Next Article in Journal
Differential Urban Heat Vulnerability: The Tale of Three Alabama Cities
Next Article in Special Issue
To What Extent Have Nature-Based Solutions Mitigated Flood Loss at a Regional Scale in the Philadelphia Metropolitan Area?
Previous Article in Journal
Community Perceptions of the Importance of Heritage Protection Relative to Other Local Government Council Operations
Previous Article in Special Issue
Stakeholder Engagement around Water Governance: 30 Years of Decision-Making in the Bogotá River Basin
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Use of Spatial Data in the Simulation of Domestic Water Demand in a Semiarid City: The Case of Campina Grande, Brazil

by
Higor Costa de Brito
1,*,
Iana Alexandra Alves Rufino
1,
Mauro Normando Macedo Barros Filho
1 and
Ronaldo Amâncio Meneses
1,2
1
Center for Technology and Natural Resources, Federal University of Campina Grande, Campina Grande 58429900, Brazil
2
Water and Sewage Company of Paraíba (CAGEPA), Campina Grande 58401150, Brazil
*
Author to whom correspondence should be addressed.
Urban Sci. 2023, 7(4), 120; https://doi.org/10.3390/urbansci7040120
Submission received: 18 September 2023 / Revised: 10 November 2023 / Accepted: 13 November 2023 / Published: 24 November 2023
(This article belongs to the Special Issue Water Resources Planning and Management in Cities)

Abstract

:
In the face of urban expansion, ensuring sustainable water consumption is paramount. This study aims to develop a domestic water demand forecast model that considers population heterogeneity and the urban area distribution in a city in the Brazilian Semiarid Region. The methodology comprises three main steps: (1) spatial data collection to identify explanatory variables for a future Land Use and Cover (LULC) model; (2) simulation of LULC data for 2030, 2040, and 2050 using the MOLUSCE plugin; and (3) estimation of domestic water demand based on projected urban area expansion and a linear regression model incorporating demographic indicators of household income, residents per household, total population, and gender. The results demonstrated a consistent LULC simulation, indicating an urban expansion of 4 km2 between 2030 and 2050, with reductions of 0.6 km2 in natural formations and 3.4 km2 in farming areas. Using LULC data, the study predicted a 14.21% increase in domestic water consumption in Campina Grande for 2050 compared to 2010, equivalent to an increase of 2,348,424.96 m3. Furthermore, the spatial analysis draws a spatial profile of water consumption among residents, highlighting the areas with the highest per capita consumption. Thus, this research offers a consistent approach to estimating water demand in regions with limited consumption data, providing valuable insights for decision-makers to consider in urban planning.

1. Introduction

Urbanization changes demographic characteristics and transforms the physical landscape of the environment. Inadequate planning can cause significant impacts on various environmental components, especially soil and water resources [1]. Economic activities unsustainably exploit land resources, resulting in an imbalance between supply and demand, driven by the urbanization process [2].
As more people migrate to urban areas for a better quality of life, human demands for water, energy, and food exert ecological pressures that contribute to climate change, pollution, biodiversity loss, and land erosion [3]. With the expansion of built-up areas, natural resources become increasingly constrained, necessitating a reevaluation of land-use processes. Policymakers, therefore, must consider scientific studies to formulate appropriate policies [4].
The 2016 United Nations New Urban Agenda (UN-Habitat) highlighted the need to adopt more enabling and facilitative approaches to analyzing urban form and extent in developing countries [5]. This movement has stimulated discussions on how best to plan cities, seeking to integrate areas and services, create sustainable population densities, and ensure optimal urban design [6,7]. One way to achieve this goal is by analyzing predictive scenarios and anticipating further possible developments for the region studied. This technique helps to understand and project how the region’s state will be at a future time and the processes leading up to this projected future [8].
Consequently, several spatiotemporal modeling, simulation, and transition potential techniques have been used to study Land Use and Land Cover (LULC). Among these are cellular automata, which can effectively simulate and represent spatially stochastic nonlinear land-use change processes [9]. Some cellular automata models successfully used for LULC analyses include Dinamica EGO [10], FLUS [11], SLEUTH [12], Artificial Neural Network-Markov Chain [13], SIMLANDER [14], and Cellular Automata-based Artificial-Neural Network (CA-ANN) [15].
Using CA-ANN to predict changes in LULC offers advantages compared to other methods. CA-ANN is excellent at modeling patterns and behaviors, which helps capture the multifaceted dynamics of growth by representing the nonlinear relationships between variables [16]. Another benefit is that CA-ANN can adapt well to datasets and scenarios without requiring knowledge of data distribution or assumptions about the data-generating process. Additionally, CA-ANN can easily integrate with models, leading to an understanding of the intricate dynamics behind urban growth patterns [17].
Such techniques may reflect the increasing demand for water in urban centers attributed to population growth and high urbanization rates. In water-scarce cities, it is crucial to establish reasonable urban boundaries to ensure efficient utilization of limited water resources and encourage sustainable economic and population expansion [18]. Therefore, this problem generates an even more significant concern in the Brazilian case, especially in the Semiarid Region, where the climatic conditions are already naturally unfavorable for maintaining an adequate water balance.
The Brazilian Semiarid Region (BSA), comprising more than 70% of the Northeast Region (NEB), has historically faced challenges such as limited water resources, high rates of internal migration due to drought, and increased poverty and social inequality [19,20]. Between 1995 and 2000, over 800,000 people left rural areas in the NEB [21]. Furthermore, a prolonged drought from 2012 to 2018 resulted in depleted reservoirs and reliance on water supply by tanker trucks in several municipalities [22]. Thus, strategic urban nodes emerged in response to intensified migratory movements, fostering significant urban growth in the BSA [23]. This region now grapples with the impacts of climate change, including desertification, which affects not only urbanization but also the overall way of life [24].
The need for more incisive urban interventions, as emphasized in the New Urban Agenda, is crucial for urban planning research. In this regard, this study aims to develop a domestic water demand forecasting model that considers the population’s heterogeneity and urban occupation, contributing to ensuring water availability for future generations in the city of Campina Grande, situated in the BSA. This effort serves as a tool for defining public policies that promote sustainable urban development in municipalities that face the challenge of monitoring water consumption data, thus increasing the capacity for future projections.

