Abstract

This work studies the seasonal effects of the surface albedo change after a tropical deforestation and its influence on the seasonal pattern of precipitation in Amazonia. Results of the climate simulation study of Costa and Foley are used to represent the precipitation changes in a scenario where all tropical forest was replaced by pasture. The simulated forest and pasture albedo are compared to both the forest and pasture albedos observed during the Anglo-Brazilian Amazonian Climate Observation Study (ABRACOS) project. The surface albedo measurements provided verification that both forest and pasture albedo show seasonal variability, with the forest albedos higher in the dry period (June–November) and lower values in the rainy period (December–May), while the pasture albedo decreases in the dry season. Most of the characteristics of the observed differences in albedo are present in the simulated albedo too, which is in part due to the ability of the land surface parameterization to represent vegetation phenology. Verification was also obtained to show that the seasonal variability in the reflected radiation anomalies, caused by the tropical deforestation, is associated with seasonal fluctuations in the precipitation anomalies. This study demonstrates that most of the spatial and seasonal variability in the simulated climate after a tropical deforestation can be explained by the difference in the radiation reflected by the surface.

1. Introduction

During the past few years, the climate change after a tropical deforestation has been studied using several approaches, including global climate modeling, regional climate modeling, theoretical approaches, and field observations. Considering only the global climate modeling efforts, it has been learned that changes in surface albedo (α′) and vegetation roughness length (z0) are the two most important biophysical properties that affect the local climate (Lean and Warrilow 1989; Hahmann and Dickinson 1997), although the reduction in latent heat flux and atmospheric feedback processes, like moist convection and cloud-radiative effects, are also important (Zeng and Neelin 1999). The change in surface albedo has long been considered the most important driver of deforestation-induced climate change, so field efforts, like the Anglo-Brazilian Amazonian Climate Observation Study (ABRACOS; Gash et al. 1996), were conducted to obtain realistic values of albedo of tropical forests and pastures, so that the climate models could be driven by realistic parameters.

Eltahir (1996) theoretically explained the changes in precipitation patterns after the conversion of tropical forests to pastures. According to him, the increase in the surface albedo after deforestation causes a reduction in the net surface radiation, which cools the upper atmosphere over the deforested area, inducing a thermally driven circulation that results in subsidence. Zeng and Neelin (1999) described the feedback mechanisms that affect the water and energy balance of an atmospheric column in response to a scenario of increased surface albedo, which include feedbacks related to moisture convergence, evaporation, and cloud cover, concluding that the precipitation change (P′) could be explained as a linear function of the change in surface albedo (α′).

Most of the studies about the climate change caused by tropical deforestation analyzed annual mean data, or one specific season. However, analysis of the results of tropical deforestation experiments indicates that the differences in precipitation are seasonal. In addition, the precipitation change over the deforested area is not uniform, and an explanation about the spatial heterogeneity has not yet been proposed. Our goal in this paper is to contribute to the discussion of the climate change after a large-scale tropical deforestation, analyzing the seasonal precipitation changes. In this paper, we evaluate how seasonal variations in surface albedo and reflected radiation affect the changes in precipitation. First, we analyze the observed seasonal and spatial variability in surface albedo and discuss the reasons for the variability. Then, we do a simple analysis relating the seasonal change in precipitation to the seasonal changes in albedo and to the seasonal variability in incoming solar radiation.

2. Seasonal variability in the albedo of tropical vegetation

We can group the most important causes of albedo variability throughout the year in four major categories: (i) temporal variations in soil and leaf reflectance, (ii) changes in the fraction of soil covered by vegetation, (iii) factors that affect multiple reflections within the canopy, and (iv) partition of the solar radiation in direct and diffuse components.

In this paper, we will consider that reflectance is the ratio of reflected to incident radiation at a single wavelength, while albedo is the integration of reflectance through the solar radiation spectrum. Within a vegetation type, the leaf reflectance can vary due to different vegetation species, to leaf water stress, to the presence of water on leaf surfaces (leaf wetness), to the nutritional status of the plant, and to the age of the leaf (Bégué et al. 1996; Bremer and Ham 1999). Soil moisture affects soil albedo and leaf water stress, so it is also linked to surface albedo. Except for leaf water stress, soil moisture, and leaf wetness, which are all dependent on the precipitation patterns, all the additional above-mentioned causes of albedo variability are not likely to be important at the large scale.

