FUTURE LAND USECOVER FLOWS IN ROMANIA. SCENARIOS BASED ON CLUE-S MODEL AND CORINE LAND COVER DATABASE [308598]

FUTURE LAND USE/COVER FLOWS IN ROMANIA. SCENARIOS BASED ON CLUE-S [anonimizat]-[anonimizat], [anonimizat], Bianca Mitrică

The present study aims to analyze the past changes and to simulate this process for a better understanding of the future tendency in land use/cover dynamics in Romania. Based on the data derived from CORINE Land Cover Database and using the CLUE-S model (the Conversion of Land Use and its Effects at Small regional extent), future land use/cover change flows were modeled. The authors calibrated the model using two scenarios to explore potential future changes. These scenarios represent the historical trends in land use/cover dynamics calculated for two time intervals: 1990–2000 (HT1) and 2000-2006 (HT2). [anonimizat]/[anonimizat] 17 [anonimizat]. The results indicate a [anonimizat], [anonimizat], the open spaces with little or no vegetation and the agricultural complex cultivation patterns and a [anonimizat]/[anonimizat], the heterogeneous agricultural areas and the natural grasslands. This study could become useful tool for monitoring and quantifying spatial and temporal tendency of land use/cover changes in order to adopt appropriate land use planning and environmental policy in line with the sustainable development principles. Furthermore, the predicted map could directly be used to assess the natural hazards models in order to evaluate different scenarios of soil erosion or landslides susceptibility at national or regional level.

Key words: future land use/[anonimizat], CLUE-S model, Romania.

Introduction

Land use/cover changes is an widespread and accelerated process (Rawat & Kumar, 2015), considered one of the most important environmental issues of global concern (Guan et al., 2011; Veldkamp & Lambin, 2001), [anonimizat], [anonimizat] (IGBP, 1999). Land use/[anonimizat] (Lambin & Geist, 2006). Land use/[anonimizat]; [anonimizat] (Dale 1997; Watson et al. 2000). Therefore, it is acknowledged that land use/land cover changes and the associated habitat loss and fragmentation are major causes of biodiversity degradation (Sala et al., 2000). [anonimizat]/cover changes have become a [anonimizat], given their implications for global environmental change (Turner et al., 1993). [anonimizat]/cover change is a critical requirement for sustainable development because these changes can lead to land use conflicts due to the need for resources and space and the capacity of the land to absorb and support these needs (EEA, 2007).

Understanding land use/cover dynamics is essential in order to predict future changes accurately, and to facilitate the development of sustainable management practices designed to preserve essential landscape functions (Lin et al., 2007; Lin et al., 2009). Moreover, the analysis of the past and future land cover is significant to thoroughly investigate two of the major research questions dealing with land cover processes (Lambin, 1997): (1) to understand where exactly land use/cover change occurs, and (2) to assess the change rates. Therefore, using multi-temporal land use/cover spatial data, numerous models have proposed to explore this process. Land use/cover models have used to assess the cumulative impact of land use changes and to develop future scenarios (Veldkamp & Lambin, 2001), which help and support land use planning and decision making (Guan et al., 2011). Predicting land use/cover changes is important for understanding and highlighting the potential landscape transformations that might occur in the near future (Marwa et al., 2015). Many spatially explicit models have developed with the purpose of explaining and predicting the locations of land use/cover changes through either empirical-statistical models (Irwin & Geoghegan, 2001; Lambin et al., 2000; Mertens & Lambin, 1997) or rule-based simulation models, particularly cellular automata (Clarke et al., 1997; Kamusoko et al., 2009), agent-based models (Evans & Kelley, 2004; Mena et al., 2011; Parker et al., 2003), and hybrid which combines estimation and simulation models (Veldkamp and Fresco, 1996). One of the most used models to predict land use/cover changes based on driving factors is CLUE-S (Conversion of Land Use/Land Cover and its Effects at Small regional extent). This model is an example of advanced statistical land use change (Veldkamp and Fresco, 1996), an empirical multi-scale land use/cover change model developed for understanding and predicting the impact of biophysical and socio-economic forces that drive land use/cover change, used and validated in a wide range of applications (e.g. Veldkamp et al. 2001; Verburg & Veldkamp 2004, Castella et al. 2007, Wassenaar et al. 2007, Jinyan et al., 2007, Luo et al., 2010, Zhu et al., 2010). Therefore, CLUE-S is a process-built modelling framework that allows the user to develop spatially explicit future land use/cover dataset based on multiple scenarios. CLUE-S model has used in land change investigations across a wide range of scales of analysis and especially in several regions from Europe, Asia or Central America. In addition, the model was applied to simulate forest dynamics and conservation (Wassenaar et al., 2007; Lin et al., 2009; Manuschevich & Beier, 2016), to model urban growth (Li et al., 2014), to simulate agricultural land abandonment (Verburg & Overmars, 2009; Renwick et al., 2013; Price et al., 2015) and to simulate extreme events (Chen et al., 2009; Zhou et al., 2013, Promper et al., 2014) or groundwater pollution related to land use/cover changes (Lin et al., 2007; Dams et al., 2008; Liu et al., 2014; Lima et al., 2015).

In the current study, we have chosen to apply the CLUE-S model in order to understand and model the spatially complex process of interactions between the biophysical/socio-economic factors and land use/cover. Modelling land use/cover changes using this method helped us answer few essential questions: (1) which biophysical and socio-economic driving factors had better contribute to the explanation of land use/cover patterns in Romania? (2) which areas will be affected by land use/cover changes? (3) which land use/cover categories will be mainly affected? (4) which land use/cover categories transition will occur and which land use/cover change flows will result? The resulted outcomes could become useful tools for monitoring and quantifying spatial and temporal land use/cover changes at national, and particularly at regional level, in order to adopt appropriate land use planning and environmental policy in line with the sustainable development principles. The analysis of the spatio-temporal land use/cover patterns could be the base for further investigations of the potential landslide or soil erosion risks. Thus, part of these outlined results have already been used as input data to evaluate the landslide hazard scenario, as well as the spatial distribution of elements at risk in relation to the present and future land use/cover. This assessment was made within the recently completed project Disaster Risk Evaluation at a National Level (RO-RISK) under the Operational Programme for Administrative Capacity (POCA).

2. Materials and methods

2.1. Description of the study area

Romania is located in the South-Eastern part of the Central Europe, at the contact with Eastern and Southern Europe (Fig. 1), at the crossroads of the main European access roads. It is a medium-sized European state covering a surface of 238,391 km2 and a population of 20,121,641 inhabitants (Census data, 2011), being the largest country in the southeast of the Continent (Bălteanu et al., 2016). The moderate temperate-continental climate, the varied landforms, the diversity of soil resources and the socio-economic conditions favour the great diversity of the land use/cover types with significant regional differences. Forests, pastures and natural grasslands dominantly cover the mountains and hills, while in the plains and tablelands arable lands prevail. In terms of structure (Romanian Statistical Yearbook 2012), the agricultural land represents 61.2% of Romania’s area, the forest-covered areas 28.5%, the underwater terrains and ponds 3.5%, the built-up areas 3.1% and other land use/cover categories 3.7% (roads and railways, degraded and unproductive lands). According to the CLC database (2012), agricultural areas amounted to 57%, forest and semi-natural areas to 34%, artificial surfaces to 6%, and wetlands and water bodies to 3%.

