Skip to main content
NCBI home page
Ambio logoLink to Ambio
. 2024 Aug 8;53(12):1722–1736. doi: 10.1007/s13280-024-02058-9

Long-term sustainability of the water-agriculture-energy nexus in Brazil’s MATOPIBA region: A case study using system dynamics

Minella Alves Martins 1,✉, David Collste 2, Francisco Gilney Silva Bezerra 1, Marcela Aparecida Campos Neves Miranda 1, André Rodrigues Gonçalves 1, Jocilene Dantas Barros 1, Manoel Ferreira Cardoso 1, Amanda Sousa Silvino 1, Taís Sonetti-González 3, Jean Pierre Henry Balbaud Ometto 1, Celso von Randow 1, Javier Tomasella 1, Ana Paula Dutra de Aguiar 1,2
PMCID: PMC11568102  PMID: 39115747

Abstract

The global demand for agricultural commodities has driven extensive land conversion to agriculture in Brazil, especially in the MATOPIBA region. This area encompasses the Rio Grande Basin, a major tributary of the São Francisco Basin that is known for expanding intensive irrigated agriculture and hydropower generation. However, recent data reveal declining precipitation and aquifer recharge, potentially exacerbating ongoing water and land conflicts. This study investigates the long-term sustainability of agricultural expansion amid the worsening water scarcity using a system dynamics model. Findings suggest that rising costs and decreasing profits due to irrigation water shortages may hinder the expansion of irrigated land. By 2040, the irrigation demand may remain partly unmet, while downstream flow and baseflow could decrease. Additionally, agricultural expansion will significantly raise energy demand, posing a developmental challenge. We suggest that ensuring the sustainability of the Rio Grande Basin depends on improved water management and exploring alternative energy sources to address existing constraints.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13280-024-02058-9.

Keywords: Agricultural expansion, Climate change, Irrigation demand, Sustainability, Water scarcity, Water management

Introduction

The agriculture sector represents about a third of the Brazilian Gross Domestic Product—GDP (CEPEA/CNA 2021). Brazilian grain production has increased from 38 million tons in 1975 to 239 million in 2020 (EMBRAPA 2018; Aragão and Contini 2021). During this period, the planted area doubled in size, demonstrating the advancement of agricultural techniques to improve production efficiency (EMBRAPA 2018).

In recent decades, the growth of agribusiness in Brazil has led to the expansion of agricultural land, through the conversion of forests and other natural habitats. The MATOPIBA region, covering the states of Maranhão, Tocantins, Piauí, and Bahia, stands out as the largest and most rapidly developing area of the agricultural frontier (Dionizio et al. 2020). The factors driving this growth include affordable land prices, suitable terrain for mechanized farming, government-subsidized credit, and improved infrastructure, all of which have attracted migrant workers (Santos et al. 2018). However, this agricultural expansion has been accompanied by a troubling trend: land seizures on untitled public lands (Calmon 2020), often linked to illegal deforestation (Valdiones et al. 2021). These activities have significant impacts on various aspects of society, the economy, and the environment.

The Southeast part of MATOPIBA is in a strategic area from the water resources perspective since it hosts the Rio Grande Basin, a crucial source for the São Francisco River. The São Francisco, in turn, is one of the most important rivers in the country, flowing across the Brazilian semiarid and six Federative States. The basin is critical to produce energy, food, and water for the region and country. Almost 60% of the drainage area of the Rio Grande Basin is underlain by the Urucuia Aquifer System—UAS. The availability of surface and groundwater resources was crucial for the development of intensive irrigated agriculture (Dionizio and Costa 2019; Marques et al. 2020). In the last 20 years (2000–2020) rainfed agricultural areas in Rio Grande Basin have grown by 82%, while irrigated areas by 140%. Simultaneously, deforestation has increased by 47% (Souza et al. 2020). Following this trend, the water used for agriculture has increased significantly, surging from 0.7 m3 s−1 in 1980 to 30 m3 s−1 in 2020. Remarkably, over 90% of the total water consumption in the basin was dedicated to irrigation purposes (ANA 2021a).

Despite the current abundance of water resources, the irregularity and seasonality of rainfall regimes make irrigation indispensable for agricultural production. According to Pousa et al. (2019), rainfall has decreased by up to 12% since 1980. Hofmann et al. (2023) have similarly identified a significant reduction in both the quantity and frequency of rain in the Cerrado region, especially during the dry season. In addition, future climate change scenarios indicate a decrease in precipitation, ranging from 100 to 250 mm Year−1 by 2050, concomitant with an increase in evapotranspiration (Martins et al. 2019; Pousa and Costa 2023). Data from the Groundwater Information System (CPRM 2022) indicate a decrease in the water levels of the UAS over the period 2011–2019 and future projections suggest a decrease in UAS recharge of up to 45% (Pereira et al. 2022).

This scenario highlights the importance of adequate knowledge of local water resources, combined with appropriate management (Eger et al. 2021). In addition, the pollution from agrochemicals and the use of small hydropower plants to support irrigation systems have compromised clean water access for riverside communities, resulting in conflicts over land and water resources (CPT 2022).

Another critical factor that impacts agricultural expansion is the availability of electrical power for irrigation infrastructure. Challenges associated with power transmission infrastructure have constrained the expansion of irrigation systems. As a result, small hydropower plants are being installed in a manner that has substantial impacts on both the local population and environment, including diverting the river and diminishing access to water for the population, as reported by the São Francisco River Basin Committee (CBHSF 2021). Considering these multiple users and the variety of sectoral interests within the Rio Grande basin, which involve economic, social, and environmental factors, the main goal of this study is to assess the long-term sustainability of agricultural expansion associated with water scarcity under climate change.

The sustainability principles used in the Rio Grande basin are based on a holistic approach that includes supporting functioning ecosystems as well as riverside communities’ water requirements, maintaining river flow for ecological health, sustaining energy supply, and ensuring the long-term sustainability of agricultural production. This approach seeks to harmonize the dynamic interactions between water use, energy production, agricultural activities, and the conservation of natural ecosystems to achieve a balanced and sustainable coexistence.

Prior studies have explored how climate change and intensive irrigation affect water availability and the impacts of changes in land use on soil and water conservation in the region (for instance, Dionizio and Costa 2019; Pousa et al. 2019; Pimenta et al. 2021; Pousa and Costa 2023). However, none of these studies have taken a comprehensive approach to explore the feedback between water resources, agriculture, and energy. The methodological approach we use is based on system dynamics modeling (Sterman 2000). System dynamics (SD) is an approach for analyzing complex feedback-driven systems in which nonlinearity usually plays a key role. SD is commonly employed in situations where formal analytical models are unavailable or challenging to develop. In such cases, simulations can be constructed by interconnecting multiple processes (see, e.g., Collste et al. 2017). This modeling approach is in this case particularly useful as a precise mathematical representation is difficult to explicitly formulate (Sušnik et al. 2012).

