2011 International SWAT Conference & Workshops
The use and results of the Soil and Water Assessment Tool in Brazil: A
review from 1999 until 2010
Luis H. P. Garbossa, PhD in hydraulics and sanitation
Santa Catarina Agricultural Research and Extension Corporation – Epagri/Ciram – Florianópolis,
SC, Brazil. E-mail: [email protected]
Luiz R. C. de Vasconcelos, undergraduate student of Sanitary and Environmental
Engineering
Santa Catarina Federal University – UFSC – Florianópolis, SC, Brazil. E-mail: [email protected]
Katt R. Lapa, PhD in hydraulics and sanitation
Santa Catarina Federal University – UFSC – Florianópolis, SC, Brazil. E-mail: [email protected]
Éverton Blainski, PhD in agronomy
Santa Catarina Agricultural Research and Extension Corporation – Epagri/Ciram – Florianópolis,
SC, Brazil. E-mail: [email protected]
Adilson Pinheiro, PhD in Water Resources
Blumenau University – Furb – Blumenau, SC, Brazil. E-mail: [email protected]
Abstract
The Soil and Water Assessment Tool (SWAT) is becoming a familiar tool for some Brazilian
students, professors and professionals. The model has been applied for several hydrological
studies and evaluations under Brazilian conditions. This research was carried out with the objective
of identifying the SWAT model applications in Brazilian watersheds, its strengths, and its
weaknesses. Over 70 publications such as theses, dissertations and articles about the use of the
model in Brazilian watersheds were analyzed. From these, 60 are referred to in this paper. Despite
the significant variability of climates and hydrological behaviors of the modeled watersheds, the
results provided by the papers demonstrate that the model had good performance in several
Brazilian regions. Based on the papers’ results, it is also possible to identify that the SWAT
performance in Brazilian watersheds permits its use as a support decision tool for municipalities,
state companies, federal institutions, basin committees, and environmental organizations.
However, there are very few reports about using the SWAT as a practical assessment tool in
Brazil. Furthermore, there is some evidence that the hindrance to use SWAT in Brazil is not
whether the model can be used here or not, but how to obtain enough quantity and quality data to
simulate a basin.
Keywords: SWAT, Brazil, review, application.
2011 International SWAT Conference & Workshops
Introduction
The approval of the Brazilian National Water Resources Policy in the year of 1997 demands tools to support
decision making. The researchers, engineers and professionals involved with the water resources
management were facing the need for the development and improvement of these tools. The hydrological
and quality water models are these tools, and one of these models is the SWAT. This model allows the
simulation of hydrological processes, sediments production and transport, and nutrients and pesticides
cycling. Moreover, it is possible to simulate different scenarios of water availability for climate changes,
which is one of the objectives of the CLIMASUL project (Study of the climatic changes in southern Brazil),
supported by MCT/Finep-CNPq. The project involves universities and research institutions from three
southern Brazil States: Paraná, Santa Catarina and Rio Grande do Sul. The CLIMASUL project aims to
evaluate the effects of climate changes in experimental watersheds using the SWAT model with input data
from HadRM3P regional climate model.
The SWAT is becoming a familiar tool for some Brazilian students, professors and professionals. The model
has been applied in several hydrological studies and evaluations in Brazilian watersheds. One of the first
registers found of the use of the Soil and Water Assessment Tool (SWAT) in Brazil dates from 1999. Since
then, the model has been used in several academic studies in various Brazilian regions. It is possible to find
articles published in scientific journals, meetings and conferences. Over 70 publications were found as
theses, dissertations and articles about studies applying SWAT in Brazilian watersheds. Despite the growing
use of deterministic hydrological models in Brazil, the use of the SWAT was restricted mainly to researches
in academic studies. These studies focused mainly on the capacity of the model to represent Brazilian
watersheds adequately. There are few reports on the use of the model for practical applications by
environmental companies, private and governmental institutions. Therefore, this situation is beginning to
change with studies that will use SWAT for environmental assessments. There is, for instance, a study of the
influence of the water contribution from watersheds in the Florianópolis Bay and its influence on shellfish
production. The SWAT will be used to simulate the contributing watersheds streamflow. This paper aims to
present the uses of the SWAT model in Brazil in the past twelve years, the results obtained using the model,
the regions where the model has been applied, the main results obtained, and the viability of using the model
in Brazil.
SWAT published papers in Brazil from 1999 until 2010
This survey has identified several papers from master theses and doctoral dissertations, to congress and
seminars publications to scientific papers about the use of SWAT in Brazil. The year of 1999 was defined as
the starting survey date due to the information availability. Furthermore, it was identified that one of the first
reports on the SWAT use in Brazil was made by Oliveira (1999). The theses and dissertations presented as
bibliographical reference were used just when no article was found. For this study 74 publications about the
use of SWAT in Brazil were found. From these, 60 are presented in this paper and 52 publications are
reported in Table 1. The publications that are found but not presented here correspond to 14 theses and
dissertations that are already reported in papers. The studies applied SWAT for several watershed sizes,
topographic conditions and different climate regions in Brazil. The smallest watershed modeled has
approximately 1.2 km2 and is located in Rio Grande do Sul, and the largest one has 29,000 km2, and is
located in the State of Mato Grosso.
2011 International SWAT Conference & Workshops
The focus of most studies was to verify the viability of using SWAT in Brazil. Most of the papers presented
the data source used as input data, the calibration and validation results and the main conclusions about its
applicability. For this study an indicator was created in order to separate the studies in three indicators. These
indicators correspond to studies focused on streamflow, sediments, and nutrients. Figure 1 presents a
summary of the proportion of studies in each class and the quantity of studies developed in the past twelve
years.
Indicator
Years
2%
Streamflow
16%
Nutrients
1999
2003
2004
2005
2006
2007
2008
2009
2010
8%
Sediment
4%
25%
17%
41%
15%
43%
13%
10%
(a)
6%
(b)
Figure 1. Percentage of each indicator (a); distribution of the SWAT publications for the period (b).
It is assumed that studies that have used the model to simulate sediments have already simulated streamflow,
even if not presented. Moreover, it is assumed that studies that presented nutrients results have already
simulated streamflow and sediments, even if not presented. It is possible to identify in Figure 1(a) the
percentage of studies for the three indicators. Almost half of the studies have modeled streamflow and
sediment. In Figure 1(b) it is possible to identify that, besides the year of 2005, the years of 2009 and 2010
are the ones with the largest number of publication about SWAT in Brazilian watersheds. This is a strong
indication of the growing use of the SWAT model in Brazil.
Based on the information gathered in the consulted bibliographical reference it was possible to identify the
Brazilian territory where SWAT applications were made. Figure 2 is a map of Brazil with the location of the
studies developed. The States of São Paulo, Santa Catarina, and Paraná are responsible for 51% of the
SWAT papers, distributed as 19%, 16% and 16%, respectively.
The South region of Brazil, which is composed of Paraná (PR), Santa Catarina (SC) and Rio Grande do Sul
(RS), is responsible for 42% of the published studies. The second region that publishes the most studies is
the Southeast (São Paulo (SP), Rio de Janeiro (RJ), Minas Gerais (MG) and Espírito Santo (ES)), with 32%.
The state with the most studies of water quality is Paraná, which is the same state where practical
applications of the model were found.
2011 International SWAT Conference & Workshops
Figure 2. Map of Brazil with states, number and indicator of SWAT modeled watersheds.
Table 1 summarizes some information about SWAT use in Brazil. The tabulation was based on the model
presented by Gassman et al. (2007). This table model was chosen because it makes it possible to compare the
results obtained in Brazil with other world basins. This table was organized by year of paper publication and
by alphabetical order.
Several studies presented statistical calibration and some others presented graphical statistics. The two
statistical parameters chosen to be present in Table 1 are NSE (Nash-Sutcliffe Efficiency Index) and R2
(Coefficient of Determination). Some of the other statistical parameters presented in the papers are detailed
in this text.
Table 1. Published papers using SWAT in Brazilian watersheds from 1999 until 2010.
Reference
Year
1
Oliveira and
Medeiros
1999
2
Barsanti et al.
2003
3
Garrido
2003
4
Machado and
Vettorazzi
2003
5
Machado et al.
2003
6
Minoti et al.
2004
7
Silva et al.
2004
8
Baldissera
2005
9
Coelho et al.
2005
10
Neves et al.
2005
11
Oliveira et al.
2005a
12
Oliveira et al.
2005b
13
Pereira et al.
2005a
14
Pereira et al.
2005b
15
Prado et al.
2005
16
Silva et al.
2005
17
Armas
2006
Watershed
(location)
Joanes River
(Bahia)
Taquarizinho river
Aquidauana river
Jiquiriçá river
(Bahia)
Marins river
(São Paulo)
Marins river
(São Paulo)
Guabirobas river
(São Paulo)
Canchim river
(São Paulo)
Cuiabá river
(Mato Grosso)
Piraquara river
(Paraná)
Bonito river
(São Paulo)
Salitre river
(Bahia)
Salitre river
(Bahia)
Jiquiriçá river
(Bahia)
Jiquiriçá river
(Bahia)
Jundiaí-Mirim river
(São Paulo)
Bonito river
(São Paulo)
Corumbataí river
(São Paulo)
Drainage
Area
(km²)
Indicator
Time Period
C = calib.
V = valid.
1993-1995
1997-2002
Best NSE and R²
C = Calibration V = Validation
Daily
Monthly
Annual
C: 0.61 and 0.79
C: 0.58 and 0.82
C:-1.03 and 0.23
V: -65.9 and 0.37
-
755.40
Streamflow
sediment
1,500.00
15,200.00
Streamflow
6,900.00
Streamflow
59.73
Sediment
C: 1999-2000
-
C: 0.83 and 0.92
-
59.73
Streamflow
C: 1999-2000
-
C: 0.92 and 0.94
-
54.12
Sediment
1999-2003
-
-
-
15.00
Sediment
1999-2003
-
-
-
29,000.00
Streamflow
C: 1994-1998
-
C: 0.77 and 0.81
-
100.00
Stream flow
Water Quality
C: 1984-1998
-
-
C: - and 0.83
-
-
223.00
----
1992-2004
-
-
-
13,470.00
Streamflow
C: 1969-1972
-
C: - and 0.80
-
13,470.00
Streamflow
C: 1977-1979
V: 1969-1973
-
C: - and 0.88
C: - and 0.70
-
6,900.00
Streamflow
C: 1993-1995
-
-
-
6,900.00
Streamflow
C: 1993-1995
V: 1997-2002
C: -2 and 0.37
V: -49 and 0.51
-
-
120.15
Sediment
-
-
-
-
47.17
Sediment
1992-2004
-
-
-
1,710.00
Streamflow
C: 1973-1984
V:1985-2003
-
C: 0.94 and 0.94
V: 0.84 and 0.92
C: 0.89 and 0.93
V: 0.27 and 0.67
2011 International SWAT Conference & Workshops
N
Reference
Year
18
Bittencourt and
Gobbi
2006
19
Moro et al.