If the albedo of the soil surface is significantly different from the vegetation albedo, changes in vegetation cover that expose the soil can be important. Tropical rainforests are usually evergreen or semideciduous, and the exposure of the soil surface underneath a tropical rainforest is rare. On the other hand, pasture grasses and crops that usually replace the tropical rainforest provide a poor cover of the soil surface. Therefore, in agricultural ecosystems, a small seasonality in the leaf area index may cause important variations in the surface albedo (Rosset et al. 1997; Giambellucca et al. 1999; Nouvellon et al. 2000).

The phenomenon of multiple reflection of the solar radiation by the canopy can also introduce a significant seasonal variability in the surface albedo. For a surface covered by a dense tropical forest, the solar beam has a low probability of penetration into the canopy and the albedo of the canopy is close to the albedo of the individual leaves. However, in partially opened canopies, solar beams penetrate the canopy, being reflected multiple times by branches, stems, soil surface, and the lower part of leaves, until a fraction finally leaves the canopy. When multiple reflections happen, the absorption of energy increases, and albedo decreases. Multiple reflections are only important when the vegetation cover is intermediate; that is, it is small enough to let direct sunlight penetrate the canopy, but high enough to reflect back a substantial portion of the upward-reflected solar radiation. Multiple reflections are more common if the solar elevation angle is higher, and if the leaf orientation is more vertical than horizontal. A detailed description of the radiative transfer in vegetative canopies can be found in Suits (1972), Kimes (1984), and Sellers (1985).

It is also expected that the albedo would be smaller under cloudy conditions. The diffuse radiation is more absorbed than the direct radiation, so, the higher the proportion of the diffuse radiation, the less important the solar angle (Pereira 1997; Song 1998; Nouvellon et al. 2000).

Here, we describe the seasonal variations of the surface albedo of tropical forests and ranch land and review the major factors that are important to the variability. Our discussions are based mainly on the data collected during the Amazon Region Micrometeorological Experiment (ARME) and ABRACOS experiments.

Figure 1 shows a typical seasonal variability of observed tropical vegetation surface albedo (solid lines). Figure 1a shows values measured at Reserva Ducke (forest) and at Fazenda Dimona, a nearby pasture, both at about 2°S, while Fig. 1b shows the data collected at Reserva Jaru (forest) and at Fazenda Nossa Senhora, a nearby pasture, both at around 10°S (data from the ABRACOS project, collected between 1991 and 1993; Culf et al. 1996). The forest albedo shows some seasonal variability, with higher values in the dry period (June–November), and lower values in the rainy period (December–May). On the other hand, the pasture albedo presents a stronger variability throughout the year, especially in Ji-Paraná, Rondonia, Brazil (Fig. 1b), with albedo values decreasing in the dry season, a consequence of the decrease in the leaf area index in the dry period (Wright et al. 1996). In Fig. 1, the combined effects of four of the most important sources of variability of surface albedo can be observed: leaf and soil moisture, leaf area index, partition of solar radiation in direct and diffuse components, and solar elevation angle.

Fig. 1.

Monthly values of surface albedo (α), measured at (a) Reserva Ducke (forest) and Fazenda Dimona (a nearby pasture), and (b) Reserva Jaru (forest) and Fazenda Nossa Senhora (a nearby pasture), during the ABRACOS project (data from Culf et al. 1996). The figure also shows simulated albedo values for the band of grid cells with the same latitude of the experimental site. The uncertainties in the estimates are on the order of 0.005 for simulated values, and 0.006 for observed values (Wright et al. 1996)

Fig. 1.

Monthly values of surface albedo (α), measured at (a) Reserva Ducke (forest) and Fazenda Dimona (a nearby pasture), and (b) Reserva Jaru (forest) and Fazenda Nossa Senhora (a nearby pasture), during the ABRACOS project (data from Culf et al. 1996). The figure also shows simulated albedo values for the band of grid cells with the same latitude of the experimental site. The uncertainties in the estimates are on the order of 0.005 for simulated values, and 0.006 for observed values (Wright et al. 1996)

Culf et al. (1995) discuss these results. They found a strong relationship between surface albedo and soil moisture, although most of it is probably related to soil moisture–correlated variables: the decrease in forest albedo in the rainy season is probably a consequence of smaller soil exposure, darker leaves (associated with the leaf water potential), higher leaf wetness, and darker humid soils, in both forest and pasture, although darker soils would be important only in the pasture.

