Interactions between rainfall, deforestation and fires during recent years in the Brazilian Amazonia

Phil. Trans. R. Soc. B (2008) 363, 1779–1785
doi:10.1098/rstb.2007.0026
Published online 11 February 2008
Interactions between rainfall, deforestation and
?res during recent years in the Brazilian Amazonia
Luiz Eduardo O. C. Araga
˜ o1,*, Yadvinder Malhi1, Nicolas Barbier1,2,
Andre Lima3, Yosio Shimabukuro3, Liana Anderson1 and Sassan Saatchi4
1Environmental Change Institute, Oxford University Centre for the Environment, University of Oxford,
Oxford OX1 3QY, UK
2Universite´ Libre de Bruxelles, Service de Botanique Syste´matique et Phytosociologie,
CP 169, 1050 Bruxelles, Belgium
3Brazilian Institute for Space Research (INPE ), Sa˜o Jose´ dos Campos, Sa˜o Paulo 12227-010, Brazil
4Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA
Understanding the interplay between climate and land-use dynamics is a fundamental concern for
assessing the vulnerability of Amazonia to climate change. In this study, we analyse satellite-derived
monthly and annual time series of rainfall, ?res and deforestation to explicitly quantify the seasonal
patterns and relationships between these three variables, with a particular focus on the Amazonian
drought of 2005. Our results demonstrate a marked seasonality with one peak per year for all
variables analysed, except deforestation. For the annual cycle, we found correlations above 90% with
a time lag between variables. Deforestation and ?res reach the highest values three and six months,
respectively, after the peak of the rainy season. The cumulative number of hot pixels was linearly
related to the size of the area deforested annually from 1998 to 2004 (r 2Z0.84, pZ0.004). During
the 2005 drought, the number of hot pixels increased 43% in relation to the expected value for a
similar deforested area (approx. 19 000 km2). We demonstrated that anthropogenic forcing, such as
land-use change, is decisive in determining the seasonality and annual patterns of ?re occurrence.
Moreover, droughts can signi?cantly increase the number of ?res in the region even with decreased
deforestation rates. We may expect that the ongoing deforestation, currently based on slash and burn
procedures, and the use of ?res for land management in Amazonia will intensify the impact of
droughts associated with natural climate variability or human-induced climate change and, therefore,
a large area of forest edge will be under increased risk of ?res.
Keywords: Amazonia; ?re; deforestation; drought; land use; climate change
1. INTRODUCTION
SST alone has been implicated as a causal factor of the
There is a growing concern about the impacts of
severe drought that affected Amazonia in 2005
climate change on the stability of ecological processes
(Marengo et al. in press).
in Amazonia, the resulting feedbacks from the local to
The impacts of reducing rainfall over Amazonia are
the global circulation system and the ensuing con-
likely to be exacerbated by the synergic interactions
sequences on plant, animal and human populations.
among other anthropogenic forcing factors such as
Some global circulation models suggest that Amazonia
deforestation and ?res (Cochrane & Laurance 2002;
may be vulnerable to extreme drying in response to
Hutyra et al. 2005). Positive feedbacks among deforesta-
circulation shifts induced by global warming (Li et al.
tion, ?res and drought have been previously reported
2006), possibly leading to a dieback of tropical
(e.g. Cochrane et al. 1999; Laurance & Williamson
rainforest with potential acceleration of global warming
2001). Drought alone is reported to reduce tree growth,
(Cox et al. 2004).
increase tree mortality (particularly in forest edges) and
Amazonian droughts have been strongly related to
increase leaf shedding. This process leads to the increase
El Nin
˜ o events, such as in 1982/1983, 1986/1987 and
of canopy openness and understorey insolation with
1997/1998 (Marengo 1992; Uvo et al. 1998; Ronchail
consequent drying of the accumulated litter. When these
et al. 2002; Marengo 2004) and more recently to the
conditions are combined with intense forest degradation
tropical Atlantic sea surface temperature (SST)
through edge effects and logging, the risk of forest ?res
anomalies associated with the Atlantic Multidecadal
can increase dramatically in Amazonia (Uhl & Kauffman
Oscillation (Li et al. 2006; Good et al. 2008; Marengo
1990; Cochrane & Schulze 1999; Cochrane et al. 1999;
et al. in press). The increase of the tropical Atlantic
Laurance & Williamson 2001; Barlow & Peres 2004;
Nepstad et al. 2004). On the other hand, large-scale forest
conversion (Nobre et al. 1991; Laurance & Williamson
* Author for correspondence (leocaragao@gmail.com).
2001; Laurance et al. 2002; Silva Dias et al. 2005; Costa
et al. 2007) and the smoke from ?res (Rosenfeld 1999;
One contribution of 27 to a Theme Issue ‘Climate change and the
fate of the Amazon’.
Ackerman et al. 2000; Artaxo et al. 2005) may promote
1779
This journal is q 2008 The Royal Society