Regarding water resources, system dynamics models have been successfully applied, for instance, to assess problems related to agricultural water management (Fernald et al. 2010), water scarcity (Sušnik et al. 2012), and water security management (Zomorodian et al. 2018). According to Phan et al. (2021), there is great potential for using SD to provide a holistic view of feedback processes and interdependencies between hydrological, social, economic, and environmental processes for decision-making in water resource management systems.

To assess the long-term sustainability of the Rio Grande Basin, this study aims to investigate the following research questions: (a) What are the possible constraints for the expansion of irrigated agriculture? (b) How do existing water usage patterns and climate change influence baseflow and downstream flow? (c) What is the energy consumption associated with irrigated agriculture, and what are the alternative strategies to ensure a sustainable energy supply to sustain economic activities?

Within this framework, the study seeks to contribute to the identification of current non-sustainable trends, understanding their root causes, and assessing appropriate management measures comprehensively, supporting strategic decision-making.

The key variables identified in the conceptual model, serving as primary indicators for sustainability, include downstream flow and base flow, as well as irrigation shortage and energy deficit. The time horizon evaluated spans from 2010 to 2040.

The paper is structured as follows: Sect. "Materials and methods" presents the Methodological approach, describing the model development and data utilized. Sect. "Results" delves into the results obtained. Following that, Sect. "Discussion" engages in a discussion elucidating how these results contribute to addressing our research questions. Finally, Sect. "Conclusion" summarizes the main conclusions drawn from this study.

Materials and methods

Case study area

Located in the southeastern part of MATOPIBA (Fig. 1), the Rio Grande Basin is considered one of the most important of the region due to its abundant water resources. It is a subbasin of Brazil's significant São Francisco River Basin, covering around 76 000 km2 and overlaying more than a third of the Urucuia Aquifer System—UAS. The UAS is mostly an unconfined groundwater system, which regulates the seasonal and interannual variations of the Grande River basin (Pousa et al. 2019) and accounts for at least 80% of the baseflow (Gonçalves et al. 2016).

Fig. 1.

Fig. 1

Open in a new tab

Study area and the main land use/cover (A) and Rio Grande River Basin in the context of MATOPIBA, São Francisco River basin and Brazil (B)

The Rio Grande River basin, located in the transition between the savanna (Cerrado) and dry forest (Caatinga) biomes, is home to approximately 470 thousand inhabitants (IBGE 2022). The topography is diverse, encompassing flat plains, plateaus, gently undulating terrain, wavy landscapes, river valleys, and elevated areas (Gonçalves and Chang 2017). While subsistence farming predominates in low-lying areas, intensive agricultural expansion is dominant to the west of the basin, due to the gentle upland topography that facilitates mechanization (Pimenta et al. 2021).

The regional climate is tropical humid (Aw according to the Köppen climate classification) (Alvares et al. 2013), featuring distinct rainy (October to March) and dry (April to September) seasons, with average annual precipitation ranging from 700 to 1300 mm (Simões et al. 2018), but characterized by irregular interannual rainfall patterns.

The UAS and Rio Grande Basin rely on rainwater infiltration in flat and elevated relief areas, with sandy latosols playing a significant role due to their porosity and permeability (Gaspar and Campos 2007). Since recharge areas correspond to the regions with more vigorous agricultural expansion, concerns about how land cover changes and the growing number of withdrawal wells might be affecting aquifer recovery have been raised in recent decades.

Model development

To build a system dynamics model of the water-agriculture-energy nexus in the Rio Grande basin, the Vensim software was used. System dynamics modeling involves various components, known as stocks (accumulations, also referred to as level variables), flows (also referred to as rate variables), and auxiliary variables (Binder et al. 2004). Mathematically, stocks are integrals and flow derivatives. The flows represent functions that typically facilitate the movement of materials in and out of these stocks, whereas the auxiliary variables act as converters influencing the rates of material flow. These components are interconnected through causal connectors, which facilitate the transfer of information within the model. There is an emphasis on circular causality, which is feedback loops that generate the dynamic behavior of systems (Forrester 1969). Feedback loops can be either reinforcing (positive feedback) or balancing (negative feedback) (Sterman 2000).

System dynamics modeling typically comprises four iterative steps: problem scoping, model conceptualization, model implementation, and test and scenario analysis (Sterman 2000). The model conceptualization and implementation are described in Sect. "Model conceptualization and implementation for the Rio Grande Basin," while Sect. "Data required for model calibration and simulation" describes the data required for model calibration and simulation. Finally, Sect. "Scenario and model analysis under the baseline scenario" presents some points to be addressed, including scenario and future projection analysis.

Model conceptualization and implementation for the Rio Grande Basin

We developed a conceptual model based on available empirical and theoretical information as well as dialogs with local stakeholders and communities, which identified key issues, causes, and interconnections in the study area. (The stakeholder dialogs used an approach that is expanding on Collste et al. 2023.) The primary concern highlighted during the meetings was the water exploitation limits of the agricultural sector, which are a source of conflicts among users and impact the water availability of the São Francisco River.

This conceptual model was developed as a causal loop diagram (CLD), illustrating the relationship between the expansion of irrigated land, agricultural water demand, energy supply, and water resource availability in the Rio Grande Basin.

By analyzing the CLD for the Rio Grande Basin, we identified the proposed key variables within the system: expansion of irrigated land, water and energy demand, and supply for irrigation. This process facilitated the development of a conceptual model that represents the subsystems comprising the Rio Grande Basin System (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Conceptual model summarizing the subsystems and key relationships within the Rio Grande Basin System. Basin System

As depicted in Fig. 2, the model considers that the expansion of irrigated land in the agricultural sector is closely influenced by the interaction between investment, costs, and profits, shaped by the extent of irrigated land, and the intensity of irrigation, which, in turn, is determined by crop water requirements. For instance, the decision to expand irrigated land in agriculture is driven by financial considerations for investments, including costs and profits. This decision-making process involves considering the initial investment required for irrigation infrastructure, ongoing costs such as maintenance and water supply, and the potential profits from increased agricultural production. Additionally, the extent of irrigated land, referring to the total area under irrigation, directly impacts its availability for agricultural use, implying higher production but also higher investment and operational costs. The intensity of irrigation, determined by factors such as crop water requirements, irrigated areas, and climate conditions, plays a crucial role in determining the amount of water used for irrigation and the withdrawal from rivers and groundwater. In this region, the primary crops are soybeans and cotton, and the implementation of irrigation facilitates the cultivation of at least two crops annually: soybeans grown from October to January, followed by cotton harvested in July. Beyond economic factors, climate change also has a significant impact on crop water demand and water supply, directly influencing the energy demand for irrigation. However, the region has experienced a persistent energy deficit, mainly attributed to the lack of a suitable electrical network to support irrigation expansion. Therefore, by considering the energy demand of irrigation systems, the total extent of irrigated land, and the existing energy supply, it is possible to make inferences about the expected energy supply deficit.