2006
20
Neves et al.
2006
21
Paiva and Paiva
2006
22
Silva et al.
2006a
23
Silva et al.
2006b
24
Machado et al.
2007
25
Rodrigues and
Reis
2007
26
Roloff et al.
2007
27
Adriolo et al.
2008
28
29
Lopes and
Kobiyama
Lopes and
Kobiyama
2008a
2008b
30
Lopes et al.
2008
31
Marchioro
2008
32
Blainski and
Garbossa
2009
33
Gibertoni et al.
2009
Watershed
(location)
Piraquara river
(Paraná)
Marins river
(São Paulo)
Bonito river
(São Paulo)
Menino Deus I
(Rio Grande do Sul)
Canchim river
(São Paulo)
Beija-Flor/Jataí river
(São Paulo)
Marins river
(São Paulo)
Coruripe river
(Alagoas)
Toledo river
Ajuricaba river
(Paraná)
Apucaraninha river
(Paraná)
M2 Experimental
(Santa Catarina)
M2 Experimental
(Santa Catarina)
M2 Experimental
(Santa Catarina)
Santa Maria river
(Rio de Janeiro)
Araranguá river
(Santa Catarina)
Antonina Bay
(Paraná)
Drainage
Area
(km²)
Indicator
Time Period
C = calib.
V = valid.
58.00
Streamflow
Phosphorus
1998-2002
59.73
Sediment
1999-2000
-
-
-
223.00
Nitrogen
Phosphorus
12 years
period
-
-
C: 0.76 and C: 0.74 and -
18.00
Streamflow
1996-1998
C: - and 0.54
C: 0.88 and 0.88
-
15.00
Nitrogen
Phosphorus
1993-2004
1993-2004
-
-
C: 0.79 and C: 0.80 and -
110.00
Sediment
-
-
-
-
59.73
Sediment
1999-2000
-
C: 0.78 and -
-
1,562.00
Sediment
2004-2006
-
-
-
65.00
16.50
Streamflow,
sediment and
quality
1998-2004
-
-
-
504.00
Sediment
C: 2000-2005
V: 1988-1999
C: 0.20 and V: -2.36 and -
C: 0.42 and V: 0.62 and -
-
8.56
Streamflow
C: 2007
C: 0.23 and 0.51
-
-
8.56
Streamflow
C: 2007
C: 0.23 and 0.51
-
-
Streamflow
Sediment
Streamflow
Sediment
C: 2006-2007
-
C: - and 0.59
-
-
C: 2006-2007
-
C: 0.72 and C: -6.11 and -
3,000.00
Streamflow
C: 2005-2007
-
C: 0.72 and -
-
C: 1975-1991
V: 1992-2007
C: 0.42 and 0.47
V: 0.53 and 0.54
------------------C: 0.41 and 0.42
V: 0.52 and 0.49
C: 0.81 and 0.65
V: 0.74 and 0.72
------------------C: 0.94 and 0.63
V: 0.92 and 0.70
-
1,597.00
Streamflow
----------------Sediment
8.56
13.50
Best NSE and R²
C = Calibration V = Validation
Daily
Monthly
Annual
C: - and 0.82
-
-
-
2011 International SWAT Conference & Workshops
N
Reference
Year
Watershed
(location)
Drainage
Area
(km²)
Indicator
Time Period
C = calib.
V = valid.
34
Lino et al.
2009
Preto river
(Santa Catarina)
1,000.00
Streamflow
C: 1993-1997
-
C: 0.51 and -
-
35
Lubitz
2009
Concórdia river
30.74
Streamflow
Sediment
Nutrients
2006-2009
C: 0.32 and C: -0.03 and *
C: 0.88 and C: 0.84 and *
-
36
Paim and
Menezes
2009
2,840.00
Streamflow
1991-2001
-
C: 0.73 and 0.95
-
37
Souza et al.
2009
788.00
Streamflow
C: 1994-1998
-
C: 0.72 and -
-
38
Uzeika
2009
C: 2002-2006
C: 0.79 and C: -5.58 and -
C: 0.84 and C: -6.53 and -
-
39
Xavier
2009
1994
-
-
-
40
Baltokoski et
al.
2010
-
C: 0.70 and C: 1.00 and -
-
41
Blainski et al.
2010
C: 0.85 and 0.85
-
42
Blainski et al.
2010
43
Bonumá et al.
2010
44
Fontes et al.
2010
45
Garbossa et al.
2010
46
Guimarães et
al.
2010
47
Lelis and
Calijuri
2010
48
Pereira et al.
2010
Tijucas river
(Santa Catarina)
A. Negro river
(Santa Catarina)
Arvorezinha
(Rio Grande do Sul)
Manso river
(Mato Grosso)
Conrado and
Pinheiro rivers
(Paraná)
Araranguá river
(Santa Catarina)
Lajeado dos
Fragosos river
(Santa Catarina)
Arroio Lino
(Rio Grande do Sul)
Jacuípe river
(Bahia)
Lajeado dos
Fragosos river
(Santa Catarina)
Riacho dos
Namorados
(Paraíba)
São Bartolomeu
river
(Minas Gerais)
Cachoeirinha
(Minas Gerais)
1.19
10,553.00
Stramflow
Sediment
Streamflow
Sediment
52.97
Streamflow
Phosphorus
2003-2005
3,000.00
Streamflow
C: 2005-2007
59.00
Streamflow
Sediment
C: 1999-2009
-
3.20
Streamflow
Sediment
2004-2005
1,895.00
Streamflow
C: 1966-1974
V: 1975-1982
59.00
Streamflow
13.00
Best NSE and R²
C = Calibration V = Validation
Daily
Monthly
Annual
-
-
C: 0.73 and -
V: 0.55 and 0.74
V: 0.57 and -0.33
-
V: 0.87 and 0.90
V: 0.70 and 0.77
C: 0.84 and 0.71
V: 0.63 and 0.39
-
C: 1999-2009
-
C: 0.73 and -
-
Streamflow
1993-2006
-
-
-
54.22
S. Runoff
Sediment
2006-2008
-
-
-
-
Sediment
1979-2008
-
-
-
* Due to table organization, the nutrients statistical parameters are presented in the text.
2011 International SWAT Conference & Workshops
N
N
Reference
Year
Watershed
(location)
Drainage
Area
(km²)
Indicator
Time Period
C = calib.
V = valid.
49
Santos et al.
2010
Apucaraninha river
504.00
Streamflow
----------------Sediment
1988-2005
50
Schultz et al.
2010
788.00
Streamflow
Sediment
1994-2004
51
Souza et al.
2010
-
Streamflow
2003-2006
C: 0.84 and -
-
-
52
Trindade et al.
2010
85.00
Nitrogen
Phosphorus
1999-2007
-
-
-
A. Negro river
(Santa Catarina)
Tocantins-Araguaia
sub-watershed
(Tocantins)
Experimental
(Espírito Santo)
Best NSE and R²
C = Calibration V = Validation
Daily
Monthly
Annual
C: 0.73 and V: 0.78 and ----------------------------------C: 0.20 and C: 0.42 and V: -2.36 and V: 0.62 and C: 0.71 and C: 0.38 and C: 0.86 and -
2011 International SWAT Conference & Workshops
2011 International SWAT Conference & Workshops
Observing the statistical results presented in Table 1 it is possible to identify in the studies that the model
runs for monthly average outputs is normally adequate to represent the watersheds. It is possible to identify
that these results were obtained for different watershed sizes and locations. However, there is great variation
in the model performance when evaluating the model output statistics on a daily basis. Most of the studies
have a poor representation of the watershed when evaluating the statistical parameters on a daily basis. This
can be identified for both small and large watersheds.
There is little statistical evaluation to validate the presented modeled watersheds. Probably this is because of
the lack of information and sufficiently long datasets.
Brazilian data sources used for SWAT modeling
In this chapter the focus is to verify the data sources used in order to obtain the necessary information to run
the model. In Brazil there are some national institutions that have some of the information needed. The main
national institutions are ANA (National Water Agency), ANEEL (National Agency for Electrical Energy),
Embrapa (Brazilian Agricultural Research Corporation), (IBGE) Brazilian Institute of Geography and
Statistics, INMET (National Institute of Meteorology), and INPE (National Institute for Space Research).
ANA owns the daily basis national network database for precipitation, streamflow, sediment, and water
quality. Besides that, the papers report several state institutions, universities and private companies
monitoring databases. These databases are from watersheds of interest and experimental watersheds,
generally with sub-daily data. These databases have provided important information to be used as input data
for the model.
The data used to run SWAT in the Joanes River watershed and Jacuípe river basin were obtained mainly
from the Bahia state company of urban development in scales of 1:100,000 (soil use and vegetation) and data
from exploratory soil surveys obtained during the RADAMBRASIL (Radar in the Amazon) project during
the decades of 1970 and 1980 (Oliveira and Medeiros, 1999; Fontes et al., 2010). Other data sources for the
Jacuípe river basin were the GETOPO 30, SRTM in the EROS DATA CENTER – USGS for DEM (Digital
Elevation Model). The weather data was obtained from Morro do Chapéu weather station. Finally the
precipitation data was gathered from ANA (Fontes et al., 2010).
For the studies developed in the Marins river watershed (Machado et al., 2003; Machado and Vettorazzi,
2003; Moro et al., 2006; Machado et al., 2007) the soil use data was obtained from an image of the SPOT
satellite, taken in 1998. The information about the soil was obtained from a scientific bulletin by the
agronomic institute of Campinas and from a project called Piracena project. The Brazilian soil classification
system from Embrapa was also used as a data source. The weather, precipitation and sediment data were
obtained from the University of São Paulo and from the Department of Water and Electrical energy.