---

1780
L. E. O. C. Araga˜o et al.
Rainfall, deforestation and ?re in Amazonia
a reduction in rainfall over these areas. This chain of
correspond to the breaking down of the signal’s variance into
events generates a positive feedback loop that increases
frequency bins. In other words, the relative strength of a
the vulnerability of Amazonia to climate change.
periodic component of a given frequency in the signal is given by
In this study, we focus on Brazilian Amazonia where
the power spectrum value at that frequency. For a pair of signals,
data are readily available. Here we used satellite-derived
one can equivalently compute a combined power spectrum (or
time series of rainfall, ?res and deforestation to explicitly
cross-spectrum), which allows exploring shared periodicities
quantify the seasonal patterns of these three variables and
(cycles yrK1) between the two signals. In this case, it is the
their relationships, with a particular focus on the 2005
covariance between signals, which can be investigated at speci?c
Amazonian drought. In addition, we investigate how
temporal scales (frequencies). The resulting cross-spectrum
rainfall and deforestation in?uence ?re dynamics at the
can be analysed in terms of amplitude and phase. The
monthly and annual time scales. Finally, we discuss how
coherence spectrum, which is the amplitude normalized
between 0 and 1, can be interpreted as a Pearson product–
climate variability and the occurrence of droughts,
moment correlation coef?cient between series, computed for
deforestation and ?res can potentially increase the
each frequency. The phase spectrum indicates the phase lag
vulnerability of Amazonia to climate change.
between signals. A strong coherence for a speci?c temporal
frequency, combined with a null phase shift, indicates a positive
correlation while a phase shift of p corresponds to a negative
2. MATERIAL AND METHODS
correlation. Variance estimates can be computed for both
(a) Rainfall, ?re and deforestation datasets
coherence and phase spectra (Diggle 1989), to allow building
We used a time series ( January 1998–December 2006) of
pointwise CIs for these estimates.
cumulative monthly precipitation (mm per month) derived
After identifying the connections among the variables,
from the tropical rainfall measuring mission data (TRMM
we conducted a regression analysis using the monthly and
3B43-v6) at 0.258 spatial resolution ( NASA 2006). The
annual data to explore the shape of the relationship between
validation of this dataset showed that TRMM product
the variables.
captures the rainfall patterns of the Amazonian region
accurately (Araga˜o et al. 2007).
The INPE-DETER (Detection of Deforested Areas in Real
Time) dataset ( INPE 2006a) was used to quantify the
3. RESULTS AND DISCUSSION
cumulative monthly area (km2) of deforested polygons
(a) Seasonality of rainfall, ?re and deforestation
(April 2004–October 2005 and March 2006–September
and their relationships
2006). Deforestation values for four missing months
Results presented in ?gure 1 show that rainfall,
( November 2005–February 2006) were estimated using
deforestation and hot pixels have a marked annual
proportional values between the subsequent months in the
periodicity. Our analysis indicates that, on average, the
previous year. In addition, the time series (1998–2005) of
dry season (rainfall below 100 mm per month, based
annual cumulative deforested area was obtained from the
on Araga˜o et al. 2007) persists from July to September
INPE-PRODES (Assessment of Deforestation in Brazilian
Amazonia) dataset (INPE 2005).
for most of the years analysed, excepting 2005, when
Hot pixel counts were derived from daily, 1 km spatial
the dry season started in June (?gure 1a), in association
resolution, NOAA-12 ( National Oceanic and Atmospheric
with the drought that struck the basin in this year. Both
Administration) database from the Brazilian Institute for
deforestation and hot pixels peaked during the dry
Space Research ( INPE ) Queimadas project (mid-1998–2005;
season in Amazonia. The major peak of deforestation is
INPE 2006b). Hot pixels are indicators of ?res and may well
observed in May (?gure 1b) while the peak of hot pixels
underestimate their occurrence owing to clouds and forest
coincides with the months of lowest rainfall, August
canopy cover, but hot pixel counts do allow the evaluation of
and September (?gure 1c).
patterns over time.
The seasonal signal of ?res observed here can be
decomposed into three distinct types: (i) areas that have
(b) Data analysis
been deforested and then burnt in the same year, (ii) areas
We extracted from the remote-sensing surfaces the monthly
that have been deforested in previous years and then
cumulative values of the area of deforested polygons and
burnt later, and (iii) ?res in natural vegetation and other
the number of hot pixels as well as the average rainfall
non-forest areas that are not included in the INPE
within the limits of the Brazilian Legal Amazonia (approx.
dataset. To investigate the relative contribution of each of
4 000 000 km2). This region includes the states of Amazonas,
these categories to the total number of hot pixels
Acre, Rondoˆnia, Roraima, Mato Grosso, Para´, Amapa´,
observed, the 2005 map of hot pixels was overlaid on
Maranha˜o and Tocantins.
the deforestation map derived from INPE-DETER data
We analysed the behaviour of the monthly rainfall,
and subdivided into land cover classes (?gure 2).
deforested area and hot pixels through time to identify possible
Our results demonstrate that ?res in deforested
seasonality in the data. As an additional support for the
areas contributed to 60% of the total number of hot
interpretation of these data, we generated four maps showing
pixel detections in 2005. Of the remaining 40% of
?rst the total cumulative deforestation in Amazonia, based on
detections, 28% occurred in forests and 12% in areas
INPE-DETER data, and subsequently the hot pixels and
counts in 2005 for each one of the three land cover classes
considered as non-forest in the INPE land cover
de?ned in the deforestation map. Time series were analysed and
classi?cation. The hot pixels in areas deforested
compared using (cross)-spectral analysis ( Priestley 1981;
during 2005 and until 2004 contributed to 8 and
Diggle 1989). This well-established approach extends the
92%, respectively, of the total number of detections
power spectra methodology to the comparison of pairs of time
in deforested areas. On the other hand, the large
series. The values in a power spectrum, computed as the
percentage of hot pixels detected in forests during 2005
squared amplitude of the Fourier transform of the signal,
was associated with the leakage of ?res from the
Phil. Trans. R. Soc. B (2008)

