1. Introduction
In recent years, more and more land has been converted for metropolitan development. This is due to increases in population and a decrease in the average household size (Brown et al., 2004). Many metropolitan cities have responded to growing concern about the problems associated with sprawling development patterns (Bengston and Youn, 2005). Such sprawling urban growth is the more dominant pattern of urban development these days and is generally considered undesirable due to its negative effects, such as the loss of open space and damage to the natural environment.
Against this background, many policies and regulations have been designed and implemented to control urban sprawl. As a result, urban containment policies have emerged as a popular means of reducing urban sprawl and preserving farmland (Dawkins and Nelson, 2002). These policies include the establishment of greenbelts of preserved lands around cities. Korea’s greenbelt system was introduced in 1971, which was the legal basis for the creation of the restricted development zone. The greenbelt in Korea may be evaluated as one of few successful greenbelt experiences in Asian countries (Yokohari et al., 2000). However, the greenbelt policy in Korea has been reformed gradually according to the urban planning for sustainable urban land-use management and development. Contemporary planning agendas like sustainable development, smart growth, and the compact city are all explicit or implicit reactions to dispersed and excessive urban expansion. The top level of the urban planning in Korea is the Urban Master Plan which outlines the direction of long-term growth and the future image of the city for sustainable development.
Since the urban planning policy requires knowledge about future urban states, the use of dynamic urban growth models enables planners to explore various what-if scenarios. In the last few decades, urban growth models have played an important role in understanding the causes, mechanisms, and consequences of urban growth dynamics. These models have provided an opportunity to explore and evaluate the urban planning policy, and have helped to visualize alternative futures. Moreover, advances in computer science and its application in urban planning have brought about new spatial modeling approaches, such as cellular automata (CA), statistical models, and multi-agent models. Among all documented dynamic models, CA are probably are the most impressive in urban growth modeling in terms of their fl exibility, simplicity in application, and their close ties to geographic information system (GIS).
The SLEUTH urban growth model is a CA-based model for simulating urban growth in order to improve the understanding of how surrounding land use changes are due to urban expansion (Clarke et al., 1997). SLEUTH has been successfully applied in various regions of the United States and the world over recent decades to simulate land-use change. Recently, SLEUTH also has been applied in South Korea to analyze the inundation of urban areas (Kim and Kim, 2014), climate variability (Kim et al., 2012) as well as land-use change (Han, 2011; Park and Ha, 2013). In addition, the greenbelt policy has been assessed for effi ciency using logistic regression (Kim and Yeo, 2008), the CA-Markov chain (Lee and Oh, 2010), and the Metronamica model (Kim, 2013). Most of the research regarding the greenbelt policy has been limited to the Seoul Metropolitan Area, and future land use has not been assessed for variation by policy changes such as urban planning and legal conservation policies.
This research focuses on modeling urban growth and land-use change under different scenarios to provide a basis for urban planning in the Busan Metropolitan City using the SLEUTH model. The specific tasks of this study include: (1) establishing three scenarios for urban growth; (2) calibrating and validating the model; (3) predicting future land-use change for the years 2020 to 2030, and (4) analyzing and evaluating the results of the prediction under each scenario.
2. Methodology
2.1. Study area
Busan Metropolitan City (hereafter Busan) was selected for the study. Busan is located on the Southeastern tip of the Korean Peninsula within latitude 35° 10 N and longitude of 129° 04 E (Fig. 1). The total area is 656 km2, which excludes some parts of the city located on an island. The period used for the basis of the historical trend was 1990–2010, and the urban growth of Busan was estimated for the year 2030. The area covered by greenbelts in Busan is 253 km2. These greenbelts are distributed across Dongnae-gu, Buk-gu, Haeundae-gu, Gangseo-gu, Geumjeong-gu and Gijang-gun. The area covered by greenbelts at Gangseo-gu and Gijang-gun accounts for 70% of the whole greenbelt area of Busan. In addition, the majority of the area covered by greenbelts except for Dongnae-gu accounts for about 50% of the administrative areas. Busan borders low mountains on the north and west, and ocean on the south and east. The most densely built-up areas of Busan are situated in a number of narrow valleys between the Nakdong and Suyeong rivers, with mountains separating some of the districts. There is a shortage of land for urban development, and has been limited by the greenbelt. Therefore, we need to secure available land by a partial release of the greenbelt and use the land efficiently through environmentally-conscious urban development.
