the uptake of carbon by recovering forests on land abandoned by shifting agriculture and also suggest reduced clearing rates based on satellite analyses". Interdecadal and Interannual Land Cover Change In Asia. The current global rate of deforestation is relatively unknown. Problems documenting it center on: (1) the lack of geographically-referenced data at high spatial resolution derived from a single consistent method for the whole of the tropics -- preferably from satellite data, (2) a paucity of quantitative information on the relative rates of new deforestation and abandonment to regrowth, and (3) detailed, annual time-series for specific periods of time from which a complete global carbon budget could be compiled in conjunction with annual atmospheric measurements. The most recent estimates of the annual rate of tropical deforestation over the last decade by the FAO is 15.4 x 106 ha year-1 (FAO 1993). The regional contributions of this global annual rate are 4.1 x 106 ha year-1 for Africa, 3.9 x 106 ha year-1 for Asia, and 7.4 x 106 ha year-1 Latin America & Caribbean. Within the Latin America & Caribbean region FAO estimated that the rate for Tropical South America was 6.2 x106 ha year-1, with Brazil alone at 3.7 x106 ha year-1. Satellite remote sensing is the only means for resolving discrepancies or quantifying temporal and spatial variations in deforestation rates. There are no reasons why satellite-based techniques cannot be applied to a large area like the Southeast Asian subcontinent, or a significant portion of the tropical forest belt, to resolve the aforementioned controversies and uncertainty, and thereby provide vastly improved forcing functions for global carbon models. In the two sections which follow we report new results of analysis of satellite data for Southeast Asia, an important region in the northern tropics, and the Brazilian Amazon, an important region of the southern tropics. Methods for satellite data analysis. To analyze what is happening in Asia, Landsat MSS and TM data were acquired for the entire Indochina region, including Thailand, Laos, Burma (Myanmar), Vietnam, and Kampuchea. These data were acquired to compile a complete mosaic of satellite data in the mid 1970s, the mid 1980s and the early 1990s in order to compute synoptic average annual deforestation rates over the last 20 years. At selected test sites we compiled multi-year data for analysis of interannual dynamics. Digital data are selected from the national or foreign archives and from new acquisitions. These data are pre-processed to standard analysis products. This includes conversion to a common format (CCT-P) from other formats (e.g. CCT-X, CCT-A, and others) and the addition of map coordinates. Each scene is registered to a Universal Transverse Mercator (UTM) map projection using satellite navigation data (includes only system corrections), and the digital image imbedded with coordinate information as UTM labeled tic marks. A 1:250,000 scale photographic three-band color composite print and a 1:1,000,000 color transparency are also produced and distributed. These pre-processed data are analyzed for deforestation rates. We use a hybrid methodology based on digital data combined with visual interpretation. Landsat scenes are first georeferenced to "north-up" by using the four corner points provided in the data descriptor record on the data tape as ground control points (GCPs) to define the linear transformation. The rectification is performed using nearest neighbor resampling technique to minimize convolution of the data. An iterative, self-organizing (ISO), unsupervised clustering technique is applied to the data with a knowledge-based classifier to produce five output classes for our analyses in these regions: forest, non-forest, water, cloud, and cloud shadow. For the ISO analysis in Southeast Asia we first create a new band which is the ratio of band 4 (NIR) to band 2 (RED). We run the unsupervised classification using only band 2 (RED) and this new band. We apply the ISO algorithm to assign 45 output clusters. This algorithm is based on initially using the statistics of the raw data to arbitrarily determine the first clusters. The statistical mean spectral values of clusters are shifted to locate the actual spectral statistics for each cluster. This step is repeated until 95% of the pixel are not reassigned to new clusters during the iterations that shift the cluster means. Assignment of individual pixels to a cluster is accomplished using a minimum spectral distance decision rule such that all pixels are assigned to one of the 45 output clusters. These output clusters are then recoded into a new classification containing up to five classes. This is done by overlaying the 45 clusters on the original imagery to aggregate the 45 spectral clusters into the final five output classes. This classification product from is plotted at 1:250,000 scale on vellum and then overlaid on a 1:250,000 scale photo product of the Landsat scene. Polygons that are mis-classified are identified and recoded in the digital geographic information system. The classification product is re-plotted