2. Methods and Data

2.1. Study Area

The municipality of Campina Grande, located in the semiarid region of the NEB (Figure 1), characterizes itself as the second largest city in the State of Paraíba, both economically and demographically. It stands as one of the largest cities in the interior of the NEB. The city has an urban population of 413,830 inhabitants, occupying an area of 591.658 km2, of which approximately 91 km2 falls within the urban perimeter. The city boasts an average altitude of 551 m and lies about 120 km from the state capital, João Pessoa [25]. According to data from the last demographic census, the municipality was divided into 524 census sectors (territorial units for the collection of census operations), classifying 65 as rural and 459 as urban [26].
The municipality faces urban water scarcity problems from growing water demand, hydraulic imbalances, poor management, and poorly planned urbanization processes [28]. Therefore, water insufficiency causes a cycle in which growing urbanization increases water demand, aggravated by a poorly planned urban environment, as illustrated in Figure 2.
With the substantial impacts of the multi-year drought between 2012 and 2018, agricultural production and human supply in the municipality were compromised. As a result of the water shortage, in critical periods, the municipality of Campina Grande began to suffer progressive rationing of water supply, where some localities began to suffer up to ten consecutive days without piped water [31]. Thus, studying water demand in a city like Campina Grande, which faces serious water scarcity problems, can contribute significantly to formulating public policies for sustainable urban development in semiarid regions.

2.2. Data Collection

The LULC images of Campina Grande were obtained from the Google Earth Engine platform [32]. The 2000, 2010, and 2020 images were acquired from the Annual Mapping of Land Use and Land Cover in Brazil Project (Mapbiomas). Mapbiomas represents an initiative involving a collaborative network of experts in biomes, land use, remote sensing, GIS, and computer specialists. These experts utilize images from Landsat satellites, each with a 30 m resolution. The entire process relies on machine learning algorithms, resulting in highly reliable products that cover the entire territorial extent of the country, provided in a free and accessible manner [33]. Up until this study, Mapbiomas had been utilizing 34 LULC classes. However, they were grouped within a GIS environment for simplification purposes, as illustrated in Table 1.
For the cellular automaton model simulation, the explanatory variables were chosen based on the research of [34] for a similar region since the motivations for urban expansion in semiarid regions occur similarly in different cities. Table 2 describes the four explanatory variables of urban growth chosen within this study and the assumptions used for each. After data collection, the products were cropped and rescaled to the exact spatial resolution (30 m) to serve as input data for the Modules for Land Use Change Evaluation (MOLUSCE) QGIS plugin and later assist in the future estimation of water demands.
The slope variable was calculated within a GIS environment using the digital elevation model of the ALOS satellite. Similarly, accessibility, proximity to urban centers, and urban expansion areas were developed based on Euclidean distances. Accessibility relied on road data supplied by the National Department of Transport Infrastructure of Brazil [35]. Proximity to urban centers was determined by measuring the distance to the municipal headquarters using information from the Brazilian Institute of Geography and Statistics (IBGE) [36]. Finally, the urban expansion zones corresponded to the distances to the urban infrastructure classes of Mapbiomas for the year 2020 (the most current mapping available at the time of the research). All the variables were reclassified to a resolution of 30 m to match the resolution of Mapbiomas. Figure 3 represents the methodological flow of the study.

2.3. Model Analysis, Prediction, and Validation

Spatiotemporal changes and the potential LULC transition between study intervals were obtained using the MOLUSCE plugin [37], which integrates with the free software QGIS, version 2.18. A probability matrix for area change and transitions was derived using the LULC data and explanatory variables. This matrix included rows and columns representing LULC categories in the initial and final years of the observed time interval.
This simulation aims to simplify the dynamics of composite urban structures and interpret them in an easily understandable way. The CA-ANN approach in the MOLUSCE plugin is considered more efficient than linear regression for potential transition modeling and future simulation [38]. The MOLUSCE plugin effectively calculates land-use change analysis and is suitable for analyzing seasonal forest and land-use change, potential transition modeling, and future scenario simulation.
The LULC for 2020 was projected using the Mapbiomas LULC maps of 2000 and 2010 along with the explanatory variables. To validate both the model and the accuracy of the prediction, the simulated 2020 LULC was compared with the Mapbiomas LULC data for the same year. The MOLUSCE plugin offers a Kappa validation technique for comparing the projected and actual LULC images. This method assesses two types of similarity: quantitative similarity, which examines the number of pixels in each class (KHisto), and spatial similarity, which checks their spatial distribution (KLoc). The Kappa statistic was calculated by multiplying KHisto and KLoc.
Using metrics that can provide a more informative and transparent approach for assessing the agreement between LULC classifications, cross-tabulation was employed to validate the simulation, as recommended by [39]. The tabulation was produced using the Semi-Automatic Classification Plugin within the QGIS environment, where two factors were analyzed: User’s Accuracy (UA) and Producer’s Accuracy (PA). UA indicates how well the model correctly identifies different land cover types. On the other hand, PA shows how well it can notice reference pixels from specific classes.
Satisfactory model validation results were achieved after conducting numerous tests. Consequently, the LULC maps for 2010 and 2020 were employed to make predictions for the LULC in 2030. Subsequently, the 2020 and 2030 (simulated) LULC maps were utilized to generate the 2040 LULC map. Finally, the simulated 2030 and 2040 maps allowed the prediction of the 2050 LULC.
During the CA-ANN learning process, 5000 random samples were utilized, 150 iterations were conducted, a neighborhood value of 3 × 3 pixels was set, a learning rate of 0.001 was implemented, 10 hidden layers were employed, and a momentum of 0.05 was applied. To summarize, selecting these parameters during the CA-ANN learning process demonstrates an attempt to strike a balance between the speed of learning, the complexity of the model, and the stability of training. The learning rate is intentionally kept low to ensure weight adjustments, whereas having 10 hidden layers indicates a network capable of capturing intricate relationships—moreover, a momentum value of 0.05 aids in enabling the optimization process to discover solutions.