The data used for Taquarizinho and Aquidauana basins was obtained from several sources. The DEM was
obtained by digitalizing topographic maps with a scale of 1:100,000 from the year 1966. The land cover
databases were created from Landsat 5 TM images for three different periods, 1966, 1985 and 1996. The soil
map was obtained from a study developed for the Alto Paraguay Basin Conservation Plan. The agricultural
management data was obtained through interviews with local agricultural organizations. The hydrological
data was obtained from ANEEL’s reports. The sediment data was obtained from other studies developed in
similar basins (Barsanti et al., 2003).
2011 International SWAT Conference & Workshops
Minoti et al. (2004) constructed the DEM from topographic maps from IBGE. The soil use information was
obtained from LANDSTAT-7 satellite images. The weather data was obtained from a weather station,
property of Embrapa.
For the study developed in the Piraquara river watershed the soil use data was obtained from the
PARANASAN project, which was an environmental sanitation project for the state of Paraná. The soil
characteristics were obtained from studies developed by Paraná Federal University. The DEM was built
from information obtained from COMEC (Coordination of the Metropolitan Curitiba Region), which is a
governmental institution. The streamflow data was obtained from ANA. Finally, the weather data was
obtained from Simepar (Paraná Technology Institute) and SUDERSHA (Paraná Water Institute), both of
them state companies (Coelho et al., 2005; Bittencourt & Gobbi, 2006)
The soil use data for Cuiabá river basin and Manso river sub-basin were obtained from the RADAMBRASIL
project. The DEM was obtained from previous studies developed in the area. The soil data was obtained
from a study organized by FEMA (Mato Grosso Environmental State Foundation). The precipitation data
and streamflow data were obtained from gauges operated by ANA. The weather data was acquired from two
meteorological stations, one from UFMT (Mato Grosso Federal University) and the other one from INMET
(Baldissera, 2005; Xavier, 2009).
The data to simulate the Salitre river basin was obtained from several sources. The soil use map has a scale
of 1:250,000. The map was obtained from the integrated management project for São Francisco Basin
supported by ANA. The DEM was constructed from topographic maps with a scale of 1:100,000. The
weather data was obtained from INMET for 3 stations and the precipitation data was obtained from ANA
from 14 gauges. The soil data was obtained from the results of the RADAMBRASIL project (Oliveira et al.,
2005a, 2005b).
The Jiquiriçá river basin soil characteristics data was obtained from a study developed for the PDRH (Water
Resources Master Plan for Bahia State) (Garrido, 2003; Pereira et al., 2005a, 2005b). The DEM was created
based on maps with the scale of 1:100,000. The soil information was obtained from documents developed by
SRHSH (Water Resources, Sanitation and Housing Secretary) during the execution of the Environmental
Recovery Program for Jiquiriça river watershed and information from Embrapa. The soil use data was based
on images from the LANDSAT-7 satellite, documents from IBGE and field surveys. The weather data was
obtained from one weather station from INMET. Finally the precipitation and streamflow data were obtained
from INMET and ANA gauges, respectively.
The topographic, land use and soil characteristics data were obtained from the Campinas Agronomic
Institute. That information was acquired during the development of the Environmental diagnosis for land use
management of the Jundiaí-Mirim river watershed in 2002 (Prado et al., 2005).
The Bonito river watershed input data was obtained from previous studies in the watershed as presented for
several papers. (Silva et al., 2005; Silva et al., 2006a; Neves et al., 2005; Neves et al., 2006; Crestana et al.,
2010). The previous studies obtained weather data from Embrapa weather station. The DEM was built from
an IBGE topographic map from 1971 with a scale of 1:50,000. Studies developed in the years of 2000 and
2003 created the soil use and soil maps used in the studies.
The soil use and soil data used to model Corumbataí river basin was obtained from the Master Plan for
conservation of the water resources developed by the Institute for Forest Studies and Researches with a grid
of 25 m x 25 m extracted from a map with a scale of 1:100,000. The DEM was obtained from PIRACENA
project (Project developed by the Center for Nuclear Energy in Agriculture, Piracicaba Campus). The
weather data was obtained from the Meteorological station of ESALQ/USP (Luiz de Queiroz Superior
2011 International SWAT Conference & Workshops
Agriculture School) and from the Meteorology and Climatology Analysis Laboratory from UNESP. The
precipitation and streamflow data were obtained from DAEE database (Department of Water and Electric
Energy of São Paulo state). The sediment and water quality data was obtained from field surveys. Armas
(2006) reported that the refinement of the model could be done with more detailed information about soil
characteristics.
The Beija-Flor and Jataí rivers sub-watersheds are located inside the Mogi-Guaçu river watershed. The
database used as input to the model was obtained from previous studies. The data quantity and quality was
available due to the projects developed in the Jataí ecological station (Silva et al., 2006b).
The Apucaraninha river watershed sediment data was obtained from previous studies in the area and an
intensive monitoring of sediment transport during the years 2004 and 2005 (Santos et al., 2006; Santos et al.,
2010). The DEM, land use, weather and soil information source was not detailed in the text.
Paiva and Paiva (2006) used weather data from an INMET weather station. The soil map was obtained from
previous studies in the area. The soil use was obtained from maps built from 1998 LANDSAT satellite
images. The DEM was built from topographic maps with a scale of 1:50,000.
The study developed in the Coruripe river basin used a DEM developed from the SRTM (Shuttle Radar
Topography Mission). The soil use was obtained from the river Coruripe Water Resources Master Plan
published in the year 2000. The soil classification and characteristics were obtained from Embrapa and
studies developed by a private company named Coruripe Plant. The weather data was obtained from
Coruripe Plant. The precipitation data was obtained from 10 precipitation gauges owned by ANA (Rodrigues
and Reis, 2007). The paper states that SWAT is an excellent tool to be used as a scientific support decision
tool. However, the model calibration was not possible due to the lack of detailed information.
The input weather data to be used in Toledo and Ajuricaba river watersheds was obtained from a weather
station owned by Simepar. The soil characteristics were based on Embrapa’s database. The other data
sources were studies developed in these watersheds during the execution of the Good Water Farming
Program supported by Itaipu Binacional.
The soil characteristics in M2 - experimental watershed were obtained from the Embrapa’s database (Lopes
and Kobiyama, 2008a). The weather database was supplied by Battistella Florestas Company and part of the
precipitation data from Epagri/Ciram (Santa Catarina Agricultural Research and Extension Corporation). The
soil use data was generated from Rio Negrinho’s municipality orthophotographs.
The soil use map for Santa Maria river was obtained from previous studies and is composed mainly by
meadows (67%) and forests (20%). The soil use map was made from photos with scales varying from
1:10,000 up to 1: 30,000. The soil information was also obtained from previous academic studies and
technical reports developed in the area. The classification was based on Embrapa’s recommendations. The
weather data was obtained from previous studies and from INMET meteorological station. The streamflow
data and the information on sediment production were acquired during the research development. There was
a hydro-sedimentologic gauge monitoring turbidity and water level in the watershed (Marchioro, 2008).
The data source used for the Tijucas river basin was obtained from several sources. The DEM was obtained
from the SRTM. The soil map was gathered from Embrapa with the scale of 1:250,000. The soil use
information obtained from the Landsat_5_TM satellite image. The precipitation and streamflow data was
acquired from ANA. The weather data was gathered from two weather stations, in Florianópolis and in São
José (Paim and Menezes, 2009).
2011 International SWAT Conference & Workshops
Gibertoni et al. (2009) obtained the soil use data from the LANDSAT ETM satellite for the year 1999. The
hydrologic and meteorological data were obtained from the Aneel and Suderhsa.
Lubitz (2009) applied the model to the Concórdia river watershed located in city of Lontras region. This
watershed has a monitoring history because it is one of the seven monitored watersheds of the Prapem and
Matasul projects. The Prapen is a project of environmental recovery and support for small farms and Epagri
is one of the companies that give support to the watershed hydrology and precipitation monitoring. Besides,
there are several scientific studies developed in the watershed. These studies have contributed to the
hydrological and meteorological data acquisition. The soil characteristics and management were obtained
from Embrapa’s and Epagri’s studies. The DEM was developed from IBGE topographic maps with the scale
of 1:50,000 from 1980, and aero-photos with the scale of 1:25,000 from 1978. The soil use information was
obtained from SPOT5 satellite images. The sediment data was obtained through field surveys.
The database used in the Araranguá river basin model was obtained from several sources. The DEM was
constructed from a topographic map by IBGE in the scale of 1:50,000. The soil map was gathered from
Embrapa and the soil use information was obtained from Fatma (Santa Catarina State Environmental
Company). The meteorological data was available on Epagri’s and Casan’s database. Finally, the
hydrological information was acquired from ANA (Blainski and Garbossa, 2009; Blainski et al., 2010).
The Preto river basin soil characteristics database was obtained from Embrapa. The DEM was constructed by
Epagri from a topographic map by IBGE with the scale of 1:50,000. The soil use was obtained based on
three LANDSAT-TM5 satellite images. The weather data was gathered from Epagri’s and Inmet’s
meteorological station. The streamflow data was obtained from a limnimetric gauge owned by ANA (Lino et
al., 2009). The database to the Altíssimo rio Negro watershed was obtained mainly from the same sources as
the Negro river basin (Souza et al., 2009; Schultz et al., 2010).
Uzeika (2009) obtained soil use data from field surveys during the research project development. The
Arvorezinha watershed soil characteristics were obtained from previous studies in the area and from EmaterRS (Agricultural Technical Assistance and Extension Corporation – Rio Grande do Sul). The DEM was
created from IBGE topographic maps with a scale of 1:50,000. The management practices input data was
organized during the development of the study. The precipitation, streamflow and sediment data started to be
collected during a governmental program named RS-Rural in 2004. After the end of the program, a
partnership among Sindifumo (Rio Grande do Sul Tobacco Trade Union), Emater-RS and UFRGS-IPH
(Hydraulics Research Institute) is monitoring the watershed.
The Lajeado dos Fragosos river watershed soil characteristics map was generated based on soil samples
analyses and classified as recommended by Embrapa. The soil use was obtained from Fatma. The hydrologic
information was obtained from Embrapa and Epagri. (Blainski et al., 2010; Garbossa et al., 2010).
Complementary information was obtained from PNMA II (National Program for the Environment II) reports.
Bonumá et al. (2010) obtained the input data from previous studies developed in the Arroio Lino watershed.
The weather and precipitation data for Riacho dos Namorados watershed were obtained from gauges
previously installed for researches. The DEM was obtained from the SRTM images (Guimarães et al., 2010).