and checked for further editing. These editing steps are repeated until the classification is completed. To produce databases covering the entire region of analysis, individual scenes are edge-matched and merged to form a seamless mosaic. The mosaic represents a complete inventory of the forest area as it existed in 1973, 1985, and 1992 for all of continental Southeast Asia. Average annual rates of deforestation are computed by difference between dates. It is not possible to compile complete regional coverage each year in order to estimate the actual annual deforestation and regrowth dynamics. In this case we establish several test sites in each region. In the test sites we co-register annual or semi-annual Landsat data and compute change detection analysis for deforested areas and areas in regrowth. For this chapter we only report on the regional analysis of continental Southeast Asia for two dates, 1973 and 1985, and our multi-date time-series analysis from 1989 to 1994 for the test site in Chiang Mai, Thailand (Figure 7.3). Results from Asia. Results for Southeast Asia are shown in Figures 7.4 and 7.5, which show the mapped extent of forest areas at two dates: ca. 1973 and 1985. The numerical summary of results are shown in Table 7.2. This is the first analysis we know of which reports forest areas and rates of forest loss for the entire region using a single consistent method from satellite data which completely cover the region. It is also the first analysis for the entire region of which we are aware which has been compiled and mapped in a digital geographic information system at a spatial resolution less than 100 meters. The rate of deforestation from the mid 1970s to the mid 1980s was ~1.4 x106 ha year-1, which is a rate close to that reported for the much larger Brazilian Amazon during the same period (Skole and Tucker 1993). The highest rate of deforestation occurs in Thailand at 0.48 x106 ha year-1, representing approximately 35% of the total for the region. It also has the highest percentage loss rate in the region; approximately 26% of the forest area in 1973 had been lost by 1985. Most of the deforestation in Thailand took place in the northern mountainous region around the province of Chiang Mai and in eastern Thailand along the Cambodian border where shifting cultivation, expansion of commercial agriculture and logging are very active. The total area of Northern Thailand is 169,000 km2 with 84,000 km2 (or 49% of the total area) in forest in 1985. By 1993 the Royal Forestry Department of Thailand (RFD) reports that the was just over 75,000 km2 (or 44%) of forest remaining. This loss of forest corresponds to an annual clearing rate of over 0.11 x 106 ha year-1. During this period from 1985 to 1993 all of Thailand had an average annual deforestation rate of just under 0.2 x106 ha year-1, hence almost 60% of the deforestation occurs in an area less than one third the size of the entire country. Table 7.2 Summary of results
from satellite data analysis of forest areas and their rate of change in
continental Southeast Asia.
Units: 106 ha or 106 ha year-1 These region-wide results are interesting in that they are higher than that reported by the World Resources Institute (~0.95 x106 ha year-1) for the period 1981-1985 (WRI 1992) and FAO (1.3 x106 ha year-1) for a later period, 1981 to 1990 (FAO 1993). This latter comparison implies that the current estimates from FAO, upon which some carbon flux studies have been based (Houghton 1994), are either too low or the rate has dropped since 1985. Our analysis of Thailand is higher than that reported by the World Resources Institute (~0.25 x106 ha year-1 ) for closed forests for a later period, 1985-1988 and approximately comparable to those reported by FAO for the later period 1981-1990. At this writing we do not have results from satellite analysis for the entire region for the period 1985 - 1993, but this analysis will be forthcoming soon. Nonetheless, the indications from these data for the entire region and for Thailand suggest that rates of deforestation peaked in the late 1980s and have declined since. Shifting cultivation is a main land use agent of forest loss in this part of Southeast Asia. United Nations’ estimates suggest that shifting cultivation supports about 28 million people in insular and continental Southeast Asia. During the early 1980s FAO estimated that shifting cultivation contributed to 49% of all the deforestation in Southeast Asia (Lanly 1983). FAO estimated that 1 x106 ha year-1 were in shifting cultivation in Thailand alone at this time. Kammerer (1989) estimated that 50,000 members of hill tribes in Northern Thailand were practicing integral shifting cultivation. In the mid-1980s it was estimated that 0.5 x106 ha year-1 of land was under some sort of shifting cultivation cycle in Northern Thailand. The test site in Chiang Mai (Figure 7.3) is in the center of this heavy deforestation pressure. Our results of multi-temporal analysis of co-registered annual data from 1986 to 1994 reveals that the area in cultivation increased by 38% from 4.5 x 103 ha in 1989 to 6.2 x 103 ha in 1994 (Figure 