• Anthes, R. A., 1984: Enhancement of convective precipitation by mesoscale variations in vegetative covering in semiarid regions. J. Climate Appl. Meteor.,23, 541–554.

  • Barradas, V. L., and J. Adém, 1992: Albedo model for a tropical dry deciduous forest in western Mexico. Biometeorology,36, 113–117.

  • Bastable, H. G., W. J. Shuttleworth, R. L. G. Dallarosa, G. Fisch, and C. A. Nobre, 1993: Observations of climate albedo, and surface radiation over cleared and undisturbed Amazonian forest. Int. J. Climatol.,13, 783–796.

  • Bastos, T. X., D. Hölscher, and T. D. de A. Sá, 1993: Climate evaluation in northeast Pará, Igarapé-Açu site. Proc. Studies on Human Impact on Forest and Floodplains in the Tropics, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agroflorestal da Amazônia Oriental, p. 19.

  • Bird, R. E., and C. Riordan, 1986: Simple solar spectral model for direct and diffuse irradiance on horizontal and tilted planes at the earth’s surface for cloudless atmospheres. J. Climate Appl. Meteor.,25, 87–97.

  • Bruijnzeel, L. A., 1990: Hydrology of Moist Tropical Forests and Effects of Conversion: A State of Knowledge Review. UNESCO, 224 pp.

  • Culf, A. D., G. Fisch, and M. G. Hodnett, 1995: The albedo of Amazonian forest and ranchland. J. Climate,8, 1544–1554.

  • Denich, M., and M. Kanashiro, 1995: Secondary vegetation in the agricultural landscape of Northeast Pará, Brazil. Management and Rehabilitation of Degraded Lands and Secondary Forests in Amazonia, J. A. Parotta and M. Kanashiro, Eds., United States Dept. of Agriculture, 12–21.

  • Dirmeyer, P. A., and J. Shukla, 1994: Albedo as a modulator of climate response to tropical deforestation. J. Geophys. Res.,99, 20 863–20 877.

  • Fisch, G., I. R. Wright, and H. G. Bastable, 1994: Albedo of tropical grass: A case study of pre- and post-burning. Int. J. Climatol.,14, 102–107.

  • Galvão, E. U. P., J. Cesar, and E. C. Italiano, 1986: Estado atual de conhecimentos sobre sistemas de producas de culturas alimentares para o Estado do Amazonas (Current state of knowledge on annual cropping systems for the State of Amazonas). Proc. Simposio do Tropico Humido III, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisas Agropecuária do Trópico Úmido, 57–69.

  • Gash, J. H. C., and W. J. Shuttleworth, 1991: Tropical deforestation: Albedo and the surface-energy balance. Climate Change,19, 123–133.

  • Giambelluca, T. W., 1996: Tropical and cover change: Characterizing the post-forest land surface. Climate Change: Developing Southern Hemisphere Perspectives, T. W. Giambelluca and A. Henderson-Sellers, Eds., Wiley and Sons, 293–318.

  • Giorgi, F., 1989: Two-dimensional simulations of possible mesoscale effects of nuclear war fires, I. Model description. J. Geophys. Res.,94, 1127–1144.

  • Hecht, S. B., 1993: The logic of livestock and deforestation in Amazonia. Bioscience,43, 687– 695.

  • Henderson-Sellers, A., R. E. Dickinson, T. B. Durbridge, P. J. Kennedy, K. McGuffie, and A. J. Pitman, 1993: Tropical deforestation: Modeling local- to regional-scale climate change. J. Geophys. Res.,98, 7289–7315.

  • ——, H. Zhang, and W. Howe, 1996. Human and physical aspects of tropical deforestation. Climate Change: Developing Southern Hemisphere Perspectives, T. W. Giambelluca and A. Henderson-Sellers, Eds., Wiley and Sons, 259– 292.

  • Hölscher, D., 1995: Wasser- und Stoffhaushalt eines agrarökosystems mit waldbrache im östlichen Amazonasgebiet (Water and nutrient budget of an agroecosystem with forest fallow in eastern Amazonia). Ph.D. dissertation, Institute für Bodenkunde und Waldernährung der Georg-August-Universität Göttingen, 134 pp.

  • Kato, M. do S. A., and M. Denich, 1993: The municipality of Igarapé-Açu-Pará study site. Proc. Studies on Human Impact on Forest and Floodplains in the Tropics, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agroflorestal da Amazônia Oriental, 7–9.

  • Leopoldo, P. R., W. Franken, E. Matsui, and E. Salati, 1982a: Estimativa de evapotranspiragco de floresta Amazonica de terra firme (Estimation of evaporation of terra firme Amazonian forest). Acta Amazonica,12 (Suppl.), 23–28.

  • ——, ——, and E. Salati, 1982b: Water balance of a small catchment area in “terra firme” Amazonian forest. Acta Amazonica,12, 333–337.

  • Lesack, L. F. W., 1993: Water balance and hydrological characteristics of a rain forest catchment in the central Amazon Basin. Water Resour. Res.,29, 759–773.

  • McGuffie, K., A. Henderson-Sellers, H. Zhang, T. B. Durbridge, and A. J. Pitman, 1995: Global climate sensitivity to tropical deforestation. Global Planet. Change,10, 97–128.

  • Moran, E. F., E. Brondizio, P. Mausel, and Y. Wu, 1994: Integrating Amazonian vegetation, land-use, and satellite data. Bioscience,44, 329–338.

  • Nascimento, C., and A. Homma, 1984: Amazônia: Meio ambiente e technologia agricola Agropecuária do Trópico Úmido, 57–69. Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisas Agropecuária do Trópico Úmido, 282 pp.

  • Nobre, C. A., P. J. Sellers, and J. Shukla, 1991: Amazonian deforestation and regional climate change. J. Climate,4, 957–988.