Figure 1. The study-area. The Development Regions and the major landform units

Romania’s territory includes eight Development Regions: North-West, Central, North-East, South-East, South-Muntenia, Bucharest-Ilfov, South-West Oltenia and West. As of 1998, they act as territorial-statistical regions without legal personality for Eurostat and for the absorption of European Structural Funds. The inter-regional disparities are visible in the variety and dynamics of the main land use/cover categories. Thus, the largest agricultural land characterises the South-Muntenia Development Region (2.47 mill. ha) and South-East Development Region (2.31 mill. ha). The largest forestlands are found in the West Development Region (1.25 mill. ha), North-East Development Region (1.18 mill. ha) and North-West Development Region (1.17 mill. ha), while the built-up areas prevail in the South-Muntenia Development Region (0.247 mill. ha) and North-East Development Region (0.243 mill. ha) where some of the biggest cities of Romania are developed. The extended areas covered by natural grasslands are in the West and North-West Development Regions (both including 0.18 mill. ha), while the waters and inland marshes spreads in the South-East Development Region (almost 0.45 mill. ha), mainly due to presence of the one of the largest delta in Europe (Danube Delta).

2.2. Modelling future land use/cover dynamics

The CLUE-S model is used for simulation of the land use/cover scenarios. It includes a spatial and a non-spatial module (Verburg et al., 2002) which combines statistical analyses and decision rules that determine the sequence of land cover types (Schaldach & Priess, 2008). Synthetically, Figure 2 describes the methodology applied to calibrate, simulate and validate the predicted land use/cover dynamics and to assess the main land use/cover flow changes until 2050. Due to the particularities of the land use/cover changes at regional level, the specific characteristics of the biophysical and socio-economic factors, and the resolution used, the model was applied for each Development Region of Romania. Finally, the simulated results were aggregated into single map to analyses and understand the spatial distribution of the future main land use/cover flows at national level.

Figure 2. The flowchart showing the methodology used to predict future main land use/cover flows

2.2.1. The CLC database

In the present study, the prediction is based on the CLC database (available at the European Environment Agency) which facilitate the detection and quantification of the past land use/cover changes, as well as modelling future land use/cover change flows. Thus, the CLC 1990, 2000 and 2006 datasets were used to assess past changes for two periods (1990-2000 and 2000-2006) and to prepare the dependents and two independents variables. According to CLC nomenclature classification, ten land use/cover categories were generalised and simulated: built-up areas, arable lands, permanent crops, pastures, scrub and/or herbaceous vegetation association, forests, open spaces with little or no vegetation, heterogeneous agricultural areas, natural grasslands and agricultural complex cultivation patterns.

2.2.2. The dependent variables: land use/cover categories

The CLC 2006 datasets were used to prepare the depended variables. Hence, for the simulation were created a binary raster with the “presence” and “absence” for each category. Due to their complexity and rapid dynamics, a few land use/cover categories (e.g. waters, inland marches, mineral extraction sites) were not included in the simulation.

2.2.3. The Explanatory factors

The factors taken into account as explanatory were obtained from different sources which included NIS, NMA and ESRI Romania. Finally, depending on their availability, 17 biophysical and socio-economic explanatory factors were included in the model (Table 1). Explanatory variables related to relief features (elevation, slope declivity, relief fragmentation) were derived from/calculated by the Digital Elevation Model (30 m) and main stream network. The distance explanatory variables were calculated using a proximity analysis based on the Minimum Euclidean Distances (buffer = 1 km): to the nearest towns and to the nearest major roads. The socio-economic factors were derived from the available census data compatible with the requirements of the simulation.

Table 1. The independent variables included in the logistic regression model

Both the driving factors and land use/cover classes for each raster cell were determined using ArcGis 10.1 and transformed into an ASCII text format (necessary for the simulation). Due to the different scale/resolutions of used spatial data and the large areas of the simulation, all spatial data were resampled at 500 x 500 m cell size for further analysis.

2.2.3. The configuration of the CLUE-S model

The CLUE-S model is divided into two modules: non-spatial and spatial. The non-spatial module calculates the demands for land use/cover based on the analyses of the explanatory factors. The spatial module translates these demands into land use/cover changes according to the probabilities and rules of different land use/cover types using a raster-based system (Verburg et al., 2002). The current simulations of the land use/cover conversion revealed the following aspects:

1. Land use/cover requirements. The extrapolation of land use/cover change trends in the near future is a common technique to calculate the land use/cover requirements. Because of the political and socio-economic changes that took place after 1990 and their relevance for the resulted spatial transformations in land use/cover pattern, two scenarios were used in order to explore potential future land use/cover changes (2007 – 2050). The first scenario (HT1) assumes that land use/cover will change based on the historical land use/cover dynamics occurred between 1990 and 2000. This interval covers the period subsequent to the fall of the communist regime (1989) largely overlapping the so-called transition period which has led to a series of radical political and socio-economic transformations acknowledged as major drivers of spatial changes. These spatial changes were primarily related to the excessive fragmentation and degradation of the productive quality of agricultural lands, the abandonment of arable lands and permanent crops which gives place to the conversion into other urban sprawl-related land use categories (e.g. residential, commercial, warehouses). Thus, the transition period resulted in important spatial changes related to deforestation and suburbanization process (Popovici et al., 2013; Grigorescu et al., 2015). The second scenario (HT2) assumes that land use/cover will change based on the historical land use/cover dynamics of the 2000–2006 interval. This interval overlaps the so-called post-transition period, which gave rise to changes mainly related to the country’s accession to the European Union and the implementation of the Common Agricultural Policy (CAP). Now important land use changes related to the conversion of agricultural lands, forests or pastures to residential, commercial and industrial areas (logistic) through deforestation and urban sprawl (suburbanisation) processes took place (Popovici et al., 2013; Grigorescu et al., 2015). According to the historical changes (Table 2), linear trends for future land use/cover dynamics were assumed.

Table 2. Annual change rate (ha) of land use/cover calculated for HT1 and HT2. Differences at the level of Development Regions of Romania

2. Conversion settings. The conversion settings for specific land use/cover types determine the temporal dynamics of the simulations (Verburg, 2010). Therefore, two sets of parameters were necessary to characterize the individual land use/cover types -conversion elasticity (ELAS) and land use/cover transition sequences or conversion matrix (TS) – configured, based on the past changes and the authors understanding of the land use/cover system in the study area. The first parameter (Table 3) was related to the reversibility, the values ranging from 0 (easy conversion) to 1 (irreversible change).

Table. 3. Elasticity (ELAS) values for simulated land use/cover classes

According to the second parameter (TS), all land use types could be converted into any land use category (TS = 1), except built-up areas which may not be converted into other categories (TS = 0). Moreover, scrub and/or herbaceous vegetation association and forests were not allowed to change to built-up areas and permanent crops.