The relationships in the model propose that decisions made by stakeholders based on the availability of land suitable for irrigation and economic returns influence the expansion or retraction of irrigated land, and consequently, the use of water in agriculture and the need for energy. In turn, water availability/water shortage in the basin influences agricultural productivity and, consequently, the profits obtained.

Data required for model calibration and simulation

We have chosen four main variables as our reference modes: Irrigated land (km2), Downstream flow (km3 Year−1), Baseflow (km3 Year−1), and Water demand for irrigation (km3 Year−1), see Fig. 3.

Fig. 3.

Fig. 3

Open in a new tab

Reference modes of behavior

We selected the period from 2011 to 2020 for model calibration considering the reference modes of Fig. 3. The simulation time step corresponded to a 1/16 calendar year (using time steps that are powers of 2 is common practice as it aligns with the binary representation used in computer models). However, the modeling primarily considers longer-term trends and annual changes and does not incorporate seasonal fluctuations. Additional information on basic data in this study can be found in the Supplementary Material.

Scenario and model analysis under the baseline scenario

Using the calibrated model, future projections were carried out based on a business-as-usual scenario. This scenario aims to account for current causal relationships and their strengths without significant changes. The evaluation of climate change incorporated rainfall projections up to 2040 to assess long-term sustainability. The main variables considered in the model that provide indications of basin sustainability are downstream flow and base flow, representing the available water in the basin to support all uses, and irrigation shortage and energy deficit, representing the potential impacts of unsustainable activities in the basin.

To propose an alternative energy scenario, we conducted a thorough assessment to quantify the energy deficit resulting from the expansion of irrigation in the region. This assessment included a discussion of both the advantages and limitations of the transition to a cleaner energy supply and replacing hydropower generation with photovoltaic sources.

Further details about the future climate scenario used in this simulation as well as the model documentation are presented in the Supplementary Material.

Results

Rio Grande Basin model structure

The Rio Grande Basin Model could be studied as comprising three subsystems: (A) the socioeconomic subsystem, which considers the socioeconomic factors influencing land expansion; (B) the water subsystem, responsible for tracking inputs and outputs in the Rio Grande Basin; and (C) the energy subsystem, which accounts for demand and alternative supply sources. Figure 4 shows the structure and feedback accounted for by the model.

Fig. 4.

Fig. 4

Open in a new tab

Stock-flow diagram of the dynamic of agricultural lands, water availability, and energy supply in the Rio Grande Basin. Rectangles represent the stock variables, pipes represent the flows, and valves represent the rate of change per unit of time of stock. The auxiliary variables act as converters, whose value can change or be derived from other parts, the arrow represents the influence of auxiliary variables, and the clouds represent sources or sinks showing the boundaries of the model. Balancing feedback loops are denoted by “B,” while reinforcing feedback loops are indicated by “R”

The Socioeconomic subsystem is composed of two stocks: available land for irrigation (km2) and Irrigated land (km2) linked by the irrigated land expansion (km2 Year−1). The primary factors influencing the rate of irrigated land expansion are the costs of land (US$ km−2) and the investments (US$ km−2).

Costs depend on available land for irrigation; more land available means lower costs. This reflects the assumption that the easiest and cheapest-to-irrigate lands are developed first. This relationship results in balancing feedback: as available land becomes scarcer, costs rise, inhibiting future expansion.

Investments are linked to profits. Profits depend on both total irrigated land (more land means more crop yields and hence profit) and water supply (water scarcity reduces profits as a consequence of constraints on yields). A reinforcing feedback loop exists, where expanding irrigated land leads to increased investments, further driving irrigation expansion.

The fulfillment of irrigation demand (km3 Year −1) is dictated by the irrigation water shortage (km3 Year −1), which measures the gap between irrigation demand and potential surface or groundwater supply. This relationship forms a balancing feedback loop, where converting more land to irrigation increases demand, nearing water supply limits, elevating the risk of inadequate supply, with reducing profits consequently.

Total irrigation demand hinges on three key factors: irrigated land area, irrigation intensity (km3 km-2 Year−1), and rainfall (km3 Year−1). This demand is the crucial link between the model’s socioeconomic and water subsystems.

The Water subsystem depends primarily on rainfall (km3 Year−1), which comes into the system through runoff (km3 Year−1) and soil infiltration (km3 Year−1). Runoff contributes to surface water, while infiltrated water replenishes the aquifer (Groundwater/Aquifer).

There are three main water stock variables: Upper soil water (Km3), Groundwater/Aquifer (km3), and Surface water (km3). Flow variables include deep percolation (km3 Year−1) connecting Upper Soil Water and Aquifer/Groundwater, and baseflow (km3 Year−1) linking Aquifer/Groundwater and surface water.

Stock variables change due to rainfall and water withdrawals, which include evaporation, surface irrigation withdrawal, and other uses, influenced by population and rainfall. Downstream flow represents the water available for riverside communities and the São Francisco River, including both surface and groundwater.

Surface water withdrawals are regulated by a water license, while groundwater withdrawals prioritize human and animal consumption over irrigation based on the aquifer drawdown level.

Finally, the Energy subsystem is impacted by the expansion of irrigated land, resulting in increased demand for electric energy. As a result, the model estimates the surplus energy requirements beyond the 2019 energy supply, based on the expansion of irrigated land and the energy needed to support the irrigation system.

Baseline model performance

The results of the baseline simulation are depicted in Fig. 5, where simulations are compared with observations for the period 2011–2020. Note, however, that in system dynamics model validity is not only a matter of the variables’ behavioral matching (behavioral validity), as much emphasis is given to finding the causal structure of the model credible (structural validity) (Barlas 1996). In line with this, a few suitable validation tests were performed. These are discussed in the Supplementary Material.

Fig. 5.

Fig. 5

Open in a new tab

Rio Grande model validation

The model effectively captures the overall trend, although it tends to overestimate downstream flow (Fig. 5A) and baseflow (Fig. 5B). In contrast, the representation of irrigated land (Fig. 5C) aligns well with empirical data. Simulating irrigation demand (Fig. 5D) presented the most significant challenge for the Rio Grande Model, but it approximated the average demand and overall trend for the entire period accurately. While observations indicate an average of 1.76 (km3 Year−1), the model simulated an irrigation demand of 1.66 (km3 Year−1). It is worth noting that the available empirical data for irrigation demand were only obtained every ten years for surface sources and from 2013 to 2020 for groundwater sources, which may introduce some annual estimation errors into our reference values. Overall, the Rio Grande Model effectively captures the primary dynamics of the system and can be valuable for understanding the implications of various policy scenarios in the study area.

Rio Grande model analysis under baseline scenario with climate change

Figure 6 illustrates the trends in the main variables assessed in the Rio Grande Model for the period 2010–2040. Figure 6A depicts the increasing trend of irrigated lands, driven by the rise in irrigation land expansion. Irrigated lands ranged from 860 km2 in 2011 to 1516 km2 in 2040. The model indicates a tendency toward stagnation from around mid-2025, owing to a sharp decrease in land conversion.

Fig. 6.

