Agricultural and Forest Meteorology 214215 (2015) 494505

Contents lists available at ScienceDirect

Agricultural and Forest Meteorology

journal homepage: www.elsevier.com/locate/agrformet

Monitoring vegetative drought dynamics in the Brazilian semiarid

region
A.P.M. Cunha , R.C. Alval, C.A. Nobre, M.A. Carvalho
Brazilian Centre for Monitoring and Warning of Natural Disasters, So Jos dos Campos, SP, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Drought is a complex natural phenomenon that can lead to reduced water supplies and can consequently
Received 13 August 2015 have substantial effects on agriculture and socioeconomic activities that cause social crises and political
Received in revised form 9 September 2015 problems. Different drought indicators are used for identifying droughts. This work explored the appli-
Accepted 18 September 2015
cability of a near-real time drought monitoring methodology using Terra-MODIS Normalized Difference
Vegetation Index (NDVI) and land surface temperature (LST) products. This approach is called the Vege-
Keywords:
tation Supply Water Index (VSWI), which integrates land surface reectance and thermal properties. The
Drought indices
results indicate that during a major drought event from 2012 to 2013, approximately 85% of the Brazilian
Drought monitoring
Vegetative drought
semiarid region was affected. The number of days of soil moisture decit, which was derived from a
Vegetation index simple water balance model and the daily interpolated precipitation, were used to verify the results. A
Semiarid correlation analysis of VSWI, precipitation and soil moisture decit shows that VSWI is closely related
to rainfall and soil water content, especially under dry conditions, and indicates that the use of VSWI
can be a suitable near-real time drought monitoring approach. The evaluation of the 20122014 drought
considering the VSWI index highlighted two major characteristics of vegetation response to drought
conditions, i.e., the recovery and memory effects of vegetation.
2015 Elsevier B.V. All rights reserved.

1. Introduction these studies have shown a newly directional change pattern of

vegetation communities in response to climate change and the
Drought is basically dened as an extended period of months consequent dryness (Zhou et al., 2011, 2013a,b, 2014).
or years, in which precipitation is less than the annual aver- Drought is considered to be among those natural disasters
age, resulting in water scarcity. Generally droughts are classied that can cause the most serious global economic and social losses
as either a meteorological drought (lack of precipitation over a (Carolwicz, 1996) and affects more people than any other natu-
region for a period of time), hydrological drought (deciencies ral disasters (Keyantash and Dracup, 2002). The effects of drought
in surface and subsurface water supplies), agricultural drought often accumulate slowly over a considerable period of time and
(deciency in water availability for crop or plant growth) or socio- may linger for years after the termination of the event, and both
economic drought (failure of water resources systems to meet the onset and end of drought are difcult to determine (Tannehill,
water demands, which impacts human activities both directly and 1947).
indirectly) (Wilhite, 2000; Yang, 2010; Son et al., 2012; Udmale In recent decades, droughts have increased in frequency and
et al., 2014). Although precipitation deciencies are important, intensity over much of the planet and can be related to climate
agricultural drought severity is usually more closely associated change (Marengo et al., 2009; Zhou et al., 2011). The percentage
with deciencies in soil moisture. The areas affected by drought of area that is affected by drought has doubled from the 1970s to
evolve gradually as the symptoms of moisture stress in plants often the early 2000s (Nagarajan, 2009). In Brazil, this phenomenon has
develop slowly. These impacts of drought on vegetation are here occurred mainly in the semiarid area of Northeast Brazil (SANEB)
referred as vegetative drought (Rulinda et al., 2012). Several studies due to uneven precipitation in space and time. Seasonal droughts
have documented that uneven temporal distributions of precipi- usually occur in the winter and spring and have a signicant impact
tation and rising temperatures have caused vegetation shifts, and on agricultural harvests. The severe droughts that are caused by
climactic variations harm the growth on plantations and cause seri-
ous social problems because a large number of people who inhabit
Corresponding author. the region truly live in a situation of extreme poverty (Marengo
E-mail address: [email protected] (A.P.M. Cunha). et al., 2009). More than 80% of the agricultural establishments in

http://dx.doi.org/10.1016/j.agrformet.2015.09.010
0168-1923/ 2015 Elsevier B.V. All rights reserved.
 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505 495