---

Rainfall, deforestation and ?re in Amazonia
L. E. O. C. Araga˜o et al.
1781
500
(a)
450
400
350
300
250
all (mm) 200
150
rainf 100
6000
50
0
5000
)2
(b)
4000
3000
2000
1000
deforested area (km
60
(c)
0
50
counts)3 40
10
30
els (×
20
10
hot pix
0
v
v
v
v
v
v
v
v
v
Jan
Feb Mar Apr
May
Jun
Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct No Dec
1998
1999
2000
2001
2002
2003
2004
2005
2006
Figure 1. Monthly time series of (a) mean rainfall (mm) derived from the TRMM dataset ( January 1998–December 2006),
(b) cumulative deforested area (km2) from the INPE-DETER dataset (April 2004–September 2006) and (c) cumulative number
of hot pixel detections from NOAA-12 dataset (May 1998–December 2006) within the limits of the Brazilian Legal Amazonia.
Dashed lines in (a) correspond to the s.d. of the mean monthly rainfall (nZ6705 pixels). Grey bars indicate the dry season length
for each year (months with rainfall below 100 mm monthK1).
deforested areas to surrounding drought-stressed forest
These results elucidated the interaction between
edges (Araga˜o et al. 2007).
climate and land-use practices, describing the timing of
The spectral analysis of the monthly data stressed a
slash and burn activities in the Brazilian Amazonia. In
clear seasonal variation in all time series analysed
summary, approximately three months after the peak of
(?gure 3a – c), as all power spectra showed a peak for a
the rainy season, deforestation reaches its highest
frequency of one cycle yrK1. Deforestation, however, is
annual values. In this case, there is a prognostic action
dominated by another peak at two cycles yrK1.
in relation to the peak of the dry season, giving time for
The coherence spectra (?gure 3d – f ) showed that
the fallen wood to dry until the driest month. After-
the correlation among all time series was higher than
wards, during the peak of dry season (minimum rainfall
90% for an annual periodicity (i.e. one cycle yrK1).
values), farmers set ?re to the dry material on the
Focusing on the coherent annual frequency (high-
ground and hot pixel values reach their maximum.
lighted by an arrow in ?gure 3g – i ), the phase spectrum
showed a phase shift of approximately p/2 for the
(b) In?uence of monthly and annual rainfall
relation between hot pixels and deforestation, meaning
and deforestation on ?re dynamics
that deforestation led the presence of hot pixels by
At the monthly time scale, deforested area increases
approximately three months (note that at one cycle yrK1,
exponentially with the decrease of rainfall assuming the
2p, p and p/2 are equivalent to 12, 6 and 3 months time
three-month lag de?ned in the spectral analysis
lag, respectively). Similarly, the relationship between
(r 2Z0.74, equation (3.1); ?gure S1 in the electronic
rainfall
and
deforestation
had
a
phase
shift
supplementary material).
of approximately p/2, which indicates that the peak of
deforestation Z 4116:55 expK0:01!rain:
ð3:1Þ
rainfall precedes the deforestation peak by three
This means that the higher rate of deforestation in
months; however, the second deforestation peak high-
April/May is strongly related to the rainfall in June/July,
lighted in the power spectra coincides with the peak of
which is the beginning of the dry season in most of the
the dry season. Finally, the comparison between
Brazilian Amazonia. Besides this, hot pixel detections
rainfall and hot pixels revealed, as expected, that
tended to increase exponentially with the decrease of
rainfall was negatively correlated (phase shift of p)
rainfall (r 2
with the number of hot pixels. Therefore, the peak of
Z0.60, equation (3.2); ?gure S1 in the
electronic supplementary material)
hot pixel detections matches the peak of the dry season
in Amazonia without time lag.
hot pixels Z 63588:21 expK0:02!rain:
ð3:2Þ
Phil. Trans. R. Soc. B (2008)

