Future land use conflicts: Comparing spatial scenarios for urban-regional planning

Abstract

Latin America’s intensive urbanization processes are triggering rapid peri-urban transformations and the expansion of cities. These include accelerated metropolization processes, urban sprawl, and the emergence of new conurbations. These changes parallel the expansion of highly profitable agricultural activities and plantations linked to international markets. This paper aims to analyze land use/cover changes between 1990 and 2050 in the Quillota Province, Valparaíso Region, Chile. Specific objectives considered (1) analyzing changes in land use/cover trajectories between 1990 and 2017, (2) simulating changes in land use/cover based on three scenarios of territorial planning to 2050 (trending, ecological planning, and spatial planning), and (3), identifying the areas most likely to be modified by urbanization and agricultural activity as a result of biodiversity loss in the study area. The Dyna-CLUE model implemented was complemented with GIS techniques for the analysis of land use/cover trajectories that allowed classifying and characterizing the most dynamic land uses/cover within the Quillota Province, such as urban land uses. The results of simulations to 2050 show a probable conurbation of medium-sized cities of Quillota-La Cruz-Calera, and future land use conflicts between peri urban-agricultural land use and plantation-natural conservation land use. The results suggest that it is essential to choose scenarios to ensure sustainable land use planning to control urban and peri-urban sprawl and protect areas of high natural value.

Keywords

Urban sprawl dyna-clue Quillota Province land use modeling

Introduction

The rapid growth of cities worldwide has prompted widespread debate on the environmental and social impacts of urbanization. In this context, the simulation of future urban growth based on spatial analysis techniques is a useful tool allowing an early assessment of the environmental impacts of the phenomenon. The development of models to analyze changes in land use is strongly influenced by the identification of spatiotemporal patterns that determine future simulation (Arsanjani, 2012; Batty, 2013; Mas et al., 2014). Urban areas currently expand contravening environmental, economic, political, and even cultural sustainability criteria (Lagrosa IV et al., 2018; Roweis, 2018), where one of their most evident effects is that they possess a dispersed and monofunctional form of expansion, replacing other forms of use and cover characterized by their diversity and mixture. To solve this problem, in recent years, there has been an increase in the use of methodologies that perform spatially explicit analyses based on future scenarios of land use/cover changes as a tool that seeks to provide answers to the impacts that the demands of urbanization would have on the rural and natural environment in the future (OECD, 2017).

In this research, we examine the Quillota Province, Valparaiso Region, Chile, which has undergone substantial changes in land use over the last 20 years. Its urban areas expanded at a rate of 68.7% between 1993 and 2017 due to urban expansion toward the periphery (MINVU, 2007, 2018). In addition, agricultural changes in Quillota Province have impacted native vegetation in the Coastal Mountain Range (Salazar et al., 2015). However, few studies have addressed the problem of modeling urban sprawl and expanding agricultural land use over the natural area in an interlinked way under different spatial scenarios.

This work proposes identifying future conflicts that could arise from the formation of the Quillota-La Calera-La Cruz urban conurbation, which is understood as the union of several urban settlements whose peripheries have merged following parallel growth, thus creating a contiguous urban area. This paper begins by analyzing the changes in land use in the Quillota Province, Valparaiso region, between 1990, 2003, and 2017. Next, we simulate land use/cover using the Dyna-CLUE model to 2050, using a trending, ecological, and spatial planning scenarios. Last, to validate the data, different local stakeholders with an influence over territorial planning and inter-district decision-making in the Quillota Province were interviewed.

Literature review: Simulation techniques

Currently, there are several programs that allow the incorporation of various techniques that simulate and project land use/cover changes, including SLEUTH (Clarke et al., 1997), DINAMICA (Soares et al., 2002), MOLAND (Engelen et al., 2007), LULCC (Moulds et al., 2015), CLUE model (Veldkamp & Fresco, 1996; Verburg et al., 1999), among others.

In particular, the CLUE model (The Conversion of Land Use and its Effects modeling framework) was developed to simulate land use/cover changes based on the relationship they have with different driving forces, including demands and competition between the various land uses at different scales (Verburg et al., 2008). This model has had several updates, including the CLUE-S model (Verburg et al., 2002) and Dyna-CLUE (Verburg and Overmars, 2009). One of the main differences between the CLUE-S model and Dyna-CLUE is that the former is a top-down model, that is, mostly using census and remote sensing data as main inputs. The latter turns out to be a hybrid model that incorporates features of top-down and bottom-up processes in a simple algorithm and incorporates, unlike the former, a neighborhood suitability model (Ren et al., 2019).

Land use/cover change models are not only determined for urban analyses but also include multiple land use/cover change dynamics, such as agricultural or forest or natural plantations. Spatially explicit models use a wide variety of techniques, including linear extrapolation processes, suitability mapping, neural networks, analysis scenarios, expert opinion, public participation, and agent-based models (Pontius et al., 2008: 14), among others. Thus, the Dyna-CLUE model has been chosen for its greater flexibility and ductility, especially for including land-use demands based on spatial and non-spatial techniques, and for results validated in several case studies (Henríquez-Dole et al., 2018).

In the case of the urban sprawl of Latin American cities, free-market policies have impacted urban growth, strengthening the role of the market and promoting investment strategies (Castro and Aliaga, 2010; Martner, 2016), often in a disorderly and fragmented manner (Borsdorf, 2003). In this sense, some research has focused on fast-growing metropolitan areas such as Santiago (Henríquez-Dole et al., 2018) or medium-sized cities (Henríquez, 2014). In all of them, emphasis has been placed on determining the main factors or driving forces that influence the increase in urban cover at the expense of agriculturally or ecologically important areas. However, little attention has been paid to the growing pressures that high-yield crops exert over surrounding natural areas (Brannstrom, 2009; Zaiatz et al., 2018).

The importance of scenarios

Scenarios are of great importance to explore the future of societies, institutions, and territory. They have been used for a long time in multiple disciplinary fields applied to political, military, economic, ecological, or climatic tasks, among others. A scenario is defined as “a postulated or projected situation or sequence of potential future events; (also) a hypothetical course of events in the past, intended to account for an existing situation, set of facts, etc.” (OECD, 2017). Thus, for example, the Intergovernmental Panel on Climate Change (IPCC), in its latest projection, foresees a scenario where the global temperature over the next 20 years will reach or exceed warming of 1.5°C (IPCC et al., 2021). In this case, a scenario is considered not only as a hypothetical future but also a valid tool to compare different action strategies. In the case of land use change scenarios, these are extraordinarily important for planning. However, there is no consensus on their definition: on the one hand, they can be understood as the storylines around a spatial modeling exercise. On the other hand, the future land uses map itself (Escobar et al., 2018). In our case, it is of interest to highlight the potential of spatially explicit scenarios through models (maps) that indicate where land use changes will occur.

Scenarios involve different degrees of uncertainty. However, in the case of urban use change, at least in regions such as Latin America, it is highly probable that the rapid urbanization dynamics will continue to advance in the coming decades, causing significant territorial impacts, especially in the areas of influence of future metropolises, new conurbations, and emerging medium-sized cities.