Northeast Brazil consist of smallholder farmers that practice sub- global scales. Researchers have concluded that the combination of
sistence agriculture (IBGE, 2009). Thousands of subsistence farmers vegetation and temperature conditions is a good indicator of soil
have seen their livelihoods wither away during drought episodes. moisture content.
Other consequences of large droughts are starvation, malnutrition, A large number of vegetation health and drought indices are
misery and migration to urban centers (rural exodus). Thus, the based on the LSTNDVI space such as Vegetation Health Index
livelihood and food security of millions of smallholder farmers are (VHI, Kogan, 1990, 1997), Temperature-Vegetation Drought Index
exposed to profound risk from drought. (TVDI, Sandholt et al., 2002), Vegetation Supply Water Index (VSWI,
In Brazil, from 1980 to 2010 approximately 15% of natural disas- Carlson et al., 1990, 1994), and Drought Severity Index (DSI, Mu
ters were related to drought conditions in the Northeast Region et al., 2013). These drought indices are often applied for arid or
(NEB). Drought affects more people than any other natural disas- semi-arid regions, and their use in humidsemi humid regions
ter; since 1980 more than 50 million people have been affected by are limited (Rhee et al., 2010). The NDVILST relationship, which
drought (Sapir and Below, 2014). The recent drought of 2012/2013 characterizes moisture and thermal conditions and the entirety
reached approximately 1300 municipalities and affected approxi- of vegetation health, has been used successfully for early drought
mately 10 million people in Brazil. Drought is, on average, Brazils detection and the estimation of crop and pasture production losses
most costly natural hazard, primarily because it causes hard for winter wheat in the USA (Kogan, 1997). The potential for stress
impacts on agriculture and livestock production. For example, the exists when the water stored in soil is insufcient to sustain the
2012/2013 drought resulted in economic losses of US $1.6 billion current growth. In the vegetation covering areas, LST can be con-
for the 10 most important crops (beans, rice, corn, cotton, bananas, sidered equal to the temperature of vegetation canopy (Liu et al.,
sugar cane, cassava, soybeans and coffee), US $1.5 billion due to 2013). The canopy temperature response can occur even when the
cattle mortality and costs of greater than US $1.5 billion in insur- plants are green because stomata closure to minimize water loss by
ance claims, according to the Brazilian Institute of Geography and transpiration results in a decreased latent heat ux (Berliner et al.,
Statistics (IBGE). 1984; Carlson et al., 1994; Yang and Merchant, 1997).
The quantication of drought is usually determined by remotely Because of the territorial expansion of the Northeast region of
sensed spectral indices and water balance simulations. Drought Brazil, it is necessary to develop methods for large-scale vegetative
indices are particularly useful for monitoring the impact of cli- drought assessment. To direct the emergency actions of the govern-
mate variability on vegetation because the spatial and temporal ment that are taken to mitigate the effects of drought, it is crucial
identication of drought episodes is extremely complex. A number to determine an appropriate and user friendly index that reects
of drought indices, including meteorological (Wilhite and Glantz, the direct impact of drought on livestock and subsistence agricul-
1985), remote sensed, hydrological and other indicators, have been ture. Thus, the main objective of this study is to evaluate possible
used to measure drought impacts (Palmer, 1965, 1968; Gibbs and indicators to monitor the impacts of drought on vegetation over
Maher, 1967; Shafer and Dezman, 1982; Kogan, 1990, 2002; McKee the NEB. In this paper, the Vegetation Supply Water Index (VSWI)
et al., 1993; Keyantash and Dracup, 2004; Bhuiyan et al., 2006; Yagci performance is compared with precipitation data and soil water
et al., 2011; Zhou et al., 2012; Du et al., 2013; Yang et al., 2013; decit. This study focuses on the diagnosis of a remote sensing
Abbas et al., 2014; Nichol and Abbas, 2015). Traditional methods indicator that is responsive to short-term environmental changes
of drought assessment and monitoring rely on rainfall data (e.g., because early warning capabilities are limited in current drought
the Palmer drought severity index (PDSI) and Standardized Pre- monitoring systems.
cipitation Index (SPI)). However, in a region where the density of
meteorological stations as well as the temporal scale of the data are
insufcient, it is impossible to monitor drought using indices that 2. Materials and methods
are based on rainfall data. In contrast, satellite-sensor data are con-
sistently available and can be used to detect the onset of drought, The study area (Fig. 1a) is located in the equatorial zone (121 S,
its duration and magnitude across large areas (Thiruvengadachari 3249 W) and covers an area of 1,800,555 km2 , which represents
and Gopalkrishna, 1993). Remote sensing has proven to be a pow- approximately 20% of Brazils territory. The limits of the study area
erful tool for evaluating the temporal and spatial aspects of drought were dened by the Superintendence for the Development of the
conditions (Johnson et al., 1993; Peters et al., 2002). Northeast (SUDENE).
Moderate Resolution Imaging Spectroradiometer (MODIS) data The semiarid area of Northeast Brazil covers an area of
play an increasingly important role in drought monitoring and 980,323 km2 and consists of 1133 municipalities and a popula-
assessment (Wan et al., 2004) because of their associated rich tion of approximately 22 million people (approximately 12% of the
spectral information, high temporal repeat cycle and convenient national population). These numbers make the Brazilian semiarid
means of data access. The Normalized Difference Vegetation Index region the most populated semiarid region in the world. In the
(NDVI), which provides a general measure of the state and health SANEB, rural areas in the interior are generally used for subsistence
of vegetation, was one of the rst remote sensing-based indicators agriculture that is primarily comprised of beans, manioc, potatoes
that was used for drought detection and monitoring. Many stud- and other crops (Cavalcanti et al., 1999). Most of the study area is
ies have reported relationships between vegetation indices, rainfall covered by mixed grasslandscroplands (Fig. 1b). Other land cover
and soil moisture (Davenport and Nicholson, 1993; Herrmann et al., types are caatinga (closed and open shrublands) and savanna (not
2005; Liu et al., 2013; Ibrahim et al., 2015). This is an important shown). In 2010, the total area of pasture and agricultural activi-
reason why NDVI is widely used in agricultural drought monitor- ties was 1,024,621 km2 , which represents 57% of the NEB territory
ing (Henricksen and Durkin, 1986; Tucker and Choudhury, 1987; (Vieira et al., 2013).
Tucker, 1989; Gutman, 1990). On the other hand, vegetation cover Northeast Brazil is characterized by a variable and irregular
condition, as sampled by vegetation index, is a relatively slow spatio-temporal distribution of precipitation. The rainfall ranges
response variable that typically adjusts only after notable crop from less than 800 mm/year in the semiarid interior to more than
damage has already occurred. In contrast, land surface tempera- 1500 mm/year in the rainy climatic zone that is mainly on the east
ture (LST) derived from thermal infrared (TIR) information can be coast (Fig. 2a). Different rainfall regimes have been identied in
considered to be a rapid response variable. LST is a good indicator of the NEB (Fig. 2b). (i) In the South-Southwest sector, the main rainy
the energy balance at the Earths surface because it is one of the key season is from November to February, and the rainfall is associated
parameters in the physics of land-surface processes on regional and with cold fronts traveling from the South Region of Brazil. (ii) In the
496 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505

Fig. 1. (a) Location of Northeast Brazil and (b) distribution of mixed grasslandscroplands areas in the study region.

Fig. 2. (a) Mean annual rainfall 19702012 and (b) wet seasons regions (FMAM: February to May; JFMA: January to April; DJFM: December to March; NDJF: November to
February; AMJJ: April to July).
 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505 497

composited land surface temperature (LST) products (MOD11A2,

MYD11A2, 1 km 1 km) from 2002 to 2014 in the study area were
acquired from the U.S. National Aeronautics and Space Administra-
tion (NASA) Land Processes Distributed Active Archive Center (LP
DAAC).
MODIS, which is carried on the Terra-Aqua satellite, is a
cost-effective sensor that covers the globe at least once a day.
MODIS, which is a successor of the Advanced Very High Reso-
lution Radiometer (AVHRR), is the primary sensor in the NASA
Earth Observing System (EOS) program for monitoring the ter-
restrial ecosystem (Justice et al., 2002) and has several advances
over AVHRR (Thenkabail et al., 2004). MODIS is more sensitive
to changes in vegetation dynamics (Huete et al., 2002) and has
been found to be a more accurate and versatile instrument for
monitoring global vegetation conditions than AVHRR (Gitelson and
Kaufman, 1998; Justice et al., 2002). Composite MODIS data have a
temporal resolution of 8 days and are available from 2000 onwards.
The 8-day, 7-band data are made available by the USGS EROS
DAAC after corrections for molecular scattering, ozone absorption
and aerosols. The data are also adjusted to nadir (sensor look-
ing straight down) and standard sun angles using bidirectional
reectance (BRDF) models (Vermote et al., 2002; Justice et al.,
2002). The products have been processed for atmospheric and geo-
metric corrections.