2.4. Estimating Domestic Water Demand

Equations were used to estimate water consumption per pixel of urbanized areas, aiming to estimate the increased water demand resulting from urban expansion in Campina Grande. Initially, a correlation was performed between the census data of the resident population in each census sector and the LULC product of Mapbiomas. It is essential to highlight that the most recent demographic census for the Brazilian territory was available in 2010; therefore, the LULC of the same year was used for comparison purposes.
Using zonal statistics, estimating the number of inhabitants per pixel of an urbanized area was possible. Based on the sanitation indicators for the year 2010, made available by the National Information System on Water Resources [30], it was found that the average daily per capita consumption for the municipality of Campina Grande was 117.49 L per inhabitant per day. Thus, the water demand per pixel in each census sector could be estimated by associating water consumption with the number of inhabitants per pixel. The population growth projection from [40] for the state of Paraíba was used to estimate the population increase in the region. For the sake of simplification, it is essential to note that this study does not address the increased water demand resulting from the region’s socioeconomic development over the years.
Several factors can influence household water consumption and result in various patterns. In this perspective, [41] developed a multiple linear regression model (Equation (1)), which demonstrated that household income, water tariff, number of residents per household, temperature, percentage of households with washing machines, total population, gender, percentage of households with piped water, and municipality GDP influence urban water consumption in Brazil.
l n C O N S U P T I O N = 1.3488 + 0.1775 l n I N C O M E 0.1456 l n P R I C E 1.0298 l n R E S I D E N T S + 0.5686 l n T E M P E R A T U R E + 0.0431 l n W A S H I N G M A C H I N E + 0.0325 l n P O P U L A T I O N + 0.4087 l n G E N D E R + 0.2272 l n P I P E D + 0.0234 l n G D P
Among the variables listed, based on demographic data made available for the entire Brazilian territory by [42], it was possible to identify sectoral variations in a city, such as household income (PRICE), residents per household (RESIDENTS), total population (POPULATION), and gender (GENDER). Household income directly influences the ability to pay for water, whereas the number of residents per household and the total population impact the water demand per person. In addition, consumption differences between men and women can also influence the water consumption profile. Thus, from the analysis of the linear regression coefficients, it was possible to establish relationships between each variable’s municipal average and each’s influence on the percentage of water consumption in the census sector through Equation (2).
C O N S U P T I O N V A R = e ( ln ( A ) + C ln ( B ) )
where
  • C O N S U P T I O N V A R is the variation in water consumption resulting from the analyzed variable;
  • A is an arbitrary constant value, used as a reference value;
  • B is the percentage that represents the increase (or decrease) in relation to the municipal average, in a whole number;
  • C is the regression coefficient of the analyzed variable;
Therefore, from the LULC forecast (performed for the years 2030, 2040, and 2050) and the spatial variables that affect domestic water demand, it became possible to estimate the increase in demand arising from the increase in the urban area class. We used the R programming language [43] to quantify the new urban areas and calculate the estimated future water demand.
To validate the results, water consumption data for the year 2020 was collected from the Water and Sewage Company of Paraíba (CAGEPA). These data correspond to the total volume of water consumed by Campina Grande, covering all uses, such as irrigation, industry, and urban and rural supply. For comparison, it was assumed that urban supply constituted 60.95% of the total consumption, based on information provided by the National Water Agency [44] regarding consumptive water uses. This data collection and analysis allowed the comparison of simulated results with the observed urban water consumption data for 2020.