Fig. 6

Open in a new tab

Baseline scenario with climate change for the main variables in the Rio Grande model: A Agriculture variables, B Economic variables, C Infrastructure and energy needs, D River flow and water for irrigation, E Surface and groundwater withdrawals

The model hypothesizes that the increase in irrigated land expansion is fueled by the low cost per unit of conversion and sound investments (Fig. 6B). As costs escalate, investments tend to decrease, as do profits.

Due to land expansion, the number of pivot irrigation systems in the simulation is increasing from 816 in 2011 to more than 1400 in 2040 to cover the newly irrigated land, as shown in Fig. 6C. However, this leads to an increase in energy demand to about 350 million kWh Year−1 in 2040, surpassing the current electric energy supply. The simulated energy deficit increases to 40 million kWh Year−1 until 2040, an increase of about 40% compared to the baseline, which would require expanding the current energy grid or exploring alternative power sources.

Regarding downstream flow and baseflow, Fig. 6D illustrates that both variables demonstrate a gradual decline. In the simulation, downstream flow decreases to 6.6 km3 Year−1 by 2040, while baseflow decreases to 5.0 km3 Year−1. Comparing the last ten years to the baseline (2011–2020) it represents a reduction of about 15% and 13%, respectively. Additionally, there is a clear increase in the simulated irrigation demand, reaching 2.9 (km3 Year −1) in 2040 but experiencing some peaks at nearly 4.0 (km3 Year−1). However, there is a corresponding increase in irrigation water shortage, assuming that the ability to meet irrigation requirements could decline in the future due to limitations in surface water availability and groundwater extraction. Over the last 15 simulated years, 30% of the total irrigation demand is unmet.

Finally, Fig. 6E illustrates the withdrawal patterns of surface and groundwater for various purposes. Surface water irrigation reaches its limit early in the simulation, making it necessary to shift toward groundwater irrigation. Given that most of the water in this basin is allocated for irrigation, other withdrawal categories from the surface remain relatively stable. Nevertheless, withdrawals from groundwater, primarily intended for human and animal consumption, exhibit significant fluctuations driven by the interannual variation in rainfall.

Discussion

The Rio Grande model illustrates how the expansion of irrigated land slows down in response to a limit, in this case, the increase in the cost per unit of agricultural area. This rise in costs is attributed to the reduced availability of land and the consequences of irrigation water shortages. By 2040, the model assumes that irrigated land will cover about 1500 km2, significantly less than the available agricultural land suitable for irrigation, as projected by the National Water Agency (~ 2700 km2) (ANA 2021b). The Agency’s expansion projections account for multiple factors, including water availability, crop requirements, environmental restrictions, infrastructure, logistics, land tenure, and agricultural suitability based on soil and topography.

In contrast, our approach involves a dynamic response, enabling year-by-year observation of proposed feedback effects, limiting the expansion of irrigated land as the potential consequences of not meeting crop requirements due to water shortages are considered. In the model, this affects costs, profitability, and subsequently, investments in the following years. Incorporating such feedback is crucial when simulating changes in land use and land cover, as it reflects how economic factors interplay with the landscape (Pimenta et al. 2021).

Overall, our study aligns with previous research that highlights a reduction in the hydrological production capacity of the basin and an increase in water demand primarily driven by the expansion of irrigation. However, our study revealed that areas potentially suitable for irrigation expansion projected by ANA for 2040 may have overestimated available water resources. This discrepancy is likely since ANA’s future scenarios did not consider the adverse effects of climate change or the feedback mechanisms considered in our model. These mechanisms could make the expansion of agricultural lands less attractive due to worsening water shortages, which could ultimately reduceprofits. Furthermore, this issue could be exacerbated if an energy deficit were incorporated as a feedback mechanism in the model.

Multsch et al. (2020) also identified limitations in expanding irrigated areas, particularly in the eastern Cerrado, where the Rio Grande Basin is situated. They suggested a substantial reduction in irrigated areas in this region and recommended converting only a small portion of current rainfed areas to irrigation to avoid reaching critical water scarcity levels.

Our results highlight that the main constraint to sustaining irrigated land expansion could be the shortage of water for irrigation. Starting in 2025, 30 to 40% of the irrigation demand may not be met as a consequence of the model assumptions. This is due to restrictions on surface and groundwater withdrawals. Moreover, surface withdrawal reached its limit in 2014, in agreement with the study of Santos (2016), indicating that the Rio Grande basin is already operating at its maximum capacity in terms of water resource usage. Consequently, there is no room to issue new authorizations (water permits) for the use or withdrawal of water for any additional activity.

In our model, aquifer water withdrawals prioritize human and animal consumption. When the aquifer level falls below a specific threshold, irrigation withdrawals are suspended to ensure water availability for human and animal needs, aligning with Brazilian water law. Despite a recent regulatory directive issued by the environmental agency establishing that the maximum extraction authorized from wells drilled in the Urucuia aquifer should not exceed 9000 m3/day (INEMA 2022), compliance with this regulation is poor due to the lack of fiscalization, as evidenced by the water levels of some wells monitored by the Groundwater Information System (CPRM 2022).

In addition, Reis (2018) demonstrated that from 2006 to 2016, the total surface water permits issued by INEMA (the environmental agency responsible for managing environmental and water resources in Bahia) exceeded about 70% of the limit established by the São Francisco River Basin Hydrographic Plan. This allowed for the short-term fulfillment of irrigation demands and the subsequent expansion of irrigated lands. However, in the long run, this practice could jeopardize the ability to meet the societal needs mandated by law.

Furthermore, Marengo et al. (2022) emphasized that the conversion of land for agribusiness expansion, combined with climate change in the MATOPIBA region, may have worsened severe drought conditions over the last decade. This highlights the growing climate challenges in this region, which have significant implications for the Brazilian economy and global food security.

Even though our simulation adhered to the water limits for surface water withdrawal, it showed a reducing trend in both baseflow and downstream flow, despite only a slight decrease in rainfall. According to the model simulation, by 2040, the downstream flow will reduce by 15%, while the baseflow will reduce by 13%. Consequently, this reduction in flow has the potential to further jeopardize downstream water availability in the entire São Francisco River basin.

Moreover, a decrease in baseflow serves as an indicator of aquifer depletion. Marques et al. (2020) observed groundwater level declines of up to 6.63 m in the Alto Grande watershed during their analysis from 1990 to 2018. The authors suggested that the reduction in precipitation due to climate oscillations alone does not fully account for the water level drawdown. Therefore, the potential overexploitation of the aquifer in this watershed could also be identified as one of the causes of the observed water depletion.

Additionally, the reduction in river flows compromises the potential for hydropower generation. Most of the small power plants in the Rio Grande Basin are run-of-river, which makes power generation entirely dependent on river flow. In this way, considering the current issues related to energy supply in the region and the downward trend in river flow projections, an alternative power supply should be considered. In our simulation, by 2040, the region requires an additional power supply of 40 million kWh. This demand could be met through photovoltaic energy, capitalizing on the region’s solar potential (Pereira et al. 2017).