2.1.2.1. Acquisition and processing MODIS data. MODIS images are

available on a daily basis, but their use involves considerable extra
processing. Time series of MODIS imagery provide near real-time,
continuous and relatively high-resolution data. These images can
Fig. 3. Spatial distribution of location of the weather stations. used to assess the development of drought and its severity in
regions with scarce and inaccurate on-the-ground meteorological
observations.
Northern sector (semiarid zone), the rainfall occurs from February
The daytime MOD11A2 and MYD11A2 products consist of 12
to May, January to April and December to May, and the rainfall
Science Data Sets (SDSs), including 8-day composite LST and the
is associated with the southward movement of the Intertropical
quality of each LST pixel. The LST values in Kelvins are encoded in a
Convergence Zone (ITCZ). (iii) In the East sector, the main rainy
16-bit unsigned integer that ranges from 7500 to 65,535. To derive
season occurs from April to August and is mainly caused by greater
the actual temperature values, a multiplication factor of 0.02, as
temperature differences between the ocean and nearby land.
stated in MODIS product manual, was used. The LST data from two
consecutive periods were averaged to generate composite LST for
2.1. Data the same periods as the NDVI to match the 16-day NDVI compos-
ite product (MOD13A2 and MYD13A2). NDVI values that were less
2.1.1. Precipitation than or equal to zero were excluded from further analysis because
The data that were used in this study include in situ meteorolog- pixel values less than or equal to zero are assumed to represent
ical data and remotely sensed data. The meteorological data were either cloud contaminated imagery or the presence of a water body.
acquired from different sources, including weather stations (Data Moreover, NDVI and LST images were ltered using a mask of the
Collection Platform) from the National Institute for Space Research water bodies that are in the NEB region. Finally, digital maps of
(INPE), National Institute of Meteorology (INMET), National Center NDVI and LST products were generated for the entire research area
for Monitoring and Early Warning of Natural Disasters (CEMADEN) over the course of 10 years (20022014).
and the State Centers of Meteorology. Fig. 3 shows the spatial distri-
bution of the 1947 weather stations for which we had data. Rigorous
quality inspection was carried out for the daily precipitation data. 2.2. Drought-monitoring products
The interpolation of the data from the weather stations into a reg-
ular grid with 5 km resolution was performed using a technique 2.2.1. Soil moisture decit
called kriging that was developed by (Matheron, 1969) and has Soil moisture condition is an important indicator for evaluating
been primarily applied in hydrology and other earth science dis- drought, reects recent precipitation and antecedent conditions
ciplines for the spatial interpolation of various physical quantities and indicates agricultural potential and available water storage
given a number of spatially distributed measurements. The spa- (Boken, 2005). Soil moisture conditions are very important in agri-
tially interpolated data were also used as an input to a water balance culture because they are used directly to assess the irrigation needs
model that was developed by PROCLIMA/INPE (Souza et al., 2001) for a variety of crops.
and that is used to estimate soil moisture and the number of days In this study, the number of days of soil moisture decit (NDD),
of soil moisture decit with a spatial resolution of 5 km. which is derived from a simple water balance model, is one of the
indicators that was used to characterize the drought. Days of soil
2.1.2. NDVI and LST moisture decit measures the number of days that a plants growth
The 16-day MODIS NDVI time series product (MOD13A2, is restricted by insufcient moisture in the soil. The NDD in each
MYD13A2, 1 km 1 km) and an 8-day average value of the pixel is calculated from the soil moisture data.
498 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505

Soil moisture is calculated daily from the water balance in the calculated by the ratio of the average 8-day LST product versus the
soil (Souza et al., 2001; Rossato et al., 2005): smoothed high-quality NDVI value in the MODIS composted data:
Qs(t) = prec(t) evap(t) DP(t) + qs0 (1) 0.02(LSTijk )
VSWIijk = (2)
0.001(NDVIijk )
The model water balance components are soil water content
at time t (Qs(t)), precipitation (prec), actual evapotranspiration For pixel i in period j for year k. Land surface temperature can
(evap), drain depth (DP) and soil water content for the previous day be considered similar to canopy surface temperature when ground
(qs0 ). This soil water balance model developed for Brazil (Rossato covered with crop (Chen et al., 2014). High VSWI values indicate a
et al., 2005), uses a new approach that takes into account the spatial high canopy temperature and low vegetation index and therefore
variability of soil characteristics. The maximum soil water storage may indicate stressed vegetation condition.
was derived from the eld capacity and the wilting point, using Small VSWI values indicate a low canopy temperature and high
a pedotransfer function (PTF). PTF allows the estimation of soil vegetation index, which represent unstressed vegetation. While
hydraulic properties from basic soil data (such as texture, particle- VSWI is characterized by varying moisture and the thermal con-
size distribution, organic C content and bulk density). Tomasella ditions of vegetation, it represents overall vegetation health.
developed specic PTFs for the prediction of soil water retention In the present study, the VSWI that is used for drought monitor-
curves for the tropical soils of Brazil (Tomasella and Hodnett, 1998; ing was modied. According to the literature, drought is apparent
Tomasella et al., 2000, 2003, 2008). when the value of the VSWI threshold exceeds 60, and the severity
Once we have the measured data, including the temperature, of drought will decline when the values are less than 60. However,
wind and radiation, the FAO-56 PenmanMonteith method (Allen this threshold can vary as a function of different factors, includ-
et al., 1998) was applied to determine reference evapotranspira- ing the vegetation cover types, local weather conditions and soil
tion. The vegetation parameters needed for the PenmanMonteith type. Thus, to avoid the use of a threshold that may contain errors,
method were those obtained from the SSiB model (Simplied Sim- in this study we considered the use of VSWI anomaly percentage
ple Biosphere, Xue et al., 1991), as cited by Dorman and Sellers (VSWIanom% ). VSWIanom% was calculated in each pixel to assess the
(1989). Calibrated vegetation parameters for semiarid Brazilian changes of the index with regard to mean conditions (VSWIanom% ).
(stress conditions) which were used, are described in Cunha et al. The VSWIanom% is used to compare a period of the index (VSWIijk )
(2009) and Cunha et al. (2013) (plant height, root depth, vegeta- with the average for several years (13 years), which reects the
tion roughness length, surface resistance, albedo, etc.). Thus, the deviation of the VSWI from the normal range:
reference evapotranspiration was calculated according to the land
cover present in the study area. For calculating the water balance, VSWIijk VSWIijk
VSWIanom% = 100 (3)
the reference evapotranspiration was transformed into actual VSWIijk
evapotranspiration, as suggested by the FAO method (Doorenbos
and Pruitt, 1977), i.e., considering that the amount of water that A positive anomaly percentage indicates different levels of vege-
plants transpire depends on the water storage in soil. tation stress, and a negative anomaly percentage means a favorable
The soil water content was calculated by combining data from condition for vegetation.
weather stations with a soil database in a geo-referenced environ- The second approach relates to the application of a vegetation
ment. The soil water decit is considered to be low when the soil lter. Thus, to assess drought impacts using the VSWI index over
moisture falls below a critical value. The critical value is dened pasture and agricultural lands (Fig. 1b), the data were extracted
as the soil moisture at the point where soils are extremely drier based on land use information from the Land Cover Land Use map
than normal and drought conditions are likely to be evident. It of the study area. This map was derived from a Landsat 7 equipped
was assumed that the critical soil matric potential threshold was with Enhanced Thematic Mapper Plus (15 m) and Landsat 5 The-
60 kPa (Feddes et al., 1988). Therefore, NDD is the sum of days matic Mapper (30 and 60 m) mosaic (Vieira et al., 2013). A total of
during which the water loss is higher than the soil water retention. 162 satellite images were used to prepare the vegetation cover map.
Additional details about these methodologies can be found in Souza The map presents the location and distribution of major vegetation
et al. (2001). types and non-vegetated land surface formations for the Northeast
Brazil Region, which includes the semiarid region.
2.2.2. Vegetation Supply Water Index (VSWI) calculation
Growing crops need continuous supplies of soil water to ensure 2.3. Comparative analysis of the spatial and temporal
harvest. Rainfall and irrigation are the main sources of soil water in characteristics drought indicators
agricultural elds. When the soil water supply is sufcient for grow-
ing crops, evapotranspiration from agricultural elds is high, which Four representative hydrological years were selected as a
leads to the observation of low surface temperatures in satellite basis to compare the MODIS image-derived and meteorological-
remote sensing images. During a drought period, the soil water sup- measured drought indices: 20102011 as a wet year and
ply is insufcient to meet the normal demands for growing crops. 20112012, 20122013 and 20132014 as dry years at different
Consequently, the stoma on crop leaves tend to close to decrease intensities.
water loss from the canopy, which leads to an apparent increase Scatter plots, Pearson correlation analyses and a correlation
of temperature in the elds. Therefore, by using the relationship matrix for the VSWI and meteorological data were developed. Aver-
between the canopy temperature change and the soil water sup- age values of VSWI over the 5 km 5 km square were used for the
ply in the elds, it is possible to develop an approach for drought statistics calculation to match the spatial resolution of the interpo-
monitoring (Gao et al., 2008; Wu et al., 2015). lated meteorological data.
VSWI has been widely applied as a drought indicator based on
the philosophy of this approach. The VSWI approach, which is an 3. Results
additive combination of NDVI and the thermal data (land surface
temperature LST), was established to detect vegetation stress, 3.1. NDVILST relationships as an indicator of drought
moisture and drought-affected areas (Zhou et al., 2013a,b). VSWI
can be used to monitor the onset, change, development and inten- NDVI for vegetative cover generally range from 0.1 to 0.8.
sity of drought and the extent of the impact on vegetation. VSWI is The values depend on the canopy architecture, density and
 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505 499