7.6). From 1989 until 1992 the cultivation area increased annually until it reached 7.4 x 103 ha in 1992. The next two years saw a drop in the total cultivated area. During the six year period 1.1 x 103 ha of land remained under cultivation, hence the percentage of area in permanent cultivation or in long cultivation cycles (greater than 6 years) varied from 14% to 23% of the total area cultivated at any given time. During this time period 66.2 x 103 ha of the over 81 x 103 ha of study area was undisturbed, hence 81% of the area was in long fallow, secondary forest or undisturbed primary forest (not likely since many of the hill tribes prefer to clear primary forest over secondary forest since it leads to higher crop yields). Figure 7.7 shows the annual extent of new clearings and abandonment areas. Over the time period of observation, the area of new clearings in any given year varied from 2.3 x 103 ha in 1990 to a peak of 3.7 x 103 ha in 1992 with a decline to 2.4 x 103 ha cleared for 1994. In 1990 and 1991 over 50% of the area in cultivation was on areas that had just been cleared. This percentage decreased every year until 1994 when only 39% of the cultivated area was cleared that year. The extent of the abandonment area also fluctuated significantly over this time period. The smallest abandonment occurred from 1990 to 1991 with only 1.8 x 103 ha abandoned, while the extent of abandonment almost doubled to 3.5 x 103 ha and 3.4 x 103 ha in 1993 and 1994, respectively. These results of the multi-temporal analysis suggest that deforestation rates are highly variable from one year to the next and that the process of deforestation involves considerable abandonment to fallow, or secondary forests. The rate of abandonment is also highly variable and sometimes not synchronous with clearing (i.e. one could have years with large amounts of fallow formation and very little new deforestation). Carbon accumulating regrowth is a predominant feature of the Asian landscape. The overall trend in the region appears to be characterized as having rates of deforestation peaking in the late 1980s and the declining somewhat in the 1990s. As this "pulse" of deforestation passes through the system as a general interdecadal trend which would suggest increased carbon emissions, the high variation in rates of clearing and abandonment (which are not necessarily occurring simultaneously) could create interannual periods in which the annual carbon flux is characteristically different than the decadal trend. More work is underway at this time to test
these ideas in other locations throughout the tropics on an interannual
basis and to develop region-wide analyses for more recent time periods,
such as 1985-1993 and 1993-1996.
Comparison with Results from Other Tropical Regions. Regional Analysis in the Brazilian Amazon. Analyses of satellite data for the Brazilian Amazon indicate that from 1978 to 1988 the average annual rate of deforestation was between 1.6-2.2 x106 ha per year-1 (INPE 1992, Skole and Tucker 1993). Our own work-in-progress for the same period suggests that the annual rate of deforestation in the Brazilian Amazon peaked during the period 1978-1986 at around 1.8 x 106 ha year-1 and then declined to ~1.4 x 106 ha year-1 from 1986 to 1993. The Brazilian Space Agency, INPE, reports annual deforestation rates from 1988-1991 for the Amazon of ~1.9, 1.4, and 1.1 x106 ha year-1 , respectively, for this 3-year time period (D.S. Alves, pers. comm.). The 1990-1991 rate was 67% less than the average rate for this 3-year census. Apparently, the deforestation rate from 1990-1991 was half of the average annual rate from 1978 through 1986. Other data from the Brazilian Amazon suggest large differences in deforestation rates and the existence of large areas of abandonment each year (Alves and Skole 1996, Skole et al. 1994). Approximately 30% of the deforested areas in the Amazon are in secondary growth. From a few site studies in which interannual time series data have been acquired, it has been shown that the abandonment rate can be as high as 82% of the deforestation rate (1991-1992 in Alves and Skole 1996). At the same time, the deforestation rate varied annually by as much as 100% with an average over 6 years of 50%. These results bolster those reported here from Southeast Asia and give strong indications that there are indeed significant influences of interannual changes in the deforestation rate on net fluxes of carbon, but without further analyses in more sites it is not possible to draw final conclusions. The Significance of Secondary Growth. Results reported here for Southeast Asia indicate extensive areas of forest fallow and secondary growth in Continental Southeast Asia exist. Similar results exist for the Legal Amazon. For example, 30% of the deforested area in the Legal Amazon were in some form of secondary growth in 1986 and 1992. This percentage represent areas over 77,000 km2 and 98,000 km2 in secondary growth in 1986 and 1992, respectively. Satellite analysis by Alves and Skole (1996) indicate that for a site in Rondonia, Brazil, secondary growth is a large (over 40% of the