Although Culf et al. (1995) did not find an important relationship between albedo and solar elevation angle in ABRACOS data, other studies indicate that albedo shows daily and seasonal cycles associated with changes in the solar elevation angle. Higher angles allow the deeper penetration of light inside a canopy, contributing to a higher absorption of the incident solar radiation, causing the albedo to be typically smaller in midlatitude summers (Giambellucca et al. 1999). Shuttleworth et al. (1984), analyzing ARME data, verified that the mean albedo (α) of the tropical rainforest depends on the solar elevation angle β, according to α = 15.09 − 0.136β + 0.00123β2 (α in percent, β in degrees). Pereira (1997) reports that, when the solar elevation angle is greater than 60°, there is a significant increase in the frequency of solar rays that penetrate deep in the canopy of a tropical rainforest, with consequent decrease in the albedo.

3. Seasonal variability in precipitation

To analyze the effects of the seasonal variability in surface albedo in the precipitation patterns, we use the results of the climate experiment of Costa and Foley (2000, hereafter CF). This study is performed using the National Center for Atmospheric Research (NCAR) Global Environmental and Ecological Simulation of Interactive System (GENESIS) climate model (version 2), coupled to the Integrated Biosphere Simulator (IBIS). GENESIS also includes an ocean model, in the form of a 50-m slab mixed layer ocean model with sea ice. Although the original experiment studied the combined effects of deforestation and double CO2 concentrations on the climate of Amazonia, only the results of the control run, that is, original vegetation cover (forest; F) and Amazonian deforestation experiment (D), in which the forest is replaced by a grassland, both at 1 × CO2, are considered in the present study. The readers can refer to CF for more details about the models and the experiment.

The CF tropical deforestation experiment has a unique characteristic that differs from other experiments of its genre. While most land surface parameterizations use prescribed albedos (either fixed or variable according to the calendar year), IBIS simulates the leaf phenology of the deciduous/semideciduous vegetation according to the climate. Thus, the effective albedo of the vegetation is calculated, rather than prescribed, and depends on the leaf area index (LAI), on the orientation and reflectance of leaves, and on the partition between direct and diffuse solar radiation (a detailed description of the albedo calculation in IBIS is given in the appendix). However, additional sources of variability of the surface albedo in the real world are not included in the computation of albedo in the model: the leaf reflectance does not vary with soil moisture, and the level of leaf wetness does not affect radiation transfer and reflection processes.

The annual mean simulated albedos, both for the forest and pasture areas, are close to the observed values (Table 1). The observed annual mean albedos are 0.133 and 0.176, for the sites of forest and pasture, respectively (Culf et al. 1996). The CF-simulated values are 0.131 and 0.167, respectively. The annual mean forest albedo varies between 0.122 and 0.147 for the three experimental sites of ABRACOS, while the simulated albedo varied spatially between 0.127 and 0.138. The spatial variability for the three pasture sites range from a minimum of 0.169 to a maximum of 0.193, while the variability of the simulated albedos throughout the deforested grid cells range from 0.163 to 0.190. In both cases, although the annual mean of the simulated albedo is close to the observed, the spatial variability of the simulated albedo is smaller than the observed one.

Table 1.

Comparison between the annual mean simulated and observed albedos, for forest and pasture areas, and range of spatial variability

Comparison between the annual mean simulated and observed albedos, for forest and pasture areas, and range of spatial variability
Comparison between the annual mean simulated and observed albedos, for forest and pasture areas, and range of spatial variability

In Fig. 1, it is shown that the simulated values of albedo usually correctly reproduce the observed values between August and November, when the canopy is dry and the LAI is low. During the wet season, forest albedo is overestimated, and pasture albedo is underestimated. This implies that the misrepresentation of albedo during this part of the year is related to the incomplete representation of the solar radiation processes associated with a wet canopy/soil. Uncertainties in the partition of the solar radiation in direct and diffuse components by the climate model can also affect these results. Simple model improvements that can enhance the representation of the vegetation albedo include the incorporation of a dependence of the leaf reflectance to leaf water stress and to the presence of water on leaf surfaces (leaf wetness), the representation of dead leaves (with remarkably different radiative properties) in the canopy (Wright et al. 1996), and the incorporation of the effect of wind in the orientation of leaves (Song 1998).