The progress of spatially explicit scenarios has been underdeveloped in developing countries, despite rapid urban growth such as that experienced in Latin America. The main contributions come primarily from the northern hemisphere and developed countries (Arsanjani, 2012; Clarke and Gaydos, 1998; Liu et al., 2014; Verburg and Overmars, 2007). Spatially explicit models are often based on Cellular Automata (CA); however, numerous models consider this type of technique or are combined with others. For several authors (Pontius et al., 2004; Sapena et al., 2017; Trisurat et al., 2010; Verburg et al., 2002), techniques associated with statistical regressions are one the best calibration methodologies, while CA’s are considered a decision model that analyzes whether the state of a pixel explicitly takes into account the current state of the neighboring pixel. Similarly, for Sapena et al. (2017), there are other specific techniques such as exogenous quantification methods where the quantification of each category is specified in the prediction of maps, independent of their location or agent-based models.

What is the importance of spatially explicit scenarios? First, they play a relevant role in early decision-making to reverse future impacts of a course of action or trends and avoid future problems. ESPON (2015) states that scenarios can be used to communicate ideas and discuss possible territorial developments, the effects of policies with territorial impact, and policy decisions that must be made. Ideally, scenario planning is supported by quantitative and qualitative techniques: the former provides figures that support all the assumptions included; while the latter facilitates the building of arguments that help represent the scenario in a narrative and realistic manner (Escobar et al., 2018). Namely, scenarios are a key input for territorial planning and governance. However, land use change scenarios in the case of Chile are less used results in instances of territorial planning or strategic environmental assessment. Hence it is estimated that their use in these or similar situations can significantly help to choose development alternatives and options that allow for comprehensive and sustainable territorial development.

Material and methods

Study area

The Quillota Province is in the Valparaiso Region, Chile, and, according to the 2017 population census, it has a population of 203,277 inhabitants. It covers an area of 1638.7 km², divided into five districts: Quillota, La Cruz, La Calera, Nogales, and Hijuelas (Supplementary figure S1).

Land use changes between 1990 and 2017

Landsat 4 Thematic Mapper (TM), Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and Landsat 8 Operational Land Imager (OLI) satellite images obtained from the Earth Explorer Platform (United States Geological Survey, 2018) were used to determine land uses for 1990, 2003, and 2017, respectively (Supplementary table S1). The images were pre-processed using geometric, radiometric, and atmospheric corrections (Pu et al., 2015). The FLASHH method (Kaufman et al., 1997), which is included in the ENVI software (ENVI, 2009), was used for radiometric and atmospheric corrections. In the case of the 1990 Landsat-4 image, a median-based low-pass smoothing filter was applied, considering a 5x5 filter matrix in order to smooth the image. In the case of geometric correction, the satellite images were adjusted according to the Chilean National Statistics Institute’s census blocks for the urban areas of Quillota, La Calera, La Cruz, Hijuelas, and Nogales, projected on datum WSG 84, Zone 19 South.

To identify the types of land use/cover, the methodology of Local Climate Zones (LCZ) developed by Stewart and Oke (2012) and proposed by the World Urban Dataset and Access Portal Tools project (WUDAPT) was applied (WUDAPT, 2015) (Supplementary table S2). This allows different urban densities and land uses to be identified. The first step consisted in identifying the Region of Interest (ROI) for the Quillota Province. This ROI allows a clip (sub-scene) to be generated in the Landsat images obtained for 1990, 2003, and 2017. The second step consisted in digitizing training sites for each period analyzed, with a total of 246 in 1990, 264 in 2003, and 312 in 2017, using random samples which are representative of the total ROI for each land use/cover, redrawing the grids of the Landsat images adjusted to the ROI to a 30-by-30 m quadrant. The third step involved supervised classifications of pre-processed satellite images using SAGA GIS (v.6.3.0) in the LCZ Classification module, which integrates the previous steps, and incorporates the Random Forest Classifier, which is applied to the multispectral Landsat images from 1990, 2003, and 2017 and whose number of trees was 6 (Bechtel et al., 2015). Last, an LCZ map was created for each period. The maps were inspected visually using Google Earth Pro (Bechtel et al., 2017) and onsite visits to assess whether the results obtained coincided with the landscape of the Quillota Province.

Calibration and model validation

Accuracy assessment of the land use maps

For the accuracy assessment of the land use maps, 300 sample points for each land use/cover type were generated using a stratified random sampling method (Congalton and Green, 2019). The application of this method implicitly assumes that each sample point, when classified, belongs completely to one of the classes in an exhaustively defined set of discrete and mutually exclusive classes. These control points were matched by visual inspection using images obtained with Google Earth Pro (GEP) (Calva et al., 2019; Hu et al., 2013; Pathan et al., 2021) using 1990, 2003, and 2017 as reference years. The GEP images for the year 1990 were supported with historical geospatial information corresponding to the land use and native forest map scale 1:50.000 of the National Forestry Corporation (Conaf-Conama, 1994), Municipal Regulatory plans for Quillota (Municipality of Quillota, 1991) and La Calera (Municipality of La Calera, 1992), the census cartography for the year 1992 (INE, 1992) and the land registry of Consolidated Urban Areas for the year 1990 of the Ministry of Housing and Urban Planning (MINVU, 2020). The GEP images for 2003 and 2017 were the high-resolution (<5 m) Quick Bird and RapidEye, respectively. Accuracy assessment statistics were then generated from the population error matrix proposed by Olofsson et al. (2013) and Pontius et al. (2008), allowing accuracy measures such as user’s accuracy, producer’s accuracy, and overall accuracy to be calculated. In addition, Kappa coefficient was used to assess the level of spatial agreement at a single time point (Varga et al., 2019).

The overall accuracy is the proportion of area classified correctly and thus refers to the probability that a randomly selected location on the map is classified correctly (FAO, 2016; Olofsson et al., 2013).

The relationship between these different indicators of land use/cover classification accuracy allows inferring that if use accuracy, producer accuracy, and overall accuracy are high, the estimated Kappa coefficient will also be high (Foody, 2002).

The accuracy levels obtained from the classification of land uses and cover in the Quillota Province reveal that the overall accuracy of the classification is 86.9% (1990), 91.1% (2003), and 90.2% (2017). Regarding urban use classifications (high, medium, and low density), one can see that, on average, user accuracy rates of over 90% are obtained, while in the case of the producer accuracy indicator, average values of 88% are obtained (Supplementary table S3).

Validation of the Dyna-CLUE model

To validate the proposed Dyna-CLUE model, Pontius et al. (2008) proposed indicators—Figure of Merit, Producer’s accuracy, and User’s accuracy—were used to validate using 2003 observed map, 2017 observed map, and 2017 simulated map. The 2017 simulated map was carried out using 2003 as the base reference year.

