i PARTICIPATORY URBAN GROWTH MODELING - INCORPORATION OF A PARTICIPATORY SIMULATION APPROACH IN MICRO- LEVEL URBAN GROWTH MODELS A Research Paper In Partial Fulfillment of the Requirements for the Degree of Master of Regional Planning by Yawen Chen [May 2024] ii CORNELL UNIVERSITY COLLEGE OF ARCHITECTURE, ART, AND PLANNING DEPARTMENT OF CITY AND REGIONAL PLANNING APPROVAL OF RESEARCH PAPER, PROFESSIONAL REPORT, or THESIS (choose one type of exit project and delete the other two along with this line) Name of Candidate: Yawen Chen First Name Middle Name/Initial Family/Last Name Graduate Field: CITY AND REGIONAL PLANNING Degree: MASTER OF REGIONAL PLANNING (M.R.P.) Title: Participatory Urban Growth Modeling - Incorporation of a Participatory Simulation Approach in micro-level urban growth models COMMITTEE SIGNATURES: Chairperson: Date: 5/21/24 Member: Date: 5/21/24 Member: Date: 5/24/24 Member: Date: Member: Date: LICENSE TO USE COPYRIGHTED MATERIAL I do hereby give license to Cornell University and all of its faculty and staff to use the above-mentioned copyrighted material in any manner consonant with, or pursuant to, the scholarly purposes of Cornell University, including lending such materials to students or others through its library services or through interlibrary services or through interlibrary loan, and delivering copies to sponsors of my research, but excluding any commercial us of such material. This license shall remain valid throughout the full duration of my copyright. ____________Yawen Chen_______________ Student Signature © 2024 Yawen Chen ABSTRACT Around the world, the concomitant impacts of urban growth on environmental and socio-economic issues are causing much concern. The development of urban growth models is to monitor and evaluate the impacts of urban growth and has transformed from mathematical implementations of macro-economic theories to micro-level simulation. Still, few urban growth models consider public participation in the modeling process. At a micro- level urban growth modeling where unique conditions of study areas necessitate a localized research framework, it is crucial to incorporate the active engagement of stakeholders throughout modeling to prove the accuracy of growth models. This paper introduces a participatory modeling approach that allows stakeholders’ engagement in urban growth models, analyzes its methodology, and elaborates on its application in urban growth models using two case studies. Results of case studies show that the introduction of the participatory modeling approach is significant for stimulating public participation, strategic and sustainable urban planning, and innovative future scenarios in micro-level urban growth modeling. iii BIOGRAPHICAL SKETCH Yawen Chen is a second-year master student who is currently studying for the Master of Regional Planning degree at Cornell University, advancing beyond the Bachelor of Engineering degree in Land Use Management obtained from China University of Geosciences (Wuhan). Her previous educational background has honed her academic and practical abilities and research interests in the inter-discipline of natural and social science. Throughout research and internship experiences, Yawen Chen has gained abundant experience in land use and spatial planning. Utilizing a combination of fine-resolution remote-sensing data and socio-economic data to assess current land use status and ecosystem services and establish future scenarios, she has collaboratively developed in China new management methods and proposed land-use pattern optimization that brings sustainability and long-term benefits in seven research programs. In her undergrad study in China, she researched the coupling degree of human-earth system interactions and received training in research programs that aim to design a more sustainable socio-ecological system. These experiences set her foundation for socio- environmental studies. Since coming to Cornell, Yawen Chen has been studying social- environmental interactions in an urban context in the concentration of Land Use and Environmental Planning. Outside of academics, Yawen Chen also emphasizes well-rounded development. She used to volunteer at the Seventh Military Games and the outbreak of COVID-19 and acted as a researcher in a program promoting educational equity among female students in an ethnic minority group in China. iv ACKNOWLEDGMENTS I hope to express my deepest appreciation to all the teachers and friends I met at Cornell University. The education I received at Cornell University and the Department of City and Regional Planning has expanded my knowledge of urban planning, especially in land and environmental planning, in both breadth and depth. Studying in this inclusive environment has fed my passion for embracing diversity in all forms and is an invaluable experience. Last, I am sincerely grateful to my exit project committee board. The completion of my research paper would not have been possible without their support and guidance. v TABLE OF CONTENTS BIOGRAPHICAL SKETCH..............................................................................................iii ACKNOWLEDGMENTS.................................................................................................. iv TABLE OF CONTENTS.................................................................................................... v LIST OF FIGURES............................................................................................................ vi LIST OF ABBREVIATIONS........................................................................................... vii PREFACE........................................................................................................................ viii 1. Introduction...................................................................................................................... 1 2. Urban Growth Models and Modeling.............................................................................. 2 3. Evolution of Urban Growth Models: From macro-level to micro-level.......................... 3 4. Incorporating Participatory Simulation: Urban Growth Models at Local Scale..............6 4.1 Doing the Right Thing or Doing Things Right.............................................................. 