Fig. 4. Long-term annual averages of NDVI for the three years: (a) 2011, (b) 2012, and (c) 2013. (d)(f) Long-term averages of LST and (g)(i) density scatter plots of LST
versus NDVI for the same sub-periods. Yellow and red colors represent higher concentration of pixels. (For interpretation of the references to color in this gure legend, the
reader is referred to the web version of this article.)
500 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505

vegetation moisture. High values (between 0.4 and 0.8) are asso-
ciated with greener and denser vegetative cover, whereas small
values (between 0.2 and 0.3) represent low vegetation cover. Actu-
ally, NDVI values of semiarid areas, as well as those of the Brazilian
semiarid area, present low values (Fig. 4) and have been demon-
strated to be strongly dependent on plant water availability in wet
months. The semiarid region is characterized by the presence of
the natural vegetation that is known as caatinga. Caatinga vegeta-
tion is composed of shrubs and small trees that are usually thorny
and deciduous and lose their leaves in the early dry season (lower
NDVI). During the dry season, the NDVI values in caatinga vege-
tation are low, while in the rainy months the values are higher
due to the response of the vegetation. Previous studies (Santos and
Shimabukuro, 1993; Gao et al., 2003; Ferreira et al., 2005) have
shown seasonal variability of vegetation in semiarid or caatinga
vegetation where NDVI values have shown large annual ampli-
tudes.
On the coast outside of the semiarid boundary, high values of
NDVI are found in regions that are covered by broadleaf evergreen
trees and deciduous forest (not shown).
Fig. 4 shows the spatial distribution of the annual average NDVI
and LST. Among the three analyzed years, 2011 presented the high-
est NDVI values and lowest LST values (Fig. 4a, b) and correspond
to a wet year. During 2012, the inverse relation was observed for
most of the LST/NDVI space, i.e., when the surface temperature
increased, the NDVI value decreased (Fig. 4c, d), which represents a
higher level of vegetation stress (less water left on the soil for plant
transpiration). During dry conditions, rising leaf temperatures are
good indicators of plant moisture stress and precede the onset of
drought. The same behavior is observed for 2013 (Fig. 4e, f). Fig. 5. Long-term average of VSWI within hydrological year (October to September).
These results are corroborated by the study performed by
Karnieli et al. (2006, 2010), who demonstrated that it is the most
common behavior in drought environments (Karnieli et al., 2006; (Fig. 5). In the long-term average of VSWI, the highest values are
Julien and Sobrino, 2009). Karnieli et al. (2010) showed that during concentrated in the semiarid region, which can be associated with
the mid-growing season in the US Great Plains and Southwest, 80% the distribution of the annual precipitation in this region (Fig. 2a).
of the cropland areas and 68% of the pasture lands are characterized Climatologically, this region is characterized by an average annual
by strong negative correlations between LST and NDVI. rainfall that does not exceed 800 mm. Furthermore, this semiarid
The averages of NDVI and LST for each year (20112013) were region is also predominantly covered with grassland and partly
used to create a density scatter plot. Fig. 4gi is scatter plots of sparse vegetation. These types of vegetation usually show low NDVI
NDVI and LST data for the entire study area, croplands and grass- values throughout the year. Throughout the long-term average, the
lands, respectively. In the scatter plots, the yellow and red colors VSWI based on the vegetation stress indicates high stress values
represent higher concentration pixels. Fig. 4 shows a high negative in regions with sparse vegetation and bare soil, which is found in
correlation between NDVI and LST. High NDVI values are reached at some parts of the study area (not shown). The bare soil region and
low LST values, which can be easily explained by the fact that their some dune areas in the study area never exhibit wet conditions.
vegetation is water limited and not temperature limited (Goward The sandy soil that is exposed to the direct solar radiation cannot
et al., 2002; Karnieli et al., 2010). Because of the inclination angle retain water, and therefore no photosynthesis activity is observed.
of the Earth, the radiation level throughout the entire study area is However, it is noteworthy that the VSWI method is applicable for
high during whole year, and radiation is most likely not a limiting monitoring drought effect in vegetation covering regions. When
factor for vegetation growth. used to the regions with bare land, the VSWI method is less ef-
In the scatter plot for 2011 (Fig. 4g), it can be observed that cient, since the higher soil background temperature would severely
the higher concentration pixels are between 0.45 and 0.7 (NDVI) interfere in the assessment (Son et al., 2012).
and 299.65 K and 310.95 K. The concentration is larger for a smaller Fig. 6 presents the spatial and temporal pattern of VSWI for
range of NDVI values and higher temperatures for 2013 and espe- four hydrological years over the study area. As can be observed on
cially for 2012 (0.250.5 for the NDVI and 306.15 K and 315.85 K the VSWI anomalies map for 20102011, the majority of the study
for temperature). From the scatter space, drier areas were charac- area was almost drought-free. The lowest VSWI values were found
terized by low NDVI and high temperature (Kogan, 2000; Karnieli against a long-term average (negative VSWI anomaly percentage),
et al., 2010). primarily in the northern sector of the region (Fig. 6e).
The VSWI anomalies in the semiarid region reached approx-
3.2. Drought assessment using VSWI imately 8%, which indicates a wet condition. This condition
agrees with the mean annual rainfall during the hydrological year
To calculate the VSWI, the daily MODIS LST was averaged to a 20102011, with an accumulation of approximately 900 mm in the
8-day composite using the same day of year for the composite of region. In addition, 21 days of soil moisture decit during the rainy
MODIS NDVI (Terra and Aqua platforms). The time series data of season were found (Fig. 6i, n).
the VSWI were used to analyze the drought occurrence trends. Overall, the spatial distribution of NDD follows the rainfall pat-
A high value of VSWI means that the canopy temperature is high tern, and regions with higher NDD are directly associated with
or that the vegetation density is low, leading to a severe drought those with the lowest cumulative precipitation.
 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505 501

Fig. 6. Long-term annual averages of VSWI, normalized VSWI anomalies, number of days of soil moisture decit and mean annual rainfall for the four hydrological years:
20102011, 20112012, 20122013 and 20132014.