The Dyna-CLUE model was validated for the Quillota Province land use/cover by overlapping three maps containing the observed levels for 2003 and 2017, along with the simulated ones for 2017. Results show that the value is 10.0% for Figure of Merit, 17.7% for User Accuracy, and 18.8% for Producer Accuracy (Supplementary figure S2).

Definition of scenarios and simulation of future land use/cover

The Dyna-CLUE model (Dynamic Conversion of Land Use and its Effects) (Verburg and Veldkamp, 2002) was used to identify land use changes between 2017 and 2050. This spatial model is based on conversion rules for existing land uses/cover in a region, distribution of the spatial probability of driving forces of land use/cover changes, and expected patterns of change for each land use type from the base year (Tian et al., 2018).

Definition of scenarios of land use/cover change

For the application of the Dyna-CLUE model, three scenarios of land use/cover change to 2050 were established: trend, spatial planning, and ecological planning. The trend scenario corresponds to the simulation of the different land uses/cover trajectories in 1990, 2003, and 2017, following the Business As Usual (BAU) trend. The spatial planning scenario corresponds to a set of policies, regulations, and instruments for spatial planning, with a focus on the urban issue. The main instruments considered are the General Law of Urbanism and Construction (LGUC) (1975), the National Urban Development Policy (2014), and the “La Campana” Intercommunal Regulatory Plan (PRI) (in approval) including the districts of Quillota, La Calera, La Cruz, Nogales, Hijuelas, Olmué, and Limache. The PRI is mandatory and defines the permitted and prohibited land uses. Using the Dyna-CLUE Locspec option, ranging from 0 to 1, the PRI urban extension areas were given a higher preference for urban change. Finally, the ecological planning scenario is based on laws, rules, and regulations that help to protect ecological conservation areas. This includes the National System of State Wildlife Protected Areas (SNASPE), the priority sites for ecological conservation defined by the Ministry of the Environment (MMA), and the areas restricted for urban development by the PRI.

Land-use demands

Calculation of land use/cover demands for the three scenarios to 2050 consisted of estimating the simple regression curve based on the observed surface in 2017. For the trend scenario, the simple linear estimate considers the growth rates for the different land uses/cover between 1990, 2003, and 2017. For the conservationist and regulated scenarios, this linear estimate was calibrated with in-person semi-structured interviews of the district and regional-level decision-makers as a function of applying spatial policies. These land-use demands are adjusted according to the 5% margin of error (Table 1).

Table 1.

Land use demand in the Quillota Province by 2050. Urban high density (UHD), Urban medium density (UMD), Urban low density (ULD), Agricultural land (AGR), Forest (FOR), Bare soil (BS), and Water bodies (WB).

	Land use demands (ha)
Land use/cover	Trend scenario	Spatial planning scenario	Ecological planning scenario
Land use/cover	Trend analysis	Local stakeholders surveys
UHD	958.4	857.8
UMD	1638.7	1511.5
ULD	1338.7	1064.2
FOR	68,871.5	71,256.3
AGR	35,250.7	33,709.7
BS		(*)
WB		(*)

Note: (*) not simulated.

Location characteristics (driving forces)

Table 2 shows the driving forces used in the model. The first step is to identify the explanatory variables of the land use change model. For this purpose, 19 variables have been proposed which have been used in similar studies (Arsanjani et al., 2013; Henríquez-Dole et al., 2018; Poelmans & Van Rompaey, 2010; Salazar et al., 2020). The second step entails applying Cramer’s V test to obtain the statistical significance of dependent variables (land use/cover) and exploratory variables (driving forces). This statistical measurement is used before applying the logistic regression model, and its purpose is to select the statistically most significant exploratory variables (p-Value <0.05) for the logistic regression model that must be applied to each land use and cover (dependent variable). Cramer’s V test generates an index from 0 to 1 and weights the power of the exploratory variable (driving forces), where values of over 0.15 are considered good, and values over 0.4 are considered very good (Eastman, 2012). The exploratory variables are then analyzed using the logistic regression model (Lagrosa IV et al., 2018).

Table 2.

Driving forces used for logistic regressions for land/cover use urban high density (UHD), urban medium density (UMD), urban low density (ULD), agricultural land (AGR), forest (FOR), and bare soil (BS).

Variables	Variables selected (X) by Cramer’s V test						Source
	UHD	UMD	ULD	FOR	AGR	BS
Proximity to commercial activities (m)	X	X	X	X	X	X	OpenStreetMap (2017)
Proximity to the main city center (m)	X	X	X	X	X	X	INE (2017)
Proximity to secondary urban center (m)	X	X	X	X	X		INE (2017)
Population density (Inhabitants/ha)	X	X	X	X	X	X	INE (2017)
Proximity to primary and secondary schools (m)	X	X	X	X	X	X	Ministry of Education (2015)
Proximity to health facilities (m)	X	X	X	X	X	X	Ministry of Health (2013)
Slope (degrees)			X		X	X	JAXA/METI (2010)
Building permit (categorical definition/km²)		X	X	X	X	X	Chilean Chamber of Construction (2016, 2017)
Housing prices (index-linked unit) (*)	X	X	X	X	X	X	Portal Inmobiliario (2017)
Housing density (housing/ha)		X	X	X	X	X	INE (2017)
Proximity to road network (m)	X	X	X		X	X	OpenStreetMap (2017)
Property plot size (ha)			X	X	X	X	CIREN (1996) and SII (1996)
Proximity to potable water operating area (m)	X	X	X	X	X	X	SISS (2016)
Soil capability (class I–VIII)	X	X	X	X	X	X	CIREN (2010)
Risk erosion (class 1–5)			X	X	X	X	CIREN (2010)
Proximity to irrigation channel (m)				X	X	X	National Irrigation Commission (2018)
Proximity to water intake (m) (**)					X	X	Ministry Public Works (2017)
Proximity to agroindustries (m) (**)			X		X	X	Ministry of Agriculture (2018)
Elevation (m)					X	X	JAXA/METI (2010)

Note: (*) The variable “Unidad de Fomento” (UF) is an inflation-pegged Chilean monetary unit that is adjusted daily, whose value as of 31 March 2019 was Ch$27,565.76, equivalent to US$36.10. (**) This corresponds to the calculation of the distance of lower cumulative cost for each pixel, to or from the source of lower-cost on a cost surface, which in this case is indicated by the DEM.

The Geographical Information System Idrisi Selva (Eastman, 2012) was used to estimate the two statistical procedures (Cramer’s V test and logistic regression), importing files in ASCII format containing georeferenced data applied to a 30x30 m resolution.

Spatial policies and restrictions

In addition to the land-use demands for the different proposed scenarios and driving forces, the proposed Dyna-CLUE model considered areas that enable and restrict the changes in land use/cover to 2050. In both cases, the main source of information was the Inter-Municipal Regulatory Plan “La Campana” (MINVU, 2015) which is a Territorial Planning Instrument at regional scale according to LGUC. In this plan, restriction areas correspond to areas of protection of resources of environmental value and spatial policies, or areas of specific location, correspond to urban extension zones established by this Inter-Municipal Regulatory Plan.