3. Logistic regression analysis. The CLUE-S model is using binomial logistic regression to define the empirical relationships between the land use/cover types and its explanatory factors. Verburg et al. (2002) provided more detailed description of the model. In the current study, the coefficients were estimated through logistic regression using the land use/cover categories (2006) as dependent variables and the biophysical and socio-economic factors as independent variables. The data were analysed using SPSS statistical software package (Statistical Package for the Social Sciences) using the stepwise procedure (forward stepwise regression) to select only the relevant explanatory factors. The response of these regression functions could then be visualized into raster probability maps, based on the location suitability, given the probability of the occurrence of a certain land use/cover type per cell. The results of the logistic regression analysis were tested by using ROC (Relative Operating Characteristics), a measure of the goodness of fit of the logistic regression model (Pontius and Schneider, 2001). In the standard ROC approach, the predictive probability map is compared with the map of the true binary event in order to assess the spatial coincidence between the event and the probability values (Mas et al., 2013). This graph displays the predictive accuracy of the logistic model, which could be evaluated using the area under the ROC curve (AUC). A completely random model gives a ROC value of 0.5, while a perfect fit results in a ROC value of 1.0. Finally, the simulation of land use/cover dynamics (2007-2050) by setting rules of land transfer, demand, and other requirement parameters was carried out using Dyna-CLUE software.

2.2.4. Spatial data accuracy assessment

The measure of the spatial agreement or accuracy between the predicted and reference data was assessed using the results of confusion matrix (error matrix), a common method used in remote sensing literature, also applied in land use modelling to compare the predicted data with real data (Congalton & Green, 2009; Geri et al., 2011; Verburg et al., 2002; Liu et al., 2009; Ahmed et al., 2013; Kumar et al., 2014). The confusion matrix includes resulted values from the cross-classification map created by overlaying the predicted land use/cover map (2012) resulted from the CLUE-S model and reference map (2012) according to the CLC database. Depending on the error matrix, different statistical measures of accuracy were calculated: Overall Accuracy (OA), Producer’s Accuracy (PA), User’s Accuracy (UA), and Kappa index (Cohen, 1960). Based on this assessment, the agreement and disagreement were calculated for each land use/cover simulated category. The agreement was considered when the pixels representing a certain land use/cover category in the predicted map overlap the pixels representing the same land use/cover category in the reference map. Disagreement was considered when the pixels representing a certain land use/cover category in the predicted map overlap the pixels representing other land use/cover categories in the reference map.

2.2.5. Establishment of the major land use/cover change flows

The predicted land use/cover changes singled out helped establishing the main change flows. The method to identify the land use/cover change flows was developed by Stott & Haynes-Young (1998), Haines-Young & Weber (2006), Weber (2007), Feranec et al. (2000, 2007, 2010) which grouped the changes by major land use processes. According to the previous methodology, Popovici et al. (2013) and Kucsicsa et al. (2015) identified eight main land cover flows to assess the past changes in land use/cover pattern in Romania at national and regional scale using CLC database. In this respect, according to the simulated land/use classes, the authors established seven main land use/cover change flows (Fig. 3).

Figure 3. The main land use/cover change flows established according to the simulated land use/cover classes

1. Urbanization and industrialization (URB) refers to the change of agricultural lands, semi natural and forestlands into built-up areas;

2. Intensification of agriculture (IA) – involves the internal conversion of agriculture, namely the transition of the agricultural land types associated to lower intensity use (e.g. pastures) into higher intensity use (arable lands or permanent crops). This flow also includes the conversion of natural lands (except forest) into agriculture lands;

3. Extensification of agriculture (EA) – describes the internal conversion of agriculture, specifically the transition of the agricultural land types associated to higher intensity use (arable lands or permanent crops) to the lower intensity use (e.g. pastures and grasslands);

4. Agricultural lands abandonment (ALAB) – defines the change of agricultural lands to semi natural areas;

5. Afforestation (AFF) – refers to a flow which deals with forest regeneration (natural or planted), more precisely the transition from agricultural lands and semi-natural vegetation to forestlands and scrub and/or herbaceous vegetation association;

6. Deforestation (DEF) – involves a flow related to forestlands changes to agricultural lands or semi natural areas;

7. Other changes (OTH) – refers to other land use changes such as: the transition from built-up to other land use/cover categories, the transition from open spaces with little or no vegetation areas to natural grasslands, the transition from natural grasslands to open spaces with little or no vegetation areas.

3. Results and discussions

3.1. Land use/cover changes in 1990-2006

The fall of the communist regime has led to a series of radical political and socio-economic changes in many Central and East European countries (Popovici et al., 2013). Studies focusing on assessment and quantification of land use/cover changes in the context of post-communist period generally acknowledged socio-economic and political conditions as major drivers of land use/cover dynamics (Ptáček, 2000; Bičík et al., 2001; Zemek et al., 2005; Kuemmerle et al., 2006; Václavík & Rogan, 2009; Bičík & Jeleček, 2009; Szilassi et al., 2010; Popovici et al., 20013; Hanganu & Constantinescu, 2015 etc.). Consequently, after 1990, agricultural lands where the main land use/cover categories with significant dynamics in Romania. According to the CLC database (Table 4), arable lands increased with 1.8%, pastures with 1.7%, while the permanent crops decreased dramatically with 8%. Forests registered a slight increase with nearly 0.6%, gaining about 41,000 ha in comparison with the 1990. Built-up areas increased steadily throughout the 1990-2006 period with 0.8%. The agricultural complex cultivation patterns also increased with 0.7% over analysed interval, while the natural grasslands and scrub and/or herbaceous vegetation association decreased with 9.3% and 8% respectively.

Table 4. Land use/cover change in Romania between 1990-2006, based on the CLC database

3.2. Prediction of land use/cover dynamics. The main land use/cover change flows

3.2.1. The explanatory variables of land use/cover pattern

The regression results of the regional analysis suggest that the spatial relations between the explanatory factors and characteristics of land use/cover vary due to ecological potential and socio-economic specifics of the each development region. Thus, according to the spatial resolution and data used, the β coefficients show that the biophysical variables have the most important contribution to explaining the current spatial pattern of land use/cover. As from socio-economic factors, regression coefficients suggest that secondary roads density, settlement density, proximity of the main roads, population growth and unemployment rate have the most significant role for the land use/cover spatial distribution. The results of the β coefficients for the most important factors determined by logistic regression are illustrated in Figure 4.