7 4.2 Methodology of Participatory Modeling........................................................................8 4.3 Applicability of PM in Urban Growth Modeling...........................................................9 5. PM in Urban Growth Modeling: Worldwide Case Studies............................................10 5.1 Strategic land use planning in Maipo River Basin, Chile........................................ 10 5.1.1 Participatory simulation approach to measure future development plans......... 11 5.1.2 The role and impacts of participatory simulation.............................................. 13 5.2 Disruptive future scenarios simulation in Henares Corridor, Madrid - Guadalajara (Spain)............................................................................................................................ 15 5.2.1 Research framework of the simulation of disruptive future scenarios of urban land use change.................................................................................................................. 16 5.2.2 The role and impacts of participatory simulation.............................................. 18 6. Conclusion......................................................................................................................20 Bibliography.......................................................................................................................22 vi LIST OF FIGURES Figure 1 Typology of methods used in PM with example work flows................................8 Figure 2Methodology of assessing land use planning scenarios.......................................11 Figure 3 Research framework of disruptive future scenarios building in simulating urban land use change.................................................................................................................. 16 vii LIST OF ABBREVIATIONS CA............................................................................................................ Cellular Automata ABMs...................................................................................................Agent-Based Models PM.................................................................................................... Participatory Modeling SBT.................................................................................................Scenario Building Team Dyna-CLUE............................................ Dynamic Conversion of Land Use and its Effects LCM......................................................................................................Land Change Model BAU........................................................................................................ Business-As-Usual PROT..............................................................Regional Land Use Plan (Spanish Acronym) LP-CA..................................................................................Land Parcel Cellular Automata SMR...................................................................................... Santiago Metropolitan Region viii PREFACE This paper introduces a participatory modeling approach in urban growth models to establish a highly explanatory, localized research framework at micro-level urban growth modeling. 1 1. Introduction Around the world, growing numbers of people are moving to urban centers for a better life quality economically, socially, and environmentally. Cities are expanding fast in response to population dynamics, income variations, transportation, land prices, housing size, public services, and amenity accessibility. At the same time, the concomitant impacts of urban growth on environmental and socio-economic issues, such as forest and farmland degradation, loss in biodiversity, climate and energy crisis, public health emergencies, and urban heat island are causing much concern (Grimm et al., 2008, Seto et al., 2011). To monitor and evaluate the impacts of urban growth, researchers have been developing urban growth models to dissect the mechanism of urban growth and extrapolate it to global cities. There are three classic types of urban growth models, the Land-Use Transportation model, the Urban Dynamic model, and the micro-level simulation models. These classic models are useful tools in forecasting urban growth patterns, but they also have distinct limits, especially in micro-level urban growth simulation. For the Land-Use Transportation model and the Urban Dynamic models, their underlying hypotheses build upon macro- economic theories and lack the capturing of real-life urban dynamics. The micro-level simulation models, on the other hand, usually build upon relatively ‘static’ modeling rules that do not vary substantially among different circumstances. Regarding micro-level urban growth simulation, social and environmental conditions of study areas are dynamic, unique, and usually lack empirical data, making typical urban growth models less applicable. Therefore, a localized research framework is necessary, from which researchers can interact with diverse groups of stakeholders, collect accurate information, and learn exactly about the decision-making processes of stakeholders. 2 In response to the need for localization in micro-level urban growth modeling, it is crucial to promote the active engagement of all groups of stakeholders in different stages of the research framework. The active participation of local stakeholders enables researchers to build retrospective feedback mechanisms between individuals and their modeling environment and implement iterative model calibration and validation in micro-level urban growth modeling, which significantly increase the accuracy of modeling results. To establish an urban growth research framework that engages local stakeholders, this paper introduces a participatory modeling framework in urban growth models that allows public participation throughout modeling. This paper first evaluates the limits of existing models in micro-level urban growth modeling and argues the necessity of incorporating pubic participation in micro-level urban growth modeling. Then, this paper utilizes two case studies to elaborate on the research framework of coupling participatory modeling with urban growth models, tests the effectiveness of the research framework, and discusses the advantages of applying the participatory modeling approach in micro-level urban growth modeling. The following section briefly introduces the mechanism of urban growth models. 