Even though the concept was restricted to planning, it is possible to find distinct models of Decision Support Systems (DSS) for hydric resources. Mohaneed et al. [3] states that Decision makers usually take their decisions based upon economical and technical analysis, considering the available alternatives. “Traditional understanding of Decision Support Systems (DSS), as tools predominantly developed by water resources experts and for the same experts, is evolving towards viewing DSSs as potential platforms for knowledge exchange among users belonging to a much broader user base, which includes the experts, the involved stakeholders and the decision makers.” Jonoski and Popescu [4] .

The mathematical optimization and simulation models have been used in studies focused on the planning of hydric resources since the decade of 1960. They are essential for interpreting and trying to predict the behavior of water bodies. Several researchers have proposed different tools, with different functionalities and capabilities, to assist the decision-making process. These tools can be mathematical models, geomatics systems, performance evaluation systems and others, and they can be used for several challenges related to water planning and management to help the decision-making. For instance:

・ Zhang et al. [5] proposes a web-based watershed management spatial decision support system based on a Geographic Information System to evaluate different scenarios for watershed planning and management;

・ Kronaveter et al. [6] proposes a Negotiation Support System applied to the management of conflicts involving the use and management of water resources;

・ Pearson et al. [7] develops a decision support framework that assists managers in the urban water industry;

・ Zhang et al. [8] proposes a multi objective decision/bargaining model based on the “satisfaction principle” which is developed for inter-basin water transfer system decision-making;

・ Purkey et al. [9] presents an overview of decision-making processes ranked based on the application of a 3S: Sensitivity, Significance and Stakeholder support;

・ Ana and Bauwens [10] presents a review of selected sewer asset management decision-support tools, which are grouped according to their functionalities and capabilities, and describes the concept behind each one;

・ Fanghua and Guanchun [11] discusses how to apply a fuzzy multi-criteria group decision-making (MCDM) model to watershed ecological risk management, presenting a case study of the Three Gorges Reservoir Area in China.

This knowledge may be employed for determining areas in which dams can be constructed and where energy can be captured and generated. In addition, it may help define the granting of rights to the use of water resources⁴, since the discharge data of the historic series are used by the government agencies which manage hydric resources in order to establish what they call reference discharges, making it possible to determine the limit discharge that can be removed from a water body which can be used in different manners due to a granting of rights.

Nevertheless, there are limitations for including non-quantitative aspects, which reduce the capacity that hydrologic models have in aiding some aspects of the planning, such as: the identification of areas that may potentially affect the resources, the identification of who is involved in each activity and the gaps of space and time coverage when gathering hydrometeorological data, which may lead to inconsistencies in the historical series.

In this case, mathematical modelling and the Decision Support System (DSS) tools are components of a wider Management System, which must provide information in a dynamic way and integrate technical and administrative routines, allowing for consultations, analyses and the elaboration of reports. In geohydroinformatics, the Management Systems based on DSS tools are Real-Time Control (RTC) systems, for they can manage a wider variety of information in the related database. In this context, the capacity of associating spatial and non-spatial data and the possibility of interacting with a mathematical model and more complex databases and, lastly, the capacity of spatializing the analysis with output graphical interfaces that are becoming increasingly more accessible make the Geographic Information Systems (GIS) an essential tool for the planning and management of hydric resources, as well as for other topics that are directly or indirectly associated with the conflicts and impacts regarding these resources.

Among the several definitions which were presented, the definition proposed by Calkins and Tomlinson, singled out by Christofoletti, emphasizes the potential of integrating technology with its stages and mentions that the GIS is a integrated set of programs (software) elaborated specifically for activities related to data manipulation. These activities include input, storage, recuperation and problems associated with data management, excluding the wide variety of descriptive and analytical processes.

In a more recent work, Burrough and Mcdonnell [12] state that a GIS is a powerful set of tools to collect, store and recuperate information by transforming and organizing data gathered from the real world into a particular set of goals.

Over the last 20 years, several authors have used GIS technology as a research tool in order to analyze and solve environmental problems. They had different approaches that can be divided into: environmental monitoring and land use: Papastergiadou et al. [13] / Bhattarai et al. [14] ; adaptation studies and modeling of extreme climate events: Amaral et al. [15] /Adger [16] /Pahl-Wostl [17] /Sena et al. [18] [19] ; modeling and monitoring the quality and pollution of water: Tsihrintzis et al. [20] /Mantzafleri et al. [21] /Singh et al. [22] ; and specially hydrologic modelling and studies applied to watersheds: Venkata et al. [23] /Bhalla et al. [24] /Coskun et al. [25] /Correia et al. [26] /Tsihrintzis et al. [27] / Jasrotia et al. [28] /Panagopoulos et al. [29] /Wiek and Larson [30] .