Land use type specific conversion settings

Land use type-specific conversion settings determine the temporal dynamics of the simulations (Verburg et al., 2002). The Dyna-CLUE model requires two sets of parameters to characterize land use/cover conversion dynamics, corresponding to elasticities and land use conversion matrix. The former is explained in detail in Integration and application of the Dyna-CLUE model. The conversion matrix showing land use/cover changes was then built. These conversion rules were developed using a 7x7 matrix, where direct relationships in land use t₁ result in a land change t₂. Land use values with the same category are assigned a value of 1. The value is 0 if there is no probability of land use type t₁ changing to land use type t₂. Thus, the probability that an agricultural use will change to urban is highly probable, that is, it has a value of 0 (easy conversion). Meanwhile, it is improbable that an urban use will change to agricultural, that is, a value of 1 (irreversible change) (Supplementary table S4).

Integration and application of the Dyna-CLUE model

For data integration into the Dyna-CLUE model, geospatial information was used to identify the demand by applying future land use/cover allocations, using the formulae proposed by Verburg et al. (2002) for total probability ( $T P R O P_{i, u}$ ) (Supplementary figure S3)

T P R O P_{i, u} = P_{i, u} + E L A S_{u} + I T E R_{u}

where

P_{i, u}

is the ideal location of each cell

i

for each land use/cover

u

, derived from the application of the logistic regression model for land use/cover for the period 1990–2017.

E L A S_{u}

corresponds to reversibility of land use change or elasticity of each land use/cover, and it is the result of the analysis on the level of understanding of each stakeholder in the territory, its value fluctuating from 0 (easy conversion) to 1 (irreversible change) (Supplementary table S5). Values were assigned by observing past land use changes in 1990–2003–2017, based on similar results obtained by Henríquez-Dole et al. (2018). Likewise,

I T E R_{u}

corresponds to the probability factor given by the annual land use/cover demand value. The proposed simulation model uses the Cellular Automata (CA) technique to measure neighborhood properties for each land use/cover using the total probability formula (

T P R O P_{i, u}

) (Verburg et al., 2004).

Results

Land use trends 1990–2017

As for land use/cover trends between 1990 and 2017, one can observe that there are significant changes in the interior of the Quillota Province, with an increase in the surface of both urban and agricultural areas. Urban land uses accounted for 1077.7 ha and 1513.4 ha in 1990 and 2003, respectively, representing an average increase of 37.2% for high-, medium, and low-density urban uses. This represents an increase of 435.7 ha in that period. However, agricultural uses, represented mainly by less intensive crops (e.g., vegetables or fruit trees), increased by a smaller percentage compared to the total area (11.2%) (Supplementary table S6).

Following the same trend, during the 2003–2017 period, changes in land use/cover distribution were marked by asymmetries in occupation. On average, urban uses increased by 121.6%, a process marked by the growth of low-density urban areas (228.4%). This exponential growth of 1763.9 ha between 1990 and 2017 is spatially reflected in new housing projects built on the outskirts of Quillota. At the same time, the city border with the northern area of La Cruz begins to blur.

Regarding changes in land use/cover in the Quillota Province, 89.2% of the surface area maintained its status between 1990 and 2017 (Supplementary table S6). High, medium, and low-density urban uses are the ones that show the least change, as is the case with water bodies and bare soil. Regarding changes in total surface area, the change in land use/cover shows that 7668.2 ha of forest cover in 1990 had been converted to agricultural use in 2017, revealing changes in the current surface area occupied by agricultural lands and an expansion in the agricultural frontier.

Another of the main changes affecting land use/cover trends corresponds to the net changes in different land uses. For the 1990–2003 period, the forested land cover fell by 3250.7 ha, one of the most significant land use changes (Supplementary figure S4). These losses can be explained by gains in agricultural and urban areas, 2914 ha and 436 ha, respectively.

This trend is quite similar for the 2003–2017 period, where the decrease in forest cover (dense vegetation) totals 4496 hectares (Supplementary table S6). However, there was also a lesser increase in agricultural land cover (10.9%) in this period.

Land use change model 2017–2050

Based on the results obtained in the analysis of trajectories in land use/cover for the system of cities in the Quillota Province, urban growth scenario maps to 2050 were simulated from a combination of conversion and restriction rules, cover demands and land uses, and the analysis of spatial variables in the Dyna-CLUE model. In general, it is expected that by 2050 over 95% of the area will not change, regardless of the proposed scenario (Figure 1).

Figure 1.

Land use/cover simulation scenarios maps by 2050.

Regarding the analysis of the logistic regression variables for each land use/cover, it can be said that the selected driving forces explain their distribution, as indicated by the high values of the ROC statistical test (>0.9). Additionally, this statistical analysis shows that high-density urban land use correlates directly with potable water production facilities (PWOA): the closer the city is to a water production facility, the higher the probability of future urban activity. However, this trend is not the same when related to medium- and low-density urban land uses, where the relationship is positive but not as strong (Supplementary table S7).

In the case of high- and medium-density urban land uses, there is a positive relationship with respect to proximity to major and secondary urban centers, as well as proximity to commercial centers and areas with higher population density. However, these same variables have a negative impact on low-density land use. At the same time, agricultural land use also shows a high correlation with proximity variables (distance to agro-industries or water intake operations), probably because they play a key role in the economic growth of the area. Other variables influencing future urbanization expansion are housing prices and proximity to road networks (positive correlation).

Aggregate changes between 2017 and 2050 were segmented to present conversions for each land use/cover type (Supplementary table S8). These changes are derived from the specifications of conversion elasticities entries in the Dyna-CLUE model. The main assumption for the different scenarios was that land converted to residential areas and industrial uses cannot be converted into other uses. Forest loss totals 4% for 1990–2003 and 5.8% for 2003–2017.

Trending scenario

For this scenario, the urban area is expected to increase by 31.4% compared to 2017. This increase is particularly detrimental to agricultural areas. The urban areas experiencing the most significant expansion correspond to medium- and low intensity. This trend is observed in the rural and peri-urban sectors of the Hijuelas and La Cruz districts, toward northern and southern Quillota Province, respectively.

Likewise, new areas with a significant urbanization potential are beginning to emerge in the rural sectors of the Hijuelas district, the area of Las Palmas de Ocoa, adjacent to the La Campana National Park, which currently has less intensive housing units in plots of over 0.5 ha. The expansion of this residential development may be due to urban development pressures arising from its proximity to the province’s main population centers and major transportation networks, in addition to the average land value.

Spatial planning scenario

This scenario was estimated with the Dyna-CLUE model, using specific localization criteria, and based on the urban development areas determined by the La Campana Inter-Municipal Regulatory Land Use Plan (MINVU, 2015). Under these criteria, it is expected that urban expansion by 2050 will tend toward a complete conurbation of the cities of Quillota-La Calera-La Cruz, with a 48.7% increase in total area compared to 2017. At the same time, agricultural areas will decline sharply, with over 1364 ha becoming urban land.