Ribeirão Cachoeirinha soil use database was constructed during the study development. The DEM was built
from IBGE topographic maps on a 1:50,000 scale. Emater provided the information about soil characteristics
with the scale of 1: 1,000,000. The weather and hydrologic data were obtained from INPE and ANA (Pereira
et al., 2010).
2011 International SWAT Conference & Workshops
The São Bartolomeu river watershed soil use data was obtained from the IKONOS satellite image, associated
with field surveys. The soil characteristics were defined by using the Embrapa soil database associated with
soil analyses, satellite images and aero-photos. The DEM was created from maps with a scale of 1:50,000
(Lelis and Calijuri, 2010).
The precipitation and streamflow data source used to model the Conrado and Pinheiro rivers watershed was
obtained from monitoring gauges installed during the PNMA II. The DEM was created from topographic
maps with a scale 1:50,000. The land use data was obtained from Landsat TM5 satellite images. The weather
information was obtained from Iapar (Paraná Agricultural Institute) and Simepar (Batlokoski et al., 2010).
Souza et al. (2010) have obtained the hydrological information from ANA. Besides that the authors have
estimated the heat flow from orbital images with the model SEBAL (Surface Energy Balance Algorithms for
Land).
The SWAT model concept is to have a model in which you can readily use available data, such as SCS
database and climatological radar data (Pereira et al., 2005a). However, the model input data to be used in
Brazilian watershed modeling is not organized to be readily used by the model. The information has to be
gathered from various sources, with a great variety of formats and levels of detail. This diversity of data
sources and difficulties can be identified in the reported researches. Paiva and Paiva (2006) also report that
there is a great need for more detailed information to be used in mathematical models.
There is an effort in order to organize a Brazilian general use database. This can be seen in the available soil
database from Embrapa and the available water resources database from ANA. But there is a great need for
more detailed information. The soil data used from Embrapa is adapted and some theoretical
parameterization is made because not all the input soil data is available. The use of SWAT in Brazilian
watersheds should be preceded by an evaluation of the available data for the region. The official
governmental institutions database should be consulted. Besides that, other regional, state and private
companies should be consulted. Based on available dataset the modeler can decide when it is possible to
apply the model for the region.
The researched papers confirm the statement of Adriolo et al. (2008). The challenge to use SWAT in Brazil
is not whether the model can be used here or not, but how to obtain enough data to simulate a basin.
Climate characteristics of the modeled watersheds
The variation of the climate in the presented watersheds is significant. The climate varies from tropical in
Brazil’s Northeast Region to temperate in the extreme southern state. The mean annual precipitation varies
from less than 500 mm up to more than 2,000 mm. The climate variation is presented here for each one of
the four Brazilian Regions that presented studies with SWAT, except for one study developed in the southern
part of the North region.
Brazil’s Northeast Climate
The Jiquiriçá river basin (Garrido, 2003; Pereira et al., 2005b) is located in a region where several weather
classifications are found. Based on the Köppen climatic system it is possible to find the following types: Af,
Am, Cwb and BSh. The basin region has climates varying from hot summer with average hottest month near
22 °C. The average annual precipitation varies from less than 800 mm in some region up to 1,200 mm for
other areas.
2011 International SWAT Conference & Workshops
The annual mean precipitation for the Jacuípe river basin varies from 500 mm up to 1,300 mm depending on
the location. The average temperature is about 23 °C (Fontes et al., 2010). The annual precipitation in the
Riacho dos Namorados watershed is 400 mm and occurs from March until June (Guimarães et al., 2010).
The Coruripe river watershed is influenced by the tropical humid climate. In the far East there is a tropical
climate and in the far West a semi-arid climate. The annual mean precipitation is approximately 1,100 mm.
The annual mean temperature is above 18°C (Rodrigues and Reis, 2007).
The region of Ribeirão Cachoeirinha watershed is located in the south portion of the Brazilian Northeast
Region. The summer and winter are well defined and the annual average precipitation varies from 1,300 mm
up to 1,700 mm (Pereira et al., 2010).
Brazil’s Central-West Climate
The Cuiabá river basin and the Manso river sub-basin are located in a region with Aw climate type by the
Köppen classification system. The average precipitation is 1,500 mm distributed in two periods: a rainy
period from December until March, and a dry one from June until August. The average temperature varies
from 12 °C up to 28 °C (Baldissera, 2005; Xavier, 2009).
Brazil’s Southeast Climate
The Bonito river watershed and sub-watersheds modeling efforts are presented in several papers (Neves et
al., 2005; Silva, et al., 2005; Neves et al., 2006; Crestana et al., 2010). The climate in the region is
considered as Cwa in the köppen international system. This corresponds to mesothermal dry winter with
temperatures varying from 18 °C up to 22 °C.
The Jundiaí-mirim watershed has an Aw climate type by the Köppen classification system, with a very wet
mesothermal climate without a well defined dry season (Prado, 2005).
The regions of the São Paulo State where Corumbataí river and Marins river watersheds are located have a
climate characterized as Cwa by the Köppen classification system. The region has a sub-tropical
mesothermal climate. It has dry winters and rainy summers. The average annual precipitation is
approximately 1,280 mm. The temperature varies from 22 °C in the summer and 17 °C in the winter.
(Armas, 2006; Moro et al., 2006).
Brazil’s South Climate
Coelho et al. (2005) states that the climate for Piraquara river watershed is sub-tropical humid and
mesothermal with fresh summers (under 22 °C) and winters with average temperatures close to 18°C. The
annual mean precipitation is 1,400 mm. The Conrado river and the Pinheiro river watershed are located in a
region with a climate labeled as Cfb in the Köppen classification system. The annual mean precipitation
(1979-2006) is 2,094 mm (Baltokoski et al., 2010).
The climate is labeled as Cfb for Köppen classification in the M2 - experimental watershed region. This
climate is characterized for being temperate, continually wet, for having no dry seasons and for having a
fresh summer. The annual average temperature varies from 15.5 °C up to 17 °C. The annual precipitation
varies from 1,360 mm up to 1,670 mm (Lopes and Kobiyama, 2008a, 2008b; Lopes et al., 2008; Souza et al.,
2009). Lino et al. (2009) modeled the Preto river basin. The basin is located in the same region as the M2 –
experimental watershed and consequently has the same climatological classification. The Concórdia river
watershed is close to the Preto river basin and has a humid subtropical climate, without a dry season and with
2011 International SWAT Conference & Workshops
a hot summer. The hottest month has a mean temperature of 22 °C. The annual average temperature varies
from 17 up to 19 °C. The annual precipitation varies from 1,320 mm up to 1,640 mm (Lubitz, 2009).
The Tijucas river basin is located in two climatic regions. The East side presents warm and wet mesothermal
summer and is labeled as Cfa in the Köppen classification. The West basin region has a mesothermal wet
climate with dry summers and is labeled as Cfb in the Köppen classification (Paim and Menezes, 2009).
The Araranguá river basin region has a climate labeled as Cfa in the Köppen classification system. The
annual average temperature is 18 °C and the annual precipitation vary from 1,220 mm up to 1,660 mm
(Blainski and Garbossa, 2009; Blainski et al., 2010).
The Santa Maria river basin is located in a region with Aw climate type by the Köppen classification system.
The tropical climate has well defined seasons. The region temperature varied from mean temperature of
19.9 °C in the coldest month and 26.5 °C in the hottest month Marchioro (2008).
The climate is defined as Cfb for Köppen classification in the Arvorezinha watershed. This climate is
characterized for being temperate, continually wet and with a fresh summer. The daily average temperature
for the hottest month is approximately 22 °C and the daily average temperature for the coldest month is 3 °C
(Uzeika, 2009).
The Lajeado dos Fragosos river watershed region has a climate labeled as Cfa in the Köppen classification
system. The annual average temperature is 18 °C and the annual precipitations vary from 1,700 mm up to
1,900 mm (Blainski et al., 2010; Garbossa et al., 2010).
The climate in the Arroio Lino region is labeled as Cfb in the Köppen classification system. There are hot
and wet summers, morning frost in the winter and well distributed precipitation along the year (Bonumá
et al., 2010).
Despite the fact of a significant variability of climates for the modeled watersheds, the results presented in
the papers demonstrate that the model had a good performance for all the climate regions. This is a strong
evidence that the climate is not a limiting factor to the model use in Brazil.
Studies developed in Brazilian watersheds
Oliveira and Medeiros (1999) wrote one of the first papers found for the 12 year period of researches using
the SWAT model in Brazil. The studied watershed is located in the Northeast Region. The main objective
was to evaluate the potential of using SWAT for watershed assessment. The authors concluded that the
model can be used for planning and management of a Brazilian watershed.
The studies in the Marins river basin tested different percentages of discretization, varying the area for
discretization from 0.1 km2 up to 2 km2. The best results were obtained using 39 sub-watersheds for the
upstream limnimetric gauge (Machado et al., 2003). This represents an average discretization area of 0.4 km2
and a NSE of 0.92 for the calibration. The authors concluded that, despite some events of super- or underestimations of sediment production, in general the model had a good performance representing the
production and transport of sediments (Machado and Vettorazzi, 2003).
A study of soil erosion in two basins of the Pantanal area was developed: a highly populated basin
denominated Taquarizinho river and one little populated, the Aquidauana river basin. The calibrated model
has shown the importance of management practices. High specific soil loss was identified comparing single
crop (45 ton/ha) versus crop rotations (5 ton/ha). The correct pasture management resulted in a decrease of
2011 International SWAT Conference & Workshops
specific soil loss from 12 ton/ha to 0.6 ton/ha. Barsanti et al. (2003) concluded that the model had a good
estimation of the effects of soil use and management practices.
Garrido (2003) modeled the Jiquiriçá river basin aiming to evaluate the SWAT applicability in order to
support the water resources management for the region. The poorly gauged basin has shown to cause
uncertainty in model calibration and validation. The obtained results may be considered inconsistent due to
the lack of available data. Garrido (2003) states that these results should be a stimulus in order to improve
the hydrologic monitoring system for the region.
The discretization level for the Marins river watershed was evaluated (Machado et al., 2007). Six
discretization levels were tested, resulting in 15, 17, 25, 33, 39 and 43 sub-watersheds. From 17 up to 33 subwatersheds no significant changes were identified in the NSE and R2 coefficients. Satisfactory results were
obtained with 39 sub-watersheds together with a sediment production increase. When using 43 subwatersheds, besides the increase in the sediment production, a significant fall occurred for the NSE
coefficient. These results are linked to the variation in the topographic characteristics of the sub-watersheds,
markedly the channel slope and length.