2. Urban Growth Models and Modeling In discussing urban growth models, it is necessary to identify the definition of ‘models’ and ‘modeling’ in our context. According to computer and software development research, a model is an abstract of a system created for some purposes. A model represents the exact and distinct copies of certain properties in a system that the creator hopes to retain, and modeling is the process of simplification, where creators pick the properties of a system (theories, for instance) and 3 develop ways to preserve them for the purpose of description, simulation, and construction plan (Kühne, 2005). Considering urban growth research, urban growth models are created to describe or predict urban growth, develop alternative scenarios, and determine urban growth impacts. They serve as effective city planning instruments, and the modeling process integrates the principles and methods in different fields. Depending on the diversity of purposes and objects of modeling, creators of urban growth models can describe the urban system in aggregated or disaggregated (namely bottom-up) ways. The variety of modeling approaches is influenced by their underlying urban growth hypotheses and the application circumstances of urban growth models. The following section discusses the methodology of three classic types of urban growth models and argues their ineffectiveness in micro-level urban growth modeling. 3. Evolution of Urban Growth Models: From macro-level to micro-level The development of the first type of urban growth model is closely related to modeling urban land use and can be dated back to the work of Alonso (1964), Mills (1967), and Muth (1969) who first encapsulated city transport, land use, and population issues into what is known as the monocentric city model, the Alonso-Muth-Mills model (Duranton et al., 2015). The Alonso-Muth-Mills model assumes that in a city with a given population and wage level, all location choices are related to their proximity to the city center. In 1964, Lowry (1964) introduced a method that modularized the interactions of land use and transportation in an analytical model based on the retail-employment loop he created and named the model Land-Use Transportation model. The Lowry model is rooted in the assumption of urban areas being a static, spatial homogeneous system, and that activities of urban systems can be 4 well simulated in a growth model under good fit (Batty, 1976). These early models are the paradigms of land-use transportation models built upon general equilibrium theory in microeconomics and the gravity principle of social physics. Within less than two decades after the development of the first type of urban growth models, new growth theories that refuted the assumptions of urban systems being static and spatial homogeneous became dominant. In 1978, the concept of ‘order by fluctuation’ was introduced to urban system modeling, where the urban system was characterized as internally unstable while evolving towards an order that allows extrapolative predictions (Allen, 1978). This marks the development of urban dynamics models, which applied the non-linear and non-equilibrium mechanisms to modeling and filled the gap in spatial and timely dynamics in growth modeling. However, the typical urban dynamics models still performed poorly in interpreting real-life urban dynamics because most of them lacked spatial disaggregation of the urban system (Alonso, 1964; Forrester, 1970; Allen, 2001) The third class of urban models allows the projections on a more individualistic level, as they focus on specific sectors or activities within the urban system and build upon the growing maturity of urban growth theories that urban is spatially dynamic and heterogeneous. Typical models in this class are the Cellular Automata (CA) models and Agent-Based Models (ABMs). CA models build upon the transition rules or possibilities of one cell to change to another state, and in land use/land cover change simulation, the ‘cell’ always refers to a land parcel unit and the ‘state’ is a land use type (White & Engelen, 1993; Clarke et al., 1997; Wu, 2002; Arsanjani et al., 2013). To apply CA models in urban growth modeling, modeling rules are based on assigning constraints of modeling, weighting parameters, stochastic disturbance 5 term, and uniform random variate to an equation that calculates the potential of a unit within an urban system to transit from one state to another. Unlike CA models, ABMs build upon the simulation of decision-making processes and individuals’ behaviors in an urban system and have no universal calculations. Despite the mathematical implementations that vary widely for different kinds of problems and the groups of agents involved in the decision-making and modeling, almost all ABMs consist of agents and their behaviors, goals, and their interactions. ABMs have excellent performance in spatial planning and urban policy-making, where they play significant roles in solving specific problems in urban growth and do well in providing reality-based reproduction on human actions (Ferber, 1999; Matthews et al., 2007; Irwin & Bockstael, 2017). Based on the discussions of typical urban growth models, we can see with their transition from macro-level to micro-level simulation, the complexity and dynamics of urban systems are well-considered in the micro-level urban growth models. However, the effectiveness of existing micro-level urban growth models in interpreting such urban complexity and dynamics is questionable. For instance, CA models excel in simplicity, flexibility, self-adjustment, and self- learning (Li, & Gong, 2016), while given the complexity and unpredictability of disaggregated urban systems, it is hard to propose transition rules that fit all. The reliance of CA models on statistical methods, data calibration, and data mining makes them lag behind the explanation of urban growth patterns at the micro-level that lacks empirical data. On the other hand, even if ABMs engage different groups of stakeholders in urban growth modeling, true modelers are still groups of people who know well about the institutional approach and academic theories behind urban growth. At a micro-level urban growth simulation, where 6 unique environmental conditions, historical and political context, and social and cultural milieu necessitate a localized research framework, classic modeling methods are not as well accurate because the design of their modeling processes cannot incorporate the thoughts, needs, and decisions of all stakeholders. Therefore, at a micro-level urban growth modeling, a localized research framework is necessary, from which researchers can interact with diverse groups of stakeholders, collect accurate information, and learn exactly about the decision-making processes of stakeholders to train urban growth models. 