  • Oguntoyinbo, J. S., 1970: Reflection coefficient of natural vegetation, crops, and urban surfaces in Nigeria. Quart. J. Roy. Meteor. Soc.,96, 430–441.

  • Pinker, R. T., 1982: The diurnal asymmetry in the albedo of tropical forest vegetation. Forest Sci.,28, 297–304.

  • ——, O. E. Thompson, and T. F. Eck, 1980: The albedo of a tropical evergreen forest. Quart. J. Roy. Meteor. Soc.,106, 551–558.

  • Rego, R. S., B. N. R. da Silva, and R. S. O. Junior, 1993: Detailed soil survey in an area in the municipality of Igarapé-Açu-Pará. Proc. Studies on Human Impact on Forest and Floodplains in the Tropics, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agroflorestal da Amazônia Oriental, 13.

  • Segal, M., R. Avissar, M. McCumber, and R. Pielke, 1988: Evaluation of vegetation effects on the generation and modification of mesoscale circulations. J. Atmos. Sci.,45, 2268–2292.

  • Shuttleworth, W. J., 1988: Evaporation from Amazonian rainforest. Proc. Roy. Soc. London, Ser. B,233, 321–346.

  • ——, 1989: Micrometeorology of temperate and tropical forest. Philos. Trans. Roy. Soc. London, Ser. B,324, 173–436.

  • ——, and Coauthors, 1984: Observations of radiation exchange above and below Amazonian forest. Quart. J. Roy. Meteor. Soc.,110, 1163–1169.

  • Sud, Y. C., G. K. Walker, J.-H. Kim, G. E. Liston, P. J. Sellers, and K.-M. Lau, 1996: Biogeophysical consequences of a tropical deforestation scenario: A GCM simulation study. J. Climate9, 3225–3247.

  • Szott, L. T., and C. A. Palm, 1986: Soil and vegetation dynamics in shifting cultivation fallow. Proc. Simposio do Tropico Humido III, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agropecuária do Trópico Úmido, Belém, Brazil, p. 360.

  • Turner, B. L., R. H. Moss, and D. L. Skole, 1993: Relating land use and global land-cover change: A proposal for an IGBP-HDP core project. International Geosphere-Biosphere Programme Rep. 24: A Study of Global Change and the Human Dimensions of Global Environmental Change, IGBP, 95 pp.

  • Uhl, C., R. Buschbacher, and E. A. S. Serrão, 1988: Abandoned pastures in eastern Amazonia, I. Patterns of plant succession. J. Ecol.,76, 663–681.

  • Watrin, O. d. S., 1994: Análise da dinâmica na paisagem agrícola da Amazônia oriental: Abordagem integrando técnicas de processamento digital de imagens e sistemas de informação geográfica (Agricultural dynamics evaluation in eastern Amazonia using digital image analysis techniques and geographic information systems). M.S. thesis, Ministério da Ciência e Tecnolocia, Instituto Nacional de Pesquisas Espaciais, 153 pp.

  • Wright, I. R., and Coauthors, 1996: Towards a GCM surface parameterisation for Amazonia. Amazonian Deforestation and Climate, J. H. C. Gash, C. A. Nobre, J. M. Roberts, and R. L. Victoria, Eds., Wiley and Sons, 474–504.

  • Yan, H., and R. Anthes, 1987: The effect of latitude on sea breeze. Mon. Wea. Rev.,115, 936–956.

  • View in gallery

    Map of Igarapé-Açu study area.

  • View in gallery

    Diurnal cycle of albedo for sites at Igarapé-Açu; curves for secondary vegetation sites are labeled by the age of vegetation.

  • View in gallery

    Cycle of albedo for an actively farmed location in the eastern Amazon, based on a typical shifting cultivation pattern for small farms in the Igarapé-Açu region (see Table 8); 10-, 8-, and 6-yr cultivation cycles are shown to illustrate the effect of increasing land pressure on the regional-mean albedo; mean albedo is obtained by integrating the area under the curve for each cycle.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 355 136 23
PDF Downloads 185 99 12

Observations of Albedo and Radiation Balance over Postforest Land Surfaces in the Eastern Amazon Basin

View More View Less
  • 1 Department of Geography, University of Hawaii at Manoa, Honolulu, Hawaii
  • | 2 Institute of Soil Science and Forest Nutrition, Göttingen University, Gottingen, Germany
  • | 3 Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisas Agroflorestal da Amaônia Oriental, Belem, Para, Brazil
  • | 4 Department of Geography, University of Hawaii at Manoa, Honolulu, Hawaii
Full access

Abstract

Regional climatic change, including significant reductions in Amazon Basin evaporation and precipitation, has been predicted by numerical simulations of total tropical forest removal. These results have been shown to be very sensitive to the prescription of the albedo shift associated with conversion from forest to a replacement land cover. Modelers have so far chosen to use an “impoverished grassland” scenario to represent the postforest land surface. This choice maximizes the shifts in land surface parameters, especially albedo (fraction of incident shortwave radiation reflected by the surface). Recent surveys show secondary vegetation to be the dominant land cover for some deforested areas of the Amazon. The characteristics of secondary vegetation as well as agricultural land covers other than pasture have received little attention from field scientists in the region. This paper presents the results of field measurements of radiation flux over various deforested surfaces on a small farm in the eastern Amazonian state of Pará. The albedo of fields in active use was as high as 0.176, slightly less than the 0.180 recently determined for Amazonian pasture and substantially less than the 0.19 commonly used in GCM simulations of deforestation. For 10-yr-old secondary vegetation, albedo was 0.135, practically indistinguishable from the recently published mean primary forest albedo of 0.134. Measurements of surface temperature and net radiation show that, despite similarity in albedo, secondary vegetation differs from primary forest in energy and mass exchange. The elevation of midday surface temperature above air temperature was found to be greatest for actively and recently farmed land, declining with time since abandonment. Net radiation was correspondingly lower for fields in active or recent use. Using land cover analyses of the region surrounding the study area for 1984, 1988, and 1991, the pace of change in regional-mean albedo is estimated to have declined and appears to be leveling at a value less than 0.03 above that of the original forest cover.