Minoti et al. (2004) and Silva et al. (2004) discuss the output results obtained by the model but there is no
detailed information about the efficiency of the model in the calibration and validation steps. These authors
focused on the discussion of the model capabilities based on the results presented. The conclusion is that the
model has great potential do be used as an assessment tool, but further investigation should be conducted.
A study was developed in the Curitiba region. The objective was to evaluate the possibility of using SWAT
as a tool to quantify environmental impacts in a watershed and support decision makers (Coelho et al., 2005).
The correlations between measured and modeled values were considered poor. Probably this problem is
related to the fact that the precipitation and streamflow data were not hydrologically reliable. Another
important piece of information is that the soil use data was from a different period of time from that of the
streamflow data. Besides some problems, the authors considered SWAT a good tool to be used in Brazilian
watersheds.
Neves et al. (2005) evaluated three methods to calculate the evapotranspiration with SWAT. The highest
evapotranspiration values were obtained by the Hargreaves method and the lowest values by the Penmanmonteith method. The authors concluded that in the study area the best method in representing the
evapotranspiration was the Penman-Monteith.
The Salitre river basin has some problems with water availability, mainly in dry months, from August to
October. Several Salitre river contributory streamlets and the Salitre river itself have no flow during dry
seasons. There was a lack of information to parameterize the model, but still the model was capable of
representing the streamflow behavior. The obtained coefficient of determination was 0.8, despite the poor
data available (Oliveira et al., 2005a). SWAT users should be aware that the available information about CN
values, outside the U.S., is sometimes not representative (Oliveira et al., 2005b).
Pereira et al., (2005a) focused on the study of sensibility analyses. They have identified that the model was
sensible to the following parameters: FFCB, SOL_AWC, SOL_Z, SOL_BD, and SOL_K, and for the
evapotranspiration processes are the CANMX and ESCO, and the superficial runoff the CN.
Pereira et al., (2005b) chose the Hargreaves method to calculate evapotranspiration, based on the results of
other studies developed in the region. To evaluate the results of the calibration and validation the authors
used five different methods on a daily basis.
2011 International SWAT Conference & Workshops
The objective of the study developed in the Jundiaí-Mirim river watershed was to evaluate the model
capacity to represent different scenarios of land use. Prado et al., (2005) concluded that the SWAT has an
adequate level of response to detect different soil uses, even for small changes. Garbossa et al. (2010) used
the model and identified that the SWAT response can simulate the compliance with the Brazilian forest law
watershed effects.
The objective of Baldissera (2005) was to evaluate the applicability of the SWAT model for flow estimation
in the Cuiabá river basin. The model was applied to simulate the present scenario and simulate the
streamflow considering that 100% of the original vegetation was recovered. The model was adequate to
represent the basin. The statistical evaluation of four streamflow gauges resulted in NSE above 0.7. The main
problem was to represent the base flow values. Probably this issue was due to the limited availability of
information on soil quality.
A study was developed in the Manso river basin, which is a sub-basin of the Cuiabá river basin. The
objective was to evaluate five different scenarios, testing the soil use of the basin for agriculture, meadow,
forest and others. The model was considered a promising model to be used as an assessment tool. For this
study the results were used as a qualitative approximation for flow values. The scenarios indicate the most
vulnerable area, but it was not possible to quantify the sediment loss (Xavier, 2009).
Silva et al. (2005) used the SWAT model to simulate the soil loss in the Bonito river watershed. Although
the statistical analysis was not presented, the authors identified that the model had adequately represented the
watershed soil loss. Moreover, the model identified the areas in the watershed that were more susceptible to
soil erosion.
The study developed in the Corumbataí river basin aimed at the evaluation of the biogeodynamic of the
pesticides employed in sugarcane production. The study first identified that the most used pesticides in the
basin are the herbicides for sugarcane production. The herbicides were monitored for water and sediment
samples. The SWAT was used to identify the critical areas in the basin for pesticides contribution. Three
different scenarios were simulated for sugar-cane tillage with different herbicides application. Armas (2006)
stated that the model was capable of simulatimg the herbicides behavior in the basin when it is adequately
calibrated and validated.
The Canchim river watershed is a sub-watershed of the Bonito river watershed. Phosphorus and nitrogen
were simulated in the sub-watershed. The model presented promising results. The deviation of the measured
and simulated data resulted in a value of -2.25%. This indicates that the simulation overestimated the
measured values. The results of the NSE were 0.79 and 0.80 for nitrogen and phosphorus, respectively (Silva
et al., 2006). Another study in the Bonito river watershed in an area of 223 km2 evaluated the nutrients entry
due to poultry activity. The deviation resulted in -2.8% and -2.4% for nitrogen and phosphorus, respectively
(Neves et al., 2006). The SWAT was able to adequately simulate the influence of the nitrogen and
phosphorus behavior on the watershed.
Silva et al. (2006b) identified the adequate model response to the soil use in the watershed. The paper
authors stated that the results are just indicative and the simulation should be improved. Still, the authors
detailed neither information about the period of time used nor the statistical coefficients results.
Seven scenarios were simulated to evaluate the sediment production in the Marins river watershed (Moro et
al., 2006). The scenerios were composed by area variations for sugar-cane production, meadow and native
cover. Despite the fact that the calibration and validation results were not presented, some conclusions were
made. The authors presented the results comparing each scenario with each other. The model output data
resulted in equal values to different scenarios. The repetition of output data was attributed to the level of the
2011 International SWAT Conference & Workshops
discretization and tolerance for land use. The results indicate that the native vegetal soil cover was a better
protection for the soil than other uses. The contribution to sediment production was strongly related to the
terrain slope and the sugar-cane culture.
The Apucaraninha river watershed was evaluated for seven different scenarios, including the present
condition. The scenarios tested different soil use conditions. The discharge NSE values for calibration and
validation were both above 0.7. The total mean observed and simulated flow values had a difference of less
than 1%. The study has shown the model ability to represent adequately the Apucaraninha river watershed.
(Santos et al., 2006; Santos et al., 2010).
A study was developed in Menino Deus I watershed to evaluate the viability of using SWAT in the region.
The watershed was adequately represented on a daily basis, with a determination coefficient value of 0.53
and correlation coefficient value of 0.73. The model results were used to build the duration curve and
compare with the duration curve built from measured data. It was possible to identify a good match between
the two curves (Paiva and Paiva, 2006).
The research developed in the Coruripe river basin established an erosion rate. This rate represents the
sediment volume produced for a constant area and soil type. It was identified that areas with sugar-cane have
a low impact on the soil erosion, with an erosion rate of 0.52. The urban areas and the meadow areas are
more susceptible to erosion, with rates of 1.65 and 1.17, respectively. The meadow occupies a significant
area in the watershed; consequently, it was responsible for 75% of the erosion (Rodrigues and Reis, 2007).
The main objective of modeling the Piraquara watershed was to test the viability of using data generated by
SWAT together with the TMDL (total maximum daily load) process. The TMDL is a process developed by
EPA (Environmental Protection Agency) to evaluate processes that cause or contribute to the loss of water
quality. The calibration of the SWAT model for Piraquara watershed was made to the streamflow data. The
calibration to the phosphorus simulation was not possible because of lack of data. There were phosphorus
data samples collected to the watershed, but the sampling interval was not adequate to calibrate the model.
The SWAT run simulated a 10 year period, from 1998 until 2008, for phosphorus discharge in the Piraquara
II reservoir. The results of the simulation were used to test the TMDL process (Bittencourt and Gobbi, 2006).
The simulations developed in the Toledo river and in the Ajuricaba river watersheds aimed to evaluate the
model applicability to the area. The presented scenarios were modeled and the results were discussed just
comparing the different scenarios with one another. Regardless of the fact that there was not model
calibration, it was identified that the model presented coherent responses. The paper states that it is an
important tool that can be used for watershed management (Roloff et al., 2007).
The monthly correlation coefficient for the Apucaraninha river watershed resulted in a value of 0.83 and 0.84
for calibration and validation, respectively. The monthly determination coefficient to the entire period is
0.65. The daily NSE results were not satisfactory. Nevertheless, the monthly and annual results were
promising. The modeled data graphical evaluation does not have a perfect adherence to the observed data but
demonstrated a similar behavior. The results for sediment are promising since the difference between the
observed and the simulated data was only 2.6%. Adriolo et al., (2008) concluded that the SWAT model can
be used in the Brazilian conditions.
The M2 Experimental watershed was modeled for streamflow data. A good adjustment was obtained but the
model over-estimated the streamflow values for dry periods. Besides, the model significantly underestimated a peak flow event of 15.5m3.s-1. The authors attribute this to an intense upstream precipitation
event that was not measured. An analysis of the sediment production was made evaluating the hydrologic
and sediment equilibrium of the watershed. Fifteen anomalous events were identified by the analyses.
2011 International SWAT Conference & Workshops
Comparing the results to SWAT output, six events are coincident with higher sediment production (Lopes
and Kobiyama, 2008a). Another study developed in the M2 Experimental watershed tested three different
precipitation inputs and infiltration methods. The simulation with the best results used the daily precipitation
data instead of the sub-daily data and, consequently, the CN method. This result was attributed to the daily
precipitation data series length, which has the double quantity of the sub-daily data. The model had
adherence to the measured streamflow data but almost all simulations over estimated the peak flow and the
mean flow (Lopes et al., 2008; Lopes and Kobiyama, 2008b).
Gibertoni (2009) evaluated the use of SWAT to simulate the Santa Maria river watershed. Moreover, a
scenario considering the compliance with the Brazilian Forest law was simulated. He concluded that the
model is an important tool to identify sensible areas to sediment loss and to evaluate alternative land use
scenarios. Different from streamflow values, the sediment production was poorly represented by the model.
Gibertoni (2009) stated that these results are associated to constant soil use modifications in the watershed
and to the database low data quality and quantity.
The Tijucas river simulation output data was compared to the data of three hydrologic gauges to verify the
model performance. The model had an adequate performance in the outlet 24 with the NSE of 0.73. The
measured data in the outlets 9 and 10 had a poor correlation with the simulated data. The lack of sediment
data became a limitation to advance in the basin modeling. Paim and Menezes (2009) concluded that their
simulation can be used to evaluate the basin in a qualitative manner, but not the quantitative output values.