4. Incorporating Participatory Simulation: Urban Growth Modeling at Micro-Level In response to the need for localization at micro-level urban growth modeling, it is crucial to promote the active engagement of all groups of stakeholders throughout different stages of the research framework. More importantly, the participation of all groups of stakeholders, especially those historically underrepresented, should be empowered, otherwise, the modeling process is likely to adhere to existing templates in projecting future development scenarios, which lead to ineffective and inequitable modeling results. This section presents a participatory modeling approach that allows public participation throughout modeling, discusses its theories, components, and typology, and critically assesses the benefits of their application at micro-level urban growth modeling for localization purposes. 4.1 Doing the Right Thing or Doing Things Right Participatory Modeling (PM) is a purposeful learning process for action that engages the implicit and explicit knowledge of stakeholders to create formalized and shared representations of reality (Voinov et al., 2018). Underlying PM, there are the interactions of 7 single-loop learning and double-loop learning, where the former describes the learning process of all participants following a fixed, single reference frame, and the latter describes an alternative learning process that accommodates questions and critiques of participants on the reference frame in the process of learning (Argyris & Schön, 2002). Putting these two learning processes into PM procedures, the retrospective feedback mechanisms between individual and organizational learning make it possible for stakeholders of PM to identify and challenge their expectations and goals of modeling, thus reconstructing their knowledge and behavioral feedback to their environment in this iterative process (Voinov et al., 2018). In addition to the knowledge pluralism that comes with PM’s limitless modeling process, stakeholders can learn not only from other participants and their environment, but also from the process of critiquing the reference frame and adjusting their beliefs, or in other words, they also learn from the way they change their thoughts and reference system. It is invaluable as participants evaluate back and forth in a decision-making process, in which they know better what they and what other stakeholders want, what trade-offs exist here, and which pathway will maximize the benefits of the whole system. Instead of following a given template to reach an agreement and do the right thing, PM enables doing the thing right. To elaborate on ‘doing the thing right’, for instance, PM widely engages the public in interpreting scientific abstractions, selecting modeling methods and tools, and modularizing social and natural environment. The collaboration on identifying driving factors of the modeling environment is another aspect of PM that helps to maximize the benefits of the whole system (Voinov et al., 2014). PM aims to create a climate in which non-professionals have accessibility to a transparent and accountable modeling process with the help of skillful researchers, and simultaneously, it embraces indigenous knowledge and societal values with 8 respect to stakeholders’ expectations on modeling. 4.2 Methodology of Participatory Modeling Given the diversification of modeling purpose and environment, PM does not have a specific paradigm, even when solving problems of the same kind. However, main tools and methods that cover most PM processes are well-articulated. Fig. 1. shows the main contents of a complete PM process (Voinov et al., 2018). Fig. 1. Typology of methods used in PMwith example work flows. Retrieved from: Voinov, A., Jenni, K., Gray, S., Kolagani, N., Glynn, P. D., Bommel, P., ... & Smajgl, A. (2018). Tools and methods in participatory modeling: Selecting the right tool for the job. Environmental Modelling & Software, 109, 232-255. According to Fig. 1., almost all PM processes involve two elements, which are fact- 9 finding and process orchestration, where modelers acquire data, information, and knowledge to exchange thoughts and refine ideas (Voinov et al., 2018). Process orchestration is the element that maximizes public involvement and conflict resolution in PM, and Role-Playing Game is a famous example of process orchestration. By engaging non-professionals or those historically excluded in decision-making to participate in PM through role-playing games (Castella et al., 2005), an equitable and justifiable modeling process is promising. The rest of the PM processes consist mainly of qualitative, quantitative, and semi-quantitative modeling, and depending on application circumstances, modelers can choose one or a combination of these modeling approaches. It is noteworthy that many PM projects include looping back from any stage, even from the most sophisticated quantitative modeling, to fact-finding and data acquisition, and sometimes to the problem definition stage (Voinov et al., 2018). 