Corresponding author address: Dr. Thomas W. Giambelluca, Dept. of Geography, University of Hawaii at Manoa, Porteus Hall 445, 2424 Maile Way, Honolulu, HI 96822.

Email: thomas@hawaii.edu

Abstract

Regional climatic change, including significant reductions in Amazon Basin evaporation and precipitation, has been predicted by numerical simulations of total tropical forest removal. These results have been shown to be very sensitive to the prescription of the albedo shift associated with conversion from forest to a replacement land cover. Modelers have so far chosen to use an “impoverished grassland” scenario to represent the postforest land surface. This choice maximizes the shifts in land surface parameters, especially albedo (fraction of incident shortwave radiation reflected by the surface). Recent surveys show secondary vegetation to be the dominant land cover for some deforested areas of the Amazon. The characteristics of secondary vegetation as well as agricultural land covers other than pasture have received little attention from field scientists in the region. This paper presents the results of field measurements of radiation flux over various deforested surfaces on a small farm in the eastern Amazonian state of Pará. The albedo of fields in active use was as high as 0.176, slightly less than the 0.180 recently determined for Amazonian pasture and substantially less than the 0.19 commonly used in GCM simulations of deforestation. For 10-yr-old secondary vegetation, albedo was 0.135, practically indistinguishable from the recently published mean primary forest albedo of 0.134. Measurements of surface temperature and net radiation show that, despite similarity in albedo, secondary vegetation differs from primary forest in energy and mass exchange. The elevation of midday surface temperature above air temperature was found to be greatest for actively and recently farmed land, declining with time since abandonment. Net radiation was correspondingly lower for fields in active or recent use. Using land cover analyses of the region surrounding the study area for 1984, 1988, and 1991, the pace of change in regional-mean albedo is estimated to have declined and appears to be leveling at a value less than 0.03 above that of the original forest cover.

Corresponding author address: Dr. Thomas W. Giambelluca, Dept. of Geography, University of Hawaii at Manoa, Porteus Hall 445, 2424 Maile Way, Honolulu, HI 96822.

Email: thomas@hawaii.edu

1. Introduction

Regional, or even global, climatic change resulting from altered land surface characteristics is one of the potential serious consequences of large-scale tropical deforestation. General circulation model experiments using an extreme postforest scenario of impoverished grassland suggest that a reduction in Amazon Basin precipitation of 20% or more could result from complete forest removal (Henderson-Sellers et al. 1993; Henderson-Sellers et al. 1996; McGuffie et al. 1995; Sud et al. 1996). To simulate land cover change in the models, shifts are made in the values of land surface parameters such as albedo (the proportion of incident solar radiation reflected by the surface), aerodynamic roughness, leaf area, and soil moisture capacity. Dirmeyer and Shukla (1994) found simulated deforestation-induced changes in precipitation to be strongly dependent on the magnitude of albedo change. Average precipitation decreased as a result of deforestation in their study only if the increase in albedo was greater than 0.03; with lesser albedo increases, rainfall increased.

The magnitude of parameter shifts depends upon the assumed replacement land cover. GCM modelers have selected impoverished grassland (e.g., Henderson-Sellers et al. 1993; Henderson-Sellers et al. 1996; McGuffie et al. 1995; Sud et al. 1996) as the postforest scenario as a means of amplifying any possible deforestation effects. A characteristic feature of most deforestation experiments done to date is an increase in surface albedo from a forest value of about 0.12 to a grassland value of about 0.19. The choice of grassland represents the most extreme possible future for deforested land, which, the modelers reason, will indicate whether any significant regional or global climate change could result from tropical deforestation. The impoverished grassland scenario also stems from a general impression, especially in the Amazon Basin, of rapid conversion of tropical forest to agricultural land and ultimately to degraded pasture (e.g., Hecht 1993). However, forest clearing in the Amazon is partly offset by the rapid return of forest vegetation on abandoned fields (Turner et al. 1993). Uhl et al. (1988) found that for sites near Paragominas in northern Pará State, Brazil, Amazon forest ecosystems recovered after abandonment, except for the most intensively used sites, comprising less than 10% of pasture land. For sites along the Transamazon Highway near Altamira, Brazil, secondary vegetation experienced the largest absolute increase in area among all land cover categories, despite a quadrupling of pasture land (Moran et al. 1994). Watrin (1994) analyzed satellite imagery for the Igarapé-Açu, Brazil, area for 1984, 1988, and 1991 and found that secondary vegetation was by far the dominant land cover. These and other studies are beginning to bring into focus a picture of a future deforested Amazon that strongly contrasts with the homogeneous impoverished grassland scenario. That is, what comes after forest in the Amazon is a highly heterogeneous, dynamic land cover mosaic, including a variety of cultivated surfaces and significant areas undergoing secondary succession leading to forest regrowth. In the future, simulations of realistic deforestation scenarios must take into consideration the variety of land covers known to follow forest cutting, including secondary vegetation at various stages of maturity. The study presented here was undertaken to provide field observations of radiative exchange characteristics, especially albedo, of postforest land covers representative of the eastern part of the Amazon Basin.