Figure 4. The graphical representation of the β values for the most important factors determined by logistic regression

These illustrate that the highest β values are given by the relief fragmentation, which has a positive contribution mainly to built-up area, heterogeneous agricultural areas and pastures, and negative contribution to arable lands. The influence of slopes mainly on land use/cover pattern is also evident. The regression model shows a strong negative contribution to arable lands, this indicating a large occurrence of this category mainly in the plains and tablelands regions. Furthermore, it is easy to understand the positive effect of slopes to forests. Because of their low accessibility for logging, a higher declivity would indicate favourable areas for forest extension. On the other hand, the correlation between elevation and land use/forest pattern is not so notable. However, the negative values of the β coefficients point out that the agricultural land classes and built-up areas occur in relation to the decreasing in altitude, while the positive values indicate suitability for forest and semi-natural areas associated with the increasing in altitude. In terms of climate features, temperature has also positive relation with the occurrence of forests, open spaces with little or no vegetation and natural grasslands. A negative relation has with the heterogeneous agricultural areas and agricultural complex cultivation patterns because their large occurrence mainly in the mountain areas. In addition, precipitation have a lower influence, indicating a negative relation to the extension of built-up areas, pastures, forests and natural grasslands, and positive with permanent crops, scrub and/or herbaceous vegetation association, open spaces with little or no vegetation, heterogeneous agricultural areas, and agricultural complex cultivation patterns. For soil, regression coefficients suggest that the total organic matter content in topsoil is an ecological factor that have also significant role for the land use/cover distribution. Generally, regressions show a positive relation with arable lands, forests and natural grasslands. However, a negative relation has with the open spaces with little or no vegetation suggesting that this land use/cover class includes bare rocks, sandy areas or degraded lands. In terms of accessibility, the secondary roads density and distance to nearest main towns are the most important factors. Generally, the positive values indicate that the built-up areas, arable lands and agricultural complex cultivation patterns is more likely to occur in areas that have high accessibility, while the scrub and/or herbaceous vegetation association, forests, heterogeneous agricultural areas and natural grasslands in areas that have low accessibility. Main roads proximity has a strong influence for built-up areas, the negative β values indicating that the transportation network triggers urban growth. Moreover, agricultural lands tend to occur closer to the main roads. Semi-natural areas and especially forest is more likely to occur in areas that have low connectivity and accessibility in terms of communication network. Settlements density is another significant factor influencing agricultural land distribution, higher values indicating a large occurrence of arable lands, permanent crops, heterogeneous agricultural areas and agricultural complex cultivation patterns. For each Development Region, the remaining explanatory factors related to socio-economic indicators have low significance in the current spatial pattern of the land use/cover. Among these, population density and unemployment rate were identified as the most insignificant driving factors in the current spatial pattern of land use/cover. However, in some cases, the regression models show that population growth and unemployment rate influence the current land use/cover pattern. Agricultural lands tend to occur in areas where the population growth and unemployment rate are low. The explanatory factors with β value = 0 were excluded from the spatial simulation.

3.2.2. Land use/cover dynamics (2007-2050) under HT1 and HT2 scenarios

Based on the past land use/cover changes in the study area, two future land use/cover scenarios have been determined. The changes in land use/cover categories until 2050 compared with 2006 (start of the simulation) are displayed in Table 5.

Table 5. Land use/cover changes (%) in Romania until 2050 compared with 2006 for two development scenarios HT1 and HT2.

Overall, all land use/cover types are likely to continue the same historical trends as in the analysed past periods, except the arable lands and forests which will decrease under HT1 and HT2 respectively, and open spaces with little or no vegetation which will increase under both scenarios. The scenarios predict an increase in built-up areas with an average of 5%. For arable lands, HT2 scenario predict an increase with 2.8%, while HT1 a decrease with 1.7%. Furthermore, the scenarios forecast a decrease in permanent crops, with maximum change for HT2 scenario (18.4%) and scrub and/or herbaceous vegetation association. The scenarios HT2 predict a decrease in heterogeneous agricultural areas with 24.4%, natural grasslands with 22.7% and agricultural complex cultivation patterns with 4.7%. According to the HT1, it is expected an increase in the afforested area with about 1.4%, while HT2 indicate a decrease with 1.8%.

Figure 5. The spatial distribution of the predicted land use/cover dynamics (2050) compared with 2006 in the relief units of Romania: A = HT1 scenario; B = HT2 scenario (white colour in the map represents not simulated land use/cover categories)

Figure 5 shows the simulated land-use maps for 2050 according to the HT1 and HT2 scenarios. One can notice that the increase in the urban areas is mostly close to main existing cities were the urban growth tendency is also now detected. Increases in arable lands are mostly expected in the plain units and in the Transylvanian Tableland, while decreases are expected in the Sucarpathians, the Moldavian Plateau and Dobrogea Plateau, the central part of the Romanian Plain and the Danube Delta and Razim-Sinoie Lagoon Complex. The model shows important decline in the permanent crops in the Subcarpathians, the Moldavian Plateau and the central part of the Romanian Plain. The increase of pastures is largely recorded in the Romanian Plain, the Dobrogea Plateau, the southern part of the Moldavian Plateau and the southern part of the Transylvanian Tableland. HT1 shows an increase of the heterogeneous agricultural areas and agricultural complex cultivation patterns in the Transylvanian Tableland and the plain units, while the HT2 shows a decrease in the Eastern Carpathians, the Getic Piedmont and the Subcarpathians. The results also show an important increase in forest area for the Southern part of the Eastern Carpathians, the Southern Carpathians and the Subcarpatians (mainly the Curvature area), while in the Eastern Carpathians, the Apuseni Mountains and the Romanian Plain significant forest loss are predicted. Moreover, an increase in natural grasslands in the northern part of the Eastern Carpathians, Subcarpathians, Getic Piedmont, Moldavian Plateau and Danube Delta and Razim-Sinoie Lagoon Complex are expected.

3.2.3. The main land use/cover change flows (2007-2050)

The analysis of the predicted data reveal that the changes in land use/cover will cover about 7% (according to the HT1) and 16% (according to the HT2) of the total simulated area. The main land use/cover changes will be in relation to the transition between agricultural lands and forest cover dynamics, with significant spatial differences between the major relief units (Fig. 6.).

Figure 6. The spatial distribution of the predicted land use/cover change flows in the relief units of Romania: A = HT1 scenario; B = HT2 scenario

Urbanization and industrialization (URB) are characteristic to all the large cities of post-communist Romania (Popovici et al., 2013). In many cases, urbanisation was replaced by suburbanization, especially in the surroundings of large and very large towns imprinting different development patterns: compact closer to the existing built-up areas, linear along the main roads etc. (Grigorescu et al., 2015; Kucsicsa and Grigorescu, 2017). Thus, the predicted urbanization process will hold 3.4% (according to the HT1) and 2.8% (according to the HT2) of the total changed area and will be developed largely at the expense of the agricultural lands (mainly arable lands, pastures and agricultural complex cultivation patterns). All scenarios predict that the most extensive urbanizing processes will be in the Romanian Plain and Transylvanian Tableland, close to the main existing urban centres: Bucharest, Ploiești, Buzău, Brăila, Galați, Sibiu, Cluj-Napoca, Târgu Mureș. Significant land use/cover changes related to urbanization close to Constanța, Iași, Baia-Mare, Oradea, Hunedoara, Craiova and Timișoara are also expected.