3. Results and Discussion

The dynamics of LULC for the next three decades in the municipality of Campina Grande were visualized utilizing the LULC classifications of Mapbiomas. The modeling satisfactorily represented the urban area for 2020, except in the southeast portion, where an additional expansion occurred in 2017 with the implementation of the Aluízio Campos Complex. This housing complex, which included over 4100 units, marked Brazil’s most significant housing complex under construction that year.
In this context, the explanatory variables used in the modeling are illustrated in Figure 4 and demonstrate the potential for urbanization caused by population increase and migration of the rural population to urban areas. Following the intensification of the migratory movement in the NEB, the urban area of Campina Grande remains strategic for human development in the state’s interior, presenting high growth rates, as discussed by [23].
The validation process performed in MOLUSCE used the 2020 LULC maps from Mapbiomas and the simulated 2020 classification from the 2000 and 2010 LULC. The value of KHisto was 0.81 while that of KLoc was 0.71. Thus, the value of the overall Kappa index was 0.58. Based on the Kappa values, the results could be considered moderate.
Analysis of the error matrix (Table 3) highlights the simulation’s performance, focusing mainly on the urban area class, representing this classification’s core. The results reveal an AP of 98.91% for this class, highlighting the model’s remarkable ability to identify urban areas accurately. In addition, the UA of 92.25% strengthens the model’s reliability in identifying urban areas in the simulated data. The overall accuracy reached 79.81%, reflecting the model’s comprehensive effectiveness and confirming that the simulation was satisfactory and reliable for the application.
There are numerous challenges for predictive modeling in BSA; among them stands out the phenology of the vegetation of the Caatinga biome, where the photosynthetic material of the vegetation is strongly related to the rainfall regime and suffers direct impacts from climate change [45]. This region’s climatic variability corroborates intra-annual variations in vegetation and agricultural management, hindering the remote classification of LULC by machine learning systems [46].
After satisfactory results were obtained in the model validation step, the spatial variables and the transition map were used to predict LULC for 2030, 2040, and 2050. Figure 5 presents the resulting maps from the simulations. At the urban edge, subsistence agriculture is common, carried out by the most vulnerable population; as urbanization advances over the territory, agriculture will make room for new construction. The simulations estimated an increase of more than 4 km2 of the urban area between 2030 and 2050. In contrast, the classes of natural formation and agriculture suffered reductions of 0.6 and 3.4 km2 in the same period, respectively.
Although modest, the increase in urban areas accompanies changes in land-use patterns, including the transition from agricultural areas and natural formations to urban areas. Supporting this observation, [47] emphasizes the complex interplay of urbanization, population concentration, and land-use intensification, which can notably affect urban water demand. This transition often leads to increased water demand due to the rise in population density and the growth of industrial and commercial activities in urban areas.
The simulations reveal a potential slowdown in urban expansion over time, partially attributed to the declining association between population growth and urban sprawl. Based on the results, the urban area is expected to increase by 2.6 km2 between 2030 and 2040 and 1.5 km2 between 2040 and 2050. This phenomenon closely links the densification of cities and the more efficient use of available interior spaces [48]. In this context, research conducted by [49] highlights the significant influence of intergovernmental competition in promoting urban densification, ultimately leading to the vertical growth of cities.
After the second half of the twentieth century, Brazilian medium-sized cities (such as Campina Grande) began to present a quick urbanization process, accompanied by verticalization, attributed to economic interests aimed at diversification of investments [50]. In this period, Campina Grande underwent urbanization plans driven by new development ideals; from this, the verticalization process intensified in the city [51]. Significantly, this urban transformation underscores its relevance, as it sheds light on the profound implications for the region’s water demand prediction based on LULC.
In this context, the LULC results highlight the viability of predictive land-use modeling, an achievement made possible by the accessibility of resources such as the Mapbiomas collection adapted for the various Brazilian biomes. This expanded horizon of resources opens up new perspectives for hydrological and urban modeling, improving the ability to understand and anticipate complex dynamics related to land use and water demand.
The urban environment is highly dynamic, and occupation varies spatially. Thus, the heterogeneous distribution of the city’s population and social factors lead to different consumption profiles among residents of the same municipality. In most Brazilian cities, no water micro-measurement data are available. For domestic demand estimation purposes, government agencies use estimates of average consumption per inhabitant based on flow and total population [30]. Figure 6, prepared from the consumption variables proposed by [41], seeks to refine the water demand estimates from LULC, considering spatial factors that alter the consumption profile of the inhabitants.
The spatial variables of consumption that most influenced water demand were RESIDENTS and PRICE, which were directly related to the population’s purchasing power. The number of residents per household, generally higher in poorer areas, tended to increase water consumption, especially in daily activities such as bathing, washing clothes and dishes, and using the toilet. Household income directly influences water consumption since families with higher purchasing power have access to goods and services that require greater water use, such as swimming pools, gardens, and bathtubs. Thus, it was found that regions further away from the urban center and low-income neighborhoods exhibited a negative variation in water demand. Simultaneously, areas with a higher presence of high-income populations experienced an increase of more than 10% in consumption.
In a region marked by frequent droughts, whose surrounding areas face a high risk of desertification triggered by climate change, as is the case of Campina Grande, a growing increase in domestic water demand puts even more pressure on surface reservoirs in the region, primarily responsible for human, industrial, and agricultural supply. In this context, Table 4 presents the simulation results of the water demand required to supply the municipality’s domestic consumption between 2020 and 2050 and compares the results with 2010, where daily consumption was 16,530,731.18 m3. Based on the results, it is estimated that there will be an increase in domestic water consumption of 2,348,424.96 m3 in 2050 compared to 2010, equivalent to a 14.21% increase.
Unfortunately, no data in the region are available to validate all the results obtained. Consumption data for 2020 from CAGEPA were the only information obtained. According to the agency, the municipality of Campina Grande consumed 17,352,517.00 m3 of water for human and industrial consumption in 2020. Also, according to the agency, in the region, water consumption is classified into four categories: residential, commercial, industrial, and public, and each of them is assigned a specific tariff. Although there are distribution networks in some parts of the rural area, they are primarily for residential use. Agriculture, on the other hand, utilizes raw water directly abstracted from the spring. In certain regions, such as along the Paraíba River, regulatory agencies authorize farmers to use water in this manner. Therefore, it is crucial to emphasize that agriculture falls outside the consumption above categories. Furthermore, the commercial consumption category encompasses industries connected to the water distribution network. This category includes all industries that rely on water supplied by the public network for their production activities.
Thus, the data obtained from CAGEPA regarding water consumption in Campina Grande during 2020 are insufficient to validate the research results fully. This inadequacy arises because the regulatory agency does not provide figures by consumption category, making it impossible to directly compare the data obtained with the study results. However, it is essential to highlight that the simulated values obtained in the research presented a variation of only 2.43% concerning CAGEPA’s data. However, the description of the consumption categories provided by the regulatory agency, the details about water distribution in rural areas, and the exclusion of agriculture as a consumption category support the consistency of the simulated values with reality.
Consequently, the water demand estimation elucidated within this study establishes a fundamental foundation for exploring the intricate relationship between land use and water demand, providing valuable insights that can shape government policies and strategies finely tailored to the nuanced requirements of urban and rural planning. These insights facilitate resource allocation and augment the efficacy of long-term sustainable development.
Hence, it is incumbent upon stakeholders to consider a holistic perspective, one that encompasses both physical and governance aspects, to ensure the continued availability of water resources for the current generation and safeguard the water needs of posterity. In this light, domestic water demand forecasting remains a dynamic research challenge, with significant opportunities for researchers to advance hybrid or specialized methodologies that account for the distinctive physical and socioeconomic characteristics of diverse regions across the globe, particularly those facing inadequate monitoring infrastructure.