2. Background

a. Radiative exchange

Radiant energy flux at the land surface can be described as
RnetKKLL
where Rnet is net radiation, K↓ is downward shortwave radiation, K↑ is shortwave radiation reflected by the surface, L↓ is downward longwave radiation, and L↑ is longwave radiation emitted by the surface (all terms are expressed in W m−2). Here, Rnet can be expressed, with emphasis on surface conditions, as
RnetαKLϵσTs4
where α is the surface albedo (ratio), ϵ is the emissivity (ratio), σ is the Stefan–Boltzmann constant (5.67 × 10−8 W m−2 K−4), and Ts is the surface temperature (K). The surface albedo is defined as
i1520-0442-10-5-919-e3

Albedo of a vegetated surface may be only half or less the reflectivity of its principal components, leaf surfaces, because of the trapping of light penetrating the upper canopy. This effect is generally considered to increase with vegetation height, giving forests the lowest albedo and grassland the highest of any class of vegetation. At high sun angles, light can penetrate to greater depths within a forest canopy. Annual and diurnal changes in sun angle, therefore, cause changes in albedo. For this reason and because of the annual cycle of foliage loss in deciduous forests, albedo is typically lowest in summer. In a tropical location with a significant annual rainfall cycle, albedo is generally observed to be highest during the dry season(s) because of foliage changes (Pinker et al. 1980; Barradas and Adém 1992). Diffuse light more effectively penetrates the vegetative canopy and is, therefore, absorbed more effectively than parallel beam radiation. When incident radiation has a high proportion of diffuse light, the sun angle effect on albedo is less important. For moist tropical regions such as the Amazon, albedo would not be expected to vary greatly during the year because annual changes in sun angle and foliage are relatively small and because a high proportion of incident radiation is diffuse under the cloudy skies of the equatorial region. However, Wright et al. (1996) observed significant annual albedo cycles in the Amazon, which appear to be related to fluctuation in soil moisture in forested areas and leaf-area variation in pasture. Nevertheless, sun-angle-related diurnal change is the most important source of temporal variation in albedo in the region.

b. Radiative exchange of tropical land areas

Albedo values measured over tropical evergreen forests are remarkably consistent (0.1225–0.134, Table 1). Shuttleworth (1989) recommended 0.12 as a large-scale average albedo for tropical and temperate evergreen forests for use in GCMs. Cleared areas have more variable radiative characteristics. Pinker (1982) found that albedo varied seasonally between 0.125 and 0.160 in a forest clearing in Thailand. Bastable et al. (1993) reported a mean albedo of 0.163 for a pasture near Manaus in central Amazonia, but they cautioned that albedo for deforested areas is likely to vary from place to place. They speculate that this unexpectedly low value may be due to darkening of the underlying soil by the ash from repeated burns at the site. In a large pasture at Marabá in the eastern part of the Brazilian Amazon, albedo dropped from 0.19 to 0.08 due to burning, then recovered to its preburn value within about 80 days. In a more recent summary of albedo measurements over Amazonian pasture, Culf et al. (1995) give monthly means ranging from 0.171 to 0.192 and an annual mean of 0.180.

Summarizing the various albedo measurements for deforested tropical areas, principally grassland, Gash and Shuttleworth (1991, 126) state that it is “likely that tropical deforestation will always result in a decrease in the energy absorbed by the surface equivalent to some 2–13% of the solar radiation before the change. A change of about 8% is probably a typical value . . .” On the basis of such statements from field scientists, climate modelers have prescribed an albedo of around 0.19. For seven recent GCM deforestation experiments examined by Culf et al. (1995), albedo settings averaged 0.124 for forest and 0.188 for the deforested surface.

3. Methods

Albedo and related measurements have been made and continue at several locations in the Amazon (Wright et al. 1996; Culf et al. 1995), but study sites have so far been limited to primary forest and pasture. Radiative characteristics of other forest replacement land covers, such as secondary vegetation, are as yet not well known. With this in mind, we took radiation measurements over a variety of land covers representative of forest replacement in the Amazon Basin. Measurement sites were all within a small farm near the village of Igarapé-Açu (1° 11.5′ S, 47° 35.8′ W), in the eastern Amazonian state of Pará (Fig. 1). This study differs from other field measurement programs in its emphasis on land-cover-related spatial variability of radiative exchange within a small area.

a. Igarapé-Açu, Pará State

Agricultural practices in the village of Igarapé-Açu are typical of those in the deforested areas of the eastern Amazon Basin. Beginning in the early part of this century, the tropical rainforest in the area surrounding Igarapé-Açu has largely been replaced by secondary vegetation, areas cleared in preparation for planting, and crops. Secondary vegetation, known locally as capoeira, grows during multiyear fallow periods in the traditional slash-and-burn cropping system of the region (Denich and Kanashiro 1995). The species composition of the secondary vegetation in the region has been shown to include around 50 to 60 woody species in a sample area of 60 m2. In total, more than 800 plant species have been found in the region (Barr 1995, personal communication). Cassava, corn, cowpea, and rice are the main crops (Kato and Denich 1993). Figure 1 shows the sites of radiation measurements at the Igarapé-Açu study area and the pattern of land use on the farm at the time of the measurements (October and November 1992). The equatorial climate of the region is characterized by a very small annual cycle of temperature, with an annual mean of about 26°C. About 2500 mm of rainfall is received annually, with the highest amounts during February through April and the minimum during September through November (Bastos et al. 1993). Soil at the site has been classified as Typic Kandiudults (Rego et al. 1993).

b. Measurement strategy

Meteorological measurements were taken using a single set of sensors (Table 2), which were moved from site to site, producing a sequence of observations for different covers. A larger temporal sample was sacrificed to obtain greater spatial variety. Measurement sites (Fig. 1) are described in Table 3. The “slashed vegetation” site refers to a field where 7-yr secondary successional vegetation had recently been cut. The slashed vegetation remained on the ground at the time of the observations. Later, after the vegetation had dried for several weeks, the field was burned. Measurements at the same site after the burn are listed under “burned slash.” Ages of secondary vegetation are based on interviews with farmers and are considered to be only approximate.

c. Observations

Measurements were taken during the driest part of the year (October and November 1992). Because observations over different land covers were not made simultaneously, day-to-day variations in cloud cover, precipitable water, and atmospheric turbidity affected the measurements over each land cover differently. Direct comparisons among sites of measurements such as reflected shortwave radiation and net radiation are not meaningful because of day-to-day differences in solar radiation. However, when examining relatively time-invariant surface properties, such as albedo or net radiation as a fraction of incident solar radiation, comparison among sites is possible.