Fig. 1.Location of study area and greenbelt
2.2. SLEUTH model
The SLEUTH model developed by Dr. Keith Clarke of UC Santa Barbara, is a modified CA model, which is capable of modeling the complex urban growth dynamics of a land use change system given a set of historical input data (Clarke and Gaydos, 1998). The name of the model comes from the abbreviation of its data input requirements which are Slope, Land use, Excluded, Urban, Transportation and Hillshade. The model is open source and runs under Unix, Linux and Cygwin, a Window-based Unix emulator (Clarke, 2008). This model is composed of an urban growth model (UGM) and the Deltatron land use model (DLM). The UGM can run independently, but the DLM is tightly coupled with the UGM (Gigalopolis, 2003).
The model calibrates the historical data input to derive a set of five coefficients which control the behavior of the system, and are predetermined by the user at the onset of every model run. The five coefficients consist of diffusion, breed, spread, slope resistance and road gravity (Clarke et al., 1997; Dietzel and Clarke, 2004). Each coefficient is standardized in a range between 0 and 100 that indicates the relative importance of each coefficient (Sakiech et al., 2014). These coefficients drive the transition rules that simulate four types of urban growth: spontaneous (of suitable slope and distance from existing centers), diffusive (new spreading center), organic (infill and edge growth), and road influenced growth (a function of road gravity and density) (Candau, 2002; Dietzel and Clarke, 2004). In addition, SLEUTH also has a self-modification rules. These rules respond to the aggregate growth rate and help to avoid linear and exponential urban growth in the model (Silva and Clarke, 2002).
SLEUTH 3.0 Beta was downloaded and compiled under Cygwin using a GNU C complier (GCC). First, a test mode was run to ensure that the model initially functioned correctly, and then the calibration and prediction phases were implemented.
2.3. Data preparation
For the statistical calibration and application of the model, SLEUTH requires the input of historic urban expansion from at least four time periods, at least two land use layers, a historic transportation network from at least two time or more time periods, a single layer containing percent topographic slope, a hillshade layer and a layer with areas excluded from urban development. Land use layers are needed to implement the DLM and are not necessary to simulate urban growth (Gigalopolis, 2003). All the input data required to execute the SLEUTH model was compiled and analyzed in a GIS.
The land use data of 1990, 1995, 2000, 2010, analyzed using Landsat TM 5, was obtained from the Water Management Information System (WMIS) and the Korean Ministry of Environment (KME). The land use data set was classified into seven types: urban, farmland, forest, grassland, wetland, bare land and water. The maps were converted into binary urban/non-urban layers to depict the profile of Busan’s dynamics since 1990. Transportation network layers were obtained from the National Transport Information Center (NTIC). The main roads (expressways, national highways, metropolitan roads, local roads) were extracted from the maps and then assigned pixel values according to the degree of development (25-100). Slope and the hillshade layers were created from a 30-m Advanced Spaceborn Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Map (GDEM). The slope layer was transformed to percent slope. In the model, the hillshade layer was used as a background image for the model image output. The excluded layer consisted of the greenbelt and the environment conservation value assessment map (ECVAM), which were obtained from the KME. The first and second legal conservation areas from ECVAM were extracted and produced two excluded layers. One of the excluded layers was assigned the value of 100 for the first legal conservation area and the greenbelt, and the value of 75 for the second legal conservation area. This was used for scenario 1. The other layer was assigned the same way, but the greenbelt was assigned the value of 0. That layer was used for scenario 2 and scenario 3
All of the input data was clipped to the same map extent and converted into 100-m resolution raster grids. The grid dimensions were 482 columns by 464 rows with the total number of pixels being 223648. All input data was converted into grayscale .gif format, a requirement of the model. Table 1 and Fig. 2 show the input data set for the SLEUTH.
Table 1.a Korean Water Management Information System; b Korean Ministry of Environment; c National Transport Information Center (Standard Nods Link); d Global Digital Elevation Map.
Fig. 2.Data sets of Busan metropolitan city from 1990 to 2010
2.4. Model calibration
The calibration process of the model is the most important phase for the capture of urban growth, and for the success of the model (KantaKumar et al., 2011). The purpose of the calibration phase is to derive a set of values for the fi ve descriptive coeffi cients that can effectively simulate urban growth during the historical time period (Rafiee et al., 2009). The most commonly used calibration process is known as the “brute force” calibration technique, and during the calibration process, a set of control coefficients is refined by three sequential calibration phases: coarse, fine and final calibrations (Silva and Clarke, 2002). This calibration technique narrows down the range of the model behavior coefficient values sequentially, leaving the set which best replicates the historical data (Clarke et al., 1996; Silva and Clarke, 2002).