4. Conclusions

Based on the results obtained in this study, the research appears as a reasonable alternative, supported by its consistent ability to predict future domestic water consumption with the help of LULC data. This achievement is precious in regions where data regarding water consumption are notably scarce, a situation frequently observed in many BSA cities. Consequently, the implications of the findings provide a replicable methodology that can be applied across the country, offering preliminary insights into future water consumption using currently available datasets.
However, as expected, the lack of official data by consumption category limited the validation of the results. Consequently, advocacy is made for future studies to proactively establish institutional partnerships to galvanize regulatory bodies into intensifying their data collection and monitoring efforts, amassing a more comprehensive dataset capable of enhancing the precision of regulatory measures and modeling techniques. Additionally, the potential for forthcoming studies to delve into integrating supplementary variables, such as climate change and economic factors, is acknowledged to develop more comprehensive water demand models.
Notwithstanding, water quantity and quality challenges necessitate multifaceted solutions, encompassing access, regulation, control, and demand reduction. These measures are pivotal in ensuring that all segments of society benefit sustainably from this invaluable resource. Governance strategies specific to this region should also integrate forward-looking aspects, addressing the perpetual surge in water demand and the repercussions of climate change on the region’s rainfall patterns.
In summary, it is fundamental that research in water resources management continues to seek solutions to the challenges faced in planning and managing water resources in urban and rural areas. Integrating different models, including those for domestic water demand forecasting based on predictive LULC change scenarios, can contribute to more efficient and sustainable management of water resources.

Author Contributions

Conceptualization, H.C.d.B. and I.A.A.R.; methodology, H.C.d.B.; software, H.C.d.B.; validation, H.C.d.B. and R.A.M.; formal analysis, H.C.d.B. and I.A.A.R.; investigation, H.C.d.B.; resources, H.C.d.B.; data curation, H.C.d.B., I.A.A.R., and M.N.M.B.F.; writing—original draft preparation, H.C.d.B.; writing—review and editing, I.A.A.R., M.N.M.B.F., and R.A.M.; visualization, H.C.d.B.; supervision, I.A.A.R. and M.N.M.B.F.; funding acquisition, I.A.A.R. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Council of Scientific and Technological Development—Brazil (CNPq) [grant numbers 140038/2022-1, 313147/2020-5]; and Paraíba State Research Foundation (FAPESQ) [grant number 23038.000845/2021-14].

Data Availability Statement

Data sharing does not apply to this article.