As discussed earlier, electricity grids in the region often face quality issues related to low power capacity and shortages, hindering the expansion of irrigation systems. Electrical grid extensions and reinforcements take years and require large public investments.

At first analysis, solar photovoltaic (PV) systems would provide the required energy for irrigation expansion at competitive costs. Using quite conservative parameters regarding the density of photovoltaic module installation and the capacity factor compared to what is typically found in PV plants in Brazil (Pereira et al. 2017; ANEEL 2024), it is possible to estimate that an area of 6.5 ha could supply the deficit energy and avoid emissions of about 124 tCO2 Year−1 (EPE 2023). Additionally, because solar radiation increases with the reduction of precipitation (associated with the reduction of cloud cover), the increase in energy usage due to the expansion of irrigation during longer dry periods can be supplied by photovoltaic generation, improving the resilience of the integrated system.

On the other hand, large-scale irrigation typically demands a steady and high-power supply, making diesel (biofuel) generators or small hydropower plants more suitable due to their reliable river flow. Photovoltaic systems face challenges in high-load off-grid applications (Herraiz et al. 2023) due to their intermittency, which is especially critical in tropical climates where solar intermittency is more pronounced (Luiz et al. 2018).

Local energy storage is economically unfeasible due to high battery costs, so the options rely upon hybrid system arrangements like diesel-PV or grid-connected-PV irrigation systems. Grid-connected PV systems may also inject excess energy into the grid (i.e., during rainy periods), reducing energy costs (Sousa et al. 2019). Granting these technologies proper incentives may play an important role and bring environmental benefits to the region’s future scenarios (Todde et al. 2019).

To sustain agriculture production over the long term in the Rio Grande Basin, improving water management is crucial. This involves implementing efficient irrigation systems, investing in precision agriculture, and helping farmers make informed decisions to reduce water and energy wastage. Additionally, updating hydroclimatic data is essential for reversing river flow trends by reviewing water allocation (Pousa et al. 2019). These rational approaches will benefit both agricultural production and long-term river sustainability.

Predictions and quantifications of water supply fluctuations are complex and related to climatic variables, anthropic actions, and spatial changes. The primary limitation of this modeling approach is its inability to account for spatially explicit changes, which are highly significant for modeling agricultural and hydrological processes. However, this can be improved by coupling a spatially explicit crop hydrological model and considering the socioeconomic feedback accounted for in this model, leveraging the strengths of both models (Jeong and Adamowski 2016; Harms et al. 2023).

Although the results show promise for water, energy, and land-use management in the basin, it is essential to acknowledge the limitations that should be considered to enhance the analysis. Evaluating water scarcity due to agricultural water use involves considering return flows (Multsch et al. 2020), which account for water losses within irrigation systems. Integrating this concept into the model acknowledges water returning to the system and the potential for groundwater contamination. Although agricultural expansion is essential for food security and economic progress, it can lead to increased pollution in surface and groundwater (Lopes et al. 2021) due to factors like excessive use of chemical fertilizers and pesticides, resulting in the leaching of contaminants into both groundwater and surface water (Pignati et al. 2017).

Furthermore, the model’s only exogenous variable is rainfall, which was provided by a climate model that includes CO2 concentrations for an intermediate emission scenario (RCP4.5). Therefore, this study does not account for the uncertainties associated with different future projections stemming from both model structures and emission scenarios. Additionally, we do not incorporate atmospheric evaporative demand from climate models; instead, we treat it as a function of soil moisture, which can reduce the expected impacts.

Finally, the economics that govern land expansion in the model are only intended to show the development of potential costs and benefits associated with land expansion relative to today but have not been vetted by local experts or compared to empirical data. The dynamics that govern land-use expansion in the model are only used to illustrate potential effects on profits that may hamper land-use expansion because of water shortages. Nevertheless, the general dynamics and relationships portrayed, including the causal assumptions, are probably not controversial (e.g., land-use expansion is positively affected by profits from irrigated land and negatively affected by water shortages). To further improve the external validity of the model, where the thresholds lie, that is, when land-use expansion is no longer profitable, could be further investigated by, for example, interviewing landowners and incorporating empirical data on costs associated with land-use expansion. The external validity of the model could also be increased by coupling it to the development of the global or national economy which now falls outside the model boundaries.

It is also worth noticing that some of the core structure of the model, as well as the resulting behavior, is an example of the system archetype Limits to Growth (Senge 1990; Meadows 2008; Moallemi et al. 2022). Limits to Growth describes a dynamic in which a process feeds on itself providing a period of rapid growth. Eventually, however, the growth runs into some limit that results to a halt, or a reversal and potentially also collapse. In the case of the Rio Grande Model, the process feeding on itself is the irrigation expansion that provides profit that is reinvested in the expansion. The limit is represented by the shortage of water for irrigation that inhibits the expansion.

Conclusion

The Rio Grande Basin is a rapidly expanding region for irrigated agriculture in Brazil, placing a growing demand for both surface and groundwater resources. In recent years, conflicts arising from water-use allocation in the area have underscored the significance of a comprehensive understanding of local water resources and the need for effective management practices to achieve better water governance.

The results suggest that water exploitation in the basin, coupled with climate change, may restrict the expansion of irrigated agriculture in the region, raising concerns about the long-term sustainability of agriculture in the basin. Despite its simplicity, the system dynamics model of the Rio Grande basin proved to be adequate for providing a baseline scenario and can be applied to simulate various situations, providing stakeholders with insights into the anticipated impacts of actions implemented in the basin. Due to its simplicity and flexibility compared with more sophisticated physically based models, it is a powerful tool for the River Basin Committee and stakeholders, assisting in the establishment of measures for sustainable water resource management, ensuring that the needs of all users are met, and mitigating existing conflicts.

Future studies should explore the possibilities of other pathways to adapt to the current conditions. For instance, the possibility of connecting the local electric subsystem to the national power grid; additional channels to ensure water supply such as the water transfer from the Tocantins basin to the west of the study area; the use of precision agriculture to improve irrigation efficiency; the implementation of an adaptative management systems that allow for expansion and contraction of the irrigated area during periods of uncertainties.

While the model could be applied to other regions, it is crucial to emphasize its contribution to enhancing our understanding of the interconnections among the water, energy, and agriculture nexus, considering the basin is an appropriate unit for water resources management and governance. Additionally, it provides evidence of how natural resource availability may constrain economic development, reinforcing the need for rational resource use to ensure sustainability. The model also contributes to the literature on what is referred to as system archetypes by providing an example of the Limits to Growth archetype—the process of irrigation expansion feeds on itself but eventually comes to a halt as it is limited by the availability of water for irrigation.