deforested areas) and a rapidly changing pool. Almost 60% of the areas in secondary growth in 1986 were re-cleared at least once by 1992, and over 55% of the deforested areas in 1986 were abandoned into secondary growth for some period of time by 1992 (Alves and Skole 1996). This analysis and other similar satellite based analyses (Steininger 1996, Foody et al. 1996, Brondizio et al. 1994, Moran et al 1994, Lucas et al. 1993, Mausel et al. 1993) are useful for characterizing the dynamics of the secondary growth resulting from tropical land use and support similar evidence reported here for Southeast Asia. These studies of deforestation from satellite data show it is a highly dynamic process of clearing, abandonment and re-clearing, and the rates at which land is cleared or abandoned are related to the land use and management system the forest farmers employ. In some cases, deforestation and secondary succession exist in tandem as a tightly coupled system in which secondary growth is continually recycled back into farmland. In other cases active land management maintains the land in agriculture, or the lack of active land management or population displacement results in long-term succession. Thus land use change influences land cover change and is integrated with dynamics of ecosystem structure, function, and response. Within a focused research program, it will be important to frame such questions as: what human land-use and land management strategies are employed in different situations and how do they control , or interact with, the dynamics of ecosystem response to disturbance? Because deforestation and abandonment have opposite effects on atmospheric carbon dioxide (c.f. uptake vs. release), we can re-state the aforementioned general question in terms specific to the carbon cycle: over time what land use strategies determine the abundance and spatial distribution of secondary growth, how do they determine the balance between clearing and regrowth rates, and how are these land use strategies in turn influenced by ecological conditions? Implications of Recent Carbon Cycle Analyses for Tropical Deforestation Rates. Analysis for specific recent years (cf. 1992-1994) reported by Keeling et al. (1996) and Ciais et al. (1995) which suggest that the tropics as a whole, and the southern tropics in particular, are not large sources of biogenic carbon are somewhat inconsistent with interdecadal results from previous terrestrial carbon models based on land use change by Houghton and Skole (1990), Houghton (1994) and others, and recent observations from satellite analyses of deforestation rates for Southeast Asia (see below) and the Amazon (Skole and Tucker 1993) which suggest that tropical deforestation has contributed to a significant net source of carbon. Based on results presented above, an explanation worth exploring further is that pervious estimates of the tropical source term were based on decadal mean values, rather than explicitly on any specific year or set of years when atmospheric measurements are made (cf. 1992-1994). This would be especially relevant if tropical land use change is characterized by large areas and rates of abandonment with significant inter-annual variation as appears to be the case in both Southeast Asia and the Amazon. Significant interannual differences in the long term trend in deforestation rates and the relative rates of deforestation and abandonment could result in periods of asynchrony where clearing rates, which relate to carbon emissions, and abandonment rates, which relate to carbon uptake are very different. Such dynamics would result in wide interannual variations in the relative contribution to the net flux from clearing vs. regrowth, and, potentially, a dampening in the decadal mean values for particular years. In this scenario two important conditions must be met: (1) large interannual departures in rates of deforestation from the decadal mean values reported by previous analyses, and (2) large areas of forest regrowth as a result of land abandonment. In years in which deforestation rates are greatly less than the decadal mean, and with the large abundance of areas of secondary succession and forest regrowth, due to high rates of abandonment in previous years, we would expect to see a significant dampening of the carbon source computed from the decadal mean data alone. These issues then beg a very interesting
suite of questions concerning the factors which cause deforestation
rates to vary from year to year, and what factors determine or control
the balance between clearing and abandonment. The resolution of such questions
depends completely on: (1) making fine temporal resolution measurements
(annually) of deforestation rates over large areas using satellite remote
sensing, (2) making fine spatial and temporal scale analyses of the dynamics
of land use and cover change in order to document if regrowth is a significant
factor, and (3) developing an improved quantitative and diagnostic capability
to determine the factors (i.e. the so-called Human Dimensions) which control
relative rates of clearing and regrowth from one year to the next.