4.3 Applicability of PM in Urban Growth Modeling Attempts to apply PM in research in natural resources management (Voinov & Bousquet, 2010), land use/land cover change (Castella et al., 2005), climate change and water management (Bach et al., 2014; Basco-Carrera et al., 2017), and social-ecological system (Gray et al., 2015) have proved the efficacy of PM in modeling specific urban sectors or problems. However, there are some challenges when applying PM at micro-level urban growth modeling. For example, the complexity of managing a PM is incomparable to that of other urban growth models. Not all stakeholders have the expertise to contribute efficiently to the scientific abstraction and interpretation in modeling, and thus non-modelers need a much longer time to familiarize themselves with the modeling process and the model calibration and validation process. This complexity can hinder the effective participation of 10 stakeholders and the translation of their knowledge and preferences into the model. Additionally, if power problems exist and those influential and historically empowered still dominate the modeling process, the participation of stakeholders in urban growth modeling will only be window-dressing. The politicizing effects of data are an example of such a power problem, where underrepresented stakeholders have limited access to the complete data that they need to participate in the PM process (Nost, 2022). Thus, to apply PM at micro-level urban growth models, specific rules must be established and incorporated into the modeling process to strengthen the effectiveness of PM. These rules must be drawn from the methodology and participation nature in PM and the local conditions of research areas. The next section examines two examples of applying PM in micro-level urban growth modeling, focusing on how PM engages public participation, coordinates public interests, and establishes specific rules to train models and facilitate localized research frameworks. 5. PM in Urban Growth Modeling: Worldwide Case Studies 5.1 Strategic land use planning in Maipo River Basin, Chile This case study is based on the Maipo River Basin, Chile. Guided by participatory instances with a group of public, private, and private stakeholders (called the Scenario Building Team (SBT)), researchers implement a land change model (LCM) called Dynamic Conversion of Land Use and its Effects (Dyna-CLUE) (Henríquez-Dole et al., 2018) to assess the long-term policy impact. In this case, a participatory simulation approach runs throughout modeling, and two policy frameworks, the business-as-usual plan, BAU, and the strategic land use plan, PROT (Spanish acronym of regional land use plan) that is jointly formulated by SBT are incorporated into different land use planning scenarios for model 11 calibration and validation. The Maipo River Basin lies between latitudes 33 and 34 South in Chile. Most of it is located in the Santiago Metropolitan Region (SMR). Santiago is the capital of Chile and the primary area of population concentration and manufacturing in the country (Henríquez-Dole et al., 2018). Affected by the continuous wave of migration and public policies that strengthened unregulated urbanization, SMR has experienced rapid urban growth in the past few decades, while such urban growth has caused many side effects on ecosystem services (Henríquez-Dole et al., 2018). 5.1.1 Participatory simulation approach to measure future development plans In this case, methodology is derived from existing research on the same area. Fig. 2. shows the methodology flowchart that utilize a participatory modeling approach (Henríquez- Dole et al., 2018). 12 Fig. 2. Methodology of assessing land use planning scenarios. Retrieved from: Henríquez- Dole, L., Usón, T. J., Vicuña, S., Henríquez, C., Gironás, J., & Meza, F. (2018). Integrating strategic land use planning in the construction of future land use scenarios and its performance: The Maipo River Basin, Chile. Land Use Policy, 78, 353-366. As the essential component of the research framework, the Scenario Building Team (SBT) is a working group that brings together several participants representing the public, private sector, and civil society related to policy formulation (Ocampo-Melgar et al., 2016). Each decision-making process in this case study is closely related to the participation of SBT. In the research framework, there are three main inputs in the Dyna-CLUE land change model, which are Land use demands, Location suitability, and Spatial conditioning. The first input, land use demands, is obtained from historical land use areas and derived from satellite images (2001-2012). Researchers use these data for model calibration and validation, and based on the documentation that SBT provides, they can extrapolate these data into the future (Henríquez-Dole et al., 2018). The second input is location suitability with its driving force. Looking back to the local stakeholders’ database, SBT can identify spatial predictors of each land use type through seminars and then utilize statistical models to calculate Location suitability. The last input is spatial conditionings, which are policy restrictions that limit or enhance land use changes in specific regions (Henríquez-Dole et al., 2018). In this case, spatial conditionings involve two policy frameworks: the business-as-usual plan, the BAU, and the strategic land use plan jointly formulated by SBT, the PROT. SBT assumes that, in the BAU scenario, there will be increasing market influence on political and economic 13 decision-making and the continuous fragmented urban expansion. In contrast, in the PROT scenario, SBT assumes that community and private participants will actively participate in land use decision-making, and the internal institutional transformation of environmental institutions will help to strengthen ecological protection. With the above three inputs, Dyna-CLUE can simulate future scenarios from 2010 to 2050. SBT also uses landscape indicators obtained in two years (2030 and 2050) to evaluate policy effectiveness and performance (Henríquez-Dole et al., 2018). 5.1.2 The role and impacts of participatory simulation Based on the research framework in 5.1.1, future land use projections under the two policy frameworks show different results. Although neither BAU nor PROT can completely prevent the conversion of agricultural land to urban land, PROT established via participatory modeling can better concentrate urban growth, promote the development of peripheral areas, and maintain more agricultural land than BAU. In addition, PROT can also promote the integration of urban and rural areas and help to protect cultural heritage and ecosystems in the Maipo River Basin (Henríquez-Dole et al., 2018). In this case study, participatory simulation runs through every stage of the research framework and plays a significant role in forming strategic land use planning. In the data acquisition stage, SBT contributes to collecting empirical data, satellite images, and stakeholder documentation. In addition, researchers hold several seminars using the ‘open space’ method to divide the influencing factors of local development into some main topics for various group discussions. Through these seminars, stakeholders of SBT collaboratively identified the key factors of urban growth modeling in the Maipo River Basin, especially those undocumented in existing databases or official documents and reports. 