Nevertheless, the simple nature of the model comes with caution of over-relying on its results or drawing strong conclusions. Finally, the model can be customized to address specific issues that may arise, facilitating its adaptability to explore different scenarios and enabling broader applications. Such an expansion would incorporate an expanded comparison with empirical data in the economic and land-use parts of the model. In principle, the model highlights the necessity of taking a systems view of problems arising around the expansion of irrigated agriculture and that there are limits to the expansion.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 609 KB) (608.7KB, pdf)

Acknowledgements

The authors would like to thank the NEXUS Project, funded by the São Paulo Research Foundation–FAPESP, under Grant Numbers 2022/00917-0 and 2017/22269-2. DC, FGBS, TSG, and APDA acknowledge support from the XPaths project—Science in Action: Intersecting Pathways to the SDGs Across Scales in Drylands., a grant from FORMAS — a Swedish Research Council for Sustainable Development (ref: 2020-00474). CvR acknowledges CNPq grant number 314780/2020-3, and JT acknowledges CNPq grant number 304695/2020-3.

Biographies

Minella Alves Martins

is a postdoctoral researcher at the National Institute for Space Research (INPE). Her research focuses on modeling impact and risk assessment for agriculture and water resources.

David Collste

is a researcher at the Planetary Boundaries group within the Stockholm Resilience Centre. His areas of expertise encompass System Dynamics modeling, sustainability science, and development economics.

Francisco Gilney Silva Bezerra

is a postdoctoral researcher at INPE. His research interests include modeling land-use and land-cover change processes, estimating emissions, and desertification.

Marcela Aparecida Campos Neves Miranda

is a postdoctoral researcher at INPE. Her research interests include eutrophication of aquatic environments, mitigation of cyanobacterial blooms, biogeochemical cycles.

André Rodrigues Gonçalves

is a technologist affiliated with the Laboratory for Modeling and Studies of Renewable Energy Resources (LABREN) at INPE. His research field includes atmospheric modeling of solar and wind resources, prediction and variability of renewable energy resources, and the impacts of future climate scenarios on energy resources.

Jocilene Dantas Barros

is a research fellow at INPE. Her primary areas of interest encompass environmental monitoring, cartography and GIS, urban vegetation, land degradation, and socio-environmental indicators.

Manoel Ferreira Cardoso

is a researcher at INPE. His primary areas of interest include computational modeling of terrestrial ecosystem dynamics as well as the analysis and observation of interactions between the surface and the atmosphere.

Amanda Sousa Silvino

is a postdoctoral researcher at INPE. Her research interests include the perceptions and values related to biodiversity conservation.

Taís Sonetti-González

is a PhD researcher at Université Libre de Bruxelles. Her research focuses on collaborative processes for regenerative sustainability in the São Francisco River Basin, employing relational and decolonial methodologies while also exploring gender issues and intersectionality in various contexts.

Jean Pierre Henry Balbaud Ometto

is senior researcher at INPE. His research interests include ecology, forestry, carbon cycle, environmental analysis, climate change.

Celso von Randow

is a senior researcher at INPE. His primary areas of interest encompass surface fluxes, Eddy Covariance, and Carbon Cycle Interaction with the Climate System.

Javier Tomasella

is senior researcher at INPE. His research fields include hydrological impacts on land-use and land-cover changes, in particularly forest conversion to pasture or cropland, including climate change impacts.

Ana Paula Dutra de Aguiar

is senior researcher at INPE and Stockholm Resilience Centre. Her research interests include environmental modeling and scenarios participatory scenarios, land-use modeling, and greenhouse emission modeling.

Declarations

Conflict of interest

All authors declare that they have no conflicts of interest.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Change history

9/25/2025

Acknowledgements section is updated.