The Araranguá river basin simulation obtained satisfactory results, with an NSE of 0.72 on a monthly basis.
It was simulated for three different scenarios in order to test the model response. The model presented
coherent responses. The Araranguá river basin has an area of intense irrigated rice farming. From August
until December a water deficit may arise depending on the precipitation intensity (Blainski and Garbossa,
2009). The hydrological modeling with SWAT can be used as a tool to evaluate the water availability and
establish a water use protocol for the region (Blainski et al., 2010).
Lino et al. (2009) modeled the Preto river watershed including two reservoirs, one 200 km2 and the other 300
km2. The existence or not of the reservoirs made no difference with an effect in the superficial runoff of 0.93
mm. This corresponds to a flow variation of 0.75%. The streamflow mean deviation resulted in - 4.96. The
model underestimated the peak flow events. The conclusion was that the model adequately represented the
watershed. However, in order to obtain better results, more detailed information about the soil and land use is
necessary.
Souza et al. (2009) considered that the streamflow was adequately represented for Altíssimo rio Negro
watershed. However, there were limitations simulating peak and recession streamflow values. The water use
permit in Brazil is issued based on the low-flow indices obtained from the flow duration curve. The
maximum use permitted in the state of Santa Catarina corresponds to 50% of the Q98. For the state of Paraná
the maximum use permitted corresponds to 50% of the Q95 (Souza et al., 2009).
The Altíssimo rio Negro watershed output was used to estimate the Q95 and Q98 low-flow values for each
sub-watershed and compared the Q95 and Q98 low-flow to measured streamflow. The model has shown that
if the water use permit was based only on the Altíssimo rio Negro watershed low-flow values, some
problems could occur. Sub-watersheds could not supply the Q95 based forecast and attend the water demand.
Schultz et al. (2010) simulated the same watershed and found satisfactory results for monthly sediment
production. The daily simulated sediment production presented certain limitations. Probably this limitation is
related to the overestimated peak flows and under-estimated recession flow also observed by Souza et al.
(2009).
2011 International SWAT Conference & Workshops
The Arvorezinha watershed modeling objective was to evaluate the model applicability and to simulate three
different soil use scenarios. The NSE was calculated for each year with the mean results for flow values. The
model poorly represented the watershed on a daily basis, except for one year, in which the NSE reached
0.79. The model had a better performance for monthly average. Uzeika (2009) concluded that the model was
adequate to represent small watersheds for streamflow on a monthly basis. But the model was not adequate
to simulate sediment production for small watershed.
Lubitz (2009) developed a study in the Concórdia stream watershed. The objective was to evaluate the use of
the model in the watershed and the results for streamflow, sediment and nutrients. The nutrients modeling
results presented an NSE of 0.20 and -0.35 for nitrite and nitrate, respectively. These results are attributed to
two major issues. There was a need for more detailed information about effluents discharges and residues
management and there were just a few punctual water quality results, which did not represent the watershed
behavior. The difficulties were similar to modeling total phosphorus and orthophosphate. The NSE resulted
in -4.76 and -0.04 for total phosphorus and orthophosphate, respectively. The model represented adequately
the watershed for streamflow and sediment on a monthly basis. SWAT users should be careful when the data
parameterization is made beyond certain limits. It can cause too much distrust about the model results.
Blainski et al. (2010) simulated different scenarios to test the model response. Three hypothetical scenarios
were created substituting meadow and pasture for reforestation, annual crop with conventional soil
preparation and annual crop with no tillage. There was no sediment calibration. It was possible to identify
that the reforestation scenario significantly reduced the soil loss in comparison to the other scenarios.
The Lajeado dos Fragosos river watershed was simulated to evaluate the effects of the Brazilian Forest law
compliance in the streamflow. Besides the NSE of 0.73, a mean error of 46 L.s-1 and a value of -0.04 for
residual mass coefficient were obtained. It was possible to identify an increase of approximately 3.5% in the
water availability if the Forest law was complied. This represents an increase of 4.000 m3.day-1. The daily
output data had a poor NSE coefficient, probably because of the low water residence time in the watershed.
(Garbossa et al., 2010).
Fontes et al. (2010) studied the use of the isotopic hydrology associated with SWAT to calibrate the model
and solve problems of hydrological data deficiency. The model over-estimated values of discharge for peak
flows and long-term dry weather. The simulation adequately represented the mean streamflow values for the
Jacuípe river basin.
Bonumá et al. (2010) applied a sensibility analysis to evaluate the effect of certain parameters on the water
balance and sediment production values and their relations. The parameters CN2 and ESCO influenced
significantly both water and sediment production. The results indicate that the CN2, ESCO and Alpha_Bf
were the most important parameters for flow calibration. The parameters Usle_P and Slope significantly
influenced the sediment production. The results obtained in the Arroio Lino watershed demonstrated the
potential of using SWAT to simulate Brazilian watersheds. These results are attributed to the quality of the
input data available for this research.
The study developed in Riacho dos Namorados aimed to evaluate the effect of DEM interpolation in SWAT
results. The DEM used was from SRTM project with 90 m. Different interpolation techniques were used,
such as Nearest Neighbors, Simple Average, Weighed Average, and Quadrant Weighed Average. Significant
differences were identified in the morph metric model parameters for each interpolation. Apparently the
interpolation with simple average was inadequate. The DEM generated from the SRTM with a resolution of
30 m has shown stability and consistence to the model results (Guimarães et al., 2010).
2011 International SWAT Conference & Workshops
Pereira et al. (2010) studied the sediment production in the Ribeirão Cachoeirinha watershed. They
compared the results obtained with the model and concluded that the sediment yield results are similar to
values found in previous studies in the same region.
Lelis and Calijuri (2010) monitored the superficial runoff and sediment production in ten representative
locations of the São Bartolomeu watershed. The results of the monitoring were used to calibrate the model.
They identified and quantified the effects of previous precipitation and precipitation intensity in the sediment
production. A comparison of simulated runoff was compared with the measured runoff. The SWAT was
efficient in representing the sediment production and scenarios simulation.
The Conrado river and the Pinheiro river watersheds were simulated to streamflow and phosphorus and the
results were acceptable. The authors concluded that a regular sampling frequency is important to be able to
calibrate the model (Baltokoski et al., 2010).
Souza et al. (2010) estimated the heat flow from orbital images with the model SEBAL. The integration of
SEBAL output with SWAT input data allowed an improvement of 0.27 in the NSE for the simulated daily
streamflow when compared to the results without the use of SEBAL heat flow estimation.
Trindade et al. (2010) developed a study in an experimental watershed located in the state of Espírito Santo.
The objective was to evaluate the model sensitiveness for seven fertilization scenarios. The conclusion was
that the model adequately simulates the nutrients behavior related to the precipitation events. There were no
evidences of the need to change the model sub-routines related to phosphorus. However, Trindade et al.
(2010) identified a strong evidence that the model over-predicts the nitrogen kinetic behavior. The study
concluded that SWAT must be used carefully when simulating nutrients in tropical watersheds. Changes in
the sub-routine responsible for nitrogen transformation may be needed.
Based on the papers’ results, it is possible to identify that the SWAT performance for Brazilian watersheds
permits its use. Furthermore, there is evidence that the limiting factor is the database quality and availability.
The minimum requirements to use the model in Brazil should be defined. Based on those requirements it will
be easier to choose when one should invest time to model a watershed using SWAT.
Practical application in Brazilian watersheds
Some of the presented papers report the use or the attempt to use SWAT for real assessments. Coelho et al.
(2005) stated that the present-day law instruments for water and watershed management in Brazil demand
tools for environmental assessment of the watersheds. Hydrological models are a great support decisions tool
for municipalities, state companies, federal institutions, basin committee and environmental organizations.
Documents about the use of SWAT for practical application in Brazil were not easily found. There are very
few reports about using the SWAT as an assessment tool in Brazil, although, some governmental and private
institutions have used the model as a tool to make decisions about soil erosion and sediment transport.
The companies that have used the model are hydroelectric energy power plants (Santos et al., 2005 apud
Souza et al., 2009). A final report (LACTEC, 2007) of a hydroelectric energy power plant project evaluated
the viability of using SWAT. The report states that studies results recommend the use of the SWAT to
evaluate the sediment transport and water quality in the watershed.
2011 International SWAT Conference & Workshops
Junqueira and Silva (2008) have evaluated three different tools to plan and manage watersheds. The Amorim
& Cordeiro and PESMU (Strategic and sustainable planning of the urban environment) tools are methods
that use a matrix, and in some aspects qualitative data rather than quantitative. The third method evaluated
was SWAT. They have concluded that the three methods are valid to support decisions about the watersheds.
However, the user should evaluate his objectives, database information, watershed size, and software and
hardware availability.
A technical plan developed by UFBA (Federal University of Bahia) has proposed the use of SWAT as a part
of the methodology to use it as a tool for the implementation of the water resources management instrument
in the state of Bahia (UFBA, 2004). It was identified that despite the fact of poor available data the model
had an acceptable adjustment level. However, non reported documents have been found about the use of the
model as a management tool in the watershed.
A study named “Influence of the Governor Parigot de Souza Power Plant in the Antonina bay siltation” was
developed with the use of SWAT. The sediment production and transport was simulated as solid discharge in
the Antonina Bay. Distinct scenarios of soils use on the upstream basin were developed. The scenarios were
the present situation, a pessimistic scenario and a most probable scenario. The values of the statistical
efficiency for flow and sediment simulation were considered satisfactory and the results could be used to
assess the different land use scenarios and support future decisions about the Antonina bay basin (Gibertoni
et al., 2009).
Conclusions
The results from the existing papers show that SWAT has a great potential to be used in Brazil as a
decision support tool by watersheds committees and, by environmental and hydrological
governmental institutions.
The use of SWAT in Brazil occurred mainly for academic purposes. Several studies have identified the
potential of using SWAT in different Brazilian regions, for different climates conditions, watershed sizes,
soil and environmental conditions. Most of the papers identify that the model presented adequate
performance modeling the watersheds. A total of 60 studies were presented in this paper. From these, 27
papers present information about statistical coefficients for monthly basis. A total of 23 papers present a
monthly NSE above 0.7 for at least one of the evaluated indicators. Few papers have identified the need for
model adaptations in order to use it in Brazil. These results are an indication of the model robustness. It is
important to continually test the model to different conditions, but also to use it as a support decision tool in
practical applications.