14 In data analysis and modeling, SBT also ranks the importance of the driving factors by 1-10 in their seminars to assign different weights to each driver to determine the data input. Particularly, SBT value population growth, deforestation, and desertification with the lowest uncertainty level based on their abstractions from the current socio-environmntal dynamics in Maipo River Basin. When assessing the impact of different spatial policies on the future land use pattern, SBT stakeholders have cooperated to create two policy frameworks, BAU and PROT, where they incorporated their interpretations of the local territorial context and dynamics into these two frameworks. Specifically, SBT derives the assumptions of the BAU scenario from the poorly-regulated real estate market and the lack of social movements in opposing housing policies in Maipo River Basin. In contrast, in the PROT scenario, SBT assumes that community and private participants will actively participate in land use decision-making, and the internal institutional transformation of environmental institutions will help to strengthen ecological protection. Compared with existing models in land use change simulation, participatory modeling of SBT helps the Dyna-CLUE model to incorporate more indigenous knowledge by evaluating a considerable number of influencing factors. Through seminars and discussion forums, all groups of stakeholders can express their views equitably and make decisions in ranking the relevancy and uncertainty of different influencing factors, and the final input of data is based on the average weight of all the evaluations of SBT members. At the same time, stakeholders learn better about the trade-offs in urban planning through the participation processes, such as the calibration of the model and the identification of the future policy framework, thus proposing a more applicable and effective strategic land use planning. Moreover, the invaluable participation of stakeholders in the modeling is of great 15 educational value and can also demystify modeling, which draws non-professionals nearer to scientific research methods and motivates them to engage in participatory urban growth simulation in the future. 5.2 Disruptive future scenarios simulation in Henares Corridor, Madrid - Guadalajara (Spain) This case study is carried out in the Henares Corridor, Madrid - Guadalajara in Spain. Affected by unexpected events such as the Spanish real estate bubble crisis (Burriel, 2011), the European immigration crisis (Hampshire, 2015), the SARS-CoV-2 pandemic (Antipova, 2021), and the recent war in Ukraine, the effectiveness of scenario planning that is close to business-as-usual projection is questionable because unexpected events tend to undermine the linear planning (Molinero-Parejo et al., 2022). Novel methodology to forecast and manage unexpected future urban scenarios is needed. This case develops an approach that combines the land parcel cellular automata (LP-CA) model and the participatory modeling process to predict possible, undesired future urban scenarios in the long term and test the practicality of the prediction method. Through the critical conception of the future urban growth trajectory, modelers generate some disruptive urban scenarios to subvert the existing simulation scenarios. By spatially expressing the future land use change trajectory under these disruptive scenarios, this case aims to supplement the blueprint for future urban growth. The research area is located in the urban industrial corridor of Enares, Madrid- Guadalajara (Spain), which has undergone significant urban transformations in recent decades (Molinero-Parejo et al., 2022). Nowadays, the Henares Corridor is characterized by small and medium-sized towns with significant industrial fabric and a variety of territorial and social dynamics (Barreira-González et al., 2019; Cantergiani & Gómez Delgado, 2020; 16 Molinero-Parejo et al., 2022). 5.2.1 Research framework of the simulation of disruptive future scenarios of urban land use change Fig. 3. Research framework of disruptive future scenarios building in simulating urban land use change. Retrieved from: Molinero-Parejo, R., Aguilera-Benavente, F., Gómez-Delgado, M., & Shurupov, N. (2023). Combining a land parcel cellular automata (LP-CA) model with participatory approaches in the simulation of disruptive future scenarios of urban land use change. Computers, Environment and Urban Systems, 99, 101895. Fig. 3. illustrates the components of the research framework in this case. The research framework consists of two parts, the LP-CA model and the participatory workshop. In the LP-CA model, all influencing factors related to the transformation potential of land parcels 17 are first collected and processed in the data preparation stage. Because the random factors representing human decision-making and behaviors and the attraction-repulsion functions representing the neighborhood effect between adjacent land parcels are difficult to quantify, researchers help non-professional modelers solve the difficulty by introducing stochastic perturbation and the vector Enrichment Factor to calculate the transition potential in different buffer zones (Molinero-Parejo et al., 2022). Accessibility, Suitability, and Zoning are the other three parameters determining the potential of a specific urban land use to transit to another use. Accessibility in this case is measured as the distance from the centroid of each land parcel to the nearest edge of the road network (streets, highways, highways, and toll roads) (Molinero-Parejo et al., 2022). Suitability of urban land use on each land parcel is assessed with geographically weighted Logistic regression (GWLR). Three main types of zoning classification are established based on the most typical land use types in the research area (urban land, undeveloped land, and protected non-urban land) as the Zoning input, and the Zoning controls the weighting assigned to each parcel based on its suitability for development according to the legal planning framework (Molinero-Parejo et al., 2022). Besides data preparation, this case also includes retrospective model validation and calibration processes within the same time period as future simulations to calibrate the LP-CA model. In the participatory workshop, 129 locals engage to identify