During the each step of the calibration process thirteen metrics are computed and these are used to determine the goodness of fit of the model calibration. The best goodness of fit murease for the model is determined by the Optimal SLEUTH Metric (OSM) to provide the most robust results for the SLEUTH calibration (Dietzel and Clarke, 2007). OSM is the product of the compare, population, edges, clusters, slope, X-mean and Y-mean metrics (Table 2). The optimal set of coefficients, based on the OSM, determined the range of values used in the subsequent phase of calibration, and the combination of coefficients with the highest OSM value in the final calibration phase was used for forecasting urban growth and land use change, after adjustment, to reflect their values at the end of the calibration period (Chaundri and Clarke, 2014; Clake, 2008).
Table 2.Source: Dietzel and Clake (2007)
3. Results and Discussion
3.1. Model calibration
The model was calibrated according to the three phases using the historical data from 1990 to 2010. Generally, the model is calibrated using hierarchical spatial resolutions, but it has been shown that using this method may lead to coefficients that do not as accurately describe the growth of the system as a calibration at full data resolution (Dietzel and Clarke, 2004). The full spatial resolution of the input layers (100-m) was applied in three calibration phases. Additionally, self-modification parameters, road gravity sensitivity, slope sensitivity, critical low, critical high, boom, and bust, for instance, were set to default values.
In the coarse phase, the entire range of coefficient values from 0 to 100 was assigned, with incremental steps of 25. For the next calibration phase, these were refined to narrower ranges selected from the top five OSM scores. In the fine calibration phase, diffusion, breed, spread, slope resistance, and road gravity parameters covered values of 0-25, 50-100, 80-100, 0-25 and 50-75, respectively. These ranges narrowed further to 0-15, 50-80, 90-100, 0-5 and 50-65, respectively, in the final calibration phase. Each calibration was successful in increasing the OSM value. The final calibration produced an acceptable top OSM value of 0.724. Then, the best coefficients in 100 Monte Carlo iterations with one step were derived as 9, 55, 92, 3 and 50 (Table 3).
Table 3.a Calibrated and averaged best parameter values for prediction of urban growth in Busan metropolitan city
The coefficient values used to predict growth were 11, 65, 100, 1 and 57 for diffusion, breed, spread, slope resistance, and road gravity, respectively. The diffusion coefficient was low given that Busan has a compact form of growth with its main urbanization occurring near the existing urban areas and urban cores. The breed coefficient is high, which reflects the high probability of the establishment of new urban centers. Also, the high score in the spread coefficient reflects the high probability of urbanization beyond the existing urban centers. For Busan, the slope resistance coefficient was so low that topography was concluded not to be a limiting factor for urban sprawl. Finally, the high road gravity coefficient showed that urban growth has been affected significantly by road networks (Table 3).
3.2. Model validation
A receiver operating characteristic (ROC) curve was used for evaluating the accuracy of the model. The ROC is an approved measurement for assessing the accuracy of binary categorical probability estimations (Pontius and Schneider, 2001). The ROC divides the probability outcomes into percentile groups from high to low probability and compares the individual probability classes with the cumulative real values. The ROC only considers the positive values estimated by the model, which in this study is all urban growth cells. To define the ROC, true positive and false positive rates are plotted for every percentile class. The result is a curve where the area under the ROC curve (AUC) is the measure of the ROC statistic. Generally, AUC value ranges from 0 to 1 and values between 0.7 and 0.9 show reliable precision (Sakieh et al., 2014; Wu et al., 2009)
The urban growth profile based off of the historical trend was simulated. In this study, the 1990 urban extended layer was set as the beginning year and the 2010 urban extended layer was set as prediction stop date. Fig. 3 presents the ROC curve (comparing the cumulative probability image of the year 2010 and the urban extended layer of the corresponding year), where the dashed line illustrates the expected ROC (50%) for a model that selects grid cells at random. The resultant AUC value confirms the validity of the calibrated model, which scored at 0.857—an exceptional performance for the AUC.