An Approach to IGBP and IHDP Research in Tropical Asia. To better quantify and understand the significance of interannual changes in the rate of deforestation and abandonment, the IGBP could undertake a research effort in Southeast Asia which is described below. It would be composed of four central activities, which could be carried out through collaboration of several Program Elements of the IGBP and IHDP, particularly LUCC, GCTE, IGAC and GAIM. Additional participation of DIS and START would be essential framework elements. Activity 1. Determine the decadal mean rate of forest loss for the region from 1986 through 1996 using a two-date complete inventory of satellite data. This work would be similar to that described in this paper but updated to include the addition of one more date circa 1996. This could readily be achieved through collaboration with satellite observation programs already in place such as the NASA Landsat Pathfinder project whose results are reported in this paper, Japanese efforts now underway, and several European Union projects such as the TREES project. This effort provide an additional inventory year in 1996 to determine the average rate of change from 1986 to 1996. Activity 2. Determine if annual rates of deforestation have been significantly different from the decadal mean rate over large areas and the region as whole. This activity would be developed to define the interannual rate of deforestation for the region as a whole using stratified sampling based on the decadal complete inventories using the results described above as a starting point. A stratified sampling scheme in Southeast Asia could be easily developed and then implemented to estimate the annual change in forest area from 1986 to 1994, the time-frame relevant to atmospheric CO2 and O2 measurements. This analysis would focus only on forest to non-forest, and provide a single region-wide estimate which can be used to interpolate annually between the average rates of forest loss estimated by Activity 1. Activity 3. Develop case studies aimed at the following question: is secondary growth important and does land cover change dynamically on an annual basis; what are the land use transition probabilities? In five to eight selected case study sites, the interprogram-element project would analyze annual time series of satellite data from 1986 to the present to calculate transition sequences and their associated probabilities for deforestation and secondary growth. Each case study would be an area as large as a single Landsat scene and could be located in the following countries of the region: Thailand, Laos, Vietnam, Indonesia, Malaysia, and the Philippines since there exists already an ongoing precursor project being developed under START and LUCC. It would be desirable to evaluate the addition of two other sites in Myanmar and Kampuchea. Each case study would focus on estimating the transition matrix for the following classes: Forest, secondary growth, paddy, upland crops, tree crops, and urban/developed. To carry out the case studies it will be necessary to solicit the collaboration of a well-established team of regional scientists through the START programs and in collaboration with established groups which have been working since 1993 as part of an IGBP/IHDP Land Use and Cover Change (LUCC) project in Southeast Asia. Activity 4. Develop diagnostic models of the deforestation process to better understand and quantify the differential controls on rates of deforestation and abandonment. The causes of deforestation have often been attributed to such things as population growth, road building, and a host of other similar monotonically varying variables (Panayotou 1991). Although population and infrastructure are important long term determinants of general trends over decades, interannual variability is not tightly linked to such parameters. Econometric variables such as variations in price structures and local market demands are likely to be better predictors of deforestation and abandonment rates over short time cycles. This activity would couple the satellite-based deforestation dataset derived from the aforementioned activities to socio-economic data to build diagnostic models of the deforestation process. Such models can provide: (1) an improved understanding of the factors which control variations in deforestation rates annually, (2) an improved understanding of the factors which determine the balance between creation of new farmland through deforestation and the creation of new farmland by clearing secondary fallow forests, (3) an improved understanding of the factors which control the balance between deforestation (the bringing of land into production) and abandonment (the taking of land out of production), and (4) a method for spatial and temporal extrapolation of the results from the activities described above. These models should be developed at two levels. The first would be focused on each of the case study sites with an emphasis on establishing the critical village-level variables, dynamics and proximate determinants. The second level of model analysis would be at the scale of countries or the region. The case studies will be used to calibrate the regional model and provide a basis and location for validation of regional results.