Concurrent incident (K↓) and reflected (K↑) shortwave radiation measurements were taken over multiple diurnal cycles at each site to capture the sun angle effect on albedo. Eppley model 8-48 and PSP measurements, sampled every 5 s and averaged and recorded each hour, were used to estimate the mean diurnal pattern of K↓ and K↑ over each land cover in the form of hourly means. Albedo values were computed for sunny and cloudy periods. Sky condition was determined on the basis of the ratio of K↓ to estimated clear-day radiation (Kcd). The SPECTRAL2 model (Bird and Riordan 1986) was used to estimate Kcd. The cloudy condition was defined arbitrarily as periods during which K↓/Kcd was less than 0.5.

4. Results and discussion

a. Radiative characteristics of deforested sites at Igarapé-Açu

Figure 2 shows that most surfaces exhibit higher early morning albedo in accordance with expectations, but many show an albedo decline in late afternoon. Pinker (1982) reported similar asymmetry in measurements over forests in Thailand. We speculate that the observed pattern may be due to the clear conditions (high direct beam content) and dew-covered (more reflective) leaves in the early morning and cloudiness (higher diffuse light content) in the late afternoon.

The albedo site means at Igarapé-Açu (Table 4) reveal that 1) shortwave reflection from advanced secondary vegetation is very similar to that measured recently for primary tropical forest in the Amazon (see Table 1), 2) albedo of initial secondary vegetation is higher than that of advanced secondary vegetation by about 0.03, 3) three sites typical of the actively used farmland in the region have an albedo in the narrow range between 0.170 and 0.176, and 4) slashed vegetation and burned slash have respective albedo values slightly higher and significantly lower than the secondary vegetation previously covering the site.

The site means are based on relatively short observation periods and are, therefore, representative only of the season. Measurements at other times of the year would likely yield somewhat different values. Culf et al. (1995) found that forest albedo at their Amazonian sample sites varied from 0.121 in March to 0.144 in August. For October and November, corresponding to the observation period for our study, mean monthly forest albedo was 0.143 and 0.141, respectively (Culf et al. 1995), higher than our advanced secondary value (0.135). Ranch land albedo in the Culf et al. (1995) summary varied from 0.171 in February and September to 0.192 in July. The October and November ranch land albedos, 0.172 and 0.177, respectively, were very similar to our measurements for agricultural sites (Table 4).

The relative amount of sunlight reflected is expected to be greater during clear-sky periods when the fraction of diffuse light is relatively low. Our observations, however, do not show this conclusively (Table 4). Albedo for cloudy periods was lower at the slashed vegetation and bare soil sites, but higher at the cropped field and secondary vegetation sites. Culf et al. (1995) were also unable to find a distinct relationship between Amazonian forest albedo and cloudiness. We believe that our observations, as well as those of Culf et al. (1995), may be due in part to the diurnal cycle in cloud cover and the resulting tendency to identify as “cloudy” the late afternoon period, a time during which the effect of low sun angle obscures the influence of higher diffuse light content.

The result that all forest replacement covers measured in Igarapé-Açu had albedo values lower than those used in deforestation simulations indicates that radiant energy absorption over real deforested surfaces is greater than those simulated in GCM experiments. That advanced secondary vegetation and primary forest are similar in terms of shortwave reflection suggests that the radiative balance of the two land covers may be similar. The ratio of net radiation to incident solar radiation (Rnet/K↓) is expected to decrease as albedo increases. For tropical forest, the ratio may be as high as 0.7 (based on data of Shuttleworth et al. 1984). From our measurements (Table 5), Rnet/K↓ for advanced secondary vegetation is significantly lower than that of primary forest. Successively lower ratios are found for initial secondary vegetation, agriculture, and slashed vegetation. Despite the fact that these sites reflect less energy than the simulated deforested landscape, these ratios are significantly lower than that predicted by GCMs for deforested land. This implies greater-than-predicted net longwave emission from the surface—that is, higher surface (canopy) temperatures. The difference between surface temperature and air temperature (TsTa) at midday is generally low, on the order of 1°C, for actively transpiring forests (Bastable et al. 1993). Table 6 shows that, for deforested land covers in Igarapé-Açu, large values of TsTa are observed, generally declining with increasing vegetation age and height. In all cases, the higher surface temperatures indicate inhibited turbulent energy exchange, a result, in part, of lower aerodynamic roughness (Giambelluca 1996).

b. Temporal change in Igarapé-Açu regional-mean albedo

Watrin (1994) used Landsat Thematic Mapper imagery to classify land cover distribution in the Igarapé-Açu region for 1984, 1988, and 1991. His results, summarized in Table 7, show that secondary vegetation is the dominant land cover of the region, pasture is the dominant active land-use type, and remaining primary forest is very small. By assigning albedo estimates to each land cover category, regional albedo can be estimated as a weighted average. Albedo values for advanced secondary vegetation, initial secondary vegetation, bare soil, and crop land were assigned on the basis of measurements reported in this study. Based on Culf et al. (1995), values of 0.134 and 0.180 were used for forest and pasture, respectively. We used the mean of measured initial and advanced secondary vegetation, 0.150, to represent intermediate secondary vegetation. Regional albedo estimated on this basis increases from 0.1531 in 1984 to 0.1556 in 1988 and 0.1562 in 1991 (Table 7). The rate of change in regional albedo during 1988–91 is estimated to be one-third of that during the 1984–88 period, reflecting stabilization of the pace of land cover change as the portion of land remaining in primary forest approaches zero.