Fig. 3ROC curve derived from the cumulative probability of the year 2011 and the urban extended layer of the corresponding year
3.3. Model scenarios and prediction
The three scenarios were set up to analyze the effect of removing the greenbelt: historical trend, development, compact. First, in the first scenario we assumed that future urban growth would occur in the same pattern as its historical trend and that the greenbelt would be maintained in the future. The historical trend of the urban growth was predicted by applying the average values of the coefficients from the calibration. Second, since the greenbelt policy has tended towards a release of the greenbelt, we assumed that future urban growth would occur without the limitation of a greenbelt and a continuation of the historical trend. The future urban growth was predicted by applying the same coefficient of scenario 1. However, an excluded layer was produced without the greenbelt (refer to excluded 2 in Fig. 2). Thirdly, according to ‘The Urban Master Planning of Busan in 2030’, the future urban area created from the released greenbelt area will be developed to realize eco-friendly development. We assumed that future urban growth would be more compact as an answer to the hypothetical greenbelt policies and the lack of land to decrease urban spreading. A compact city is represented by high density, centralized development and spatial mixture of functions (Chin, 2002). In the model, breed and diffusion are the coefficients which mainly describe the trends of urban sprawl. The breed coefficient determines how likely a newly generated detached settlement is to begin its own growth cycle. The diffusion coefficient determines the overall dispersiveness nature of the outward distribution (Leao et al., 2004). The breed and diffusion coefficients were reduced to half (from 65 and 100 to 33 and 50, respectively). The rest of the coefficients and the excluded layer were applied the same as for scenario 2.
Fig. 4 presents the urban extent of Busan under the three scenarios. According to scenario 1 (the historical trend-base scenario), the urban area expanded about 22.0% from 2010 to 2030. However, scenario 2 (the development scenario) showed the most increase in urban growth, showing that the total urban area will expand from 175.96 km2 in 2010 to 246.25 km2 in 2030, a 39.9% increase. Scenario 3 (the compact development scenario) showed the smallest increase in future urban growth as compared to scenario 2 because of a reduced score of breed and spread coefficients to dictate the compact growth. Under this scenario, the urban area was expanded 34.1% from 2010 to 2030 and saved 10.28 km2 of lands from development (Table 4). Therefore, it can be concluded that the anticipated extent of urban growth under scenario 3 can lead to minimum consumption of vacant lands and the secureness of available land.
Fig. 4.Predicted future urban map from the SLEUTH model
Table 4.a Rate of change compared to urban area in 2010 (175.96 km2)
4. Summary and Conclusions
We predicted the impact of removing the greenbelt on urban growth in Busan using the SLEUTH model. We successfully calibrated the model based on historical data from the years 1990-2010, according to the best value of OSM in the final calibration, which was 0.724. In addition, the application of ROC statistics showed an AUC value of 0.857. The model was successful in predicting changes in Busan. We produced three scenarios (historical trend, development, and compact development) and predicted urban growth under three scenarios.
According to scenario 1 (historical trend), the urban area expanded 21.8% from 2010 to 2020 but this increase reached a plateau after 2020 because of a shortage of land for urban development due to the greenbelt. Scenario 2 showed the most increase in urban growth, showing that the total urban area expanded 39.9% from 2010 to 2030. The urban area increased rapidly in comparison to scenario 1 and occurred in Garnseo-gu and Gijang-gun because of the high slope gradients of the released areas. According to scenario 3, the urban area expanded from 175.96 km2 in 2010 to 235.97 km2 in 2030, a 34.1% increase. The urban area under this scenario showed the smallest increase in comparison to scenario 2 and saved 10.28 km2 of lands from development. In this scenario, as well as in scenario 2, urban development has largely started to encroach on farmland in Gangseo-gu, a considerable change in comparison to scenario 1.
The purpose of the greenbelt is to control urban sprawl and protect natural environments. Although the greenbelt will release or relax, the proper function of the greenbelt should be maintained to develop urban sustainably and to become eco-friendly. Therefore, it can be concluded from our results that future urban growth can be controlled by compact development, even if this results in some release of the greenbelt. The anticipated extent of the urban area under compact development can lead to minimum consumption of vacant lands and the secureness of available land.
However, urban growth is a complex process which is also affected by various socioeconomic factors and can be controlled by land use planning and municipality decision. The scenarios used in this study represented the results only as examples that planners can have available for urban growth strategies. In addition, the results of modelling using the SLEUTH model or any other models do not exactly match reality and at the best produce approximations. Any future research should into account the actual legislation and land use changes taking place, in conjunction with the scenarios represented here. The results produced through this research can serve as a decision support tool and aid urban planners and policy makers.