seven disruptive narrative storylines of the land use and traffic evolution of the Enares Corridor in 2050 through a semi-structured interview seminar that is combined with wild cards for trend transition points caused by emergencies (Molinero-Parejo et al., 2022; Soria-Lara et al., 2021). Important quantitative information in these narrative storylines as input for modeling 18 scenarios is also identified in the participatory seminar. Finally, this case combines the model framework of LP-CA with the disruptive future scenarios built by the participatory workshop to calculate the transition potential and future allocation of urban land use. 5.2.2 The role and impacts of participatory simulation Coupling the LP-CA model with the participatory process, the simulation of the Henares Corridor from 2018 to 2050 shows the following trends: (1) significant land use changes from multi-family residential to single-family residential; (2) urban land loss as a result of the abandonment of built-up areas in the city center; (3) urban growth in the periphery areas will be intensified and cities in the metropolitan area expand rapidly (Molinero- Parejo et al., 2022). These prediction results are in line with the narrative characteristics of the input simulation model established by the participating workshop, such as the isolation of urban land use and the transformation from multi-family residential to single-family residential, and the fact that the city of Alcalá de Henares has suffered a sharp decline in population in the past few decades (Molinero- Parejo et al., 2022). The participation process of this case is mainly manifested in the following aspects. First, through semi-structured interviews, local stakeholders and researchers collaboratively established the disruptive narrative storylines needed for future scenario simulation, thus, these narratives stem not just from urban planners but also from local people’s understandings of the urban growth characteristics and the conception of unexpected events in the future. For example, in envisioning future transportation, the business-as-usual scenario usually assumes that due to the increasing oil prices and the deterioration of the global environment, motorized vehicle will be losing their dominance in the future, and 19 teleworking will become the mainstream. However, in the participatory workshop, researchers asked non-professional participants aged 18 - 32 to share their preferences for daily work, leisure, tourism, and communities in the future, and then derive a future scenario from these ideas. The result shows that based on non-professional stakeholders’ envisions, motorized vehicles will still dominate in Henares Corridor in the future, and teleworking will be a marginal option. Further, stakeholders represented by urban planners, transportation planners, real estate developers, civil engineering lecturers, and environmental consultants determine the quantitative input information of these narrative storylines and establish a dynamic map that can spatially express the characteristics of these narratives (Molinero-Parejo et al., 2022). In addition, the decision of the three elements of the LP-CA model, the quantification of Accessibility, Suitability, and Zoning are also determined through consultation by stakeholder groups. Finally, the iterative validation of the adapted LP-CA model includes the analysis, design, and discussion related to the spatial configuration of urban land use and transportation networks (Molinero-Parejo et al., 2022) and is guided by experts (Molinero- Parejo et al., 2022). The highly explanatory simulation results prove the efficacy of participatory modeling in depicting the spatial representation of complex urban dynamics (Molinero-Parejo et al., 2022). This innovative urban growth modeling approach can help urban planners better prepare to respond to the changing, unexpected future. It is worth mentioning that not all stakeholders are capable of building models, but in the form of seminars, stakeholders, with the help of researchers, understand the operational principles of LP-CA, collect the necessary professional knowledge, and convert it into the 20 input of the LP-CA model, and improve information quality by integrating the human component (Molinero-Parejo et al., 2022). The participatory workshop ensures public participation in the urban growth simulation, and the spatial allocation of urban land use and transportation networks happening in the workshop also enables stakeholders to better understand the mechanism of urban land use change and participate in the modeling process. 6. Conclusion Through the review of urban growth models, the introduction of the participatory modeling framework, and the analyses of the application participatory modeling in urban growth modeling, it is evident that the addition of participatory modeling has greatly improved micro-level urban growth modeling and filled the gap of public participation in classic urban growth models. The case study of Henares Corridor proves that the research framework that engages stakeholders, residents, and experts to propose disruptive future scenarios, weigh different driving factors, and implement iterative model calibration helps to provide urban planners with a broader perspective on future urban growth. The case study of Maipo River Basin shows that the policy framework of urban growth determined in participatory seminars by the public sector, private sector, and civil society performs better in solving the current ecological and demographic issues than the business-as-usual policy framework, enabling urban planners to formulate strategic and strategic land use planning and balance trade-offs between stakeholders. Also, both cases solved the operational difficulties of non- professionals by promoting the interaction among stakeholders and experts at the participatory seminars. Besides the above strengths of the micro-level participatory urban growth model, that it 21 helps provide urban planners with a broader perspective on future urban growth scenarios and establish locality-specific research framework to help solve problems efficiently, stakeholders’ participation itself is of great educational value and can demystify scientific research methods to motivate stakeholders to engage actively in participatory urban growth simulation in the future. There are still limits to coupling