---

1782
L. E. O. C. Araga˜o et al.
Rainfall, deforestation and ?re in Amazonia
(a)
(b)
(c)
(d )
Figure 2. Maps of the Brazilian Amazonia showing (a) the total cumulative deforested area based on the INPE-DETER dataset
until 2004 ( yellow) and in 2005 (red), and the annual cumulative number of hot pixel detections in 2005 from NOAA-12 dataset
over (b) areas deforested until 2004, (c) areas deforested in 2005 and (d ) forested areas in 2005.
Conversely, we did not ?nd a strong relationship
(c) Interactions between land-use and climate
between hot pixels and deforested area at the monthly
change and the vulnerability of Amazonia
time scale.
In the last decade Amazonia experienced two droughts,
Despite the fact that deforestation may not be a
in 1997/1998 and 2005. Both droughts caused
major predictor of hot pixel counts, either spatially
signi?cant rainfall anomalies and hydrological stress,
(Cardoso et al. 2003) or temporally at a monthly scale,
signi?cantly increasing the number of ?res detected
we found a strong linear relationship (r 2Z0.84,
over this region (Araga˜o et al. 2007). The areas affected
pZ0.004, equation (3.3)) between the annual cumu-
by ?res are expected to become more vulnerable to
lative number of hot pixels and the size of the area
recurrent ?res (Uhl & Kauffman 1990; Cochrane &
deforested annually from 1998 to 2004
Schulze 1999; Nepstad et al. 1999).
hot pixels Z 8:50 !deforestationK69570:59:
ð3:3Þ
The interaction between land-use and climate
We attributed the linear trend observed between hot
change is likely to generate a positive feedback (e.g.
pixels and deforestation from 1998 to 2004 to the
Cochrane et al. 1999), increasing the vulnerability of
expansion of pastures for cattle ranching and large
Amazonia to climate change, and have signi?cant
areas of mechanized agriculture (Morton et al. 2006) in
effects on the global carbon cycle. For example, the
the southern part of Brazilian Amazonia. The expan-
estimated global ?ux of CO2 to the atmosphere from
sion of mechanized agriculture was mainly driven by
land-use change was 1.6 (0.5–2.7) Pg C yrK1 for the
the area planted by soya bean crops in Amazonia that
1990s, 22% of total anthropogenic emissions (Denman
increased from 1
et al. 2007). The Brazilian Amazon alone might yield a
!106 ha
in 1990 to 7!106 ha
in 2005, with an expansion rate of 17% yrK1 from
net ?ux of carbon from the biosphere to the atmosphere
2000 to 2005 (Costa et al. 2007). However, during the
of 0.1–0.4 Pg C yrK1, due to land-use change
2005 drought, the effect of rainfall de?cit overtook
(Houghton et al. 2000). This is equivalent to 6–25%
the in?uence of land-use change on hot pixel
of the total carbon emissions from land-use changes.
dynamics (?gure 4).
These emissions can overtake the sink of carbon
Phil. Trans. R. Soc. B (2008)