The following organogram demonstrates in a simplified manner how the three operational levels of the SIG are organized as a DSS tool (Figure 1).

In the structure level, the databases are fed and organized, including all information that may be georeferenced. The interface level includes the visual presentation of the information, and several layers of information can be overlayed according to the selected topics and the specific goal of analysis. Lastly, in the user level there are analyses based on the alfanumeric information and other information provided by the layers.

The Amazon River Basin is the largest hydrographic basin in the world, with an area of

6,100,000 km². Its drainage area includes parts of Brazil, Bolivia, Colombia, Ecuador, Guyana, Peru, Venezuela and Suriname (ACTO). The Amazon Basin is extremely important for the climate dynamics and the hydrologic cycle of the planet. It has approximately 16% of the sweet surface water of the planet and, consequently, it plays a significant role on the rainfall regime and evapotranspiration of South America and the world. The basin is also one of the most humid regions on Earth, with average precipitation between 2300 and 2460 mm・year [2] [31] .

The rivers that compose the Amazon Basin can be hierarchized and grouped into sub-basins by employing the Otto Pfafstetter methodology―which is adopted officially by Brazil to define the national hydrographic division and also by the United States Geological Survey (USGS) to elaborate the global map of hydrographic basins―to define the main rivers and the influence of the relief in each country of the Amazon Cooperation Treaty Organization (ACTO)⁵.

Figure 2 shows the hydrographic division of the Amazon River Basin and its sub- basins, with levels 1 and 2 respectively according to the criteria established by Otto Pfafstetter. The main sub-basins of the Amazon Basin are the ones from the following rivers: Solimões, Madeira, Negro, Xingu and Tapajós.

Regarding the area called Amazon Region, with South America as a reference, some topics must be discussed. Firstly, it is necessary to point out that, within this context, some concepts and ideas must be debated according to the suggestions of Corrêa [32] , for the use of the term region is not done in a standardized way; it is very complex. It is possible to say that there are some difficulties in identifying the limits of the region, especially because there are several criteria that can be considered. These criteria even change from country to country―each country uses different criteria to define their “Amazon” regions (Figure 3 and Figure 4).

In this context, the Amazon and the topics related to it are discussed so widely nowadays that, at least in the common sense, the region is a spatial entity without any geographic delimitations. When the Amazon is mentioned, it is common to overlook the different spatial outlines which can be used to define it. Nevertheless, in theoretical, political and especially quantitative terms, it is important to define the territorial matrix used as reference in order to discuss any theme related to the Amazon.

In this sense, it is possible to define at least four territorial outlines for the Amazon. The first one, determined by the hydrographic basin, includes an area of 7 million・km², 4.8 of which are in Brazilian territory. The second possibility is to consider the vegetation cover, which is estimated to be 40% of the total area of the basin. Another possible outline is to consider the regional division of the country―in this case, the Amazon would include the North Region, which is composed of the states of Acre, Amapá,

Amazonas, Rondônia, Roraima, Pará and Tocantins, and part of the Central-West Region (the state of Mato Grosso). The total area of theses states is 4.7 million・km², which indicates that this political-administrative division was based on natural geographical criteria, such as the area of the Amazon Basin which is inside the country. The last alternative is to consider the “Legal Amazon”, which was created in 1955 by a legal provision (Law 1806, 1953) and includes an area of 5,217,423 km². It was created to help the economic and territorial planning defined by the Amazon Valorization Plan approved two years before.

According to Eva and Huber [33] , although this situation does not create any problems on a national level, it may create issues for integrating data and statistics (in content or spatial dimension) on a regional level. Therefore, studies about the regionalization of the Amazon must also be done.

The Basic Geographic Information System of the Waters of the Amazon Basin on the scale of 1:5,000,000 was named GeoAmazonas (Geographic Information System of the Amazon River Basin). The initial versions (Beta version in December 2006, with the final version being delivered in February 2007) were built and updated using the software SIG ArcView 3.2a. This version was rebuilt and updated in 2011, using the software SIG ArcGIS 9.3.

The GeoAmazonas GIS was built following these stages:

1) Bibliographical and Cartographic Survey;

2) Conceptual Modelling of Data;

3) Data Acquisition and Metadata Creation;

4) Field Research;

5) Building and Spatial Integration of the Data (Implementation):

5.1) Treatment of Graph and Tabular Data;

5.2) Tests and Error Assessment;

6) Interface with the User: Production and Assessment;

7) Storage, Consultations and Analyses.