New urban areas are at the same time emerging within the city system of the Quillota Province, such as the urban areas of Nogales, Hijuelas, and the satellite city of San Pedro, in the Quillota district. The main driving forces triggering the expansion of the Quillota-La Calera-La Cruz conurbation system are proximity to transportation networks (CH-60, mainly), principal and secondary urban centers, and, lastly, to education and health centers.

Ecological planning scenario

This scenario would produce fewer urban expansion areas than the other scenarios by 2050, while the development of agricultural areas would prevail. Compared to 2017, urban land will increase by 18.4%. Nevertheless, this increase in the area produces fragmentation of low-density urban areas toward the eastern part of Hijuelas. The impact of elasticity changes proposed for this scenario shows a difference of 616.4 ha (average) between the regulated and trend scenarios compared to the conservation scenario.

Proximity to transportation networks and urban centers (main and secondary) continue to be the driving factors behind the urban expansion in the Quillota Province. Meanwhile, the Quillota-La Calera-La Cruz conurbation continues to grow, feeding on adjacent industrial land and rural residential plots in the Hijuelas area.

Discussion

The urban growth in the Quillota Province will concentrate in satellite areas due to the loss of agricultural land to urban residential areas, particularly evident in areas in the south and east of the Province. This trend projecting the Dyna-CLUE simulation model on the form of land use occupation is consistent with the projected demand for land to 2050 as expressed in interviews by different local stakeholders of Quillota Province. Thus, regarding the future of the Quillota-La Calera-La Cruz urban system, the Valparaiso Region Urban Development Division of the Ministry of Housing and Urban Planning (MINVU, 2015) estimates that urban land will increase by 80%–2050.

In addition, net changes in land use/cover relate directly to rural areas. Supplementary table S9 shows the growing pressure to change agricultural land use in favor of new urban areas between 1990 and 2017, 1618.4 ha of urban land use in 2017 was agricultural in 1990. A clear example reflecting this relationship is the rise of two major land use-related problems derived from agricultural activities. The first refers to horizontal vegetable plantations and the vertical avocado plantations on mountain slopes. Avocados are sold both inside and outside the province (Negrete and Hidalgo, 2009). The second land use-related problem is caused by cement factories. The extraction of aggregates and the operation of quarries have led to the degradation of forests at altitudes of over 600 m in the La Calera and Hijuelas districts, preventing these areas from being reused or reconverted into environmental protection or agricultural areas. It should be noted that the expansion of the agricultural frontier has a relationship on the urbanization process of surrounding cities. The reasons are that urban land uses—mainly the middle—and working-class housing units—compete with highly profitable agricultural land uses.

The Dyna-CLUE model is used to recreate various scenarios based on how space is being constructed (Mas et al., 2014), where the most significant impacts affecting urban space uses can be found in the trend and regulated scenarios. The trend scenario simulates the dynamics of urban growth in the context of the current neoliberal model, that is, without regulatory restrictions. Indeed, urban pressure is expressed in the potential creation of new settlements on rural lands, such as new urban medium-density areas in San Pedro (Quillota) and low-density areas in Ocoa (Hijuelas), and El Melón (Nogales), both with great natural value. Urbanization in rural areas can be explained by legal loopholes such as Law Decree Nº 3.516. Gated communities and low-density residential developments in rural and peri-urban areas are clear examples of this legal subterfuge. In the case of spatial planning and ecological planning scenarios, the Dyna-CLUE model proposes a more contained and compact urban growth since it favors urbanization within urban extension areas (PRI) and excludes restriction areas considering instruments of biodiversity protection (SNASPE and MMA). At the local level, there are district regulatory plans (PRC), but these were not considered because they are outdated.

The generation of urban growth scenarios to 2050 illustrates other tensions in the use of space at the regional level related to the creation of new land administration areas on the conurbation scale (Hidalgo et al., 2009). Likewise, one of the main effects is the expansion of the agricultural frontier, particularly avocado plantations, into forested areas of the Coastal Mountain Range (Nunes de Oliveira et al., 2017). It should be pointed out that from 1990 to 2050, the agricultural frontier would rise in altitude by an average of 87 m and reach a maximum of 200 m above sea level. This process is particularly marked in the trend scenario (Supplementary figure S5).

In turn, this process could affect areas in the basin with significant ecological and conservation value. For example, in the interviews carried out with key stakeholders in the Quillota Province’s territorial planning, they stated that: “…Avoid real estate development in the northern sector to keep it as a green lung (rural areas) for the city and avoid plot subdivisions of 5000 m². This is happening intensively in the area of Porvenir (north sector) and Pocochay (east sector)” (Local assessor, 12/06/2018). It is possible to observe pressures on the sectors of Ocoa (District of Hijuelas), El Melón (District of Nogales), and La Campana National Park (south of the District of Hijuelas), one of the most critical areas on the regional and national levels, due to its ecological importance and its condition as a deciduous forest hotspot in Chile’s central region (Salazar et al., 2015).

To properly apply the Dyna-CLUE model, the public participation process that stakeholders interrelate with is key to the creation of urban growth scenarios, understood within the non-spatial variables that the spatially explicit simulation models consider. Calibration and validation of the Dyna-CLUE model enable the inclusion of stakeholders in the regional land plan. Thus, asymmetries in information on the type and future land use demands are reduced.

Internationally, studies that have applied the CLUE model have shown that by simulating and projecting changes in land use/cover, incorporating both socio-economic and bio-geographic variables, it is possible to estimate diverse scenarios successfully. For example, business-as-usual scenarios, regulated scenarios, strategic planning scenarios, among others (Aduah and Mantey, 2020; Henríquez-Dole et al., 2018; Kang et al., 2019; Luo et al., 2010; Pokojska, 2019; Price et al., 2015; Salazar et al., 2020; Verburg et al., 2002; Xu et al., 2013). These cases demonstrate the importance of comparing scenarios with different degrees of urbanization and planned scenarios (different objectives) for better decision-making.

Conclusion

The work demonstrates the potential contribution and viability of spatially explicit models to support decision-making in urban-regional planning and to explore different regional futures (Jantz et al., 2004). In our analyzed case, the choice of not planning and continuing with a BAU scenario foresees an inorganic, fragmented, and conflicting urban growth, especially for the low-density urban use with the rest of the land uses in the Quillota province. On the other hand, the spatial planning scenario shows a consolidation of existing urban patches and an emerging process of conurbation between La Calera-La Cruz-Quillota, and even Hijuelas, which will require a new governance model for its proper management. Finally, the ecological planning model emphasizes the containment of urban growth and the enhancement and protection of the areas of high natural value. Based on these scenarios, better decisions can be made, and irreparable damage associated with land use change can be avoided, especially the impact generated by the increase in urban use, which is a practically irreversible process.

The decision to choose any of these scenarios requires adopting sustainability criteria that allow better management of land use dynamics in order to avoid future spatial conflicts. Such planning must lead to effective sustainability, actions and objects that support harmonious uses between economic interests, community positioning, and public policies at national, regional, and local scales. The use of planning ensures the reduction of margins of errors and conflict among these actors, especially in rural areas where rapid urbanization processes occur, often in an infiltrated and diffuse manner. Thus, territorial planning should be more strategic in order to anticipate future land use problems and, above all, not regularize urbanization processes already underway.