Alves, D.S. and D.L Skole. 1996. Characterizing land cover dynamics using multi-temporal imagery. Int J. Remote Sensing 17(4): 835-839. Bacastow, R.B. and C.D. Keeling. 1981. Atmospheric carbon dioxide concentration and the observed airborne fraction. In, Carbon Cycle Modeling, B. Bolin (ed.), SCOPE 16, John Wiley and Sons, New York. Bolin, B. 1983. The carbon cycle. In, The Maojor Biogeochemical Cycles and Their Interactions, B. Bolin and R.B. Cook (eds.), SCOPE 21, John Wiley and Sons, New York. Brondizio, E.S., Moran, E.F., Mausel, P., Wu, Y. 1994. Land use change in the Amazon estuary: Patterns of Cabocolo settlement and landscape management. Human Ecology. 22:249-278. Ciais, P., P.P. Tans, J.W.C. White, M. Trolier, R.J. Francey, J.A. Berry, D. A. Randall, P.J. Sellers, J. Collatz, D.S. Schimel. 1995a. Partitioning of ocean and land uptake of CO2 as inferred by d 13C measurements from the NOAA Climate Monitoring and Diagnostics Laboratory Global Air Sampling Network. Journal of Geophysical Research 100(D3):5051-5070. Ciais, P., P.P. Tans, M. Trolier, J.W.C. White, and R.J. Francey. 1995b. A large northern hemisphere terrestrial CO2 sink indicated by the 13C/12C ratio of atmospheric CO2. Science. 269:1098-1102. Detwiler, R.P, and C.A.S. Hall. 1988. Tropical forests and global carbon budget. Science. 239:42-47. FAO. 1993. Forest resources assessment 1990 - Tropical Countries. United Nations Food and Agriculture Organization. Rome. Foody, G.M., Palubinskas, G., Lucas, R.M., Curran, P.J., Honzak, M. 1996. Identifying terrestrial carbon sinks: Classification of successional stages in regenerating tropical forest from Landsat TM data. Remote Sensing of Environment. 55:205-216. Houghton, R.A. 1994. Emissions of carbon from land use change. In, The Carbon Cycle. T.M.L Wigley and D. Schimel (eds.). Cambridge University Press, New York. Houghton, R.A., R.D. Boone, J.R. Fruci, J.E. Hobbie, J.M. Melillo, C.A. Palm, B.J. Peterson, G.R. Shaver, G.M. Woodwell, B. Moore, D.L. Skole, and N. Myers. 1987. The flux of carbon from terrestrial ecosystems to the atmosphere in 1980 due to changes in land use: geographic distribution of the global flux. Tellus. 39B:122-39. Houghton, R.A., R.D. Boone, J.M. Melillo, N. Myers, C.A. Palm, D.L. Skole, B. Moore and G.M. Woodwell, 1985. Net flux of carbon from tropical forests in 1980. Nature 316(6029):617-620. Houghton, J.T., L.G. Meira Fihlo, J. Bruce, H. Lee, B.A. Chandler, E. Haites, N. Harris, and K. Maskell (eds.). 1995. Climate Change 1994. Radiative forcing of climate change and an evaluation of the IPCC IS92 emission scenarios. Cambridge University Press. 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, Oak Ridge National Laboratory, Ork Ridge, Tennessee, 144 pp. Houghton,R.A. and D.L Skole. 1990. Carbon. In, The Earth as Transformed by Human Action. Turner et al. (eds.) Cambridge University Press. New York. Houghton, R.A., J.E. Hobbie, J.M. Melillo, B. Moore, B.J. Peterson, G.R. Shaver, and G.M. Woodwell. 1983. Changes in the carbon content of terrestrial biota and soils between 1860 and 1980: a net release of CO2 to the atmosphere. Ecological Monographs 53:235-62. INPE. 1992. Deforestation in Brazilian Amazonia. Instituto Nacional de Pesquisas Espaciais, Sao Paulo, Brazil. Kammerer, C.A. 1989. Of labels and laws: Thailand's resettlement and repatriation policies. Cultural Survival Quarterly. 12(4):7-12. Keeling, C.D. 1960. The concentration and isotopic abundance of carbon dioxide in the atmosphere. Tellus 12:200-203 Keeling, C.D. 1994. Global historical CO2 emissions. pp. 501-504. In, Boden, T.A., D.P. Kaiser, R.J. Sepanski, and F.W. Stoss (eds.), Trends ‘93: A Compendium of Data on Global Change. ORNL/CDIAC-65, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge Tennessee, USA. Keeling, C.D., R.B. Bacastow, A.F. Carter, S.C. Piper, T.P. Whorf, M. Heimannn, W.G. Mook, and H. Roeloffzen. 