Figure 2 presents the monthly observed variability of the difference between the pasture and forest albedos (α′ = αDαF), for two ABRACOS sites (Manaus, Brazil, at 2.6°S, and Ji-Paraná, at 10.4°S), and the difference in the surface albedo simulated by the GENESIS/IBIS climate model, for the grid cells centered at latitudes 2.25° and 11.25°S. Three characteristics are evident in the observed values in Fig. 2: strong seasonality of the albedo difference, greater differences for higher latitudes and, for each paired site, the smallest differences happened in the months of September and October. Although the average simulated α′ is smaller than the average observed α′, the same characteristics are present in the albedos simulated by CF. However, since the IBIS parameterization does not include all sources of variability of surface albedo, and only represents the average characteristics for a whole grid cell, it cannot simulate as much heterogeneity as in reality, and therefore the variability of the simulated surface albedo is much smaller than the variability of the observed albedo.

Fig. 2.

Monthly values of difference in surface albedo (α′ = αDαF), observed at Manaus and Ji-Paraná, and the corresponding values simulated by the GENESIS/IBIS climate model. Uncertainties in the differences between albedos are on the order of 0.01

Fig. 2.

Monthly values of difference in surface albedo (α′ = αDαF), observed at Manaus and Ji-Paraná, and the corresponding values simulated by the GENESIS/IBIS climate model. Uncertainties in the differences between albedos are on the order of 0.01

We then examine the role that the seasonal variability of albedo might have in the local climate after a tropical deforestation. In our analysis seasonal changes in precipitation are not correlated with seasonal changes in albedo (ρ = 0.005). However, the incident radiation varies considerably at the seasonal scale and also spatially. To represent both the variability in incident radiation (S0) and in surface albedo (α), we then plot in Fig. 3 the precipitation changes after deforestation (P′) against changes in reflected radiation after deforestation (Sr), where Sr = S0 × α. In Fig. 3, each point represents an average of 3 months over a band of latitude. The regression analysis between P′ and Sr shows that 28% of the variance in the precipitation change after a tropical deforestation in Amazonia is explained by the variance in reflected radiation.

Fig. 3.

Relationship between the change in the precipitation pattern (P′), in mm day−1, and the change in energy reflected by the surface [Sr = (S0α)′]. Each point represents an average of 3 months over a band of latitude. For more details, refer to CF 

Fig. 3.

Relationship between the change in the precipitation pattern (P′), in mm day−1, and the change in energy reflected by the surface [Sr = (S0α)′]. Each point represents an average of 3 months over a band of latitude. For more details, refer to CF 

The climate change after a tropical deforestation (represented here by P′) is linked to Sr by biophysical, dynamical, and radiative mechanisms. Deforestation is characterized by an increment in albedo, a decrease in atmospheric turbulence (lower z0), a decrease in LAI, and a decrease in root depth (zR). Initially, consider only the nonradiative effects of deforestation (which is equivalent to assume α′ = 0): decreases in z0, LAI, and zR all contribute to a decrease in latent heat flux, and an increase in surface temperature, atmospheric instability, cloudiness, and precipitation.

As the reflected radiation increases (due to increases in surface albedo and to cloud-radiative feedbacks), surface temperature does not change, but surface latent and sensible heat fluxes decrease due to reduced radiation absorbed by the surface, resulting in a cooling of the atmospheric column, which induces a thermally driven circulation that results in subsidence, with subsequent reduction in convection, cloudiness, and precipitation. The cloud-radiative feedback is an important component of the entire process, as it modulates the amount of incident solar radiation at the surface.

Our results agree with the theoretical ideas of Zeng and Neelin (1999), who proposed that the precipitation change after a tropical deforestation (P′) could be explained by a linear function of the product of the incident solar radiation (S0) and albedo change (α′), being proportional to other parameters related to feedback mechanisms. Their equations, however, do not anticipate an increase in the precipitation for small increments in the product S0 × α, as our results demonstrate.

Our results also agree with the conclusions of Dirmeyer and Shukla (1994). Modeling the influence of the magnitude of surface albedo change on the climate, they demonstrated that the precipitation increases for α′ smaller than 0.03, and decreases for α′ greater than this threshold. Our study extends their results, showing that there is also a threshold in reflected radiation that separates positive from negative anomalies in precipitation.

4. Conclusions

This study arrives at two different conclusions. Initially, we verified that there are seasonal and spatial variations in the difference between the surface albedo of the pasture and of the forest. While the forest albedo has a relatively small variability throughout the year, showing little sensitivity to the solar elevation angle and the leaf and soil moisture, the pasture albedo is more sensitive to these factors, showing strong variability throughout the year. Reasons that could explain the higher variability in pasture albedo include height or density of vegetation, proportion of bare soil exposed, and the leaf elevation angle.