Acknowledgments

This work has been supported by the following Brazilian research agencies: FAPESQ, CNPq and CAPES.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Patra, S.; Sahoo, S.; Mishra, P.; Mahapatra, S.C. Impacts of urbanization on land use /cover changes and its probable implications on local climate and groundwater level. J. Urban Manag. 2018, 7, 70–84. [Google Scholar] [CrossRef]
  2. Liu, Y.; Yang, Y.; Li, Y.; Li, J. Conversion from rural settlements and arable land under rapid urbanization in Beijing during 1985–2010. J. Rural Stud. 2017, 51, 141–150. [Google Scholar] [CrossRef]
  3. Ahmed, Z.; Zafar, M.W.; Ali, S. Danish Linking urbanization, human capital, and the ecological footprint in G7 countries: An empirical analysis. Sustain. Cities Soc. 2020, 55, 102064. [Google Scholar] [CrossRef]
  4. Xu, F.; Wang, Z.; Chi, G.; Zhang, Z. The impacts of population and agglomeration development on land use intensity: New evidence behind urbanization in China. Land Use Policy 2020, 95, 104639. [Google Scholar] [CrossRef]
  5. Caprotti, F.; Cowley, R.; Datta, A.; Broto, V.C.; Gao, E.; Georgeson, L.; Herrick, C.; Odendaal, N.; Joss, S. The New Urban Agenda: Key opportunities and challenges for policy and practice. Urban Res. Pract. 2017, 10, 367–378. [Google Scholar] [CrossRef]
  6. Carneiro, E.; Lopes, W.; Espindola, G. Urban Land Mapping Based on Remote Sensing Time Series in the Google Earth Engine Platform: A Case Study of the Teresina-Timon Conurbation Area in Brazil. Remote Sens. 2021, 13, 1338. [Google Scholar] [CrossRef]
  7. Yadav, G.; Mangla, S.K.; Luthra, S.; Rai, D.P. Developing a sustainable smart city framework for developing economies: An Indian context. Sustain. Cities Soc. 2019, 47, 101462. [Google Scholar] [CrossRef]
  8. Gaur, S.; Bandyopadhyay, A.; Singh, R. Projecting land use growth and associated impacts on hydrological balance through scenario-based modelling in the Subarnarekha basin, India. Hydrol. Sci. J. 2021, 66, 1997–2010. [Google Scholar] [CrossRef]
  9. Okwuashi, O.; Ndehedehe, C.E. Integrating machine learning with Markov chain and cellular automata models for modelling urban land use change. Remote Sens. Appl. Soc. Environ. 2021, 21, 100461. [Google Scholar] [CrossRef]
  10. Rodrigues, H.; Soares-Filho, B. A Short Presentation of Dinamica EGO. In Geomatic Approaches for Modeling Land Change Scenarios; Springer: Cham, Switzerland, 2018; pp. 493–498. [Google Scholar]
  11. Liu, X.; Liang, X.; Li, X.; Xu, X.; Ou, J.; Chen, Y.; Li, S.; Wang, S.; Pei, F. A future land use simulation model (FLUS) for simulating multiple land use scenarios by coupling human and natural effects. Landsc. Urban Plan. 2017, 168, 94–116. [Google Scholar] [CrossRef]
  12. Saxena, A.; Jat, M.K. Capturing heterogeneous urban growth using SLEUTH model. Remote Sens. Appl. Soc. Environ. 2019, 13, 426–434. [Google Scholar] [CrossRef]
  13. Pahlavani, P.; Askarian Omran, H.; Bigdeli, B. A multiple land use change model based on artificial neural network, Markov chain, and multi objective land allocation. Earth Obs. Geomat. Eng. 2017, 1, 82–99. [Google Scholar] [CrossRef]
  14. Hewitt, R.; Díaz Pacheco, J.; Moya Gómez, B. A Cellular Automata Land Use Model for the R Software Environment (Weblog). Available online: https://simlander.wordpress.com (accessed on 29 November 2019).
  15. Abbas, Z.; Yang, G.; Zhong, Y.; Zhao, Y. Spatiotemporal Change Analysis and Future Scenario of LULC Using the CA-ANN Approach: A Case Study of the Greater Bay Area, China. Land 2021, 10, 584. [Google Scholar] [CrossRef]
  16. Gharaibeh, A.; Shaamala, A.; Obeidat, R.; Al-Kofahi, S. Improving land-use change modeling by integrating ANN with Cellular Automata-Markov Chain model. Heliyon 2020, 6, e05092. [Google Scholar] [CrossRef] [PubMed]
  17. Gantumur, B.; Wu, F.; Vandansambuu, B.; Tsegmid, B.; Dalaibaatar, E.; Zhao, Y. Spatiotemporal dynamics of urban expansion and its simulation using CA-ANN model in Ulaanbaatar, Mongolia. Geocarto Int. 2022, 37, 494–509. [Google Scholar] [CrossRef]
  18. Liu, S.; Wang, N.; Xie, J.; Jiang, R.; Zhao, M. Optimal Scale of Urbanization with Scarce Water Resources: A Case Study in an Arid and Semi-Arid Area of China. Water 2018, 10, 1602. [Google Scholar] [CrossRef]
  19. INSA––Instituto Nacional do Semiárido Portal INSA. Available online: https://portal.insa.gov.br/noticias/1070-nova-delimitacao-expande-o-semiarido-ate-omaranhao-73-novos-municipios-foram-.incluidos#:~:text=Comoacréscimodasnovas,demográficade25hab%2Fkm2 (accessed on 29 December 2020).
  20. Dias de Jesus, A. From fighting against droughts to coexisting with the semiarid. Int. J. Semiarid 2021, 4, 188–197. [Google Scholar] [CrossRef]
  21. Delazeri, L.M.M.; da Cunha, D.A.; Couto-Santos, F.R. Climate change and urbanization: Evidence from the Semi-arid Region of Brazil. Rev. Bras. Estud. Reg. Urbanos 2018, 12, 129–154. [Google Scholar]
  22. de Brito, Y.M.A.; Rufino, I.A.A.; Braga, C.F.C.; Mulligan, K. The Brazilian drought monitoring in a multi-annual perspective. Environ. Monit. Assess. 2021, 193, 31. [Google Scholar] [CrossRef]