Small farmers in Igarapé-Açu practice a form of shifting cultivation that is the dominant agricultural system in tropical America, including the Amazon Basin. The basic elements of shifting cultivation in this region (Nascimento and Homma 1984; Galvão et al. 1986; Szott and Palm 1986) are 1) forest felling and burning, 2) 1 to 4 yr of annual crop cultivation, 3) gradual or abrupt abandonment as the land reverts to secondary vegetation, and 4) the initiation of another felling–burning–cultivation–fallow cycle after a 3–25-yr fallow period. Using farmers accounts and the relative land area occupied by secondary vegetation of different ages and agricultural land, a typical crop–fallow rotation can be constructed for small farms in the eastern Amazon. Accounts and land cover distribution suggest that farmers in Igarapé-Açu clear and burn land for planting annual crops for 2 consecutive years, after which land is left in fallow for about 8 yr. Fields are typically prepared by cutting vegetation and allowing it to dry for about 1 month, burning it, and planting a field crop. Fisch et al. (1994) found that the albedo of burned land in the Amazon recovered preburn values within about 80 days of burning. We use this period as the time necessary to reach the pasture or crop albedo value. These assumptions lead to the land cover cycle summarized in Table 8. Albedo values, assigned on the basis of our measurements and published data, are also shown in Table 8. In Fig. 3 we have plotted the scenario of land cover change and albedo given in Table 8 to illustrate the estimated temporal cycle in albedo for an actively farmed location in the eastern Amazon. To illustrate the effect of increasing land pressure, Fig. 3 also shows albedo change for 8- and 6-yr cycles. Mean albedo can be obtained for a site by integrating over a whole cycle. Assuming that the 10-yr cycle represents current practices in the region, mean albedo of land in the crop–fallow cycle is seen to be 0.16. Factoring in a 5% forest cover at an albedo of 0.134, the regional albedo becomes 0.1587, reasonably close to the estimate of 0.1562 made on the basis of relative land cover (Table 7).

5. Conclusions

The radiative characteristics of deforested land in active use in Igarapé-Açu are significantly different from those of primary forest. Higher albedo and higher daytime surface temperature result in significantly lower Rnet [see Eq. (1)]. This implies a reduction in turbulent energy transfer—that is, lower evaporation and sensible energy flux—and an increase in radiative transfer—that is, a larger fraction of incident solar radiation at the surface is disposed of by reflection and emission. Evaporative flux from young secondary vegetation at the same study area was estimated to be 1364 mm yr−1 (Hölscher 1995), within the range of estimates of annual evaporation for Amazonian primary forest (Lesack 1993; Leopoldo et al. 1982a, Leopoldo et al. 1982b; Shuttleworth 1988) and close to the average of the most reliable estimates of tropical lowland forest evaporation, 1430 mm yr−1 (Bruijnzeel 1990). However, because Rnet is higher in eastern than central Amazonia, where primary forest evaporation estimates were made (probably due to differences in cloudiness), these evaporation estimates likely represent a reduction in the fraction of net radiation used for evaporation: evaporation accounted for only 80% of Rnet over young secondary vegetation at Igarapé-Açu in comparison with 90% of Rnet for central Amazonian primary forest evaporation (Hölscher 1995). These changes in the disposition of incident radiative energy are likely to have important effects on local and regional atmospheric processes.

Secondary vegetation is the most extensive land cover in deforested areas of the Amazon Basin. Albedo of emerging secondary vegetation declines with age from as high as 0.173 to a value indistinguishable from that of the original forest within about 10 yr. Despite the relatively small difference between forest and secondary vegetation albedo, significant disruptions of energy and mass fluxes are indicated by changes in net radiation and surface temperature. For actively used land surfaces, such effects on energy and mass exchange are large and highly variable in time and space.

The highest albedo of deforested surfaces in Igarapé-Açu was 0.176 for mature cassava, with typical values about 0.160 to 0.170 for secondary vegetation. This represents a smaller shift than is commonly used to represent effects of deforestation in GCM experiments. Lowering the expected albedo increase by 3 to 4 percentage points is very significant in that it represents a reduction of 15% to 20% in the amount of simulated shortwave energy reflection from deforested areas. The estimated trend in regional-mean albedo indicates that deforestation-related albedo increase will level at a value less than 0.03 above the original forest value. Dirmeyer and Shukla (1994) found that an albedo increase of at least 0.03 was required to bring about precipitation reduction in simulations of tropical deforestation. According to their model results, a basinwide albedo change similar to that suggested by the Igarapé-Açu measurements would result in no significant change in regional precipitation.

Deforested land surfaces are represented in GCM deforestation experiments as uniform expanses of impoverished grassland. However, the dominant impression of the landscape in the Igarapé-Açu study area is one of great spatial diversity. Aggregation of surface properties necessary in GCMs is known to be a major source of uncertainty in model results. Failure to account for the effects of subgrid-scale landscape heterogeneity by prescribing a uniform deforested land cover may prevent future climatic impacts of deforestation from being predicted by the models. For example, mesoscale effects of land surface contrasts, as yet not treated in GCMs, have been shown to have significant influences on atmospheric dynamics (Anthes 1984; Yan and Anthes 1987; Segal et al. 1988; Giorgi 1989). Perhaps the focus of efforts to predict climatic impacts of deforestation should be on the effects of increasing land surface heterogeneity, rather than on a scenario of wholesale replacement of one homogenous surface by another.

Acknowledgments

The research reported here was initiated and carried out with the support and cooperation of numerous individuals associated with Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisas Agroflorestal da Amazônia Oriental (EMBRAPA/CPATU), Belém, Pará, Brazil; the joint German–Brazilian program Secondary Forests and Fallow Vegetation in Eastern Amazon Region, Function and Manipulation (SHIFT), and the Institute of Soil Science and Forest Nutrition and Institute of Agronomy in the Tropics, University of Göttingen, Göttingen, Germany; and the Geography Department, University of Hawaii at Manoa. We are especially grateful to Nilza Pacheco (EMBRAPA/CPATU), Tatiana Sá and Milton Kanashiro (EMBRAPA/CPATU and SHIFT), Manfred Denich (Institute of Agronomy in the Tropics and SHIFT), and Rose King (Geography Department). We are deeply indebted to the Ferreira family in Igarapé-Açu who gave generously of their knowledge of the land and allowed us to take measurements on their farm. Financial support for the study was provided by Instituto Interamericano de Cooperação para a Agricultura Escritório no Brasil. Fieldwork in Igarapé-Açu was carried out while the lead author served as an agrometeorological consultant at EMBRAPA/CPATU.