In addition, we verified that the seasonal and spatial fluctuations in the precipitation anomalies caused by the tropical deforestation are linearly related to seasonal and spatial variability in the differences of the reflected radiation. The decrease of the precipitation over the deforested area is proportional to the reflected radiation increment, consistent with the results of Dirmeyer and Shukla (1994) and Zeng and Neelin (1999). Since the incident solar radiation and the change in surface albedo depend on the latitude, our analysis helps explain the regional patterns of precipitation changes over the deforested area.

At the seasonal scale, we observed that precipitation increases when the change in surface-reflected radiation is smaller than a threshold, and decreases when the change in reflected radiation is greater than the threshold. Dirmeyer and Shukla (1994) arrived at a similar conclusion analyzing annual mean results. Our results are more generic, though, because the variability in the reflected radiation includes not only the variability in surface albedo, but also the variability in incident solar radiation related to time of the year, latitude, and cloud feedbacks.

The conclusions of this work indicate that climate models that incorporate a module of dynamical vegetation have the potential to realistically represent seasonal and spatial variations in the patterns of precipitation associated with tropical deforestation. The inclusion of parameterizations that allow the leaf reflectance and scattering coefficients to vary with soil moisture and with the level of leaf wetness may improve the simulations of the seasonal variability of the surface albedo.

Acknowledgments

This research was supported by the Brazilian Agency CNPq.

REFERENCES

REFERENCES
Bégué
,
A.
,
J. L.
Roujean
,
N. P.
Hanan
,
S. D.
Prince
,
M.
Thawley
,
A.
Huete
, and
D.
Tanré
,
1996
:
Shortwave radiation budget of Sahelian vegetation: 1. Techniques of measurement and results during HAPEX–Sahel.
Agric. For. Meteor.
,
79
,
79
96
.
Bremer
,
D. J.
, and
J. M.
Ham
,
1999
:
Effect of spring burning on the surface energy balance in a tallgrass prairie.
Agric. For. Meteor.
,
97
,
43
54
.
Costa
,
M. H.
, and
J. A.
Foley
,
2000
:
Combined effects of deforestation and doubled atmospheric CO2 concentrations on the climate of Amazonia.
J. Climate
,
13
,
18
34
.
Culf
,
A. D.
,
G.
Fisch
, and
M. G.
Hodnett
,
1995
:
The albedo of Amazonian forest and ranch land.
J. Climate
,
8
,
1544
1554
.
Culf
,
A. D.
,
J. L.
Esteves
,
A. O.
Marques Filho
, and
H. R.
Rocha
,
1996
:
Radiation, temperature and humidity over forest and pasture in Amazonia.
Amazonian Deforestation and Climate, J. H. C. Gash et al., Eds., John Wiley and Sons, 175–191
.
Dickinson
,
R. E.
,
A.
Henderson-Sellers
,
P. J.
Kennedy
, and
M. F.
Wilson
,
1986
:
Biosphere–atmosphere transfer scheme (BATS) for the NCAR Community Climate Model.
NCAR Tech. Note NCAR/TN-275+STR, National Center for Atmospheric Research, Boulder, CO, 69 pp
.
Dirmeyer
,
P. A.
, and
J.
Shukla
,
1994
:
Albedo as a modulator of climate response to tropical deforestation.
J. Geophys. Res.
,
99
,
20863
20877
.
Eltahir
,
E. A. B.
,
1996
:
Role of vegetation in sustaining large-scale atmospheric circulations in the tropics.
J. Geophys. Res.
,
101
,
4255
4268
.
Gash
,
J. H. C.
,
C. A.
Nobre
,
J. M.
Roberts
, and
R. L.
Victoria
,
Eds.,
.
1996
:
Amazonian Deforestation and Climate.
John Wiley & Sons, 611 pp
.
Giambelluca
,
T. W.
,
J.
Fox
,
S.
Yarnasarn
,
P.
Onibutr
, and
M. A.
Nullet
,
1999
:
Dry-season radiation balance of land covers replacing forest in northern Thailand.
Agric. For. Meteor.
,
95
,
53
65
.
Hahmann
,
A. N.
, and
R. E.
Dickinson
,
1997
:
RCCM2–BATS model over tropical South America: Applications to tropical deforestation.
J. Climate
,
10
,
1944
1964
.
Kimes
,
D. S.
,
1984
:
Modeling the directional reflectance from complete homogeneous vegetation canopies with various leaf orientation distribution.
J. Opt. Soc. Amer.
,
A1
.
725
737
.
Lean
,
J.
, and
D. A.
Warrilow
,
1989
:
Simulation of the regional climatic impact of Amazon deforestation.
Nature
,
342
,
411
413
.
Nouvellon
,
Y.
,
A.
Bégué
,
M. S.
Moran
,
D. L.
Seen
,
S.
Rambal
,
D.
Luquet
,
G.
Chehbouni
, and
Y.
Inoue
,
2000
:
PAR extinction in shortgrass ecosystems: Effects of clumping, sky conditions and soil albedo.
Agric. For. Meteor.
,
105
,
21
41
.
Pereira
,
A. R.
,
1997
:
Radiation regime of tropical rain forest.
Rev. Bras. Agrometeor.
,
5
,
i
viii
.
Pollard
,
D.
, and
S. L.
Thompson
,
1995
:
Use of a Land-Surface-Transfer Scheme (LSX) in a global climate model—The response to doubling stomatal-resistance.
Global Planet. Change
,
10
,
129
161
.
Rosset
,
M.
,
M.
Riedo
,
A.
Grub
,
M.
Geissmann
, and
J.
Fuhrer
,
1997
:
Seasonal variation in radiation and energy balances of permanent pastures at different altitudes.
Agric. For. Meteor.
,
86
,
245
258
.
Sellers
,
P. J.
,
1985
:
Canopy reflectance, photosynthesis, and transpiration.
Int. J. Remote Sens.
,
6
,
1335
1372
.
Sellers
,
P. J.
,
Y.
Mintz
,
Y. C.
Sud
, and
A.
Dalchor
,
1986
:
A Simple Biosphere Model (SiB) for use within general circulation models.
J. Atmos. Sci.
,
43
,
505
531
.
Shuttleworth
,
W. J.
, and
Coauthors
.
1984
:
Observations of radiation exchange above and below amazonian forest.
Quart. J. Roy. Meteor. Soc.
,
110
,
1163
1169
.
Song
,
J.
,
1998
:
Diurnal asymmetry in surface albedo.
Agric. For. Meteor.
,
92
,
181
189
.
Suits
,
G. H.
,
1972
:
The calculation of the directional reflectance of a vegetative canopy.
Remote Sens. Environ.
,
2
,
117
123
.
Wright
,
I. R.
, and
Coauthors
.
1996
:
Torwards a GCM surface parameterization for Amazonia.
Amazonian Deforestation and Climate, J. H. C. Gash et al., Eds., John Wiley and Sons, 473–504
.
Zeng
,
N.
, and
J. D.
Neelin
,
1999
:
A land–atmosphere interaction theory for the tropical deforestation problem.
J. Climate
,
12
,
857
872
.