Intensification of agriculture (IA) represents 21.7% (according to the HT1) and 35% (according to the HT2) of the total changed area. This flow will be principally related to the transition of the agricultural land types associated to lower intensity use (pastures and heterogeneous agricultural areas) into higher intensity use (arable land). The arable land extension will occur mainly in the Banat and Crișana Plain and Hills, the Transylvanian Tableland, the western part of the Romanian Plain and the main depressionary areas of the Eastern Carpathians. Therefore, notable conversion from arable lands, agricultural complex cultivation patterns and natural grassland into permanent crops is expected in the Romanian Plain and the Getic Piedmont. The intensification of agriculture will be also related to the losses of scrub and/or herbaceous vegetation association, especially in the northern part of the Eastern Carpathians, the Apuseni Mountains, the Banat and Crișana Hills and the Romanian Plain.

Extensification of agriculture (EA) is the widest expected land use/cover change flow, which will represent 34.1% (according to the HT1) and 35.1% (according to the HT2) of the total changed area. The transition of the agricultural land types associated to higher intensity use (arable lands and permanent crops) into lower intensity use (mainly pastures and heterogeneous agricultural areas) will occur mainly in the Romanian Plain, the Moldavian Plateau and the Dobrogea Plateau. Therefore, in the western part of the Romanian Plain and the Transylvanian Tableland, converse on from the permanent crops into arable lands is expected.

Agricultural lands abandonment (ALAB) represent 4.6% (according to the HT1) and 1.7% (according to the HT2) of the total changed area and will prevail in the Getic Piedmont, Dobrogea Plateau and Danube Delta and Razim-Sinoie Lagoon Complex where transition of the arable land and pastures into the natural grasslands is expected.

Afforestation (AFF) corresponds to 20% (according to the HT1) and 9.2% (according to the HT2) of the total changed area. The scenarios suggest that the most distinct changes in favour of forest will be in the Southern and Eastern Carpathians and Subcarpathians (Curvature area). In these areas, the expansion of the existing forest-covered areas will be in relation to the human intervention (planting after logging or forests blow-downs, and forest fires) and natural development related to the conversion from the existing transitional vegetation to forest or connected to the tendency of pastures to be abandoned.

Deforestation (DEF). In the 1990-2006 period, the deforestation was one of the most important land use/cover change flows (37%) mostly in the Carpathian Mountains, connected to logging and various natural extreme events (e.g. forests blow-downs, avalanches, forest fires), but also related to the retrocession of forestlands, which contributed to illegal logging (Popovici et al., 2013). In the future, the model suggests that the decrease in forest cover will be largely related to the extension of arable lands and pastures mainly in the plain areas, the Getic Piedmont, the Transylvanian Tableland and the Moldavian Plateau. In addition, significant deforestation related to the natural grasslands extension in the Apuseni Mountains and in few relief units from the Eastern Carpathians is expected.

Other changes (OTH) representing only about 0.1% for both scenarios, refers to the transition between natural grasslands to open spaces with little or no vegetation. These types of changes will mainly affect some areas from the mountains units and the Danube Delta and Razim-Sinoie Lagoon Complex.

Overall, the analysis of these land use/cover change flows, can be deducted that the urbanization and industrialization, intensification of agriculture and deforestation might indicate the degradation of the environmental conditions in many regions of Romania. Therefore, significant increase in open spaces with little or no vegetation can reflect the expansion of the unused and degraded lands. At the same time, the extensification of agriculture could represent an important step for agricultural lands abandonment. This could be in relation to the degradation/abandonment of land improvement works (e.g. irrigation systems) or to the decrease tendency of the animal stock after 1990, as well as the declaration of some areas as protected. Important changes are also expected in terms of forest cover dynamics. First, afforestation in the subalpine forest-natural grasslands ecotone suggests that the upper forest limit is expected to have a positive trend, mainly in the high mountain units. This phenomenon, also currently observed (Fig. 7), could be mainly explained through its relation to the declining of the grazing activities and/or the declaration of a large number of restricted protected areas. It can be appreciated that this altitudinal forest shift could also be connected to the complex ecological processes triggered by climate change.

Figure 7. The tendency of rising the upper forest limits in the high mountains of the Eastern Carpathians (photo: Kucsicsa, 2017)

On the other hand, in the hilly and mountain units, the deforestation will be associated with the current logging tendency in the more accessible areas, favouring the extension of pastures and natural grasslands and, consequently, the potential for the occurrence of new landslide and soil erosion hotspots. Therefore, according to the results of the national-level evaluation of the landslide hazard in relation to the future land use/cover changes (RO-RISK Project, 2016), significant increase of landslides susceptibility is expected, especially in the Subcarpathians, the southern part of the Transylvanian Tableland and the central part of the Moldavian Plateau, areas which are currently affected by a great number of landslides on extended territories. In contrast, the decrease of the landslides susceptibility is expected in the Getic Piedmont, the Banat and Crișana Hills and in several mountain units (especially from southern part of the Eastern Carpathians), triggered by the extension of the forest cover to the detriment of pastures and natural grasslands.

3.2.4. Model accuracy

Generally, the statistical validation, tested using ROC method, shows a good accuracy of the models. Overall, the AUC values (Table 6) indicate a good fit between the predicted land use/cover categories and the real (observed) data, with greatest prediction ability for Central, South Muntenia and West Development Regions. Except pastures, in most cases, the AUC is equal or greater than 0.75, suggesting a good correlation and strong capacity to explain the land use/cover patterns by the selected driving factors.

Table 6. AUC values for the simulated land use/cover types. Differences at the Development Regions of Romania

The spatial accuracy assessment suggests that the model captures the trends in land use/cover changes, mainly in relation to the past changes during the 2000-2006 period. In this respect, it can deducted that land use/cover changes related to the pre-accession to the European Union (also called post-transition period) better explain the current spatial distribution and future tendency in land use/cover dynamics. Therefore, the simulation for arable lands, forests, built-up areas and natural grasslands had the greatest accuracy. In contrast, the permanent crops, scrub and/or herbaceous vegetation associations and the open spaces with little or no vegetation categories recorded the lowest accuracy. Figure 8 shows that the best results in terms of accuracy assessment were obtained for the plain regions and the Dobrogea Plateau where the percentage of correct prediction of simulated land use/cover categories is 91-94. This is the result of the good agreement mainly for the arable lands, pastures and built-up areas. Therefore, in the mountain regions the percentage of correct prediction is about 84. Here, significant forestlands and grasslands were accurately predicted. Notable disagreements were obtained for the forestlands (in the Apuseni Mountains and the Eastern Carpathinas), the arable lands (in Getic Piedmont and Moldavian Plateau), the pastures and grasslands (in the Subcarpathians and the Eastern Carpathians) and the built-up areas (in the Romanian Plain).

Figure 8. Cross-classification maps resulted from the overlay process of the predicted land use/cover map and reference map (2012): A = HT1; B = HT2

The statistical measures, calculated depending on the confusion matrix (Table 7), demonstrate again that the best accuracy was obtained according to the HT2 scenario (OA = 85.81%), mainly for heterogeneous agricultural areas (UA = 93.5%; PA = 76.5%), forests (UA = 88.6%; PA = 89.0%) and natural grasslands (UA = 85.1%; PA = 77.4%). Overall, the Kappa values are 0.80 for HT1 and 0.81 for HT2, indicating a substantial spatial accuracy between the predicted land use/cover categories and the real land use/cover categories (2012).