  23. de Espindola, G.M.; Costa Carneiro, E.L.N.; Façanha, A.C. Four decades of urban sprawl and population growth in Teresina, Brazil. Appl. Geogr. 2017, 79, 73–83. [Google Scholar] [CrossRef]
  24. Marengo, J.A.; Cunha, A.P.M.A.; Nobre, C.A.; Ribeiro Neto, G.G.; Magalhaes, A.R.; Torres, R.R.; Sampaio, G.; Alexandre, F.; Alves, L.M.; Cuartas, L.A.; et al. Assessing drought in the drylands of northeast Brazil under regional warming exceeding 4 °C. Nat. Hazards 2020, 103, 2589–2611. [Google Scholar] [CrossRef]
  25. IBGE––Instituto Brasileiro de Geografia e Estatística Estimativas da População. Available online: https://www.ibge.gov.br/estatisticas/sociais/populacao/9103-estimativas-de-populacao.html?=&t=downloads (accessed on 10 March 2021).
  26. IBGE––Instituto Brasileiro de Geografia e Estatística Malha Municipal. Available online: https://www.ibge.gov.br/geociencias/organizacao-do-territorio/malhas-territoriais/15774-malhas.html?=&t=acesso-ao-produto (accessed on 5 January 2022).
  27. SUDENE––Superintendência do Desenvolvimento do Nordeste Mapas e Shapefile. Available online: http://antigo.sudene.gov.br/delimitacao-do-semiarido (accessed on 1 May 2022).
  28. Cordão, M.J.d.S.; Rufino, I.A.A.; Barros Ramalho Alves, P.; Barros Filho, M.N.M. Water shortage risk mapping: A GIS-MCDA approach for a medium-sized city in the Brazilian semi-arid region. Urban Water J. 2020, 17, 642–655. [Google Scholar] [CrossRef]
  29. MapBiomas Project Collection [6.0] of the Annual Land Use Land Cover Maps of Brazil. Available online: mapbiomas.org (accessed on 12 December 2021).
  30. SNIS––Sistema Nacional de Informações Sobre Saneamento Painel de Saneamento. Available online: http://appsnis.mdr.gov.br/indicadores/web/agua_esgoto/mapa-agua (accessed on 1 May 2022).
  31. Miranda, L.I.B. de A Crise Hidrica e a Gestão das Águas Urbanas na Bacia Hidrográfica do Rio Paraíba. In Proceedings of the XVII ENANPUR, São Paulo, Brazil, 22–26 May 2017. [Google Scholar]
  32. Gorelick, N.; Hancher, M.; Dixon, M.; Ilyushchenko, S.; Thau, D.; Moore, R. Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 2017, 202, 18–27. [Google Scholar] [CrossRef]
  33. Souza Junior, C.; Azevedo, T. MapBiomas General Handbook; Mapbiomas: São Paulo, Brazil, 2017; pp. 1–23. [Google Scholar]
  34. de Brito, H.C.; Rufino, I.A.A.; Djordjević, S. Cellular automata predictive model for man-made environment growth in a Brazilian semi-arid watershed. Environ. Monit. Assess. 2021, 193, 323. [Google Scholar] [CrossRef]
  35. DNIT (Departamento Nacional de Infraestrutura de Transportes) DNIT Portal. Available online: http://www.dnit.gov.br/mapas-multimodais (accessed on 4 December 2019).
  36. IBGE––Instituto Brasileiro de Geografia e Estatística IBGE Cidades. Available online: https://cidades.ibge.gov.br/ (accessed on 11 March 2021).
  37. Asia Air Survey. Next GIS MOLUSCE-Quick and Convenient Analysis of Land Cover Changes. Available online: https://nextgis.com/blog/molusce (accessed on 15 February 2022).
  38. El-Tantawi, A.M.; Bao, A.; Chang, C.; Liu, Y. Monitoring and predicting land use/cover changes in the Aksu-Tarim River Basin, Xinjiang-China (1990–2030). Environ. Monit. Assess. 2019, 191, 480. [Google Scholar] [CrossRef]
  39. Camacho Olmedo, M.T.; García-Álvarez, D. Basic and Multiple-Resolution Cross-Tabulation to Validate Land Use Cover Maps. In Land Use Cover Datasets and Validation Tools; Springer International Publishing: Cham, Switzerland, 2022; pp. 99–125. [Google Scholar]
  40. IBGE––Instituto Brasileiro de Geografia e Estatística Projeções da População. Available online: https://www.ibge.gov.br/estatisticas/sociais/populacao/9109-projecao-da-populacao.html (accessed on 14 March 2022).
  41. da Veiga, F.; Kalbusch, A.; Henning, E. Drivers of urban water consumption in Brazil: A countrywide, cross-sectional study. Urban Water J. 2022, 1–9. [Google Scholar] [CrossRef]
  42. IBGE––Instituto Brasileiro de Geografia e Estatística Censo Demográfico. 2010. Available online: https://censo2010.ibge.gov.br/resultados.html (accessed on 21 December 2021).
  43. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2013. [Google Scholar]
  44. ANA––Agência Nacional de Águas. Manual de Usos Consuntivos da Água no Brasil, 1st ed.; Ministério do Desenvolvimento Regional: Brasília, Brazil, 2019; ISBN 978-85-8210-057-8. [Google Scholar]
  45. Medeiros, R.; Andrade, J.; Ramos, D.; Moura, M.; Pérez-Marin, A.M.; dos Santos, C.A.C.; da Silva, B.B.; Cunha, J. Remote Sensing Phenology of the Brazilian Caatinga and Its Environmental Drivers. Remote Sens. 2022, 14, 2637. [Google Scholar] [CrossRef]
  46. Cunha, J.; Nóbrega, R.L.B.; Rufino, I.; Erasmi, S.; Galvão, C.; Valente, F. Surface albedo as a proxy for land-cover clearing in seasonally dry forests: Evidence from the Brazilian Caatinga. Remote Sens. Environ. 2019, 2019, 111250. [Google Scholar] [CrossRef]
  47. Wang, Y. Urban land and sustainable resource use: Unpacking the countervailing effects of urbanization on water use in China, 1990–2014. Land Use Policy 2020, 90, 104307. [Google Scholar] [CrossRef]
  48. Teller, J. Regulating urban densification: What factors should be used? Build. Cities 2021, 2, 302–317. [Google Scholar] [CrossRef]