---

Rainfall, deforestation and ?re in Amazonia
L. E. O. C. Araga˜o et al.
1783
(a)
(b)
(c)
0.30
0.25
0.20
wer 0.15
po 0.10
0.05
0
(d )
(e)
( f )
1.0
y
0.8
0.6
0.4
coherenc
0.2
0
( g)
(h)
(i )
/ 2
0
phase – /2
– 0
1
2
3
4
0
1
2
3
4
0
1
2
3
4
cycles per year
cycles per year
cycles per year
Figure 3. Spectral analysis on the monthly time series showing (a– c) the power spectra (Fourier periodograms) of (a) hot pixels
and deforestation (nZ30), (b) rainfall and deforestation (nZ30) and (c) rainfall and hot pixels (nZ103). The value at each
temporal frequency gives the relative strength of the corresponding periodic component in each series. (d – f ) The coherency
spectra for the relationship between (d ) hot pixels and deforestation, (e) rainfall and deforestation and ( f ) rainfall and hot pixels
are shown. The value for each frequency gives the correlation (similar to the Pearson’s correlation coef?cient) between the
corresponding periodic components in both signals. ( g – i ) The phase spectra between ( g) hot pixels and deforestation,
(h) rainfall and deforestation and (i ) rainfall and hot pixels. Here, the values indicate the lag between the periodic components in
both signals at each frequency (at one cycle yrK1, 2p, p and p/2 are equivalent to 12, 6 and 3 months time lag, respectively). The
dashed lines in (d–i ) indicate the 95% bilateral pointwise CI computed using a four-month smoothing window. The arrows
indicate the phase shift for the coherent annual frequency. (a) Light grey circles, ?res; dark grey squares, deforestation. (b) Light
grey circles, rain; dark grey squares, deforestation. (c) Light grey circles, rain; dark grey squares, ?res.
calculated for the undisturbed ecosystems in this region
20
drought
( Nepstad et al. 1999; Barlow & Peres 2004; Malhi &
)
Phillips 2004; Phillips et al. 2008).
–1
The effect of deforestation on ?re impacts is likely to
yr 17
2004
2005
be exacerbated by drought events, which may become
more frequent under some climate change scenarios
counts 14
2002
2003
4
( Timmermann 1999; Cox et al. 2004; Li et al. 2008).
Based on the relationship found between deforested
11
pastures and mechanized crops
area and hot pixels (?gure 4, equation (3.3)), we
expansion (e.g. soyabean)
els (×10
1999
investigated the impact of drought and deforestation on
8
2001
?re patterns, not considering any political and
1998
hot pix
2000
= 0.84
economical variables that may in?uence ?re dynamics
5
in the region. We estimated that during the 2005
15
20
25
30
drought, the number of hot pixels (160 464 detections)
deforested area (×103 km2 yr –1)
were 43% higher than the expected value for a similar
Figure 4. Linear regression between the annual cumulative
deforested area (approx. 19 000 km2). Using equation
number of hot pixels and the annual cumulative deforested
(3.3), we calculated the expected values under ‘normal’
area between 1998 and 2004 derived from the INPE-
and ‘dry’ conditions to estimate the impact of drought
PRODES dataset (nZ7, pZ0.004). It shows the linear ?t,
with increased deforestation on hot pixel counts. We
indicated by the grey arrow and the coef?cient of determina-
found that the rate of hot pixel detection per kilometre
tion (r 2). Note that 2005 is not included in the regression due
square of deforested area annually would double under
to its anomalous characteristic as a function of the drought.
conditions similar to the 2005 drought. Moreover, the
difference between the number of hot pixels in normal
consequently lead to the increase of CO2 emissions to
and dry conditions increases linearly with the increase
the atmosphere due to biomass burning.
of deforested area at a rate of 6.3 detections per
kilometre square of deforested area annually (?gure S2
in the electronic supplementary material). Based on
4. CONCLUSIONS
these estimations, one can anticipate that the increased
Our results stress a clear seasonality and synergic
rate of hot pixel counts under drought conditions is
interaction between climate, deforestation and ?res.
likely to increase the area of forests affected by ?res and
We demonstrated here that anthropogenic forcing,
Phil. Trans. R. Soc. B (2008)