Table 7. The agreement statistic indicators for the predicted data calculated for the entire study area

The disagreement between the simulated land use/cover scenarios and the current could be mainly related to the: spatial data used in the simulation, namely the accuracy of the CLC database; the different scales of the explanatory factors taken into account; the availability of socio-economic statistical data, i.e. only at LAU2; the unavailability of other relevant socio-economic or biophysical data related to the simulated land use/cover categories; the lack of important dynamic driving factors (e.g. demographic or climate scenarios); the absence of human-induced or natural unpredicted events, especially related to forest cover (e.g. illegal logging, forests blow-downs and forest fires); the extrapolation of a short historical trends of land use/cover changes to simulate a longer period (44 years); unexpected the political and institutional measures with influence on the land use/cover changes (e.g. Land Laws, land use planning, increasing the number of measures for afforestation projects, especially for degraded lands); the environmental policies (e.g. declaration of a large number of protected areas, as of 2000, where the trend of extension of the agricultural lands and deforestation rate has decreased, mainly inside the national parks rather than outside their boundaries; the ecological reconstruction strategies and measures).

4. Conclusions and perspectives

The future land use/cover change flows (2007-2050) were simulated using CLUES-S model based on the CLC database. The results indicate that the future land use/cover dynamics will vary, with notable differences at regional level. Thus, the geographical area of the built-up areas, arable lands, pastures, open spaces with little or no vegetation and agricultural complex cultivation patterns will increase significantly, together with decline in permanent crops, scrub and/or herbaceous vegetation association, forests, heterogeneous agricultural areas and natural grasslands. These changes mainly indicate a decline of the environmental conditions; the expansion of the unused and degraded lands; the tendency of the agricultural lands to be abandoned; a positive trend of the upper forest limits in the subalpine forest-natural grasslands ecotone; the continuous trend of forest logging in the accessible areas and consequently, with high potential for the occurrence of new landslide and soil erosion hotspots.

Land use/cover changes simulations have important contribution to studies related to the preparation, development and, to a lesser extent, evaluation of large-scale spatial plans and strategies (Koomen et al., 2008). In this context, the present study could become useful tool for the monitoring and quantifying the spatial and temporal tendency of urban growth, agricultural lands and forest or semi-natural areas at national and regional levels in order to adopt appropriate land use planning and environmental policy in agreement with the sustainable development principles. Therewith, the complex and dynamic process of land use/cover changes links natural and human systems (Koomen, 2007), which, in the context of natural hazard and risk assessment, this linkage is a key issue (Promper et al., 2014). For further analyses, the predicted results can be also directly implemented into natural hazards models aimed at evaluating different scenarios of soil erosion or landslides susceptibility at national or regional level.

Because of uncertainty related to input data, the outcomes of the current research should be considered as a preliminary step for further simulations, requiring additional variables (e.g. climate and socio-economic scenarios) based on which, a better simulation of the future land use/cover dynamics could be achieved. However, these results indicate possible future land use/cover dynamics and, subsequently, the main land use/cover change flows related to the present biophysical and socio-economic driving factors according to the conditions of each specified scenario.

Acknowledgements

The current results are part of the RO-RISK Project carried out under the Administrative Capacity Operational Programme (Disaster Risk Assessment at national level – SIPOCA 30) and the project “Geographical research of the Romanian Danube Valley”- fundamental studies made under the research plan of the Institute of Geography, Romanian Academy. The authors would like to thank the IVM Institute for Environmental Studies (http://www.ivm.vu.nl/en/Organisation/departments/spatial-analysis-decision-support/Clue) for its support of the full CLUE soft version and the European Environment Agency for the provision of land use/cover database (http://www.eea.europa.eu/publications/COR0-landcover).

References

Administrative Capacity Operational Programme. 2016. RO-RISK Project – Disaster Risk Assessment at national level, SIPOCA 30.

Ahmed, B., Ahmed, R., Zhu, X. 2013. Evaluation of Model Validation Techniques in Land Cover Dynamics. SPRS Int. J. Geo-Inf., 2, 577-597.

Bălteanu, D., Dumitrașcu, M., Geacu, S., Mitrică, B., Sima, M. (eds.). 2016. Romania. Nature and Society. Romanian Academy Publishing House, 690 p., in Romanian.

Bičík, I., Jeleček, L., Štìpánek, V. 2001. Land-Use Changes and Their Social Driving Forces in Czechia in the 19th and 20th centuries. Land Use Policy, 18, 65–73.

Bičík, I., Jeleček, L. 2009. Land use and landscape changes in Czechia during the period of transition 1990–2007. Geografie – Sbornik ČGS, 114(4), 263–281.

Castella, J-C., Pheng, K.S., Dinh, D., Verburg, P.H., Thai, H.C. 2007. Combining top-down and bottom-up modelling approaches of land use/cover change to support public policies: application to sustainable management of natural resources in northern Vietnam. Land Use Policy, 24, 531–545.

Chen, Y., Xu, Y.,Yin, Y. 2009. Impacts of land use change scenarios on storm-runoff generation in Xitiaoxi basin, China. Quaternary International, 208 (1), 121–128.

Clarke, K.C., Hoppen, S., Gaydos, L. 1997. A self-modifying cellular automation model of historical urbanization in the San Francisco Bay Area. Environment and Planning B, 24, 247-261.

Cohen, J. 1960. A coefficient of agreement for nominal scale. Educational and Psychological Measurement, 20, 37–46.

Congalton, R.G., Green, K. 2009. Assessing the accuracy of remotely sensed Data: Principles and Practices (2nded.). Taylor and Francis Group, LLC, 200 p.

Dale, V.H. 1997. The relationship between land-use change and climate change. Ecol. Appl. 7(3), 753–769.

Dams, J., Woldeamlak, S.T., Batelaan, O. 2008. Predicting land-use change and its impact on the groundwater system of the Kleine Nete catchment, Belgium. Hydrol. Earth Syst. Sci., 12 (6), 1369–1385.

EEA. 2007. Land-use scenarios for Europe: Qualitative and quantitative analysis on a European scale. Luxembourg: European Environmental Agency (EEA). EEA Technical report 9/2007.

Evans, T.P., Kelley, H. 2004. Multi-scale analysis of a household level agent-based model of land cover change. Journal of Environmental Management, 72, 57-72.

Feranec, J., Šủri, M., Otahel, J., Cebecauer, T., Kola, J., Soukup, T., Zdenkova, D., Waszmuth, J., Vajdea, V., Vijdea, A., Nitica, C. 2000. Inventory of major landscape changes in the Czech Republic, Hungary, Romania and Slovak Republic 1970s–1990s. International Journal of Applied Earth Observation and Geoinformation, 2, 129–139.

Feranec, J., Otahel, J. 2001. Land cover of Slovakia. Veda, Bratislava, 124 p.

Feranec, J., Hazeu, G., Christensen, S., Jaffrain, G. 2007. CORINE land cover change detection in Europe (case studies of the Netherlands and Slovakia). Land Use Policy, 24, 234–247.