Other considerations include the participation processes in the intermediate stages of the Dyna-CLUE model, as it can result in better adjustments for the selection of driving forces, spatial policies, restrictions, determining future land use demands, and final validation. On the other hand, one of the main limitations of the Dyna-CLUE is the difficulty of modeling urban growth in height and not only horizontally. In addition, it is important to include new driving forces in the model, such as top-down variables (i.e., environmental variables linked to climate change) and bottom-up variables (i.e., type of real-estate owner). It is also important to consider the Regional Land Use Plan (PROT) of Valparaíso in the next steps modeling.

Why are spatially explicit scenarios not a widely used procedure in countries such as Chile? Traditional and legal planning mechanisms do not consider them, although they offer great potential and advantages for urban-regional planning. Moreover, they are particularly useful in the Strategic Environmental Assessment of territorial planning instruments, such as the PRI and PRC, to be able to evaluate and choose the best alternatives and development options.

Supplemental Material

Supplemental Material—Future land use conflicts: Comparing spatial scenarios for urban-regional planning

Supplemental Material for Future land use conflicts: Comparing spatial scenarios for urban-regional planning by Cristian Henríquez, Mauricio Morales, Jorge Qüense and Rodrigo Hidalgo in Environment and Planning B: Urban Analytics and City Science

Footnotes

Acknowledgments

The authors thank the support received by FONDECYT project no.1180268 and no. 1220688 ANID/Chile and by CEDEUS, ANID/FONDAP 15110020. Special thanks to Lenin Henríquez-Dole and Paulina Contreras. We would especially like to thank the districts of La Calera and La Cruz, the Urban Development Department of the Valparaiso Region Urban Development Undersecretariat of the Ministry of Housing and Urban Planning, and the Research Department of the Chilean Chamber of Construction A.G. for their collaboration. Finally, we are very grateful for the relevant comments of the peer-reviewers.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Centro de Desarrollo Urbano Sustentable (15110020), Fondo Nacional de Desarrollo Científico y Tecnológico (1180268), Fondo Nacional de Desarrollo Científico y Tecnológico (1220688).

ORCID iD

Cristian Henríquez

Mauricio Morales

Supplemental material

Supplemental material for this article is available online.

References

Aduah

Mantey

(2020) Modelling potential future urban land use changes in the sekondi-takoradi metropolitan area of ghana. Ghana Journal of Technology 4(2): 26–32.

Arsanjani

(2012) Dynamic Land-Use/Cover Change Simulation: Geosimulation and Muti Agent-Based Modelling. New York, NY: Springer Publication.

Arsanjani

Helbich

Kainz

, et al. (2013) Integration of logistic regression, Markov chain and cellular automata models to simulate urban expansion. International Journal of Applied Earth Observation and Geoinformation 21(1): 265–275. doi: 10.1016/j.jag.2011.12.014.

Batty

(2013) The New Science of Cities. Massachusetts, MA: Massachusetts Institute of Technology.

Bechtel

Alexander

Böhner

, et al. (2015) Mapping local climate zones for a worldwide database of the form and function of cities. ISPRS International Journal of Geo-Information 4(1): 199–219. DOI: 10.3390/ijgi4010199

Bechtel

Demuzere

Sismanidis

, et al. (2017) Quality of crowdsourced data on urban morphology—The human influence experiment (HUMINEX). Urban Science 1(2): 15. DOI: 10.3390/urbansci1020015

Borsdorf

(2003) Cómo modelar el desarrollo y la dinámica de la ciudad latinoamericana. Eure 29(86): 37–49.

Brannstrom

(2009) South America’s neoliberal agricultural frontiers: places of environmental sacrifice or conservation opportunity. Ambio 38(3): 141–149. DOI: 10.1579/0044-7447-38.3.141

Calva

Golubov

Mandujano

, et al. (2019) Assessing google earth pro images for detailed conservation diagnostics of mangrove communities. Journal of Coastal Research 92(SI) Coastal Education and Research Foundation: 33–43.

10.

Castro

Aliaga

(2010) Evaluación de la pérdida de suelo, asociada al proceso de expansión urbana y reconversión productiva. Geografia Norte Grande 45: 41–49.

11.

Chilean Chamber of Construction (2016) Permisos de edificación 2016. Santiago de Chile: Gerencia de Estudios, CChC.

12.

Chilean Chamber of Construction (2017) Permisos de edificación 2017. Santiago de Chile: Gerencia de Estudios, CChC.

13.

CIREN (1996) Propiedades rurales para las comunas de La Calera, La Cruz, Hijuelas y Nogales. Santiago de Chile: Ministerio de Agricultura.

14.

CIREN (2010) Clase de Capacidad de Uso de los Suelos: Región de Valparaíso. Santiago de Chile: Ministerio de Agricultura.

15.

Clarke

Gaydos

(1998) Loose-coupling a cellular automaton model and GIS: long-term urban growth prediction for San Francisco and Washington/Baltimore. International Journal of Geographical Information Science 12(7): 699–714. DOI: 10.1080/136588198241617

16.

Clarke

Hoppen

Gaydos

(1997) A self-modifying cellular automaton model of historical urbanization in the San Francisco Bay area. Environment and Planning B: Planning and Design 24(2): 247–261. DOI: 10.1068/b240247

17.

Conaf-Conama (1994) Catastro y Evaluación de Los Recursos Vegetacionales Nativos de Chile. Santiago, Chile: Región de Valparaíso.

18.

Congalton

Green

(2019) Assessing the Accuracy of Remotely Sensed Data. Principles and Practices. Boca Raton, Florida: Routledge.

19.

Eastman

(2012) Software Idrisi Selva. Massachusetts, MA: Clark University.

20.

Engelen

Lavalle

Barredo

, et al. (2007) The MOLAND modelling framework for urban and regional land-use dynamics. In Modelling Land-Use Change. Springer, pp. 297–319.

21.

ENVI (2009) Atmospheric Correction Module: QUAC and FLAASH User`s Guide. IDL.

22.

Escobar

Van Delden

Hewitt

, et al. (2018) LUCC scenarios. In: Mas J-FPaegelow M (ed), María Teresa COGeomatic Approaches for Modeling Land Change Scenarios. Cham, Switzerland: Springer, pp. 81–97.

23.

ESPON (2015) Territorial Scenarios for Europe, Working Paper. Annex to the ESPON Policy Brief “Territorial Scenarios for Europe Towards 2050”. ESPON.

24.

FAO (2016) Map Accuracy Assessment and Area Estimation: A Practical Guide. Rome: FAO.

25.

Foody

(2002) Status of land cover classification accuracy assessment. Remote Sensing of Environment 80(1): 185–201. DOI: 10.1016/S0034-4257(01)00295-4

26.