REFERENCES

  • Anthes, R. A., 1984: Enhancement of convective precipitation by mesoscale variations in vegetative covering in semiarid regions. J. Climate Appl. Meteor.,23, 541–554.

  • Barradas, V. L., and J. Adém, 1992: Albedo model for a tropical dry deciduous forest in western Mexico. Biometeorology,36, 113–117.

  • Bastable, H. G., W. J. Shuttleworth, R. L. G. Dallarosa, G. Fisch, and C. A. Nobre, 1993: Observations of climate albedo, and surface radiation over cleared and undisturbed Amazonian forest. Int. J. Climatol.,13, 783–796.

  • Bastos, T. X., D. Hölscher, and T. D. de A. Sá, 1993: Climate evaluation in northeast Pará, Igarapé-Açu site. Proc. Studies on Human Impact on Forest and Floodplains in the Tropics, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agroflorestal da Amazônia Oriental, p. 19.

  • Bird, R. E., and C. Riordan, 1986: Simple solar spectral model for direct and diffuse irradiance on horizontal and tilted planes at the earth’s surface for cloudless atmospheres. J. Climate Appl. Meteor.,25, 87–97.

  • Bruijnzeel, L. A., 1990: Hydrology of Moist Tropical Forests and Effects of Conversion: A State of Knowledge Review. UNESCO, 224 pp.

  • Culf, A. D., G. Fisch, and M. G. Hodnett, 1995: The albedo of Amazonian forest and ranchland. J. Climate,8, 1544–1554.

  • Denich, M., and M. Kanashiro, 1995: Secondary vegetation in the agricultural landscape of Northeast Pará, Brazil. Management and Rehabilitation of Degraded Lands and Secondary Forests in Amazonia, J. A. Parotta and M. Kanashiro, Eds., United States Dept. of Agriculture, 12–21.

  • Dirmeyer, P. A., and J. Shukla, 1994: Albedo as a modulator of climate response to tropical deforestation. J. Geophys. Res.,99, 20 863–20 877.

  • Fisch, G., I. R. Wright, and H. G. Bastable, 1994: Albedo of tropical grass: A case study of pre- and post-burning. Int. J. Climatol.,14, 102–107.

  • Galvão, E. U. P., J. Cesar, and E. C. Italiano, 1986: Estado atual de conhecimentos sobre sistemas de producas de culturas alimentares para o Estado do Amazonas (Current state of knowledge on annual cropping systems for the State of Amazonas). Proc. Simposio do Tropico Humido III, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisas Agropecuária do Trópico Úmido, 57–69.

  • Gash, J. H. C., and W. J. Shuttleworth, 1991: Tropical deforestation: Albedo and the surface-energy balance. Climate Change,19, 123–133.

  • Giambelluca, T. W., 1996: Tropical and cover change: Characterizing the post-forest land surface. Climate Change: Developing Southern Hemisphere Perspectives, T. W. Giambelluca and A. Henderson-Sellers, Eds., Wiley and Sons, 293–318.

  • Giorgi, F., 1989: Two-dimensional simulations of possible mesoscale effects of nuclear war fires, I. Model description. J. Geophys. Res.,94, 1127–1144.

  • Hecht, S. B., 1993: The logic of livestock and deforestation in Amazonia. Bioscience,43, 687– 695.

  • Henderson-Sellers, A., R. E. Dickinson, T. B. Durbridge, P. J. Kennedy, K. McGuffie, and A. J. Pitman, 1993: Tropical deforestation: Modeling local- to regional-scale climate change. J. Geophys. Res.,98, 7289–7315.

  • ——, H. Zhang, and W. Howe, 1996. Human and physical aspects of tropical deforestation. Climate Change: Developing Southern Hemisphere Perspectives, T. W. Giambelluca and A. Henderson-Sellers, Eds., Wiley and Sons, 259– 292.

  • Hölscher, D., 1995: Wasser- und Stoffhaushalt eines agrarökosystems mit waldbrache im östlichen Amazonasgebiet (Water and nutrient budget of an agroecosystem with forest fallow in eastern Amazonia). Ph.D. dissertation, Institute für Bodenkunde und Waldernährung der Georg-August-Universität Göttingen, 134 pp.

  • Kato, M. do S. A., and M. Denich, 1993: The municipality of Igarapé-Açu-Pará study site. Proc. Studies on Human Impact on Forest and Floodplains in the Tropics, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agroflorestal da Amazônia Oriental, 7–9.

  • Leopoldo, P. R., W. Franken, E. Matsui, and E. Salati, 1982a: Estimativa de evapotranspiragco de floresta Amazonica de terra firme (Estimation of evaporation of terra firme Amazonian forest). Acta Amazonica,12 (Suppl.), 23–28.

  • ——, ——, and E. Salati, 1982b: Water balance of a small catchment area in “terra firme” Amazonian forest. Acta Amazonica,12, 333–337.

  • Lesack, L. F. W., 1993: Water balance and hydrological characteristics of a rain forest catchment in the central Amazon Basin. Water Resour. Res.,29, 759–773.

  • McGuffie, K., A. Henderson-Sellers, H. Zhang, T. B. Durbridge, and A. J. Pitman, 1995: Global climate sensitivity to tropical deforestation. Global Planet. Change,10, 97–128.