Signs of droughts impacts (positive anomalies) appeared locally high VSWI values in most part of the study area (Fig. 6b). Positive
in some areas in the western region and in the center of the study VSWI anomalies reached 9% in the semiarid region (Fig. 6f). This
area. Some of these positive anomaly percentage points (Fig. 6e) result is also consistent with the observed cumulative precipita-
that were located on the western part of the study area correspond tion and NDD elds. Approximately 54 days of soil moisture decit
to agricultural areas. It can be noted that the general situation of were observed, with a cumulative precipitation of approximately
ground surface during the hydrological year 20102011 was very 500 mm in the region (Fig. 6j, o).
suitable for agricultural cropping and livestock production in most Signs of drought intensied during the hydrological year
parts of the semiarid region. 20122013, when the vegetation experienced stress and a loss
Fig. 6 indicates that the study area experienced drought con- of health. Note in Fig. 6g that most parts of the study area were
ditions during the hydrological year 20112012. According to the subjected to a spatially extensive drought (approximately 65%
VSWI index, approximately 50% of the Northeast Region suffered of the NEB and 85% of the SANEB, Fig. 7), primarily in the cen-
vegetative drought conditions, especially in the semiarid region. tral and northern part of the semiarid region. The positive VSWI
Based on the annual average, 20112012 was characterized by anomalies reached 15.4%. Approximately 60% of the study area
502 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505

Fig. 7. Percentage of grassland and/or croplands areas affected by droughts according to VSWI index in NEB and SANEB Regions (hydrological years 20042014).

consists of extensive livestock farming and agriculture, therefore inclusion could interfere with the meteorological and hydrologi-
the occurrence of drought results in reductions in livestock pro- cal conditions and maintain normal vegetation health in spite of
duction and/or crop yield. poor rainfall or water-stress, which can cause a large response
Finally, the impact of the drought became less intense during delay.
the hydrological year 20132014, especially in the northern part of Overall, the correlations between VSWI, precipitation and the
the semiarid region; it was reduced to 40% of the NEB. The VSWI NDD time series had different characteristics in the ve sub-regions
anomalies reached 7.7% on average (Fig. 6h), which followed the and rainfall regime (Table 1). Note that in the NovemberFebruary
cumulative precipitation of 700 mm and the NDD of 30 in the rainy period, the VSWI variance, which is explained by the precipitation
season (Fig. 6m, q). and the number of days of soil moisture decit, is lower than in the
other four sub-periods. The highest correlation was found between
3.3. Validation of drought-monitoring products the precipitation and NDD for the rainy seasons of all three years.
This pattern was expected because the precipitation data were used
To identify which indicators (precipitation and NDD) were most to calculate NDD (direct dependence). In this case, it is necessary to
strongly related to the spatial distributions of VSWI in the differ- consider that errors in the precipitation data can propagate into the
ent rainy seasons, a Pearson correlation analysis was computed NDD or even errors that are associated with the estimation of the
for each pixel in the study area, stratied by sub-period and with evapotranspiration because the interpolated data were also used
VSWI as the dependent variable. The results of the analysis are to estimate it.
summarized in Table 1. The analysis was performed only for With regard to the VSWI index, there is a signicant linear cor-
lands that were covered by pastures and/or crops, and areas of relation with the precipitation during the JFMA wet season of 2013
natural vegetation, urban areas and water bodies were excluded (0.78), and lowest correlation was observed for the FMAM wet sea-
from the analysis. Those areas were excluded because their son of 2011 (0.12).

Table 1
Average temporal Pearson correlation coefcient (r2 ), computed between VSWI, precipitation (Prec) and soil moisture decit (NDD) derived from simple water balance model
over grasslandscroplands areas.

2011 2012 2013

VSWI Prec NDD VSWI Prec NDD VSWI Prec NDD

DJFM
VSWI 1 1 1
Prec 0.37 1 0.60 1 0.53 1
NDD 0.31 0.82 1 0.53 0.86 1 0.53 0.89 1

NDJF
VSWI 1 1 1
Prec 0.40 1 0.30 1 0.16 1
NDD 0.37 0.82 1 0.43 0.54 1 0.25 0.48 1

JFMA
VSWI 1 1 1
Prec 0.35 1 0.75 1 0.78 1
NDD 0.40 0.84 1 0.68 0.89 1 0.77 0.89 1

FMAM
VSWI 1 1 1
Prec 0.16 1 0.60 1 0.72 1
NDD 0.12 0.69 1 0.59 0.88 1 0.69 0.89 1

AMJJ
VSWI 1 1 1
Prec 0.25 1 0.59 1 0.23 1
NDD 0.24 0.75 1 0.63 0.83 1 0.13 0.48 1

DJFM, December to March; NDJF, November to February; JFMA, January to April; FMAM, February to May; AMJJ, April to July; NDD, number of days of soil moisture decit;
Prec, accumulated precipitation; VSWI, Vegetation Supply Water Index.
 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505 503