References

  1. Alvares, C.A., J.L. Stape, P.C. Sentelhas, J.L.M. de Gonçalves, and G. Sparovek. 2013. Köppen’s climate classification map for Brazil. Meteorologische Zeitschrift. 22: 711–728. 10.1127/0941-2948/2013/0507. [Google Scholar]
  2. ANA-National Water Agency. 2021. Water Usage by Municipality. Retrieved 17 March 2023, from https://dadosabertos.ana.gov.br/datasets/1bab1b61623d4594866d47582ccd7181_9/about (In Portuguese).
  3. ANA-National Water Agency. 2021. Atlas of Irrigation–Water Use in Irrigated Agriculture. 2nd Ed. Retrieved 25 November 2022, from http://atlasirrigacao.ana.gov.br/ (In Portuguese).
  4. ANEEL‐National Electric Energy Agency. 2024. Relação de empreendimentos de Geração Distribuída. Retrieved 06 May 2024, from https://dadosabertos.aneel.gov.br/dataset/relacao-de-empreendimentos-de-geracao-distribuida (In Portuguese).
  5. Aragão, A., and E. Contini. 2021. Agriculture in Brazil and the World: A Summary of the Period from 2000 to 2020. EMBRAPA SIRE. Retrieved 10 October 2023, from https://www.embrapa.br/documents/10180/62618376/O+AGRO+NO+BRASIL+E+NO+MUNDO.pdf (In Portuguese).
  6. Barlas, Y. 1996. Formal aspects of model validity and validation in system dynamics. System Dynamics Review 12: 183–210. [Google Scholar]
  7. Binder, T., A. Vox., S. Belyazid, H. Haraldsson and M. Svensson. 2004. Developing system dynamics models from causal loop diagrams. In Proceedings of the 22nd International Conference of the System Dynamics Society. Oxford. Retrieved 12 May 2023, from https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=cf00b9084b05ba357bf0c5fa7a5b9cc1b5695015.
  8. Calmon, D. 2020. Shifting frontiers: The making of Matopiba in Brazil and global redirected land use and control change. The Journal of Peasant Studies. 10.1080/03066150.2020.1824183. [Google Scholar]
  9. CBHSF-São Francisco River Basin Committee. 2021. The construction of a new Small Hydroelectric Plant (PCH) on the Rio Grande raises concerns among residents and traditional communities. Retrieved 04 November 2023, from https://cbhsaofrancisco.org.br/noticias/novidades/construcao-de-nova-pch-no-rio-grande-causa-preocupacao-em-moradores-e-comunidades-tradicionais/ (In Portuguese).
  10. CEPEA/CNA-Center for Advanced Studies on Applied Economics–CEPEA, Brazilian Confederation of Agriculture and Livestock–CAN. 2021. Gross Domestic Product (GDP) of the Brazilian agribusiness from 1996 to 2021. Retrieved 13 March 2023, from https://ww.cepea.esalq.usp.br/br/pib-do-agronegocio-brasileiro.aspx (In Portugues).
  11. Collste, D., M. Pedercini, and S.E. Cornell. 2017. Policy coherence to achieve the SDGs: Using integrated simulation models to assess effective policies. Sustainability Science 12: 921–931. 10.1007/s11625-017-0457-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Collste, D., A.P. Aguiar, Z. Harmáčková, D. Galafassi, L. Pereira, O. Selomane, and S. van der Leeuw. 2023. Participatory pathways to the Sustainable Development Goals: Inviting divergent perspectives through a cross-scale systems approach. Environmental Research Communications. 10.1088/2515-7620/acce25. [Google Scholar]
  13. CPRM-Mineral Resources Research Company. 2022. SIAGAS: Groundwater Information System. Retrieved 30 November 2022, from http://siagasweb.cprm.gov.br/layout/ (In Portuguese).
  14. CPT-Pastoral Land Commission. 2022. Land and Water Conflits. Retrieved 15 March 2023, from https://www.cptnacional.org.br/ (In Portugues).
  15. Dionizio, E.A., and M.H. Costa. 2019. Influence of land use and land cover on hydraulic and physical soil properties at the Cerrado agricultural frontier. Agriculture 9: 24. 10.3390/agriculture9010024. [Google Scholar]
  16. Dionizio, E.A., F.M. Pimenta, L.B. Lima, and M.H. Costa. 2020. Carbon stocks and dynamics of different land uses on the Cerrado agricultural frontier. PLoS ONE 15: e0241637. 10.1371/journal.pone.0241637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Eger, G.Z.S., G.C. Silva Junior, E.A.G. Marques, B.R.C. Leão, D.G.T.B. da Rocha, T.E. Gilmore, L.G.H. Amaral, J.A.O. Silva, et al. 2021. Recharge assessment in the context of expanding agricultural activity: Urucuia Aquifer System, western State of Bahia Brazil. Journal of South American Earth Sciences.10.1016/j.jsames.2021.103601. [Google Scholar]
  18. EMBRAPA-Brazilian Agricultural Research Corporation. 2018. VISION 2030: The Future of Brazilian Agriculture. Retrieved 13 March 2023, from https://www.embrapa.br/visao/trajetoria-da-agricultura-brasileira (In Portugues).
  19. EPE-Energy Research Company. Balanço Energético Nacional. 2023. Retrieved 06 May 2024, from https://www.epe.gov.br/pt/publicacoes-dados-abertos/publicacoes/balanco-energetico-nacional-2023 (In Portuguese).
  20. Fernald, A.G., S.Y. Cevik, C.G. Ochoa, V.C. Tidwell, J.P. King, and S.J. Guldan. 2010. River hydrograph retransmission functions of irrigated valley surface water-groundwater interactions. Journal of Irrigation and Drainage Engineering. 136: 823–835. 10.1061/(ASCE)IR.1943-4774.0000265. [Google Scholar]
  21. Forrester, J.W. 1969. Urban Dynamics. Cambridge: The M.I.T. Press. [Google Scholar]
  22. Gaspar, M. T. P., and J.E.G. Campos. 2007. The Urucuia Aquifer System. Revista Brasileira de Geociências 37. Sociedade Brasileira de Geologia: 216–226. 10.25249/0375-7536.200737s4216226. (In Portuguese).
  23. Gonçalves, R.D., B.Z. Engelbrecht, and H.K. Chang. 2016. Hydrological analysis of historical data in the Rio Grande Basin (BA): Contribution of the Urucuia Aquifer System. Águas Subterrâneas 30: 190–208. 10.14295/ras.v30i2.28514. (In Portuguese). [Google Scholar]
  24. Gonçalves, R. D. and H. K. Chang. 2017. Hydrogeological model of the Urucuia Aquifer System in the Rio Grande Basin (BA) Geociências, 36:205–220. Retrieved 11 October 2023, from https://www.periodicos.rc.biblioteca.unesp.br/index.php/geociencias/article/view/11132/8152. (In Portuguese).
  25. Harms, J.Z., J.J. Malard-Adam, J.F. Adamowski, A. Sharma, and A. Nkwasa. 2023. Dynamically coupling system dynamics and SWAT+ models using Tinamït: application of modular tools for coupled human–water system models. Hydrology and Earth System Sciences 27: 1683–1693. 10.5194/hess-27-1683-2023. [Google Scholar]
  26. Herraiz, J.I., R.H. Almeida, M. Castillo-Cagigal, and L. Narvarte. 2023. Experimental performance evaluation of a PV-powered center-pivot irrigation system for a three-year operation period. Energies 16: 3654. 10.3390/en16093654. [Google Scholar]
  27. Hofmann, G.S., R.C. Silva, E.J. Weber, A.A. Barbosa, L.F.B. Oliveira, R.J.V. Alves, H. Hasenack, V. Schossler, et al. 2023. Changes in atmospheric circulation and evapotranspiration are reducing rainfall in the Brazilian Cerrado. Scientific Reports 13: 11236. 10.1038/s41598-023-38174-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. IBGE-Brazilian Institute of Geography and Statistics. 2022. Brazilian Census of 2022. Rio de Janeiro: IBGE. Retrieved 10 September 2023, from https://censo2022.ibge.gov.br/ (In Portuguese).
  29. INEMA-Institute for the Environment and Water Resources. 2022. Normative Instruction nº 003 of June 30, 2022. Retrieved 08 July 2024, from http://www.inema.ba.gov.br/wp-content/files/INEMA_-_Instrucao_Normativa_003_-_Procedimentos_Administrativos_e_Critrios_de_Autorizao_para_Perfurao_de_Poos_-_APPO__no_aqufero_Urucuia.pdf. (In Portuguese).