  49. Fan, J.; Zhou, L. Three-dimensional intergovernmental competition and urban sprawl: Evidence from Chinese prefectural-level cities. Land Use Policy 2019, 87, 104035. [Google Scholar] [CrossRef]
  50. Casaril, C.C.; Fresca, T.M. Verticalização urbana brasileira: Histórico, pesquisadores e abordagens. Rev. Faz Ciência 2000, 9, 169. Available online: https://e-revista.unioeste.br/index.php/fazciencia/article/view/7535/5570 (accessed on 29 December 2020).
  51. Souza, F.C. de Verticalização Urbana: Um novo formato de cidade moderna nos séculos XX e XXI–Um estudo de caso sobre Campina Grande/PB (1960–2012). Geoconexões Online 2021, 1, 14–28. [Google Scholar] [CrossRef]
Figure 1. Census sectors in the municipality of Campina Grande. Source: Adapted from [26,27].
Figure 1. Census sectors in the municipality of Campina Grande. Source: Adapted from [26,27].
Urbansci 07 00120 g001
Figure 2. Urbanized area and urban water demand in Campina Grande. Source: Adapted from [29,30].
Figure 2. Urbanized area and urban water demand in Campina Grande. Source: Adapted from [29,30].
Urbansci 07 00120 g002
Figure 3. Methodology overview.
Figure 3. Methodology overview.
Urbansci 07 00120 g003
Figure 4. LULC maps used in the validation step and explanatory variables. (A) LULC Mapbiomas 2020. (B) Simulated LULC for 2020. (C) Slope. (D) Urban expansion zones. (E) Proximity to urban centers. (F) Accessibility.
Figure 4. LULC maps used in the validation step and explanatory variables. (A) LULC Mapbiomas 2020. (B) Simulated LULC for 2020. (C) Slope. (D) Urban expansion zones. (E) Proximity to urban centers. (F) Accessibility.
Urbansci 07 00120 g004
Figure 5. Simulated LULC (2030, 2040, and 2050) and area composition (km2) compared to that observed in 2020.
Figure 5. Simulated LULC (2030, 2040, and 2050) and area composition (km2) compared to that observed in 2020.
Urbansci 07 00120 g005
Figure 6. Household water demand from 2010 to 2050 and spatial variables of consumption: POPULATION (A), RESIDENTS (B), GENDER (C), and PRICE (D).
Figure 6. Household water demand from 2010 to 2050 and spatial variables of consumption: POPULATION (A), RESIDENTS (B), GENDER (C), and PRICE (D).
Urbansci 07 00120 g006
Table 1. LULC classes used.
Table 1. LULC classes used.
LULC TypeMapbiomas Classes
Natural FormationForest, non-forest natural formation, beach, dune, other non-vegetated areas
Agriculture and Livestock *Grazing, agriculture, forestry, mosaic of agriculture and grazing, mining
Water BodyRiver, lake, ocean, aquaculture
Urbanized AreaUrbanized area
* The study area has no record of mining and forestry activities.
Table 2. Explanatory variables and their assumptions.
Table 2. Explanatory variables and their assumptions.
Explanatory VariableHypothesis
SlopeLow-slope surfaces are more attractive for human activities (crops, livestock, or settlements). Usually, flat surfaces allow for most activities without the need for high earth movement (cuts and embankments), which makes it more accessible for settlements or agricultural and livestock activities.
AccessibilityThe more accessible an area is (for example, near roads and highways), the greater the chance that sites will become occupied.
Proximity to urban centersHistorically, settlements begin or spread near water bodies through water availability. The ease of access to the resource also makes agricultural and cattle-raising practices feasible in nearby regions.
Urban Expansion ZonesThroughout the urbanization process of cities, urban areas commonly redefine their boundaries. Thus, places where urbanization processes are identified more pronouncedly attract real estate investments that contribute to the urbanization in the vicinity.
Table 3. Accuracy of the classes in the simulation.
Table 3. Accuracy of the classes in the simulation.
Natural FormationFarmingWaterUrban Area
Producer’s accuracy (%)75.8577.8237.5198.91
User’s accuracy (%)61.7887.5891.5392.25
Table 4. Simulated domestic water demand.
Table 4. Simulated domestic water demand.
YearDomestic Water Consumption (m3/Year)Consumption Growth Compared to 2010 (%)
202017,773,950.837.52
203018,339,904.7810.94
204018,681,267.7613.01
205018,879,156.1414.21
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Brito, H.C.d.; Rufino, I.A.A.; Barros Filho, M.N.M.; Meneses, R.A. Use of Spatial Data in the Simulation of Domestic Water Demand in a Semiarid City: The Case of Campina Grande, Brazil. Urban Sci. 2023, 7, 120. https://doi.org/10.3390/urbansci7040120

AMA Style

Brito HCd, Rufino IAA, Barros Filho MNM, Meneses RA. Use of Spatial Data in the Simulation of Domestic Water Demand in a Semiarid City: The Case of Campina Grande, Brazil. Urban Science. 2023; 7(4):120. https://doi.org/10.3390/urbansci7040120

Chicago/Turabian Style

Brito, Higor Costa de, Iana Alexandra Alves Rufino, Mauro Normando Macedo Barros Filho, and Ronaldo Amâncio Meneses. 2023. "Use of Spatial Data in the Simulation of Domestic Water Demand in a Semiarid City: The Case of Campina Grande, Brazil" Urban Science 7, no. 4: 120. https://doi.org/10.3390/urbansci7040120

Article Metrics

Back to TopTop