1989. A three-dimensional model of atmospheric CO2 transport based on observed winds: 1. Analysis of observational data. pp. 165-236. In: Aspects of climate varianbility in the Pacific and the Westerns Americas. D.H. Peterson (Ed.). Geophysical Monograph, 55, AGU, Washington, USA. Keeling, C.D., T.P. Whorf, M. Whalen and J. van der Plicht. 1995. Interannual extremes in the rate of rise of atmospheric carbon dioxide since 1980. Nature 375: 666-670 Keeling, C.D., R.B. Bacastow, A.E. Bainbridge, C.A. Ekdahl, P.R. Guenther, L.S. Waterman, and J.F.S. Chin. 1976. Atmospheric carbon dioxide variations at Mauna Loa Observatory, Hawaii. Tellus 28(6):538-51. Keeling, C.D. and T.P. Whorf. 1994. Atmospheric CO2 records from sites in the SIO air sampling network. pp. 16-26. In, Boden, T.A., D.P. Kaiser, R.J. Sepanski, and F.W. Stoss (eds.), Trends ‘93: A Compendium of Data on Global Change. ORNL/CDIAC-65, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge Tennessee, USA. Keeling, R.F., S.C. Piper, and M. Heimann. 1996. Global and hemispheric CO2 sinks deduced from changes in atmospheric O2 concentration. Nature 382:218-221 Keeling, R.F. and S.R. Shertz 1992. Seasonal and interannual variations in atmospheric oxygen and implications for the global carbon cycle. Nature. 358:723-727. Lanly, J.P. 1983. Assessment of the forest resources of the tropics. FAO Forestry Abstracts: Review Article. 44:287-317. Lucas, R.M., M. Honzak, G.M. Foody, P.J. Curran, C. Corves. 1993. Characterizing tropical secondary forests using multi-temporal Landsat sensor imagery. International Journal of Remote Sensing. 14:3061-3067. Marland, G. and R.M. Rotty. 1984. Carbon dioxide emissions from fossil fuels: a procedure for estimation of results for 1950-82. Tellus 36(b):232-61. Marland, G., R.J. Andres, and T.A. Boden. 1994. Global, regional, and national CO2 emissions. pp. 505-584. In, Boden, T.A., D.P. Kaiser, R.J. Sepanski, and F.W. Stoss (eds.), Trends ‘93: A Compendium of Data on Global Change. ORNL/CDIAC-65, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge Tennessee, USA. Mausel, P., Y. Wu, Y. Li, E.F. Moran, E.S. Brondizio. 1993. Spectral identification of successional stages following deforestation in the Amazon. Geocarto International. 4:61-71. Moore, B., R.D. Boone, J.E. Hobbie, R.A. Houghton. J.M. Melillo. B.J. Peterson, G.R. Shaver, C.J. Vorosmarty, and G.M. Woodwell. 1981. A simple model for analysis for the role of terrestrial ecosystems in the global carbon budget. pp. 365-386. In Carbon Cycle Modeling. B. Bolin (Ed.). Moran, E.F., E. Brondizio, P. Mausel, Y. Wu. 1994. Integrating Amazonian vegetation, land-use, and satellite data. Bioscience. 44:329-338. Neftel, A., E. Moor, H. Oeschger, and B. Stauffer. 1985. Evidence from polar ice cores for the increase in atmospheric CO2 in the past two centuries, Nature. 315:45-47. Panayotou, T., 1991. Population Change and Land Use in Developing Countries: The Case of Thailand. HIID Working Paper. Cambridge, Massachusetts. Quay, P.D., B. Tilbrook, C.S. Wong. 1992. Oceanic uptake of fossil fuel CO2: Carbon-13 evidence. Science. 256:74-79. Richards, J.F. 1990. Land transformation. pp. 163-179. In: Turner, B.L. III, W.C. Clark, R.W. Kates, J. F. Richards, J. T. Mathews, and W.B. Meyer (Eds). The Earth as transformed by human action: Global and regional changes in the biosphere over the past 300 years. Cambridge University Press. Richards, J.F. 1984. Global patterns of land conversion. Environment. 26(9):6-38. Sarmiento, J. 1993. Atmospheric CO2 stalled. Nature. 365:697-698. Schimel, D., I.G. Enting, M. Heimann, T.M.L. Wigley, D. Raynaud, D. Alves, and U. Seigenthaler 1995. CO2 and the carbon cycle. In, Houghton, J.T., L.G. Meira Fihlo, J. Bruce, H. Lee, B.A. Chandler, E. Haites, N. Harris, and K. Maskell (eds), Climate Change 1994. Radiative Forcing of Climate Change and an Evaluation of the IPCC IS92 Emission Scenarios. Cambridge University Press. Siegenthaler, U. and H. Oeschger. 