  30. Jeong, H., and J. Adamowski. 2016. A system dynamics based socio-hydrological model for agricultural wastewater reuse at the watershed scale. Agricultural Water Management 171: 89–107. 10.1016/j.agwat.2016.03.019. [Google Scholar]
  31. Lopes, G.R., M.G.B. Lima, and T.N.P. dos. Reis. 2021. Maldevelopment revisited: Inclusiveness and social impacts of soy expansion over Brazil’s Cerrado in Matopiba. World Development 139: 105316. 10.1016/j.worlddev.2020.1053.
  32. Luiz, E.W., F.R. Martins, A.R. Gonçalves, and E.B. Pereira. 2018. Analysis of intra-day solar irradiance variability in different Brazilian climate zones. Solar Energy 167: 210–219. 10.1016/j.solener.2018.04.005. [Google Scholar]
  33. Marengo, J.A., J.C. Jimenez, J.C. Espinoza, A.P. Cunha, and L.E.O. Aragão. 2022. Increased climate pressure on the agricultural frontier in the Eastern Amazonia-Cerrado transition zone. Scientific Reports 12: 457. 10.1038/s41598-021-04241-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Marques, E.A.G., G.C. Silva Junior, G.Z.S. Eger, M.A. Ilambwetsi, R. Pousa, T.N. Generoso, J. Oliveira, and J.N. Júnior. 2020. Analysis of groundwater and river stage fluctuations and their relationship with water use and climate variation effects on Alto Grande watershed, Northeastern Brazil. Journal of South American Earth Sciences 103: 102723. 10.1016/j.jsames.2020.102723. [Google Scholar]
  35. Martins, M.A., J. Tomasella, and C.G. Dias. 2019. Maize yield under a changing climate in the Brazilian Northeast: Impacts and adaptation. Agricultural Water Management 216: 339–350. 10.1016/j.agwat.2019.02.011. [Google Scholar]
  36. Meadows, D. 2008. Thinking in Systems: A Primer. White River Junction, Vermont: Chelsea Green Publishing. [Google Scholar]
  37. Moallemi, E.A., S.H. Hosseini, S. Eker, L. Gao, E. Bertone, K. Szetey, and B.A. Bryan. 2022. Eight archetypes of sustainable development goal (SDG) synergies and trade-offs. Earth’s Future 10: e2022EF002873. 10.1029/2022EF002873. [Google Scholar]
  38. Multsch, S., M.S. Krol, M. Pahlow, A.L.C. Assunção, A.G.O.P. Barretto, Q.J. de van Lier, and L. Breuer. 2020. Assessment of potential implications of agricultural irrigation policy on surface water scarcity in Brazil. Hydrology and Earth System Sciences. 24: 307–324. 10.5194/hess-24-307-2020. [Google Scholar]
  39. Pereira, E. B., F.R. Martins, A.R. Gonçalves, R.S. Costa, F.L. Lima, R. Rüther, S.L. Abreu, G.M. Tiepolo, S.V. Pereira, and J.G. Souza. 2017. Brazilian Atlas of Solar Energy. 2.ed. São José dos Campos: INPE, 2017. 80p. Retrieved 20 September 2023, from 10.34024/978851700089 (In Portuguese).
  40. Pereira, B.H.F., C. Dereczynski, G.C. da Silva Junior, and E.A.G. Marques. 2022. Projected climate change impacts on groundwater recharge in the Urucuia aquifer system Brazil. International Journal of Climatology 42: 8822–8838. 10.1002/joc.7773. [Google Scholar]
  41. Phan, T.D., E. Bertone, and R.A. Stewart. 2021. Critical review of system dynamics modelling applications for water resources planning and management. Cleaner Environmental Systems 2: 100031. 10.1016/j.cesys.2021.100031. [Google Scholar]
  42. Pimenta, F.M., A.T. Speroto, M.H. Costa, and E.A. Dionizio. 2021. Historical Changes in Land Use and Suitability for Future Agriculture Expansion in Western Bahia Brazil. Remote Sensing 13: 1088. 10.3390/rs13061088. [Google Scholar]
  43. Pousa, R., M.H. Costa, F.M. Pimenta, V.C. Fontes, V.F.A. de Brito, and M. Castro. 2019. Climate change and intense irrigation growth in Western Bahia, Brazil: The urgent need for hydroclimatic monitoring. Water 11: 933. 10.3390/w11050933. [Google Scholar]
  44. Pousa, R., and M.H. Costa. 2023. Interactions between large-scale and mesoscale processes define long-term rainfall variability and availability of water resources in Western Bahia, Brazil. International Journal of Climatology 43: 3416–3432. 10.1002/joc.8036. [Google Scholar]
  45. Pignati, W., F. Lima, S. Lara, M. Correa, J. Barbosa, L. Leão, and M. Pignatti. 2017. Spatial distribution of pesticide use in Brazil: a strategy for Health Surveillance. Ciência e Saúde Coletiva 22: 3281–3293. 10.1590/1413-812320172210.17742017. (In Portugues). [DOI] [PubMed] [Google Scholar]
  46. Reis, P. A. G. 2018. Evaluation of flow contributions from the Rio Grande watershed to the São Francisco River channel to meet multiple uses. University of Bahia. Salvador. 163p. Retrieved 10 August 2023, from https://maasa.ufba.br/sites/maasa.ufba.br/files/dissertacao_de_mestrado_-_polyana_alcantara_galvao.pdf. (In Portugues).
  47. Santos, C. A. P. 2016. Environmental Education as a Tool for Conflict Mitigation in the Rio Grande Watershed, Western Bahia. HOLOS. 8: 156–172. Retrieved 5 April 2023, from 10.15628/holos.2016.2837. (In Portuguese).
  48. Santos, C.A.P., E.E. Sano, and P.S. Santos. 2018. Agricultural frontier and dynamic of land use and ocuppation in western bahia. ACTA Geográfica, Boa Vista. 12: 17–32 (In Portuguese). [Google Scholar]
  49. Senge, P. M. (1990). The fifth discipline: The art and practice of the learning organization (1. Currency paperback ed). Currency Doubleday.
  50. Simões, Y.D.S., E.H.B.C. Silva, and H.A. de Araújo. 2018. Rainfall zoning of Bahia State, Brazil: An update proposal. Ambiente e Agua - an Interdisciplinary Journal of Applied Science 13: 1. 10.4136/ambi-agua.2171. [Google Scholar]
  51. Sousa, M.A., J. Alves Júnior, A.W.P. Evangelista, D. Casaroli, and M. Mesquita. 2019. Estimativa de viabilidade econômica do uso de energia fotovoltaica em pivô central no estado de Goiás. Revista Engenharia Na Agricultura - REVENG 27: 22–29. 10.13083/reveng.v27i1.848.
  52. Souza Jr, C.M., J.Z. Shimbo, M.R. Rosa, L.L. Parente, A.A. Alencar, B.F.T. Rudorff, H.H.M. Matsumoto, L.G. Ferreira, et al. 2020. Reconstructing three decades of land use and land cover changes in Brazilian biomes with landsat archive and earth engine. Remote Sensing 12: 2735. 10.3390/rs12172735. [Google Scholar]
  53. Sterman, J.D. 2000. Business dynamics - systems thinking and modeling for a complex world. Boston: McGraw-Hill. [Google Scholar]
  54. Sušnik, J., L.S. Vamvakeridou-Lyroudia, D.A. Savić, and Z. Kapelan. 2012. Integrated system dynamics modelling for water scarcity assessment: case study of the Kairouan region. Science of the Total Environment 440: 290–306. 10.1016/j.scitotenv.2012.05.085. [DOI] [PubMed] [Google Scholar]
  55. Todde, G., L. Murgia, P.A. Deligios, R. Hogan, I. Carrelo, M. Moreira, A. Pazzona, L. Ledda, et al. 2019. Energy and environmental performances of hybrid photovoltaic irrigation systems in Mediterranean intensive and super-intensive olive orchards. Science of the Total Environment 651: 2514–2523. 10.1016/j.scitotenv.2018.10.175. [DOI] [PubMed] [Google Scholar]
  56. Valdiones, A. P., P. Bernasconi, V. Silgueiro, V. Guidotti, F. Miranda, J. Costa, R. Rajão, B. Manzolli. 2021. Illegal Deforestation and Conversion in the Amazon and Matopiba: lack of transparency and access to information. Retrieved 04 October 2023, from https://www.icv.org.br/website/wp-content/uploads/2021/05/icv-relatorio-ing-v1-1-1.pdf.
  57. Zomorodian, M., S.H. Lai, M. Homayounfar, S. Ibrahim, S.E. Fatemi, and A. El-Shafie. 2018. The state-of-the-art system dynamics application in integrated water resources modeling. Journal of Environmental Management. 227: 294–304. 10.1016/j.jenvman.2018.08.097. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary file1 (PDF 609 KB) (608.7KB, pdf)

Articles from Ambio are provided here courtesy of Springer

RESOURCES