Feranec, J., Jaffrain, G., Soukup, T., Hazeu, G. 2010. Determining changes and flows in European landscapes 1990–2000 using CORINE land cover data. Applied Geography, 30, 19–35.

Feranec, J., Soukup, T., Hazeu, G., Jaffrain, G. 2016. European Landscape Dynamics: CORINE Land Cover Data. CRC Press, Boca Raton, USA, p. 337.

Geri, F., Amici, V., Rocchini, D. 2011. Spatially-based accuracy assessment of forestation prediction in a complex Mediterranean landscape. Applied Geography, 31, 881–890.

Grigorescu, I., Kucsicsa, G., Mitrică, B. 2015. Assessing spatio-temporal dynamics of urban sprawl in the Bucharest Metropolitan Area over the last century. In: Bičík et al. (eds.), Land Use/Cover Changes in Selected Regions in the World – Volume XI (IGU-LUCC Research Reports), 19-28.

Guan, D., Li, H., Inohae, T., Su, W., Nagaie, T., & Hokao, K. (2011). Modeling urban land use change by the integration of cellular automaton and Markov model. Ecological Modelling, 222, 3761-3772.

Haines -Young, R., Weber, J.L. 2006. Land accounts for Europe 1990–2000. Towards integrated land and ecosystem accounting. European Environment Agency, Copenhagen, 107 p.

Halmy, M.W.A, Gessler, P.E., Hicke, J.A., Salem, B.B. 2015. Land use/land cover change detection and prediction in the north-western coastal desert of Egypt using Markov-CA. Applied Geography, 63, 101-112.

Hanganu, J., Constantinescu, A. 2015. Land cover changes in Romania based on Corine Land Cover inventory 1990–2012. Rev. Roum. Géogr./Rom. Journ. Geogr., 59 (2), 111–116.

IGBP. (1999). Land-Use and Land-Cover Change (LUCC) implementation strategy. IGBP Report No. 48/IHDP Report No 10. Stockholm (Sweden): International Geosphere-Biosphere Programme (IGBP) e International Human Dimension Programme on Global Environmental Change (IHDP).

Irwin, E.G., & Geoghegan, J. 2001. Theory, data, methods: developing spatially explicit economic models of land use change. Journal of Agriculture Ecosystems and Environment, 85(1-3), 7-24.

Jinyan, Z., Xiangzheng, D., Qun’ou, J., Nana, S. 2007. The application of system dynamics and CLUE-S model in land use change dynamic simulation: a case study in Taips County, Inner Mongolia of China. Proceedings of the Conference on System Science, Management Science & System Dynamics, 2781–2790.

Kamusoko, C., Aniya, M., Adi, B., Manjoro, M. 2009. Rural sustainability under threat in Zimbabwe-Simulation of future land use/cover changes in the Bindura district based on the Markov-cellular automata model. Applied Geography, 29, 435-447.

Koomen, E. 2007. Modelling land-use Change: Progress and applications. Springer.

Koomen, E., Rietveld, P., de Nijs, T. 2008. Modelling land-use change for spatial planning support. Ann. Reg. Sci., 42, 1–10.

Kucsicsa, G., Balteanu, D., Popovici, E-A., Damian, N. 2015. Land use/cover changes along the Romanian Danube Valley. In: Bičík et al. (eds.), Land Use/Cover Changes in Selected Regions in the World – Volume XI (IGU-LUCC Research Reports), 7-18.

Kucsicsa, G, Grigorescu, I. 2017. Urban growth related to distance explanatory factors in Bucharest Metropolitan Area. Spatial and temporal assessment using logistic regression, Journal of Urban Planning and Development, ASCE Library (in print)

Kuemmerle, T., Radeloff, V.C., Perzanowski, K., Hostert, P. 2006. Crossborder Comparison of Land Cover and Landscape Pattern in Eastern Europe Using a Hybrid Classification Technique. Remote Sensing of Environment, 103, 449–464.

Kuemmerle, T., Hostert, P., Radeloff, V.C., Perzanowski, K., Kruhlov, I. 2007. Post-Socialist Forest Disturbance in the Carpathian Border Region of Poland, Slovakia, and Ukraine. Ecological Applications, 17(5), 1279–1295.

Kumar, R., Nandy, S., Agarwal, R, Kushwaha, S.P.S. 2014. Forest cover dynamics analysis and prediction modeling using logistic regression model. Ecological Indicators, 45, 444–455.

Lambin, E.F. (1997). Modelling and monitoring land-cover change processes in tropical regions. Progress in Physical Geography, 21, 375-393.

Lambin, E.F., Rounsevell, M.D.A., Geist, H.J. 2000. Are agricultural land-use models able to predict changes in land-use intensity? Agriculture, Ecosystems, and Environment, 82, 321-331.

Lambin, E.F., Geist, H.J. (eds.). 2006. Land Use and Land Cover Change. Local Processes and Global Impacts. Berlin, Heidelberg, New York: Springer, 222 p.

Li, W., Wu, C., Zang, S. 2014. Modeling urban land use conversion of Daqing City, China: A comparative analysis of ‘‘top-down’’ and ‘‘bottom-up’’ approaches. Stoch Environ Res Risk Assess, 28 (4), 817–828.

Lima, M. L., Romanelli, A., Massone, H.E. 2015. Assessing groundwater pollution hazard changes under different socio-economic and environmental scenarios in an agricultural watershed. Science of the Total Environment, 530, 333-346.

Lin, P-Y., Nien-Ming Hong, N-M., Pei-Jung Wu, P-J., Lin, C-J. 2007. Modeling and assessing land-use and hydrological processes to future land-use and climate change scenarios in watershed land-use planning. Environ. Geol., 53, 623–634.

Lin, Y.P., Verburg, P.H., Chang, C.R., Chen, H.Y., Chen, M.H. 2009. Developing and comparing optimal and empirical land-use models for the development of an urbanized watershed forest in Taiwan. Landscape and Urban Planning, 92 (3), 242–254.

Liu, M., Hu, Y., Chang, Y., He, X., Zhang, W. 2009. Land Use and Land Cover Change Analysis and Prediction in the Upper Reaches of the Minjiang River, China. Environmental Management, 43, 899–907.

Liu, M., Li, C., Hu, Y., Sun, F., Xu, Y., Chen, T. 2014. Combining CLUE-S and SWAT models to forecast land use change and non-point source pollution impact at a watershed scale in Liaoning Province, China. Chinese Geographical Science, 24 (5), 540–550.

Luo, G., Yin, C., Chen, X., Xu, W., Lu, L. 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.

Manuschevich, D., Beier, C.M. 2016. Simulating land use changes under alternative policy scenarios for conservation of native forests in south-central Chile. Land Use Policy, 51, 350–362.

Mas, J-F., Filho, B.S., Pontius Jr, R.G., Gutiérrez, M.F., Rodrigues, H. 2013. A Suite of Tools for ROC Analysis of Spatial Models. ISPRS Int. J. Geo-Inf., 2, 869–887.