  • Moran, E. F., E. Brondizio, P. Mausel, and Y. Wu, 1994: Integrating Amazonian vegetation, land-use, and satellite data. Bioscience,44, 329–338.

  • Nascimento, C., and A. Homma, 1984: Amazônia: Meio ambiente e technologia agricola Agropecuária do Trópico Úmido, 57–69. Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisas Agropecuária do Trópico Úmido, 282 pp.

  • Nobre, C. A., P. J. Sellers, and J. Shukla, 1991: Amazonian deforestation and regional climate change. J. Climate,4, 957–988.

  • Oguntoyinbo, J. S., 1970: Reflection coefficient of natural vegetation, crops, and urban surfaces in Nigeria. Quart. J. Roy. Meteor. Soc.,96, 430–441.

  • Pinker, R. T., 1982: The diurnal asymmetry in the albedo of tropical forest vegetation. Forest Sci.,28, 297–304.

  • ——, O. E. Thompson, and T. F. Eck, 1980: The albedo of a tropical evergreen forest. Quart. J. Roy. Meteor. Soc.,106, 551–558.

  • Rego, R. S., B. N. R. da Silva, and R. S. O. Junior, 1993: Detailed soil survey in an area in the municipality of Igarapé-Açu-Pará. Proc. Studies on Human Impact on Forest and Floodplains in the Tropics, Belém, Brazil, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agroflorestal da Amazônia Oriental, 13.

  • Segal, M., R. Avissar, M. McCumber, and R. Pielke, 1988: Evaluation of vegetation effects on the generation and modification of mesoscale circulations. J. Atmos. Sci.,45, 2268–2292.

  • Shuttleworth, W. J., 1988: Evaporation from Amazonian rainforest. Proc. Roy. Soc. London, Ser. B,233, 321–346.

  • ——, 1989: Micrometeorology of temperate and tropical forest. Philos. Trans. Roy. Soc. London, Ser. B,324, 173–436.

  • ——, and Coauthors, 1984: Observations of radiation exchange above and below Amazonian forest. Quart. J. Roy. Meteor. Soc.,110, 1163–1169.

  • Sud, Y. C., G. K. Walker, J.-H. Kim, G. E. Liston, P. J. Sellers, and K.-M. Lau, 1996: Biogeophysical consequences of a tropical deforestation scenario: A GCM simulation study. J. Climate9, 3225–3247.

  • Szott, L. T., and C. A. Palm, 1986: Soil and vegetation dynamics in shifting cultivation fallow. Proc. Simposio do Tropico Humido III, Empresa Brasileira de Pesquisa Agropecuária, Centro de Pesquisa Agropecuária do Trópico Úmido, Belém, Brazil, p. 360.

  • Turner, B. L., R. H. Moss, and D. L. Skole, 1993: Relating land use and global land-cover change: A proposal for an IGBP-HDP core project. International Geosphere-Biosphere Programme Rep. 24: A Study of Global Change and the Human Dimensions of Global Environmental Change, IGBP, 95 pp.

  • Uhl, C., R. Buschbacher, and E. A. S. Serrão, 1988: Abandoned pastures in eastern Amazonia, I. Patterns of plant succession. J. Ecol.,76, 663–681.

  • Watrin, O. d. S., 1994: Análise da dinâmica na paisagem agrícola da Amazônia oriental: Abordagem integrando técnicas de processamento digital de imagens e sistemas de informação geográfica (Agricultural dynamics evaluation in eastern Amazonia using digital image analysis techniques and geographic information systems). M.S. thesis, Ministério da Ciência e Tecnolocia, Instituto Nacional de Pesquisas Espaciais, 153 pp.

  • Wright, I. R., and Coauthors, 1996: Towards a GCM surface parameterisation for Amazonia. Amazonian Deforestation and Climate, J. H. C. Gash, C. A. Nobre, J. M. Roberts, and R. L. Victoria, Eds., Wiley and Sons, 474–504.

  • Yan, H., and R. Anthes, 1987: The effect of latitude on sea breeze. Mon. Wea. Rev.,115, 936–956.

Fig. 1.
Fig. 1.

Map of Igarapé-Açu study area.

Citation: Journal of Climate 10, 5; 10.1175/1520-0442(1997)010<0919:OOAARB>2.0.CO;2

Fig. 2.
Fig. 2.

Diurnal cycle of albedo for sites at Igarapé-Açu; curves for secondary vegetation sites are labeled by the age of vegetation.

Citation: Journal of Climate 10, 5; 10.1175/1520-0442(1997)010<0919:OOAARB>2.0.CO;2

Fig. 3.
Fig. 3.

Cycle of albedo for an actively farmed location in the eastern Amazon, based on a typical shifting cultivation pattern for small farms in the Igarapé-Açu region (see Table 8); 10-, 8-, and 6-yr cultivation cycles are shown to illustrate the effect of increasing land pressure on the regional-mean albedo; mean albedo is obtained by integrating the area under the curve for each cycle.

Citation: Journal of Climate 10, 5; 10.1175/1520-0442(1997)010<0919:OOAARB>2.0.CO;2

Table 1.

Albedo of tropical land surfaces.

Table 1.
Table 2.

Sensors and data recorders.

Table 2.
Table 3.

Characteristics of measurement sites at Igarapé-Açu. Map code refers to Fig. 1.

Table 3.
Table 4.

Summary of albedo measurements at Igarapé-Açu.

Table 4.
Table 5.

Ratios of net radiation to incident solar radiation for sites in Igarapé-Açu.

Table 5.
Table 6.

Midday surface temperature–air temperature differences for sites in Igarapé-Açu.

Table 6.
Table 7.

Estimation of regional-mean albedo for 1984, 1988, and 1991, estimated as area-weighted mean of albedo values assigned to each land cover category; areas based on land cover data of Watrin (1994).

Table 7.
Table 8.

Typical small farm land cover and albedo cycle in Igarapé-Açu.

Table 8.
Save