The researched papers confirm the statement of Adriolo et al. (2008). The challenge to use SWAT in Brazil
is not whether the model can be used here or not, but how to obtain enough data to simulate a watershed. The
use of models such as SWAT should be addressed to the effective data availability (Garrido et al., 2003).
This will prevent that SWAT output results in generic conclusions
Probably the greatest hindrance in spreading its use in Brazil is the database availability. Besides that issue,
the model needs specialized professionals that sometimes are not available in environmental and regulating
governmental agencies and watershed committees.
2011 International SWAT Conference & Workshops
Acknowledgements
The authors would like to thank the institutions that have supported, in some way, the development of this
study. They are CNPq (National Council for Scientific and Technological Development), Epagri (Santa
Catarina Agricultural Research and Extension Corporation), Furb (Blumenau University) and
MCT/Finep/Ação Transversal - Previsão de Clima e Tempo - Edital 04/2008, convênio
01.08.0568.00FINEP.
References
Adriolo, M. V., Santos, I., Gibertoni, R. C., Camargo, A. S. G. 2008. Calibração do modelo SWAT para a
produção e transporte de sedimentos. In VI Simpósio Brasileiro sobre Pequenas e Médias Centrais
Hidrelétricas. Belo Horizonte – Minas Gerais. Brasil.
Armas, E. D. 2006. Biogeodinâmica de herbicidas utilizados em cana-de-açúcar (Saccharum spp.) na subbacia do rio Corumbataí. PhD diss. Piracicaba, São Paulo: Universidade de São Paulo, Escola Superior de
Agricultura Luiz de Queiroz.
Baldissera, G. C. 2005. Aplicabilidade do modelo de simulação hidrológica SWAT (Soil and Water
Assessment Tool), para a bacia hidrográfica do rio Cuiabá/MT. MS theses. Cuiabá, Mato Grosso:
Universidade Federal de Mato Grosso, Instituto de Ciências Exatas e da Terra.
Baltokoski, V., Tavares, M. H. F., Machado, R. E., Oliveira, M. P. 2010. Calibração de modelo para a
simulação de vazão e de fósforo total nas sub-bacias dos rios Conrado e Pinheiro – Pato Branco (PR).
Revista Brasileira de Ciência do Solo. 34: 253-261.
Barsanti, P., Disperati, L., Marri, P., Mione, A. 2003. Soil erosion evaluation and multitemporal analysis in
two Brazilian basins. In 2nd International SWAT Conference. Bari. Italy.
Bittencourt, S. and Gobbi, E. F. 2006. Carga máxima de fósforo admissível ao reservatório Piraquara II, uma
aplicação do processo TMDL. Nota. Revista Brasileira de Ciências do Solo. 30: 595-603.
Blainski, E. and Garbossa, L. H. P. 2009. Estimativa do regime hidrológico em uma bacia hidrográfica com
base na modelagem hidrológica dinâmica e distribuída. In XVIII Simpósio Brasileiro de Recursos Hídricos.
Campo Grande – Mato Grosso do Sul. Brasil.
Blainski, E., Garbossa, L. H. P., Maluta, S. 2010. Aplicação do modelo hidrológico SWAT (Soil Water
Assessment Tool) para a simulação da perda de solo e da disponibilidade hídrica em uma bacia hidrográfica
não instrumentada. In X Simpósio de Recursos Hídricos do Nordeste. Fortaleza – Ceará. Brasil.
Blainski, E., Silveira, F. A., Conceição, G., Garbossa, L. H. P. 2010. Utilização do modelo SWAT (Soil and
Water Assessment Tool) para estudos na microbacia hidrográfica do rio Araranguá / SC. In Tecnologias
para o Uso Sustentável da Água em Regadio, 617-626. 1 st ed. Torrecillas, A., Palácios, E., Victória, F. B.,
Pereira, L. S., Garcia, M., Paredes, P. Lisboa, Portugal. Colibri.
Bonumá, N. B., Reichert, J. M., Minella, J. P., Barros, C. A. P., Rodrigues, M. F., Buarque, D. C. 2010.
Balanço hídrico e sua relação com a modelagem da produção de sedimentos em uma pequena bacia
hidrográfica rural. In IX Encontro Nacional de Engenharia de Sedimentos. Brasília – Distrito Federal. Brasil.
2011 International SWAT Conference & Workshops
Coelho, A. C. P., Hardt, L. P. A., Fernandes, C. V. S. 2005. Agregação de novas variáveis ao processo de
planejamento urbano e regional sob a perspectiva de gestão dos recursos hídricos. In XVI Simpósio
Brasileiro de Recursos Hídricos. ABRH. João Pessoa – Paraíba. Brasil.
Crestana, S., Minoti, R. T., Neves, F. F. 2010. Modelagem e simulação aplicadas à avaliação dos impactos
da perda de solo e dos dejetos de suínos na qualidade da água de microbacias. In Planejamento ambiental do
espaço rural com ênfase para microbacias hidrográficas: Manejo de recursos hídricos, ferramentas
computacionais e educação ambiental. 167-199. 1st ed. Gomes, M. A. F. and Pessoa, M. C. P. Brasília, DF:
Embrapa Informação Tecnológica.
Fontes, A. S., Oliveira, M. Q. C., Medeiros, Y. D. P. 2010. Calibração do modelo SWAT em bacias do semiárido com auxílio da hidrologia isotópica. In X Simpósio de Recursos Hídricos do Nordeste. Fortaleza –
Ceará. Brasil.
Garbossa, L. H. P., Blainski, E., Justen, J. G K., Vasconcelos, L. R. 2010. Potencial de uso de modelo
hidrológico de base física na avaliação da legislação florestal e seus impactos nos cenários hidrológicos. In X
Simpósio de Recursos Hídricos do Nordeste. Fortaleza – Ceará. Brasil.
Garrido, J. M. 2003. Aplicação de modelo matemático de simulação com utilização de SIG à bacia do rio
Jiquiriçá – Bahia. MS theses. Brasília, Distrito Federal: Universidade de Brasília, Departamento de
Engenharia Civil e Ambiental.
Gassman, P. W., Reyes, M. R., Green, C. H., Arnold, J. G. 2007. The soil and water assessment tool:
historical development, applications and future research directions. Transactions of the American Society of
Agricultural and Biological Engineers. 50(4): 1211-1250.
Gibertoni, R. F. C., Santos, I., Müller, I. I., Pereira, P. S. 2009. Modelagem da produção e transporte de
sedimentos em bacias hidrográficas do litoral paranaense: o caso da bacia do rio Nhundiaquara. In XVIII
Simpósio Brasileiro de Recursos Hídricos. Campo Grande – Mato Grosso do Sul. Brasil.
Guimarães, J. C. N., Lunguinho, R. L., Srinivasan, V. S., Rufino, I. A. A., Taveira, I. M. L. M. 2010. A
influência dos métodos da interpolação na geração dos modelos numéricos de terreno e os resultados da
simulação pelo modelo SWAT. In IX Encontro Nacional de Engenharia de Sedimentos. Brasília – Distrito
Federal. Brasil.
Junqueira, C. A. R. and Silva, R. S. 2008. Avaliação ambiental aplicável a bacias hidrográficas no meio
urbano: Análise dos métodos Amorim & Cordeiro, PESMU e SWAT. Revista de Estudos Ambientais. 10(2):
6-23.
LACTEC. 2007. Relatório Final – Atendimento aos requisitos da LP 9589 – UHE Mauá. Consórcio
energético Cruzeiro do Sul. Available at: http://www.usinamaua.com.br/. Accessed 08 February 2011.
Lelis, T. A., Calijuri, M. L. 2010. Modelagem hidrossedimentológica de bacia hidrográfica na região sudeste
do Brasil, utilizando o SWAT. Revista Ambiente & Água – An Interdisciplinary Journal of Applied Science.
5(2): 158-174.
Lino, J. F. L., Corseuil, C. W., Kobiyama, M., Malutta, S., Uda, K., Bortolotto, N. L., Santos, I. 2009.
Análise da vazão da bacia hidrográfica do rio Preto com o modelo SWAT. In XVIII Simpósio Brasileiro de
Recursos Hídricos. Campo Grande – Mato Grosso do Sul. Brasil.
2011 International SWAT Conference & Workshops
Lopes, N. H. Y. and Kobiyama, M. 2008a. Análise do equilíbrio hidrossedimentológico em uma microbacia
experimental com o modelo SWAT e as características geomorfológicas. In VII Simpósio Nacional de
Geomorfologia e II Encontro Latino-Americano de Geomorfologia. Belo Horizonte – Minas Gerais. Brasil.
Lopes, N. H. Y. and Kobiyama, M. 2008b. Análise da produção de sedimentos e da qualidade da água em
uma microbacia experimental com SWAT. In: VIII Encontro Nacional de Engenharia de Sedimentos.
Campo Grande – Mato Grosso do Sul. Brasil.
Lopes, N. H. Y., Godoy, F. B., Kobiyama, M. 2008. Análise da produção de água e de sedimento em uma
microbacia experimental com a aplicação do modelo SWAT. In VI Simpósio de Engenharia Ambiental. Serra
Negra – São Paulo. Brasil.
Lubitz, E. 2009. Avaliação da aplicação do modelo hidrológico SWAT à bacia do ribeirão Concórdia –
Lontras, SC. MS theses. Blumenau, Santa Catarina: Universidade Regional de Blumenau, Programa de Pósgraduação em Engenharia Ambiental.
Machado, R. E. and Vettorazzi, C. A. 2003. Simulação da produção de sedimento para a microbacia
hidrográfica do ribeirão Marins (SP). Revista Brasileira de Ciências do Solo. 27: 735-741.
Machado, R. E., Vettorazzi, C. A. and Cruciani, D. E. 2003. Simulação de escoamento em uma Microbacia
Hidrográfica Utilizando Técnicas de Modelagem e Geoprocessamento. Revista Brasileira de Recursos
Hídricos. 8(1): 147-155.
Machado, R. E., Xavier, A. C., Cruciani, D. E., Vettorazzi, C. A., 2007. Efeito do nível de subdivisão em
uma bacia hidrográfica na simulação da produção de sedimentos. Revista Brasileira de Recursos Hídricos.
12(4): 49-58.