1987. Biospheric CO2 emissions during the past 200 years reconstructed by deconvolution of ice core data. Tellus. 39B: 140-154. Siegenthaler, U., J.L. Sarmiento 1993. Atmospheric carbon dioxide and the ocean. Nature. 365:119-125. Skole, D.L., W.H. Chomentowski, W.A. Salas, and A.D. Nobre. 1994. Physical and human dimensions of deforestation in Amazonia. BioScience 44(5):314-322 Skole, D.S., C.J. Tucker. 1993. Tropical deforestation and habitat fragmentation in the Amazon: Satellite data from 1978 to 1988. Science. 260:1905-1910. Steininger, M.K. 1996. Tropical secondary forest growth in the Amazon: age, area and change estimation with Thematic Mapper data. International Journal of Remote Sensing. 17:9-17. Tans, P.P., I.Y. Fung, and T. Takahashi.
1990. Latitudinal distribution of sources and sinks of atmospheric carbon
dioxide. Science. 247:1431-1438.
Thoning, K.W., P.P. Tans, and L.S. Waterman. 1994. Atmospheric CO2 records from sites in the NOAA/CMDL continuous monitoring network. pp. 28-40. In, Boden, T.A., D.P. Kaiser, R.J. Sepanski, and F.W. Stoss (eds.), Trends ‘93: A Compendium of Data on Global Change. ORNL/CDIAC-65, Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, Oak Ridge Tennessee, USA. Turner, B.L. III and W. B. Meyer. 1994. Global land-use and land-cover change: an overview. pp.3-11. In: Meyer, W.B. and B.L. Turner III (Eds) Changes in Land Use and Land Cover: A Global Perspective. Cambridge University Press. Turner, B.L. III, W.C. Clark, R.W. Kates, J. F. Richards, J. T. Mathews, and W.B. Meyer (Eds). 1990. The Earth as transformed by human action: Global and regional changes in the biosphere over the past 300 years. Cambridge University Press. WRI. 1992. World Resources 1992-1993.
A Report by the World Resources Institute in collabroation with the United
Nations Environment Programme and the United Nations Development Programme.
Oxford University Press, New York.385 pp.
paper presented at the Fourth Science Advisory Council of the International Geosphere Biosphere Programme Beijing, China October 1995
Draft version 3 January 1997
Figure 1: The contemporary global carbon cycle derived from estimate s provided in Schimel et al. (1995) and Houghton and Skole (1990). Units are 1015 g C year-1. Figure 2: Comparison between estimates of the biotic flux of carbon from direct analysis of changes in land cover and deconvultion of ice core data (from Houghton and Skole 1990). The solid line is the net flux of carbon from a model of land use change, while the dotted and dashed lines are results from deconvolution of ice core data using an outcrop ocean model and a simple box-diffusion ocean model, respectively. Since the outcrop model removes atmospheric carbon more efficiently than a diffusion model, a larger biotic flux can be accommodated. Figure 3: Study area in Southeast Asia, also showing the location of detailed test sites, with particular reference to the test site in Chian Mai, northern Thailand. The gridded tiles show the area of coverage by Landsat data for continental Southeast Asia which is reported in this paper. Each grid cell represents the location of a Landsat scene. Several of the test sites are shown, but only the Chiang Mai results are reported here. Figure 4: Total forest area in continental Southeast Asia as derived from analysis of Landsat satellite data for 1973. The areas shaded in green are intact forests. Figure 5: Total forest area in continental Southeast Asia as derived from analysis of Landsat satellite data for 1985. The areas shaded in green are intact forests. Figure 6: Change in area in active agriculture over time as derived from multi-temporal analysis of Landsat data for the test site in Chiang Mai, northern Thailand. Figure 7: Annual rates of formation
of new clearings and new areas of abandoned fallow or secondary growth
areas in the test site in Chiang Mai, northern Thailand.
|