APPENDIX

Determination of Surface Albedo in IBIS

The solar radiation model in IBIS is solved every time step to provide the solar heating rates of the upper and lower canopy leaves and stems, and soil surface. All calculations are performed separately for two different bands: visible and near-infrared.

Within each vegetation layer, a two-stream approximation is used for solar radiation. This plane-parallel approach accounts for multiple reflections by leaves and stems, and radiation trapping by dense canopies. The formulation follows Dickinson et al. (1986), Sellers et al. (1986), and Pollard and Thompson (1995). The equations that represent the radiative transfer through the canopy are solved analytically for the two layers underlain by the soil, considered to be a nonspecular surface of known albedo, using prescribed incoming downward direct and diffuse radiation fluxes as boundary conditions. The canonical solution is applied first to the lower layer underlain by soil to obtain the effective albedo of the lower layer + soil system, and then to the upper layer underlain by the lower layer + soil system, yielding the overall surface albedo needed for the atmospheric model radiation computations.

Although complex, the representation of the solar radiation processes in the canopy by IBIS is still incomplete. For example, the leaf/stem scatter and backscatter coefficients depend on the fraction of the leaves and stems that is covered by snow, but is independent of the level of leaf wetness. A rigorous incorporation of leaf wetness in the model would increase its complexity considerably, because of the specular reflective properties of the liquid water.

Footnotes

Corresponding author address: Dr. Marcos Heil Costa, Federal University of Viçosa Av. P. H. Rolfs, s/n CEP 36571-000, Viçosa MG, Brazil. Email: mhcosta@ufv.br