Henríquez-Dole

Usón

Vicuña

, et al. (2018) Integrating strategic land use planning in the construction of future land use scenarios and its performance: The Maipo River Basin, Chile. Land Use Policy 78: 353–366.

27.

Henríquez

(2014) Modelando El Crecimiento de Ciudades Medias. Hacia Un Desarrollo Urbano Sustentable. Santiago, Chile: Ediciones Universidad Católica de Chile.

28.

Hidalgo

Arenas

Monsalve

(2009) La conurbación la serena - coquimbo: problemas y desafios para su transformación metropolitana. In: Hidalgo

De Mattos

Arenas

(eds), Chile: Del País Urbano Al País Metropolitano. Santiago, Chile: Serie GEOlibros No12-Colección Eure Libros, pp. 161–186.

29.

Xia

, et al. (2013) Exploring the use of google earth imagery and object-based methods in land use/cover mapping. Remote Sensing 5(11): 6026–6042. DOI: 10.3390/rs5116026

30.

INE (1992) Cartografía Censo Nacional de Vivienda y Población 1992. Santiago, Chile: Instituto Nacional de Estadísticas, Ministerio de Economía, Fomento y Turismo.

31.

INE (2017) Cartografía Censo Nacional de Vivienda y Población 2017. Santiago de Chile: Instituto Nacional de Estadísticas, Ministerio de Economía, Fomento y Turismo.

32.

IPCC (2021) Summary for policymakers. In: Maycock

Waterfield

Yelekçi

, et al. (eds), Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, MA: Cambridge University Press, pp. 3–32. DOI: 10.1017/9781009157896.001

33.

Jantz

Goetz

Shelley

(2004) Using the SLEUTH urban growth model to simulate the impacts of future policy scenarios on urban land use in the Baltimore-Washington metropolitan area. Environment and Planning B: Planning and Design 31(2): 251–271. DOI: 10.1068/b2983

34.

JAXA/METI (2010) ALOS PALSAR RTC Data . Available at: https://asf.alaska.edu/ (accessed March 2018).

35.

Kang

Sui

Yang

, et al. (2019) Spatial pattern consistency among different remote-sensing land cover datasets: a case study in northern Laos. ISPRS International Journal of Geo-Information 8(5): 201. DOI: 10.3390/ijgi8050201

36.

Kaufman

Wald

Remer

, et al. (1997) The MODIS 2.1-μm channel-correlation with visible reflectance for use in remote sensing of aerosol. IEEE Transactions on Geoscience and Remote Sensing 35(5): 1286–1298.

37.

Lagrosa IV

Zipperer

Andreau

(2018) Projecting land-use and land cover change in a subtropical urban watershed. Urban Science 2: 1–23.

38.

Liu

Feng

Pontius

(2014) Spatially-explicit simulation of urban growth through self-adaptive genetic algorithm and cellular automata modelling. Land 3(3): 719–738. DOI: 10.3390/land3030719

39.

Luo

Yin

Chen

, et al. (2010) Combining system dynamic model and CLUE-S model to improve land use scenario analyses at regional scale: a case study of sangong watershed in Xinjiang, China. Ecological Complexity 7(2): 198–207.

40.

Martner

(2016) Expansión dispersa, ciudad difusa y transporte: el caso de Querétaro, México. Eure 42(125): 31–60. DOI: 10.4067/S0250-71612016000100002

41.

Mas

Kolb

Paegelow

, et al. (2014) Inductive pattern-based land use/cover change models: a comparison of four software packages. Environmental Modelling and Software 51: 94–111. doi: 10.1016/j.envsoft.2013.09.010

42.

Ministry of Agriculture (2018) Localization of agro-industrial projects. Available at: https://www.ide.cl/ (accessed March 2018).

43.

Ministry of Education (2015) Establecimientos de educación escolar. Available at: https://www.ide.cl/ (accessed March 2018).

44.

Ministry of Health (2013) Establecimientos de salud de Chile. Available at: https://www.ide.cl/ (accessed March 2018).

45.

Ministry Public Works (2017) Localización de bocatomas en la Región de Valparaíso. Available at: https://www.ide.cl/ (accessed March 2018).

46.

MINVU (2007) Medición de la superficie ocupada por las ciudades de Chile de más de 15.000 habitantes: 1993-2003. Santiago, Chile: Observatorio Urbano, Ministerio de Vivienda y Urbanismo.

47.

MINVU (2015) Anteproyecto Plan Regulador Intercomunal ‘La Campana’, Región de Valparaíso. Valparaíso, Chile: Servicio de Evaluación Ambiental.

48.

MINVU (2018) Metodología para medir el crecimiento urbano de las ciudades de Chile (CEHU del Ministerio de Vivienda y Urbanismo Ed). Santiago, Chile: MINVU.

49.

MINVU (2020) Área Urbana de la Ciudades Chilena 1993-2020. Santiago, Chile: Ministerio de Vivienda y Urbanismo.

50.

Moulds

Buytaert

Mijic

(2015) An open and extensible framework for spatially explicit land use change modelling: the lulcc R package. Geoscientific Model Development 8(10): 3215–3229. DOI: 10.5194/gmd-8-3215-2015

51.

Municipality of La Calera (1992) Municipal Regulatory Plan of La Calera .

52.

Municipality of Quillota (1991) Municipal Regulatory Plan of Quillota .

53.

National Irrigation Commission (2018) Localización de canales de regadío, Región de Valparaíso. Available at: https://www.ide.cl/ (accessed March 2018).

54.

Negrete

Hidalgo

(2009) Barrios cerrados y procesos socioespaciales en las ciudades del valle de aconcagua. In: Hidalgo

De Mattos

Arenas

(eds), Chile: Del País Urbano Al País Metropolitano. Santiago, Chile: Serie GEOlibros No12-Colección EURE Libros, pp. 323–346.

55.

Nunes de Oliveira

Abílio de Carvalho Júnior

Trancoso Gomes

, et al. (2017) Deforestation analysis in protected areas and scenario simulation for structural corridors in the agricultural frontier of Western Bahia, Brazil. Land Use Policy 61: 40–52. doi: 10.1016/j.landusepol.2016.10.046

56.

OECD (2017) The Governance of Land Use in OECD Countries: Policy Analysis and Recommendations. Paris: OCDE.

57.

Olofsson

Foody

Stehman

S V

, et al. (eds) (2013) Remote sensing of environment making better use of accuracy data in land change studies: Estimating accuracy and area and quantifying uncertainty using strati fi ed estimation. Remote Sensing of Environment 129: 122–131. doi: 10.1016/j.rse.2012.10.031

58.

OpenStreetMap (2017). OpenStreetMap Data Extracts. Available at: https://www.openstreetmap.org/ (accessed 5 May 2018).

59.

Pathan

Ashwini

Sil

(2021) Spatio-temporal variation in land use/land cover pattern and channel migration in Majuli River Island, India. Environmental Monitoring and Assessment. Springer International Publishing, 193(12). doi: 10.1007/s10661-021-09614-w

60.