4. Discussions availability), and a very weak association between the rainfall and
vegetation condition occurs under those rainy conditions. Because
Drought is an ecological phenomenon that manifests itself hydrometeorological data such as precipitation and soil moisture
in the reduction of agricultural production and causes social content that are collected by surface observation stations often
crises and political problems. Different drought indicators are possess poor spatial resolution, especially in remote regions with
used to identify droughts. However, the impacts of the choice difcult access. Thus, it is important to note that the rainfall data
of a certain indicator for drought characterization (e.g., severity, were interpolated at a spatial resolution of 5 km, and it is there-
frequency and duration of droughts) are not fully understood. fore necessary to consider that there may be inconsistencies in the
Generally, the drought measuring indicators are not well corre- precipitation data if the weather stations have a low spatial density.
lated to one another. Therefore, it is quite common that when
one drought index identies drought at a particular place, another 5. Conclusions
drought index indicates a normal condition at the same place and
time. Considering the geographical complexity and irregularity of the
Overall, the results demonstrate that the empirical LSTNDVI climate of Northeast Brazil, the objective of this study was to exam-
relationship (VSWI) can be effectively exploited as an indicator of ine the spatial and temporal characteristics of vegetative drought
spatial temporal characteristics of water stress conditions in the in the Northeast Region using the remote sensing-based VSWI. The
whole study area. Comparing the two years of intense drought results have demonstrated that:
(20112012 and 20122013), it can be noted that the positive VSWI
anomalies were more intense in 20122013 than in 20112012 The empirical LSTNDVI relationship can be effectively exploited
(Fig. 6f, g). On the other hand, based on the precipitation and as an indicator of spatial temporal characteristics of water stress
NDD indicators, the hydrological year 20112012 presented more conditions in the Brazilian semiarid region. VSWI index efciently
extreme dry condition characteristics than the hydrological year identied vegetated areas (crop/pasture) that were affected by
20122013. For the hydrological year 20122013, the cumulative the 2012 and 2013 drought. It is highlighted that the use of VSWI
precipitation was approximately 600 mm, whereas the NDD was anomaly percentage is recommended once the same index val-
42 in the rainy season (Fig. 6l, p). ues might not denote the same drought severity, and different
One explanation for these inconsistencies might be the mem- values might have the same drought severity. VSWI is time-
ory effect of vegetation. In arid and semiarid lands, rainfall is dependent and usually region-specic, which reects different
seasonal, which affects vegetation directly in the same season and physiological adaptations and partially different ecological fac-
indirectly in the immediate next season through the memory tors. Furthermore, VSWI is strongly recommended for use during
effect (Martiny et al., 2005; Philippon et al., 2005) and recov- plant growing seasons (wet seasons).
ery effect (the difculty of vegetation to recover from previous The results also highlight interannual persistence effects in veg-
drought conditions, Martiny et al., 2005). The memory effect is etation dynamics, as shown by the complex mechanisms of the
dened as the capacity of semiarid ecosystems to benet from recovery and memory effects.
water surpluses, which indicates that there clearly exists a memory The study showed that the semiarid region is more affected
of the past precipitation at a time lag of one-year (Schwinning and by drought stress conditions; however, during the 20122013
Sala, 2004; Martiny et al., 2005). droughts, areas outside of the semiarid boundary also presented
Drought can reduce resiliency, rendering plants to be more vul- drought impacts. This nding indicates that the behavior of
nerable to a recurring disturbance. Therefore the mean VSWI level drought is very erratic. This has large implications for planners
of one year preceded by drought may also be higher, such as that and policy makers who are actively engaged in drought mitiga-
seen in the hydrological year 20122013 that was preceded by a tion and preparedness.
dry hydrological year (20112012). Therefore vegetation can be
durably affected by a drought if the drought is preceded by another Due to the different sources of information and principles that
dry year (recovery effect); in addition, rainfall in the immediately are used for drought indices, monitoring results that are obtained
previous season also inuences the thermal condition of the vege- from indicators often have certain differences. It can be determined
tation. that any single index is not sufcient for precisely depicting drought
Furthermore, the hydrological year 20102011 was a wet year characteristics. Thus, the combined use of different indicators at
that may have stimulated the photosynthetic activity of the vege- the same time or indices that integrate various sources of informa-
tation during the following year; hence the intensity of the drought tion may achieve results that are more consistent with the actual
was less severe in 20112012 than in 20122013. In other words, situation.
vegetation can still develop even with low rainfall if the previ-
ous year was wet (memory effect). This memory effect was also References
shown by Philippon et al. (2005) in the Sahel and by Martiny et al.
(2005) in three semiarid regions in Africa. Abbas, S., Nichol, J.E., Qamer, F.M., Xu, J., 2014. Characterisation of drought
In summary, the hydrological year 20112012 might have been development through remote sensing: a case study in Central Yunnan, China.
Remote Sens. 6 (6), 49985018.
less impacted by drought because it was preceded by a wetter Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop Evapotranspiration
year, and the hydrological year 20122013 might have been more Guidelines for Computing Crop Water Requirements. FAO Food and
impacted by drought because it was preceded by a drier hydrolog- Agriculture Organization of the United Nations, Rome.
Berliner, P., Oosterhuis, D.M., Green, G.C., 1984. Evaluation of the infrared
ical year. thermometer as a crop stress detector. Agric. For. Meteorol. 31, 219230.
The impacts of the vegetation stress for the 20112014 drought Boken, V.K., 2005. Agricultural drought and its monitoring and prediction: some
condition were also felt outside of the semiarid boundary. This is concepts. In: Boken, V.K., Crackenll, A.P., Heathcote, R.L. (Eds.), Monitoring and
Predicting Agricultural Drought: A Global study. Oxford University Press, New
an indication that the entire Northeast Brazil region is experiencing
York.
severe water stress, including the western extremity of the region, Bhuiyan, C., Singh, R.P., Kogan, F.N., 2006. Monitoring drought dynamics in the
where the climate is equatorial and the precipitation ranges from Aravalli region (India) using different indices based on ground and remote
1000 to 1200 mm per year (Marengo et al., 2009). sensing data. Int. J. Appl. Earth Obs. Geoinf. 8, 289302.
Carlson, T.N., Perry, E.M., Schmugge, T.J., 1990. Remote estimation of soil moisture
In general, the highest correlation between VSWI, NDD and pre- availability and fractional vegetation cover for agricultural elds. Agric. For.
cipitation occurs under dry conditions (there is less soil moisture Meteorol. 52, 4569.
504 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505