Mena, C.F., Walsh, S.J., Frizzelle, B.G., Yao, X., Malanson, G.P. 2011. Land use change on household farms in the Ecuadorian Amazon: Design and implementation of an agent-based model. Applied Geography, 31(1), 210-222.

Mertens, B., Lambin, E.F. 1997. Spatial modeling of deforestation in Southern Cameroon. Applied Geography, 17, 143-162.

Otahel, J., Feranec, J., Suri, M., Cebecauer, T. 2002. Land cover of Slovakia: assessment of its changes, In: Bicík I et al. (eds.): Land Use/Land Cover Changes in the Period of Globalization. Proceedings of the IGU-LUCC International Conference, Prague, 2001. KSGRR PrF UK, Prague, 100–105.

Parker, D.C., Manson, S.M., Janssen, M.A., Hoffmann, M.J., Deadman, P. 2003. Multi-agent systems for the simulation of land-use and land-cover change: a review. Annals of the Association of American Geographers, 93(2), 314-337.

Pontius Jr, G.R. Schneider, C.L. 2001. Land-cover change model validation by an ROC method for the Ipswich watershed, Massachusetts, USA. Agriculture, Ecosystems and Environment, 85, 239-248.

Popovici, E-A., Bălteanu, D., Kucsicsa, G. 2013. Assessment of changes in Land-Use and Land-Cover pattern in Romania using Corine Land Cover database. Carpathian Journal of Earth and Environmental Sciences, 8 (4), 195–208.

Price, B., Kienast, F., Seidl, I., Ginzler, C., Verburg, P.H., Bolliger, J. 2015. Future landscapes of Switzerland: Risk areas for urbanisation and land abandonment. Applied Geography, 57, 32–41.

Promper, C., Puissant, A., Malet, J-P., Glade, T. 2014. Analysis of land cover changes in the past and the future contribution to landslide risk scenarios. Applied Geography, 53, 11-19.

Ptáček P. 2000. Networking and Local Culture: Local Community Transformation after 1989 on the Example of Olomouc, Czech Republic. Acta Universitatis Palackianae–Geographica, 36, 59–64.

Rawat, J.S., Kumar, M. 2015. Monitoring land use/cover change using remote sensing and GIS techniques: A case study of Hawalbagh block, district Almora, Uttarakhand, India. The Egyptian Journal of Remote Sensing and Space Science, 18(1), 77-84.

Renwick, A., Jansson, T., Verburg, P.H., Revoredo-Giha, C., Britz, W., Gocht, A., McCracken, D. 2013. Policy reform and agricultural land abandonment in the EU. Land Use Policy, 30 (1), 446–457.

Sala, O.E., Chapin, F.S., Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R., Huber-Sanwald, E., Huenneke, L.F., Jackson, R.B., Kinzig, A., Leemans, R., Lodge, D.M, Mooney, H.A., Oesterheld, M., Poff, N.L., Sykes, M.T., Walker, B.H., Walker, M., Wall, D.H. 2000. Global biodiversity scenarios for the year 2100. Science, 287, 1770-1774.

Szilassi, P., Jordan, G., Kovacs, F., Van Rompaey, A., Van Dessel, W. 2010. Investigating the link between soil quality and agricultural land use change. A case study in the Lake Balaton catchment, Hungary. Carpathian Journal of Earth and Environmental Sciences, 5(2), 61-70.

Schaldach, R., Priess, J.A. 2008. Integrated models of the land system: a review of modelling approaches on the regional to global scale. Living Reviews in Landscape Research, 2(1), 1863-7329.

Stott, A., Haines-Young, R. 1998. Linking land cover, intensity of use and botanical diversity in an accounting framework in the UK. In: Uno, K. & Bartelmus, P. (eds.), Environmental accounting in theory and practice, Dordrecht: Kluwer, 245–260.

Turner, B.L. II, Moss, R.H., Skole, D. 1993. Relating Land Use and Global LandCover Change: A Proposal for an IGBPHDP Core Project. International Geosphere-Biosphere Programme, Stockholm, Sweden.

Václavík, T., Rogan, J. 2009. Identifying Trends in Land Use/Land Cover Changes in the Context of Post-Socialist Transformation in Central Europe: A Case Study of the Greater Olomouc Region, Czech Republic. GIScience & Remote Sensing, 46, 1, 54–76.

Veldkamp, A., Lambin, E.F. (2001). Predicting land-use change. Agriculture, Ecosystems and Environment, 85, 1-6.

Verburg, P.H. 2010. The CLUE Modeling Framework – course material, Institute for Environmental Studies. University Amsterdam, 53 p.

Verburg, P.H., Soepboer, W., Veldkamp, A., Limpiada, R., Espaldon, V., Mastura, S.S. 2002. Modeling the spatial dynamics of regional land use: the CLUE-S model. Environmental Management, 30(3), 391-405.

Verburg, P.H., Veldkamp, A. 2004. Projecting land use transitions at forest fringes in the Philippines at two spatial scales. Landscape Ecology, 19, 77–98.

Verburg, P.H., Overmars, K.P. 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.

Veldkamp, A., Fresco, L.O. 1996. CLUE: a conceptual model to study the conversion of land use and its effects. Ecological Modeling, 85, 253–270.

Veldkamp, A, Verburg, P.H., Kok, K., Koning, G.H.J., Priess, J., Bergsma, A.R. 2001. The need for scale sensitive approaches in spatially explicit land use change modelling. Environmental Modeling and Assessment, 6, 111–121.

Watson, R.T., Noble, I.R., Bolin, B., Ravindranath, N.H., Dokken, D.J. 2000. Land use, land-use change, and forestry. A special report of the Intergovernmental panel on climatic change. Cambridge University Press, Cambridge.

Wassenaar, T., Gerber, P., Verburg, P.H., Rosales, M., Ibrahim, M., Steinfeld, H. 2007. Projecting land use changes in the Neotropics: The geography of pasture expansion into forest. Global Environmental Change, 17 (1), 86–104.

Weber, J.L. 2007. Implementation of land and ecosystem accounts at the European environment agency. Ecological Economics, 61(4), 695–707.

Willems, E., Lemments, T., Buffaria, B. 2005. Utilisation of CLC 90&2000 data for monitoring the impact of CAP developments on the rural landscape. In: Trends of some agro-environmental indicators in the European Union (Report EUR 21565), 15–349.

Zemek, F., Heøman, M., Mašková, Z., Kvìt, J. 2005. Multifunctional Land Use – a Chance of Resettling Abandoned Landscapes? (A case study of the Zhùøí territory, the Czech Republic). Ecology, 24(1), 96–108.

Zhou, F., Xu, Y., Chen, Y., Xu, C.Y., Gao, Y., Du, J. 2013. Hydrological response to urbanization at different spatio-temporal scales simulated by coupling of CLUE-S and the SWAT model in the Yangtze River Delta region, Journal of Hydrology, 485, 113–125.

Zhu, Z., Liu, L., Chen, Z., Zhang, J., Verburg, P.H. 2010. Land-use change simulation and assessment of driving factors in the loess hilly region: a case study as Pengyang County. Environmental Monitoring and Assessment, 164 (1–4), 133–142.