Poelmans

Van Rompaey

(2010) Complexity and performance of urban expansion models. Computers, Environment and Urban Systems 34(1): 17–27. doi: 10.1016/j.compenvurbsys.2009.06.001

61.

Pokojska

(2019) Simulating land use change in the southern part of warsaw metropolitan area with application of dyna-CLUE model. Geografie-Sbornik CGS 124(1): 57–82. DOI: 10.37040/geografie2019124010057

62.

Pontius

Huffaker

Denman

(2004) Useful techniques of validation for spatially explicit land-change models. Ecological Modelling 179(4): 445–461. DOI: 10.1016/j.ecolmodel.2004.05.010

63.

Pontius

Boersma

Castella

, et al. (2008) Comparing the input, output, and validation maps for several models of land change. Annals of Regional Science 42(1): 11–37. DOI: 10.1007/s00168-007-0138-2

64.

Portal Inmobiliario (2017) Información de oferta inmobiliaria en la Región de Valparaíso. Available at: https://www.portalinmobiliario.com/ (accessed 5 May 2018).

65.

Price

Kienast

Seidl

, et al. (2015) Future landscapes of Switzerland: risk areas for urbanisation and land abandonment. Applied Geography 57: 32–41. doi: 10.1016/j.apgeog.2014.12.009

66.

Landry

Zhang

(2015) Evaluation of atmospheric correction methods in identifying urban tree species with worldview-2 imagery. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 8(5): 12–1897.

67.

Ren

Lü

Comber

, et al. (2019) Spatially explicit simulation of land use/land cover changes: current coverage and future prospects. Earth-Science Reviews 190: 398–415. doi: 10.1016/j.earscirev.2019.01.001

68.

Salazar

Moreira-Muñoz

Del Río

(2015) La Campana-Peñuelas biosphere reserve in central Chile: threats and challenges in a peri-urban transition zone. Eco. Mont 7(1): 66–71. DOI: 10.1553/eco.mont-7-1s66

69.

Roweis

(2018) Urban planning in early and late capitalist societies: Outline of a theoretical perspective. In: Dear, M and Scott, A (eds) Urbanization and Urban Planning in Capitalist Society: 159–177. doi: 10.4324/9781351068000

70.

Salazar

Henríquez

Sliuzas

, et al. (2020) Evaluating spatial scenarios for sustainable development in Quito, Ecuador. ISPRS International Journal of Geo-Information 9(3): 141. doi: 10.3390/ijgi9030141

71.

Sapena

Ruiz

Joó

(2017) Modelos de simulación de expansión urbana a partir de imágenes de satélite: Adecuación al análisis temporal de la fragmentación de los usos del suelo. Nuevas plataformas y sensores de teledetección. XVII Congreso de la Asociación Española de Teledetección, Murcia, España, October 2017, pp. 311–314.

72.

SII (1996) Propiedades urbanas de la provincia de Quillota. Santiago de Chile. Servicio de Impuestos Internos.

73.

SISS (2016) Áreas de operación de agua potable. Available at: https://www.ide.cl/ (accessed March 2018).

74.

Soares

Cerqueira

Pennachin

(2002) DINAMICA - a stochastic cellular automata model designed to simulate the landscape dynamics in an Amazonian colonization frontier. Ecological Modelling 154: 217–235.

75.

Stewart

Oke

(2012) Local climate zones for urban temperature studies. American Meteorological Society 93: 1880–1900.

76.

Tian

XIA

Wenbin

Qingbo

, et al. (2018) Modeling the spatio-temporal changes in land uses and its impacts on ecosystem services in Northeast China over 2000–2050. Journal of Geographical Sciences 28(11): 1611–1625.

77.

Trisurat

Alkemade

Verburg

(2010) Projecting land-use change and its consequences for biodiversity in northern Thailand. Environmental Management 45(3): 626–639. DOI: 10.1007/s00267-010-9438-x

78.

United States Geological Survey (2018) Earth Explorer. Available at: https://earthexplorer.usgs.gov (accessed 5 May 2018).

79.

Varga

Pontius

Singh

, et al. (2019) Intensity analysis and the figure of merit’s components for assessment of a cellular automata – markov simulation model. Ecological Indicators 101(3): 933–942. doi: 10.1016/j.ecolind.2019.01.057

80.

Veldkamp

Fresco

(1996) CLUE-CR: an integrated multi-scale model to simulate land use change scenarios in costa rica. Ecological Modelling 91(1–3): 231–248. DOI: 10.1016/0304-3800(95)00158-1

81.

Verburg

Veldkamp

Limpiada

, et al. (2002) Modeling the spatial dynamics of regional land use: the CLUE-S model. Environmental Management 30(3) Springer: 391–405.

82.

Verburg

Overmars

(2007) Dynamic simulation of land-use change trajectories with the clue- s model. Modelling Land-Use Change 321–335. doi: 10.1007/978-1-4020-5648-2_18

83.

Verburg

De Nijs

TCM

Van Eck

, et al. (2004) A method to analyse neighbourhood characteristics of land use patterns. Computers, Environment and Urban Systems 28(6): 667–690. DOI: 10.1016/j.compenvurbsys.2003.07.001

84.

Verburg

Overmars

(2009) Combining top-down and bottom-up dynamics in land use modeling: exploring the future of abandoned farmlands in Europe with the dyna-Clue model. Landscape Ecology 24(9): 1167–1181. DOI: 10.1007/s10980-009-9355-7

85.

Verburg

Veldkamp

(2002) Modeling the spatial dynamics of regional land use: the CLUE-S model. Environmental Management 30(3): 391–405. DOI: 10.1007/s00267-002-2630-x

86.

Verburg

De Koning

GHJ

Kok

, et al. (1999) A spatial explicit allocation procedure for modelling the pattern of land use change based upon actual land use. Ecological Modelling 116(1): 45–61. DOI: 10.1016/S0304-3800(98)00156-2

87.

Verburg

Soepboer

Limpiada

, et al. (2002) Modeling the spatial dynamics of regional land use: the CLUE-S model. Environmental Management 30(3): 391–405. DOI: 10.1007/s00267-002-2630-x

88.

Verburg

Eickhout

Meijl

(2008) A multi-scale, multi-model approach for analyzing the future dynamics of European land use. Annals of Regional Science 42(1): 57–77. DOI: 10.1007/s00168-007-0136-4

89.

WUDAPT (2015) World urban database and access portal tools project. Available at: http://www.wudapt.org (accessed 5 May 2018).

90.

Zhaoxue

Huimin

, et al. (2013) Land-use planning for urban sprawl based on the CLUE-S model: a case study of Guangzhou, China. Entropy 15: 3490–3506.

91.

Zaiatz

APSR

Zolin

Vendrusculo

, et al. (2018) Agricultural land use and cover change in the cerrado/amazon ecotone: a case study of the upper teles pires river basin. Acta Amazonica 48(2): 168–177. DOI: 10.1590/1809-4392201701930

92.

Ministry of Agriculture (2018) Localización de proyectos agroindustriales. Available at: https://www.ide.cl/ (accessed March 2018).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.77 MB