・ Stage 1 (Bibliographical and Cartographic Survey)

The first stage included activities needed for the development of the proposed product. These activities considered a wide survey of data and information sources, including graphic and tabular bases. Some of the institutions consulted and/or visited were: Servicio Nacional de Meteorología e Hidrologia (SENAMHI); National Water Agengy/Ministry of the Environment (ANA/MMA); Instituto de Hidrología, Meteorología y Estudios Ambientales (IDEAM); Consejo Nacional de Recursos Hídricos (CNRH); Guyana Water Authority/Hydraulic Research Division (GWA/HRD); Instituto Nacional de Recursos Naturales (INRENA); Ministry of Public Works/Hydraulic Research Division (MPW/HRD); Dirección General de Cuencas Hidrográficas/Ministerio del Ambiente y de los Recursos Naturales (DGCH/MARN); Management of the Ecological- Economical Zoning of the State of Acre; Brazilian Electricity Regulatory Agency (ANEEL); Mineral Technology Center (CETEM); CT-HIDRO―Water Resources Sector Fund; Brazilian Agriculture Research Corporation (EMBRAPA); National Indian Foundation (FUNAI); Global Land Cover 2000 Project (GLC, 2006)―Map of Land Use in South America; Retis Group (Federal University of Rio de Janeiro-UFRJ); Brazilian Institute of Geography and Statistics (IBGE); RADAM Project/IBGE; Amazon Protection System; Amazon Surveillance System; The Smithsonian Atlas of Amazon [34] ; Atlas mondial du développement durable [35] ; Federal University of Acre (UFAC).

・ Stage 2 (Conceptual Modelling of Data)

It was elaborated based on the bibliographical and cartographical survey, considering the topics and spatial outlines delimitated on the goals previously determined for the GeoAmazonas Geographic Information System. Subsequently, it was necessary to represent the structure of the information in the system, i.e., the types of data and how they were related to one another, as the following example shows (Figure 5).

The building of the GeoAmazonas GIS on this stage had two different approaches which were fundamental: one approach focused on the mapping the anthropic part and the other on the mapping of natural resources. The integration of both approaches using a GIS may lead, for instance, to an analysis of integrated vulnerability, representing the coexistence or spatial overlay of anthropic action and floodable areas or indigenous villages and deforestation. After the subsystems were created, the entities of each group were defined, as well as the main relationships between them.

・ Stage 3 (Data Acquisition and Metadata Creation)

After the conceptual model was created, the third stage involved the acquisition of data from the subsystems which are part of the GeoAmazonas GIS. They were related following the proposed conceptual model. Nevertheless, Stage 3 considered several cartographic data (digital cartographic bases, orbital imaging of medium spatial resolution, etc.).

Therefore, the system has information layers on the following themes:

1) Anthropic Action-fundamental modifications in land use in the Amazon Region, highlighting what occurs in the Brazilian part of the Amazon;

2) Floodable Areas;

3) Hydrographic Basins;

4) Main Cities;

5) Boundary Strip;

6) Hydrography;

7) Geographic Boundaries;

8) Ocean;

9) Relief;

10) Indigenous Land;

11) Land Use;

12) Vegetation;

13) Highways/Transportation roads;

14) Energy;

15) Mining;

16) Water: Supply/Sewage.

During Stage 3, the metadata were also produced. They are grouped according to the subsystems which were created.

・ Stage 4 (Field Research)

The field research is done in a area or areas of interest in the Amazon Region, with the goal of collecting and acquiring available data and information, as well as verifying if the database matches local reality. This stage was done when the GIS was being built and also when it was being updated. The following countries were visited: Bolivia, Brazil, Guyana, Peru and Venezuela.

・ Stage 5 (System Building and Spatial Integration of the Data Implementation)

Stage 5 was divided into two sub-stages. The first one (Treatment of Graph and Tabular Data) had the goal of gathering the surveyed data/information in order that integrated spatial analyses and thematic maps could be made. The second sub-stage (Tests and Error Assessment) compared the GIS, its bases and the results with the in loco reality.

・ Stage 6 (Interface with the User: Production and Assessment)

This stage is focused on the interaction the user has with the system. The way to access the system is through the GIS, which allows the user to communicate with the system by requesting consultations and viewing results.

・ Stage 7 (Storage, Consultations and Analyses)

Lastly, some files were organized, some consultations were made (for instance, the boundaries and the calculation of the area of the Amazon Basin inside Venezuelan territory were verified due to a request of the Environment Coordination of the ACTO) and analyses were done considering the outlines of the region, country or basin (for instance, spatial analysis regarding occupation and deforestation and spatio-temporal analyses regarding the evolution of the deforestation in the Acre River Basin). With this stage, a series of thematic maps was generated, which will be presented in the results.