Carlson, T.N., Gillies, R.R., Perry, E.M., 1994. A method to make use of thermal Kogan, F.N., 1990. Remote sensing of weather impacts on vegetation in
infrared temperature and NDVI measurements to infer surface soil water non-homogeneous areas. Int. J. Remote Sens. 11 (8), 14051419.
content and fractional vegetation cover. Remote Sens. Rev. 9, 161173. Kogan, F.N., 1997. Global drought watch from space. Bull. Am. Meteorol. Soc. 78,
Carolwicz, M., 1996. Nature hazards need not lead to natural disasters. EOS 77 (16), 621636.
149153. Kogan, F.N., 2000. Contribution of remote sensing to drought early warning. In:
Cavalcanti, N.B., Oliveira, C.A.V., Brito, L.T.L., Resende, G.M., 1999. Avaliaco do uso Wilhite, D.A., Sivakumar, M.V.K., Wood, D.A. (Eds.), Early Warning Systems for
de tcnicas de captaco de gua de chuva na regio semi-rida do Nordeste Drought Preparedness and Drought Management. Proc. Expert Group Meeting,
brasileiro. Rev. Bras. Eng. Agrc. Ambient 3 (3), 403407. 57 September 2000, Lisbon, Portugal. World Meteorological Organization,
Chen, Y.Y., You, Y.S., Tang, Y.H., 2014. Study and apply on model of monitoring and Geneva, pp. 86100.
assessing early and mid-summer drought multi-parameter a case study in Kogan, F.N., 2002. World droughts in the new millennium from AVHRR-based
Chongqing. In: International Conference on Water Resource and Vegetation Health Indices. Eos Trans. Am. Geophys. Union 83 (48), 562563.
Environmental Protection: WREP2014. Liu, J., Wu, J., Wu, Z., Liu, M., 2013. Response of NDVI dynamics to precipitation in
Cunha, A.P.M.A., Alvala, R.C.S., Correia, F.W.S., Kubota, P.Y., 2009. Calibration of the the BeijingTianjin sandstorm source region. Int. J. Remote Sens. 34,
Simplied Simple Biosphere Model SSiB for the Brazilian Northeast. AIP 53315350.
Conf. Proc. 1100, 641. Marengo, J.A., Schaeffer, R., Pinto, H.S., Zee, D.M.W., 2009. Mudancas climticas e
Cunha, A.P.M.A., Alval, R.C.S., Sampaio, G., Shimizu, M.H., Costa, M.H., 2013. eventos extremos no Brasil. FBDS, Rio de Janeiro.
Calibration and validation of the integrated biosphere simulator (IBIS) for Martiny, N., Richard, Y., Camberlin, P., 2005. Interannual persistence effects in
Brazilian semi-arid region. J. Appl. Meteorol. Climatol. 52, 27532770. vegetation dynamics of semi-arid Africa. Geophys. Res. Lett. 32, L24403, http://
Davenport, M.L., Nicholson, S.E., 1993. On the relation between rainfall and the dx.doi.org/10.1029/2005GL024634.
Normalized Difference Vegetation Index for diverse vegetation types in East McKee, T.B., Doesken, N.J., Kleist, J.,1993. The relation of drought frequency and
Africa. Int. J. Remote Sens. 14, 23692389. duration to time scales. In: Proceedings of the Eighth Conference on Applied
Doorenbos, J., Pruitt, W.O., 1977. Guidelines for Predicting Crop Water Climatology. American Meteorological Society, Boston, pp. 179184.
Requirements, FAO-ONU, Rome, Irrigation and Drainage Paper no. 24 (rev.)., Matheron, G., 1969. Le krigeage universel. Technical Report 1. Paris School of
pp. 144. Mines. Cah. Cent. Morphol. Math., Fontainebleau.
Dorman, J.L., Sellers, P.J.A., 1989. Global climatology of albedo, roughness length Mu, Q., Zhao, M., Kimball, J.S., McDowell, N.G., Running, S.W., 2013. A remotely
and stomatal resistance for atmospheric general circulation models as sensed global terrestrial drought severity index. Bull. Am. Meteorol. Soc. 94,
represented by the Simple Biosphere Model (SiB). J. Clim. Appl. Meteorol. 28, 8398.
833855. Nagarajan, R., 2009. Drought Assessment. Springer, Dordrecht, Netherlands.
Du, L., Tian, Q., Yu, T., Meng, Q., Jancso, T., Udvardy, P., Huang, Y., 2013. A Nichol, J.E., Abbas, S., 2015. Integration of remote sensing datasets for local scale
comprehensive drought monitoring method integrating MODIS and TRMM assessment and prediction of drought. Sci. Total Environ. 505, 503507.
data. Int. J. Appl. Earth Obs. Geoinf. 23, 245253. Palmer, W.C., 1965. Meteorological Drought. Research Paper No. 45. U.S.
Feddes, R.A., Kabat, P., Van Bakel, P.J.T., Bronswijk, J.J.B., Halbertsma, J.M., 1988. Department of Commerce Weather Bureau, Washington, DC.
Modelling soil water dynamics in the unsaturated zone: state of art. J. Hydrol. Palmer, W.C., 1968. Keeping track of crop moisture conditions, nationwide: the
100, 69111. new Crop Moisture Index. Weatherwise 21, 156161.
Ferreira, O.E., Beltro, N.E.M., Knig, A., 2005. Efeitos da aplicaco de gua Peters, A.J., Walter-Shea, E.A., Jii, L., Vina, A., Hayes, M., Svoboda, M.D., 2002.
residuria e nitrognio sobre o crescimento e produco do algodo herbceo. Drought monitoring with NDVI-based standardized vegetation index.
Rev. Bras. Oleagin. Fibro. 9 (01/03), 893902. Photogramm. Eng. Remote Sens. 68 (1), 7175.
Gao, X., Huete, A.R., Didan, K., 2003. Multisensor comparisons and validation of Philippon, N., Mougin, E., Jarlan, L., Frison, P.-L., 2005. Analysis of the linkages
MODIS vegetation indices at the semiarid Jornada Experimental Range. IEEE between rainfall and land surface conditions in the West African monsoon
Trans. Geosci. Remote Sens. 4 (10). trough CMAP, ERS-WSC, and NOAA-AVHRR data. J. Geophys. Res. 110, D24115,
Gao, M., Qin, Z., Zhang, H., Lu, L., Zhou, X., Yang, X., 2008. Remote sensing of http://dx.doi.org/10.1029/2005JD006394.
agro-droughts in Guangdong province of China using MODIS Satellite Data. Rhee, J., Im, J., Carbone, G.J., 2010. Monitoring agricultural drought for arid and
Sensors 8 (8), 46874708. humid regions using multi-sensor remote sensing data. Remote Sens. Environ.
Gibbs, W.J., Maher, J.V., 1967. Rainfall Deciles as Drought Indicators. Bureau of 114, 28752887.
Meteorology Bulletin No. 48. Commonwealth of Australia, Melbourne. Rossato, L., Tomasella, J., Alval, R.C.S., 2005. Avaliaco da Umidade do Solo no
Gitelson, A., Kaufman, Y., 1998. MODIS NDVI optimization to t the AVHRR data Brasil durante o Episdio El Nino (1982/83). Rev. Bras. Agromet. 13 (1),
series spectral considerations. Remote Sens. Environ. 66 (3), 343350. 143153.
Goward, S.N., Xue, Y., Czajkowski, K.P., 2002. Evaluating land surface moisture Rulinda, C.M., Dilo, A., Bijker, W., Stein, A., 2012. Characterising and quantifying
conditions from the remotely sensed temperature/vegetation index vegetative drought in East Africa using fuzzy modelling and NDVI data. J. Arid
measurements: an exploration with the simplied simple biosphere model. Environ. 78, 169178.
Remote Sens. Environ. 79, 225242. Sandholt, I., Rasmussen, K., Andersen, J., 2002. A simple interpretation of the
Gutman, G., 1990. Towards monitoring droughts from space. J. Clim. 3, 282295. surface temperature/vegetation index space for assessment of surface
Henricksen, B.L., Durkin, J.W., 1986. Growing period and drought early warning in moisture status. Remote Sens. Environ. 79, 213224.
Africa using satellite data. Remote J. Sens. 7, 15811608. Santos, J.R., Shimabukuro, Y.E., 1993. O sensoriamento remoto como indicador das
Herrmann, S.M., Anyamba, A., Tucker, C.J., 2005. Recent trends in vegetation fenofases dos cerrados brasileiros: estudo de caso com dados AVHRR/NOAA.
dynamics in the African Sahel and their relationship to climate. Glob. Environ. Simpsio Brasileiro de Sensoriamento Remoto, 8, 1993, Curitiba. Anais, 2.
Change 15, 394404. MCT/INPE, So Jos dos Campos, pp. 249257.
Huete, A., Didan, K., Miura, T., Rodriguez, E.P., Gao, X., Ferreira, L.G., 2002. Overview Sapir, D.G., Below, R., 2014. EM-DAT: International Disaster Database. Universit
of the radiometric and biophysical performance of the MODIS vegetation Catholique de Louvain, Brussels, Belgium www.emdat.be.
indices. Remote Sens. Environ. 83, 195213. Schwinning, S., Sala, O.E., 2004. Hierarchy of responses to resource pulses in arid
IBGE, 2009. O censo agropecurio 2006 e a agricultura familiar no Brasil. MDA, and semi-arid ecosystems. Oecologia 141, 211220.
Braslia. Shafer, B.A., Dezman, L.E., 1982. Development of a Surface Water Supply Index
Ibrahim, Y.Z., Balzter, H., Kaduk, J., Tucker, C.J., 2015. Land degradation assessment (SWSI) to assess the severity of drought conditions in snowpack runoff areas.
using residual trend analysis of GIMMS NDVI3g soil moisture and rainfall in In: Proceedings of the Western Snow Conference, Fort Collins, CO,
sub-Saharan West Africa from 1982 to 2012. Remote Sens. 7, 54715494. pp. 164175.
Johnson, G.E., Achutuni, V.R., Thiruvengadachari, S., Kogan, F.N., 1993. The role of Son, N.T., Chen, C.F., Chen, C.R., Chang, L.Y., Minh, V.Q., 2012. Monitoring
NOAA satellite data in drought early warning and monitoring: selected case agricultural drought in the Lower Mekong Basin using MODIS NDVI and land
studies. In: Wilhite, D.A. (Ed.), Drought Assessment, Management and surface temperature data. Int. J. Appl. Earth Obs. Geoinf. 18, 417427.
Planning: Theory and Case Studies. Kluwer Academic Publishers. Souza, S.S., Tomasella, J., Gracia, M.G., Menezes, P.C.P., Amorin, M.C., Pinto, C.A.M.,
Julien, Y., Sobrino, J.A., 2009. Global land surface phenology trends from GIMMS 2001. O Programa de Monitoramento Climtico em Tempo Real na rea de
database. Int. J. Remote Sens. 30, 34953513. atuaco da SUDENE: PROCLIMA. Bol. Soc. Bras. Meteor. 25, 1524.
Justice, C., Townshend, J., Vermote, E., Masuoka, E., Wolfe, R., Saleous, N., Roy, D., Tannehill, I.R., 1947. Drought, Its Causes and Effects. Princeton University Press,
Morisette, J., 2002. An overview of MODIS, its data processing and products for Princeton, NJ, USA.
terrestrial science applications. Remote Sens. Environ. 83 (12), 315. Thiruvengadachari, S., Gopalkrishna, H.R., 1993. An integrated PC environment for
Karnieli, A., Bayasgalan, M., Bayarjargal, Y., Agam, N., Khudulmur, S., Tucker, C.J., assessment of drought. Int. J. Remote Sens. 14, 32013208.
2006. Comments on the use of the Vegetation Health Index over Mongolia. Int. Thenkabail, P.S., Gamage, N., Smakhin, V., 2004. The Use of Remote Sensing Data
J. Remote Sens. 27, 20172024. for Drought Assessment and Monitoring in South West Asia. IWMI Research
Karnieli, A., Agam, N., Pinker, R.T., Anderson, M., Imhoff, M.L., Gutman, G.G., Panov, Report # 85. IWMI, Colombo, Sri Lanka, 25 pp.
N., Goldberg, A., 2010. Use of NDVI and land surface temperature for drought Tomasella, J., Hodnett, M.G., 1998. Estimating soil water retention characteristics
assessment: merits and limitations. J. Clim. 23, 618633. from limited data in Brazilian Amazonia. Soil Sci. 163, 190202.
Keyantash, J., Dracup, J.A., 2002. The quantication of drought: an evaluation of Tomasella, J., Hodnett, M.G., Rossato, L., 2000. Pedotransfer functions for the
drought indices. Bull. Am. Meteorol. Soc. 83, 11671180. estimation of soil water retention in Brazilian soils. Soil Sci. Soc. Am. J. 64,
Keyantash, J., Dracup, J.A., 2004. An aggregate drought index: assessing drought 327338.
severity based on uctuations in the hydrologic cycle and surface water Tomasella, J., Pachepsky, Y., Crestana, S., Rawls, W.J., 2003. Comparison of two
storage. Water Resour. Res. 40, W09304, http://dx.doi.org/10.1029/ techniques to develop pedotransfer functions for water retention. Soil Sci. Soc.
2003WR002610. Am. J. 67, 1085.
 A.P.M. Cunha et al. / Agricultural and Forest Meteorology 214215 (2015) 494505 505