참고문헌
- Bengston, D. and Youn, Y. (2005), Seoul’s greenbelt: an experiment in urban containment, In: Bengston, D., (tech. ed.) Policies for Managing Urban Growth and Landscape Change: A Key to Conservation in the 21st Century (Gen Tech Rep NC-265). US Department of Agriculture, Forest Service, North Central Research Station, St. Paul, pp. 27-34.
- Brown, D.G., Page, S.E., Riolo, R., and Rand, W. (2004), Agent-based and analytical modeling to evaluate the effectiveness of greenbelts, Environmental Modeling & Software, Vol. 19, No. 12, pp. 1097-1109. https://doi.org/10.1016/j.envsoft.2003.11.012
- Candau, J.T. (2002), Temporal Calibration Sensitivity of the SLEUTH Urban Growth Model, Master's thesis, Department of Geography, University of California, Santa Barbara, CA, 116p.
- Chaudhuri, G. and Clarke, K.C. (2014), Temporal accuracy in urban growth forecasting: a study using the SLEUTH model, Transactions in GIS, Vol. 18, No. 2, pp. 302-320. https://doi.org/10.1111/tgis.12047
- Chin, N. (2002), Unearthing the roots of urban sprawl: a critical analysis of form, function and methodology, Centre for Advanced Spatial Analysis, University College of London, London, 47p.
- Clarke, K.C. (2008), A decade of cellular urban modeling with SLEUTH: unresolved issues and problems, In: Brail, R.K. (ed.) Planning Support Systems for Cities and Regions, Lincoln Institute of Land Policy, Cambridge, MA, pp. 47-60.
- Clarke, K.C. and Gaydos, L.J. (1998), Loose–coupling a cellular automaton model and GIS: long–term urban growth prediction for San Francisco and Washing–ton/Baltimore, International Journal of Geographical Information Science, Vol. 12, No. 7, pp. 699-714. https://doi.org/10.1080/136588198241617
- Clarke, K.C., Hoppen, S., and Gaydos, L. (1996), Methods and techniques for rigorous calibration of a cellular automaton model of urban growth, Third International Conference/Workshop on Integrating GIS and Environmental Modeling, Santa Fe, New Mexico.
- Clarke, K.C., Hoppen, S., and Gaydos, L. (1997), A self-modifying cellular automaton model of historical urbanization in the San Francisco Bay area, Environmental and Planning B: Planning and Design, Vol. 24, No. 2, pp. 247-261. https://doi.org/10.1068/b240247
- Dawkins, C.J. and Nelson, A.C. (2002), Urban containment policies and housing prices: an international comparison with implications for future research, Land Use Policy, Vol. 19, No. 1, pp. 1-12. https://doi.org/10.1016/S0264-8377(01)00038-2
- Dietzel, C. and Clarke, K.C. (2004), Replication of spatio-temporal land use patterns at three levels of aggregation by an urban cellular automata, In: Cellular Automata, Springer, Berlin Heidelberg, pp. 523-532.
- Dietzel, C. and Clarke, K.C. (2007), Toward optimal calibration of the SLEUTH land use change model, Transactions in GIS, Vol. 11, No. 1, pp. 29-45. https://doi.org/10.1111/j.1467-9671.2007.01031.x
-
Gigalopolis, (2003), Project Gigalopolis: urban and land cover modeling, University of Santa Barbara, Santa Barbara, CA,
http://www.ncgia.ucsb.edu/projects/gig/. - Han, H. (2011), Development of Future Land Use Scenarios Consistent with Climate Change Storyline, Research Report 2011–05, Korea Environment Institute, pp. 1-153. (in Korean with English abstract)
- KantaKumar, N.L., Sawant, N.G., and Kumar, S. (2011), Forecasting urban growth based on GIS, RS and SLEUTH model in Pune metropolitan area, International Journal of Geomatics and Geosciences, Vol. 2, No. 2, pp. 568-579.