Marchioro, E. 2008. Modelagem hidrossedimentológica na bacia do córrego Santa Maria: Subsídios à
aplicação de práticas de conservação de água e solo no noroeste fluminense. PhD diss. Rio de Janeiro, Rio de
Janeiro: Universidade Federal do Rio de Janeiro, Programa de pós-graduação em geografia.
Mercuri, E. G. F., Deppe, F., Lohmann, M., Simões, K. 2009. Metodologia da geração de dados de entrada e
aplicação do modelo SWAT para bacias hidrográficas brasileiras. In XIV Simpósio Brasileiro de
Sensoriamento Remoto. Natal – Rio Grande do Norte. Brasil.
Minoti, R. T., Silva, F. G. B., Angelotti Netto, A., Machado, R. E., Tsuhako, E. M., Cestana, S. 2004.
Determinação preliminar do aporte de sedimentos na microbacia do ribeirão Guabirobas (Estado de SP). In
VI Encontro Nacional de Engenharia de Sedimentos. ABRH. Vitória – Espírito Santo. Brasil.
Moro, M., Cruciani, D. E., Vettorazzi, C. A., Machado, R. E. 2006. Aplicação do modelo SWAT para
simulação da produção de sedimentos em uma microbacia hidrográfica em cenários alternativos de uso do
solo. In VII Encontro Nacional de Engenharia de Sedimentos. ABRH. Porto Alegre – Rio Grande do Sul.
Brasil.
Neves, F. F., Silva, F. G. B., Crestana, S. 2005. Comparação entre três métodos de evapotranspiração
potencial aplicados a duas sub-bacias hidrográficas de Descalvado – SP. In XVI Simpósio Brasileiro de
Recursos Hídricos. ABRH. João Pessoa – Paraíba. Brasil.
Neves, F. F., Silva F. G. B., Crestana, S. 2006. Uso do modelo AVSWAT na avaliação do aporte de
nitrogênio (N) e fósforo (P) aos mananciais de uma microbacia hidrográfica contendo atividade avícola.
Engenharia Sanitária e Ambiental. 11(4): 311-317.
2011 International SWAT Conference & Workshops
Oliveira, M. Q. C. and Medeiros, Y. D. P. 1999. Aplicação do modelo SWAT na avaliação de impactos
decorrentes da modificação no uso do solo em bacias hidrográficas. In XIII Simpósio Brasileiro de Recursos
Hídricos. ABRH. Belo Horizonte – Minas Gerais. Brasil.
Oliveira, R. C., Oliveira, M. Q. C., Medeiros, Y. D. P. 2005a. Aplicação do modelo SWAT em região semiárida. In XVI Simpósio Brasileiro de Recursos Hídricos. ABRH. João Pessoa – Paraíba. Brasil.
Oliveira, M. Q. C., Oliveira, R. C., Medeiros, Y. D. P. 2005b. Hydrologic Modeling of a Semi-Arid Region
in Brazil. In 3nd International SWAT Conference. Duebendorf. Switzerland.
Paim, J. B. and Menezes, J. T. 2009. Estimativa do balanço sedimentar da bacia do rio Tijucas (SC-Brasil) a
partir da aplicação do modelo hidrológico SWAT. Revista Geográfica Acadêmica. 3(2): 5-14.
Paiva, R. C. D. and Paiva, E. M. C. D. 2006. Simulação hidrológica com modelo SWAT na Bacia Menino
Deus I, representativa da Mata Atlântica. In I Simpósio de Recursos Hídricos do Sul-Sudeste. Curitiba –
Paraná. Brasil.
Pereira, J. C. D., Silva, F. G. B., Dupas, F. A., Paes, F. S. 2010. Estudo de perda de solo na bacia
hidrográfica do ribeirão da Cachoeirinha-MG, com a finalidade de obter estimativas aproximadas para o
município de Santa Rita do Sapucaí-MG. In IX Encontro Nacional de Engenharia de Sedimentos. Brasília –
Distrito Federal. Brasil.
Pereira, J. M. G., Eid, N. J., Campana, N. A. 2005a. Aplicação de modelo matemático complexo de
simulação com base em dados disponíveis. In XVI Simpósio Brasileiro de Recursos Hídricos. ABRH. João
Pessoa – Paraíba. Brasil.
Pereira, J. M. G., Campana, N. A., Eid, N. J. 2005b. Processamento de dados para utilização de modelo
matemático de simulação à bacia do rio Jiquiriçá - Bahia. In XVI Simpósio Brasileiro de Recursos Hídricos.
ABRH. João Pessoa – Paraíba. Brasil.
Prado, T. B. G., Moraes, J. F. L., Adami, S. F. 2005. Evolução do uso das terras e produção de sedimentos na
bacia hidrográfica do rio Jundiaí-Mirim. In XVI Simpósio Brasileiro de Recursos Hídricos. ABRH. João
Pessoa – Paraíba. Brasil.
Rodrigues, D. F. and Reis, R. S. 2007. Aplicação do modelo SWAT na estimativa da produção de
sedimentos na bacia hidrográfica do rio Coruripe / AL. In XVII Simpósio Brasileiro de Recursos Hídricos.
ABRH. São Paulo – São Paulo. Brasil.
Roloff, G., Bley Jr, C., Souza, M. L. P. 2007. Modelo SWAT: Potencial de uso como auxiliar na gestão de
pequenas bacias hidrográficas sem dados históricos. In XVII Simpósio Brasileiro de Recursos Hídricos.
ABRH. São Paulo – São Paulo. Brasil.
Santos, I., Andriolo, M. V., Gibertoni, R. C., Kobiyama, M. 2010. Use of the SWAT model to evaluate the
impact of different land use scenarios on discharge and sediment transport in the Apucaraninha River
watershed, southern Brazil. In Symposium of the International Commission on Continental Erosion –
Sediment Dynamics for a Changing Future. Warsaw. Poland.
Santos, I., Andriolo, M. V., Gibertoni, R. C., Pereira filho, D. L. B. 2005. Projeto CGER 036 Desenvolvimento de modelo de avaliação da qualidade de água e transporte de sedimentos para pequenas e
médias centrais hidrelétricas: Modelagem da erosão e transporte de sedimentos. 91. LACTEC – Instituto de
tecnologia para o desenvolvimento. Curitiba-PR. Brasil.
2011 International SWAT Conference & Workshops
Santos, I., Andriolo, M. V., Gibertoni, R. C., Pereira Filho, D. L., Nishida, E., Silva, R. C. 2006. Simulação
com modelo SWAT das condições hidrológicas e de produção e transporte de sedimentos na bacia do rio
Apucaraninha – PR. In VI Simpósio Nacional de Geomorfologia – Geomorfologia Tropical e Subtropical:
processos, métodos e técnicas. Goiânia – Goiás. Brasil.
Schultz, G. B., Souza, R. M., Santos, I. 2010. Simulação da produção e transporte de sedimentos na bacia do
Altíssimo rio Negro com o modelo SWAT. In IX Encontro Nacional de Engenharia de Sedimentos. Brasília
– Distrito Federal. Brasil.
Silva, F. G. B., Minoti, R. T., Reis, J. A. T., Tsuhako, E. M., Angelotti Netto, A.Cestana, S. 2004. Simulação
da produção de sedimentos em uma sub-bacia hidrográfica experimental com o modelo SWAT. In VI
Encontro Nacional de Engenharia de Sedimentos. ABRH. Vitória – Espírito Santo. Brasil.
Silva, F. G. B., Neves, F. F., Crestana, S. 2005. Avaliação da perda de solo em sub-bacias hidrográficas da
região de Descalvado-SP a partir do modelo AVSWAT. In XVI Simpósio Brasileiro de Recursos Hídricos.
ABRH. João Pessoa – Paraíba. Brasil.
Silva, F. G. B., Neves, F. F., Crestana, S. 2006a. Simulação de parâmetros de qualidade de água em bacias
hidrográficas com o uso do modelo AVSWAT. In VIII Simpósio Ítalo-Brasileiro de Engenharia Sanitária e
Ambiental. ABES. Fortaleza – Ceará. Brasil.
Silva, F. G. B., Pioltine, V., Minotti, R. T., Crestana, S., Dupas, F. A. 2006b. Estimativa preliminar de perda
de solo com o uso do modelo AVSWAT para bacias da região de Luís Antônio – S.P. In VII Encontro
Nacional de Engenharia de Sedimentos. ABRH. Porto Alegre – Rio Grande do Sul. Brasil.
Souza, A. M. L., Castro, N. M. R., Kaufmann, V., Vitorino, M. I. 2010. Estimativa dos fluxos de calor a
partir de imagens orbitais para aplicação na modelização hidrológica. In XVI Congresso Brasileiro de
Meteorologia. Belém – Pará. Brasil.
Souza, R. M., Santos, I., Kobiyama, M. 2009. O modelo SWAT como ferramenta para a gestão de recursos
hídricos: avaliação de vazões de outorga na bacia hidrográfica do altíssimo rio Negro. In XVIII Simpósio
Brasileiro de Recursos Hídricos. Campo Grande – Mato Grosso do Sul. Brasil.
Trindade, P. B. C. B., Coutinho, L. F. N., Caiado, M. A. C., Heatwole, C. D. 2010. Evaluation of nutrient
modeling by SWAT under tropical conditions. In 21st Century Watershed Technology: Improving Water
Quality and Environment Conference Proceedings. Universidad EARTH, Costa Rica.
UFBA. 2004. Gerenciamento dos recursos hídricos do semi-árido do estado da Bahia: Volume II –
Subprojeto enquadramento de rio intermitente – estudo de caso do rio Salitre. Departamento de Engenharia
Ambiental. Salvador-BA.
Uzeika, T. 2009. Aplicabilidade do modelo SWAT (Soil and Water Assessment Tool) na simulação da
produção de sedimentos em uma pequena bacia hidrográfica rural. MS theses. Porto Alegre, Rio Grande do
Sul: Universidade Federal do Rio Grande do Sul, Programa de Pós-graduação em Recursos Hídricos e
Saneamento Ambiental.
Xavier, F. V. 2009. Contribuições metodológicas ao estudo da produção e distribuição espacial de
sedimentos na bacia hidrográfica do rio Manso utilizando o modelo AVSWAT. MS theses. Cuiabá, Mato
Grosso: Universidade Federal de Mato Grosso, Instituto de Ciências Exatas e da Terra.
Download

The use and results of the Soil and Water Assessment Tool in Brazil