---

1784
L. E. O. C. Araga˜o et al.
Rainfall, deforestation and ?re in Amazonia
such as land-use changes, is decisive in determining the
REFERENCES
seasonality and the annual patterns of ?re occurrence.
Ackerman, A. S., Toon, O. B., Taylor, J. P., Johnson, D. W.,
Moreover, drought events can increase signi?cantly the
Hobbs, P. V. & Ferek, R. J. 2000 Effects of aerosols on
number of ?res in the region even with decreased
cloud albedo: evaluation of Twomey’s parameterization of
deforestation rates. We may expect that the ongoing
cloud susceptibility using measurements of ship tracks.
deforestation, currently based on slash and burn
J. Atmos. Sci. 57, 2684 –2695. (doi:10.1175/1520-0469
(2000)057!2684:EOAOCAO2.0.CO;2)
procedures, and the use of ?res for land management
Alves, D. 2008 Some perspectives on the human dimensions
in Amazonia will intensify the impact of droughts
of environmental change in Amazonia. Phil. Trans. R. Soc.
associated with natural climate variability or human-
B 363, 1903–1909. (doi:10.1098/rstb.2007.0020)
induced climate change, and therefore a large area of
Araga˜o, L. E. O. C., Malhi, Y., Roman-Cuesta, R. M.,
forest edges will be under increased risk of ?re.
Saatchi, S., Anderson, L. O. & Shimabukuro, Y. E. 2007
The impacts of ?res on the Amazonian region could
Spatial patterns and ?re response of recent Amazonian
be reduced with the support of governments, since ?res
droughts. Geophys. Res. Lett. 34, L07701. (doi:10.1029/
here are mainly induced by human activities and could
2006GL028946)
be avoided and/or diminished by: the introduction of
Artaxo, P., Gatti, L. V., Leal, A. M. C., Longo, K. M.,
?re-free land management techniques; reinforcement
Freitas, S. R., Lara, L. L., Paulisquevis, T. M., Procopio,
of monitoring, controlling and application of the
A. S. & Rizzo, L. V. 2005 [Atmospheric chemistry in
Amazonia: the forest and the biomass burning emissions
current Brazilian legislation to illegal ?res; creation of
controlling the composition of the Amazonian atmos-
protected areas ( Nepstad et al. 2006); and environ-
phere]. Acta Amazon. 35, 185–196. [In Portuguese.]
mental education programmes.
Barlow, J. & Peres, C. A. 2004 Ecological responses to El
Some initiatives, such as the creation of the
Nin
˜ o-induced surface ?res in central Brazilian Amazonia:
Extractive Reserves statutes, the Pilot Program to
management implications for ?ammable tropical forests.
Conserve the Brazilian Rain Forest (PPG7), and
Phil. Trans. R. Soc. B 359, 367–380. (doi:10.1098/
Ecological–Economic Zoning (ZEE ) project have
rstb.2003.1423)
been implemented in Brazil in attempts to approach
Cardoso, M. F., Hurtt, G. C., Moore, B., Nobre, C. A. &
the ideas of sustainable development and territorial
Prins, E. M. 2003 Projecting future ?re activity in
planning in Amazonia (Alves 2008). On a smaller scale,
Amazonia. Global Change Biol. 9, 656–669. (doi:10.
the project ‘Roc¸a sem Queimar’, meaning farming
1046/j.1365-2486.2003.00607.x)
without the use of ?res, led by a non-governmental
Cochrane, M. A. & Laurance, W. F. 2002 Fire as a large-scale
edge effect in Amazonian forests. J. Trop. Ecol. 18,
organization in association with local farmers, is a
311–325.
pioneering experience that has been used as an
Cochrane, M. A. & Schulze, M. D. 1999 Fire as a recurrent
alternative to the traditional slash and burn process in
event in tropical forests of the eastern Amazon: effects on
some Amazonian municipalities in the Xingu region
forest structure, biomass, and species composition.
(Silva et al. 2006).
Biotropica 31, 2–16.
Mechanized agriculture and crop plantations are not
Cochrane, M. A., Alencar, A., Schulze, M. D., Souza, C. M.,
normally accompanied by subsequent ?res (Eva &
Nepstad, D. C., Lefebvre, P. & Davidson, E. A. 1999
Lambin 2000), which can potentially reduce the risk of
Positive feedbacks in the ?re dynamic of closed canopy
forest ?res in Amazonia. Despite its negative ecological
tropical forests. Science 284, 1832–1835. (doi:10.1126/
implications, this would be an important factor to be
science.284.5421.1832)
considered once the ZEE of the Amazonian region has
Costa, M. H., Yanagi, S. N. M., Souza, P., Ribeiro, A. &
Rocha, E. J. P. 2007 Climate change in Amazonia caused
been well stabilized. However, it is important to bear in
by soybean cropland expansion, as compared to caused by
mind that burning practices are common for some
pastureland expansion. Geophys. Res. Lett. 34, L07706.
crops, such as sugar cane, which is largely used for
(doi:10.1029/2007GL029271)
biofuel production in Brazil.
Cox, P. M., Betts, R. A., Collins, M., Harris, P. P.,
Ultimately, this study showed that the time series
Huntingford, C. & Jones, C. D. 2004 Amazonian forest
used, including monthly deforestation data, provides a
dieback under climate-carbon cycle projections for the
high temporal-resolution description of the interactions
21st century. Theor. Appl. Climatol. 78, 137–156. (doi:10.
between land-use dynamics and climate. This infor-
1007/s00704-004-0049-4)
mation must be included in the current models for better
Denman, K. L. et al. 2007 Couplings between changes in the
understanding of the impact of climate change in the
climate system and biogeochemistry. In Climate change
Amazonian ecosystems. However, intensive measure-
2007: the physical science basis. Contribution of Working
ments of the carbon dynamics at ?re-affected forests and
Group I to the Fourth Assessment Report of the Intergovern-
mental Panel on Climate Change (eds S. Solomon, D. Qin,
the accurate mapping of the area and damage degree of
M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M.
burnt forests are still required for the total quanti?cation
Tignor & H. L. Miller), pp. 499–587. Cambridge, UK:
of carbon emissions from ?res in this region.
Cambridge University Press.
The data used in this study were acquired as part of the
Diggle, P. 1989 Time series: a biostatistical introduction. Oxford,
TRMM project jointly sponsored by Japan’s National Space
UK; New York, NY: Oxford University Press.
Development Agency ( NASDA) and the US National
Eva, H. & Lambin, E. F. 2000 Fires and land-cover change in
Aeronautics and Space Administration ( NASA) Of?ce of
the tropics:a remote sensing analysis at the landscape
Earth Sciences. We thank the INPE PRODES, DETER and
scale. J. Biogeogr. 27, 765–776. (doi:10.1046/j.1365-2699.
Queimadas programs for making their data and images freely
2000.00441.x)
available. This work was supported by a Natural Environ-
Good, P., Lowe, J., Collins, M. & Moufouma-Okia, W. 2008 An
ment Research Council Urgency grant ( NE/D01025X/1).
objective Tropical Atlantic SST gradient index for studies of
Phil. Trans. R. Soc. B (2008)