- Kim, D. (2013), Simulating urban growth with the Metronamica model: a case study on the Seoul Metropolitan Area, Korea, Journal of Korea Spatial Planning Review, Vol. 78, pp. 3-17. (in Korean with English abstract) https://doi.org/10.15793/kspr.2013.78..001
- Kim, H., Jeong, J., and Kim, Y. (2012), Analysis of climate variability under various scenarios for future urban growth in Seoul Metropolitan Area (SMA), Korea, Journal of Korean Society for Atmospheric Environment, Vol. 28, No. 3, pp. 261-272. (in Korean with English abstract) https://doi.org/10.5572/KOSAE.2012.28.3.261
- Kim, J. and Yeo, C. (2008), Measuring greenbelt policy effects through the urban growth prediction model, Journal of Korea Planning Association, Vol. 43, No. 3, pp. 211-223. (in Korean with English abstract)
- Kim, M. and Kim, U. (2014), Inundation analysis of urban area considering climate change using SLEUTH model, Journal of Korean Society of Hazard Mitigation, Vol. 14, No. 3, pp. 277-290. (in Korean with English abstract) https://doi.org/10.9798/KOSHAM.2014.14.3.277
- Leao, S., Bishop, I., and Evans, D. (2004), Simulating urban growth in a developing nation's region using a cellular automata-based model, Journal of Urban Planning and Development, Vol. 130, No. 3, pp. 145-158. https://doi.org/10.1061/(ASCE)0733-9488(2004)130:3(145)
- Lee, S. and Oh, K. (2010), Analyzing the effect of greenbelt on urban growth and environmental pollution by cellular automata method, Journal of Korea Planning Association, Vol. 45, No. 3, pp. 193-208. (in Korean with English abstract)
- Park, I. and Ha, S. (2013), Land use change prediction of Cheongju using SLEUTH model, Journal of Environmental Impact Assessment, Vol. 22, No. 1, pp. 109-116. (in Korean with English abstract) https://doi.org/10.14249/eia.2013.22.1.109
- Pontius, R.G. and Schneider, L.C. (2001), Land-cover change model validation by an ROC method for the Ipswich watershed, Massachusetts, USA, Agriculture, Ecosystem and Environment, Vol. 85, No. 1, pp. 239-248. https://doi.org/10.1016/S0167-8809(01)00187-6
- Rafiee, R., Mahiny, A.S., Khorasani, N., Darvishsefat, A.A., and Danekar, A. (2009), Simulating urban growth in Mashad City, Iran through the SLEUTH model (UGM), Cities, Vol. 26, No. 1, pp. 19–26. https://doi.org/10.1016/j.cities.2008.11.005
- Sakieh, Y., Amiri, B.J., Danekar, A., Feghhi, J., and Dezhkam, S. (2014), Simulating urban expansion and scenario prediction using a cellular automata urban growth model, SLEUTH, through a case study of Karaj City, Iran, Journal of Housing and the Built Environment, pp. 1-21 (Published online: 23 December 2014).
- Silva, E.A. and Clarke, K.C. (2002), Calibration of the SLEUTH urban growth model for Lisbon and Porto, Portugal. Computers, Environment and Urban Systems, Vol. 26, No. 6, pp. 525–552. https://doi.org/10.1016/S0198-9715(01)00014-X
- Wu, X., Hu, Y., He, H.S., Bu, R., Onsted, J., and Xi, F. (2009), Performance evaluation of the SLEUTH model in the Shenyang metropolitan area of northeastern China, Environmental Modeling and Assessment, Vol. 14, No. 2, pp. 221-230. https://doi.org/10.1007/s10666-008-9154-6
- Yokohari, M., Takeuchi, K., Watanabe, T., and Yokota, S. (2000), Beyond greenbelts and zoning: a new planning concept for the environment of asian mega-cities, Landscape and Urban Planning, Vol. 47, No. 3, pp. 159-171. https://doi.org/10.1016/S0169-2046(99)00084-5
피인용 문헌
- Simulating Land Use Change in the Seoul Metropolitan Area after Greenbelt Elimination Using the SLEUTH Model vol.2017, pp.1687-7268, 2017, https://doi.org/10.1155/2017/4012929
- The Economic Efficiency of Urban Land Use with a Sequential Slack-Based Model in Korea vol.9, pp.1, 2017, https://doi.org/10.3390/su9010079
- A Review of Driving Factors, Scenarios, and Topics in Urban Land Change Models vol.9, pp.8, 2020, https://doi.org/10.3390/land9080246
- Analysis of the effect of environmental protected areas on land-use and carbon storage in a megalopolis vol.133, pp.None, 2015, https://doi.org/10.1016/j.ecolind.2021.108352