Tomasella, J., Hodnett, M.G., Cuartas, L.A., Nobre, A.D., Waterloo, M.J., Oliveira, S.M., International Conference on Recent Advances in Space Technologies (RAST),
2008. The water balance of an Amazonian micro-catchment: the effect of 911 June, pp. 137141, http://dx.doi.org/10.1109/RAST.2011.5966808.
interannual variability of rainfall on hydrological behavior. Hydrol. Proc. 22, Yang, W., Merchant, J.W., 1997. Impacts of upscaling techniques on land cover
21332147. representation in Nebraska, USA. Geocarto Int. 12 (1), 2739.
Tucker, C.J., Choudhury, B.J., 1987. Satellite remote sensing of drought conditions. Yang, W., (MSc thesis) 2010. Drought Analysis Under Climate Change by
Remote Sens. Environ. 23, 243251. Application of Drought Indices and Copulas. Civil and Environmental
Tucker, C.J., 1989. Comparing SMMR and AVHRR data for drought monitoring. Int. J. Engineering, Portland State University.
Remote Sens. 10, 16631672. Yang, S., Xu, Y., Gao, Y., 2013. Summer drought assessment based on Vegetation
Udmale, P., Ichikawa, Y., Manandhar, S., Ishidaira, H., Kiem, A., 2014. Supply Water Index in Chongqing. Adv. Intell. Syst. Res., http://dx.doi.org/10.
Farmers perception of drought impacts, local adaptation and administrative 2991/rsete.2013.231.
mitigation measures in Maharashtra State, India. Int. J. Disaster Risk Sci. 10, Zhou, G.Y., Wei, X.H., Wu, Y.P., Liu, S., Zhang, D., Zhang, Q., Liu, J., Meng, Z., Wang,
250269. C., Chu, G., Liu, S., Tang, X., Liu, X., 2011. Quantifying the hydrological responses
Vermote, E.F., Saleous, N., Justice, C., 2002. Atmospheric correction of the MODIS to climate change in an intact forested small watershed in Southern China.
data in the visible to middle infrared: rst results. Remote Sens. Environ. 83 Glob. Change Biol. 17, 37363746.
(12), 97111. Zhou, G.Y., Peng, C.H., Li, Y.L., Liu, S., Zhang, Q., Tang, X., Liu, J., Yan, J., Zhang, D.,
Vieira, R.M.S.P., Cunha, A.P.M.A., Alval, R.C.S., Carvalho, V.C., Ferraz Neto, S., Chu, G., 2013a. A climate change-induced threat to the ecological resilience of
Sestini, M.F., 2013. Land use and land cover map of a semiarid region of Brazil a subtropical monsoon evergreen broadleaved forest in south China. Glob.
for meteorological and climatic models. Rev. Bras. Meteorol. 28, 129138. Change Biol. 19, 11971210.
Wan, Z., Zhang, Y., Zhang, Q., Li, Z., 2004. Quality assessment and validation of the Zhou, G.Y., Houlton, B.Z., Wang, W., Huang, W., Xiao, Y., Zhang, Q., Liu, S., Cao, M.,
MODIS global land surface temperature. Int. J. Remote Sens. 25, 261274. Wang, X., Wang, S., Zhang, Y., Yan, J., Liu, J., Tang, X., Zhang, D., 2014.
Wilhite, D.A., Glantz, M.H., 1985. Understanding the drought phenomenon: the Substantial reorganization of Chinas tropical and subtropical forests: based on
role of denitions. Water Int. 10 (3), 111120. the permanent plots. Glob. Change Biol. 20, 240250, http://dx.doi.org/10.
Wilhite, D.A., 2000. Drought as a natural hazard: concepts and denitions. In: 1111/gcb.12385.
Wilhite, D.A. (Ed.), Drought: A Global Assessment, vols. 1 and 2. Routledge Zhou, L., Zhang, J., Wu, J., Zhao, L., Liu, M., L, A., Wu, Z., 2012. Comparison of
Publishers, London (chapter 1). remote sensed and meteorological data derived drought indices in
Wu, J., Zhou, L., Mo, X., Zhou, H., Zhang, J., Jia, R., 2015. Drought monitoring and Mid-Eastern China. Int. J. Remote Sens. 33, 17551779.
analysis in China based on the Integrated Surface Drought Index (ISDI). Int. J. Zhou, L., Wu, J., Zhang, J., Leng, S., Liu, M., Zhao, L., Zhang, F., Shi, Y., 2013b. The
Appl. Earth Obs. Geoinf. 41, 2333. integrated surface drought index (ISDI) as an indicator for agricultural drought
Xue, Y., Sellers, P.J., Kinter, J.L., Shukla, J., 1991. A simplied biosphere model for monitoring: theory, validation, and application in Mid-Eastern China. IEEE J.
global climate studies. J. Climate 4, 345364. Sel. Top. Appl. Earth Obs. Remote Sens. 6, 12541262.
Yagci, A.L., Deng, M., Di, L., Han, W., Peng, C., 2011. Vegetation index based
technique for global agricultural drought monitoring. In: 2011 5th