---

Rainfall, deforestation and ?re in Amazonia
L. E. O. C. Araga˜o et al.
1785
South Amazon dry season climate variability and change.
Goddard Space Flight Center Earth Science, Greenbelt,
Phil. Trans. R. Soc. B 363, 1761–1766. (doi:10.1098/
MD. See http://disc.gsfc.nasa.gov/data/datapool/TRMM_
rstb.2007.0024)
DP/01_Data_Products/02_Gridded/07_Monthly_Other_
Houghton, R. A., Skole, D. L., Nobre, C. A., Hackler, J. L.,
Data_Source_3B_43/.
Lawrence, K. T. & Chomentowski, W. H. 2000 Annual
Nepstad, D. C. et al. 1999 Large-scale impoverishment of
?uxes or carbon from deforestation and regrowth in the
Amazonian forests by logging and ?re. Nature 398,
Brazilian Amazon. Nature 403, 301–304. (doi:10.1038/
505–508. (doi:10.1038/19066)
35002062)
Nepstad, D., Lefebvre, P., Da Silva, U. L., Tomasella, J.,
Hutyra, L. R., Munger, J. W., Nobre, C. A., Saleska, S. R.,
Schlesinger, P., Solorzano, L., Moutinho, P., Ray, D. &
Vieira, S. A. & Wofsy, S. C. 2005 Climatic variability and
Benito, J. G. 2004 Amazon drought and its implications
vegetation vulnerability in Amazonia. Geophys. Res. Lett.
for forest ?ammability and tree growth: a basin-wide
32, 4. (doi:10.1029/2005GL024981)
analysis. Global Change Biol. 10, 704 –717. (doi:10.1111/
Instituto Nacional de Pesquisas Espaciais ( INPE ) 2005
j.1529-8817.2003.00772.x)
PRODES:
assessment
of
deforestation
in
Brazilian
Nepstad, D. et al. 2006 Inhibition of Amazon deforestation
Amazonia. See http://www.obt.inpe.br/prodes/index.html.
and ?re by parks and indigenous lands. Conserv. Biol. 20,
Instituto Nacional de Pesquisas Espaciais ( INPE) 2006a
65–73. (doi:10.1111/j.1523-1739.2006.00351.x)
DETER: detection of deforested areas in real time. See
Nobre, C. A., Sellers, P. J. & Shukla, J. 1991 Amazonian
http://www.obt.inpe.br/deter/dados/.
deforestation and regional climate change. J. Clim.
Instituto Nacional de Pesquisas Espaciais ( INPE) 2006b
4, 957–988. (doi:10.1175/1520-0442(1991)004!0957:
Queimadas project. See http://www.cptec.inpe.br/queima-
ADARCCO2.0.CO;2)
das/.
Phillips, O. L., Lewis, S. L., Baker, T. R., Chao, K. &
Laurance, W. F. & Williamson, G. B. 2001 Positive feedbacks
Higuchi, N. 2008 The changing Amazon forest. Phil.
among forest fragmentation, drought, and climate change
Trans. R. Soc. B 363, 1819–1827. (doi:10.1098/rstb.2007.
in the Amazon. Conserv. Biol. 15, 1529–1535. (doi:10.
0033)
1046/j.1523-1739.2001.01093.x)
Priestley, M. B. 1981 Spectral analysis and time series. London,
Laurance, W. F. et al. 2002 Ecosystem decay of Amazonian
UK; New York, NY: Academic Press.
forest fragments: a 22-year investigation. Conserv. Biol. 16,
Ronchail, J., Cochonneau, G., Molinier, M., Guyot, J. L.,
605–618. (doi:10.1046/j.1523-1739.2002.01025.x)
Chaves, A. G. D., Guimaraes, V. & de Oliveira, E. 2002
Li, W. H., Fu, R. & Dickinson, R. E. 2006 Rainfall and its
Interannual rainfall variability in the Amazon basin and
seasonality over the Amazon in the 21st century as
sea-surface temperatures in the equatorial Paci?c and the
assessed by the coupled models for the IPCC AR4.
tropical Atlantic Oceans. Int. J. Climatol. 22, 1663–1686.
J. Geophys. Res. Atmos. 111, 14.
(doi:10.1002/joc.815)
Li, W. H., Fu, R., Juarez, R. I. N. & Fernandes, K. 2008
Rosenfeld, D. 1999 TRMM observed ?rst direct evidence of
Causes of recent changes of rainfall variabilities and
smoke from forest ?res inhibiting rainfall. Geophys. Res.
implications to the future climate in the Amazon region.
Phil. Trans. R. Soc. B 363, 1767–1772. (doi:10.1098/rstb.
Lett. 26, 3105–3108. (doi:10.1029/1999GL006066)
2007.0022)
Silva, M. M., Lovato, P. E. & Vieira, I. C. 2006 Projeto Roc¸a
Malhi, Y. & Phillips, O. L. 2004 Tropical forests and global
sem Queimar: uma proposta de manejo agroecolo´gico
atmospheric change: a synthesis. Phil. Trans. R. Soc. B 359,
para a regia˜o da Transamazoˆnica—Para´. Revista Brasileira
549–555. (doi:10.1098/rstb.2003.1449)
de Agroecologia 1, 803–806.
Marengo, J. A. 1992 Interannual variabilty of surface climate
Silva Dias, M. A. F., Cohen, J. C. P. & Gandu, A. W. 2005
in the Amazon basin. Int. J. Climatol. 12, 853–863.
[Clouds, rain and biosphere interactions in Amazon]. Acta
(doi:10.1002/joc.3370120808)
Amazon. 35, 215–222. [In Portuguese.]
Marengo, J. A. 2004 Interdecadal variability and trends of
Timmermann, A. 1999 Detecting the nonstationary response
rainfall across the Amazon basin. Theoret. Appl. Climatol.
of ENSO to greenhouse warming. J. Atmos. Sci. 56,
78, 79–96.
2313–2325.
(doi:10.1175/1520-0469(1999)056!2313:
Marengo, J. A. et al. In press. The drought of Amazonia in
DTNROEO2.0.CO;2)
2005. J. Clim.
Uhl, C. & Kauffman, J. B. 1990 Deforestation, ?re
Morton, D. C., DeFries, R. S., Shimabukuro, Y. E.,
susceptibility, and potential tree responses to ?re in the
Anderson, L. O., Arai, E., Espirito-Santo, F. D., Freitas,
eastern Amazon. Ecology 71, 437–449. (doi:10.2307/
R. & Morisette, J. 2006 Cropland expansion changes
1940299)
deforestation dynamics in the southern Brazilian Amazon.
Uvo, C. B., Repelli, C. A., Zebiak, S. E. & Kushnir, Y. 1998
Proc. Natl Acad. Sci. USA 103, 14 637–14 641. (doi:10.
The relationships between tropical Paci?c and Atlantic
1073/pnas.0606377103)
SST and northeast Brazil monthly precipitation. J. Clim.
NASA 2006 Monthly 0.258!0.258 TRMM and other
11, 551–562. (doi:10.1175/1520-0442(1998)011!0551:
sources rainfall. NASA Distributed Active Archive Center,
TRBTPAO2.0.CO;2)
Phil. Trans. R. Soc. B (2008)

---

Document Outline

• Interactions between rainfall, deforestation and fires during recent years in the Brazilian Amazonia
  • Introduction
  • Material and methods
    • Rainfall, fire and deforestation datasets
    • Data analysis
  • Results and discussion
    • Seasonality of rainfall, fire and deforestation and their relationships
    • Influence of monthly and annual rainfall and deforestation on fire dynamics
    • Interactions between land-use and climate change and the vulnerability of Amazonia
  • Conclusions
  • The data used in this study were acquired as part of the TRMM project jointly sponsored by Japans National Space Development Agency (NASDA) and the US National Aeronautics and Space Administration (NASA) Office of Earth Sciences. We thank the INPE PROD...
  • References

---