the urban growth model with PM. Some of the driving factors of modeling decided by participatory seminars cannot be quantified, making it hard to establish rules or parameters through modeling to explain the spatial dynamics of such driving factors. The future urban growth policy proposed to maximize public interests may sometimes worsen some pressing problems. In addition, the coupled urban growth model is much more time-consuming and costly than classic urban growth models. However, there are still solutions to these limits if we follow some basic modeling rules. Here, I propose three rules in micro-level participatory urban growth modeling. First, stakeholders should engage in every stage of modeling and decision-making, and their participation has to be empowered. This rule helps to solve local power problems and strengthen community cohesion. Second, to prove the efficacy and efficiency of public participation, researchers/experts need to provide the necessary support throughout modeling because they can teach stakeholders to use alternative mathematical implementations and maximize overall benefits. Third, the local government should actively intervene in modeling because they can provide enough resources and strive for financial support nationally and globally. To apply these modeling rules in micro-level urban growth simulation, we are likely to solve the potential limits. Under careful design and specific modeling rules, the introduction of PM into urban 22 growth modeling is undoubtedly significant for stimulating public participation and community outreach, strategic and sustainable urban planning, and innovative future scenarios in micro-level urban growth modeling. 23 Bibliography Alonso, W. (1964). Location and land use: toward a general theory of land rent. Harvard university press. Allen, P. M., & Sanglier, M. (1978). Dynamic models of urban growth. Journal of Social and Biological Structures, 1(3), 265-280. Allen, P. M. (2001). Cities and Regions as Self-organizing Systems. Models of Complexity. System Dynamics Review, 17(4), 365-371. Argyris, C., & Schön, D. A. (2002). Apprentissage organisationnel: théorie, méthode, pratique. De Boeck Supérieur. Arsanjani, J. J., Helbich, M., Kainz, W., & Boloorani, A. D. (2013). Integration of logistic regression, Markov chain and cellular automata models to simulate urban expansion. International Journal of Applied Earth Observation and Geoinformation, 21, 265-275. Antipova, T. (2020, May). Coronavirus pandemic as black swan event. In International conference on integrated science (pp. 356-366). Cham: Springer International Publishing. Batty, M. (1976). Urban modelling (pp. 20-48). Cambridge: Cambridge University Press. Bousquet, F., and C. Le Page. 2004. Multi-agent simulations and ecosystem management: a review. Ecological Modelling. 176(3/4):313-332. Bach, P. M., Rauch, W., Mikkelsen, P. S., McCarthy, D. T., & Deletic, A. (2014). A critical review of integrated urban water modelling–Urban drainage and beyond. Environmental modelling & software, 54, 88-107. Burriel, E. L. (2016). Subversion of land-use plans and the housing bubble in Spain. In Urban Challenges in Spain and Portugal (pp. 14-31). Routledge. Basco-Carrera, L., Warren, A., van Beek, E., Jonoski, A., & Giardino, A. (2017). Collaborative modelling or participatory modelling? A framework for water resources management. Environmental Modelling & Software, 91, 95-110. Barreira-Gonzalez, P., Aguilera-Benavente, F., & Gomez-Delgado, M. (2019). Implementation and calibration of a new irregular cellular automata-based model for local urban growth simulation: The MUGICA model. Environment and Planning B: Urban Analytics and City Science, 46(2), 243-263. Clarke, K. C., Hoppen, S., & Gaydos, L. (1997). A self-modifying cellular automaton model of historical urbanization in the San Francisco Bay area. Environment and planning B: Planning and design, 24(2), 247-261. Castella, J. C., Trung, T. N., & Boissau, S. (2005). Participatory Simulation of Land-Use Changes in the Northern Mountains of Vietnam: the Combined Use of an Agent-Based Model, a Role-Playing Game, and a Geographic Information System. Ecology and Society, 10(1). Cantergiani, C., & Delgado, M. G. (2020). Simulación del crecimiento urbano con AMEBA: Modelo Basado en Agentes para la ocupación residencial. Boletín de la Asociación de Geógrafos Españoles, (86). Duranton, G., & Puga, D. (2015). Urban land use. In Handbook of regional and urban economics (Vol. 5, pp. 467-560). Elsevier. Forrester, J. W. (1970). Urban dynamics. IMR; Industrial Management Review (pre-1986), 11(3), 67. Ferber, J (1999). Multi-agent systems: an introduction to distributed artificial intelligence. Addison- Wesley Longman, Harlow, UK, 509 pp. Grimm, N. B., Faeth, S. H., Golubiewski, N. E., Redman, C. L., Wu, J., Bai, X., & Briggs, J. M. (2008). Global change and the ecology of cities. Science, 319(5864), 756-760. Gray, S. A., Gray, S., De Kok, J. L., Helfgott, A. E., O'Dwyer, B., Jordan, R., & Nyaki, A. (2015). Using fuzzy cognitive mapping as a participatory approach to analyze change, preferred states, and perceived resilience of social-ecological systems. Ecology and Society, 20(2). Hampshire, J. (2015). Europe’s Migration Crisis. Political Insight, 6(3), 8-11. Henríquez-Dole, L., Usón, T. J., Vicuña, S., Henríquez, C., Gironás, J., & Meza, F. (2018). Integrating strategic land use planning in the construction of future land use scenarios and its performance: The Maipo River Basin, Chile. Land Use Policy, 78, 353-366. Irwin, E. G., & Bockstael, N. E. (2017). Interacting agents, spatial externalities and the evolution of residential land use patterns. In The Economics of Land Use (pp. 359-382). Routledge. Kühne, T. (2005).What is a Model?. In Dagstuhl Seminar Proceedings. Schloss Dagstuhl-Leibniz- Zentrum für Informatik. Li, X., & Gong, P. (2016). Urban growth models: progress and perspective. Science bulletin, 61(21), 1637- 24 1650. Lowry, I. S. (1964). A model of metropolis. Muth, R. F. (1969). Cities and Housing. University of Chicago Press. Mills, E. S. (1967). An Aggregative Model of Resource Allocation in a Metropolitan Area. The American Economic Review, 57(2), 197–210. http://www.jstor.org/stable/1821621 Matthews, R. B., Gilbert, N. G., Roach, A., Polhill, J. G., & Gotts, N. M. (2007). Agent-based land-use models: a review of applications. Landscape Ecology, 22, 1447-1459. Molinero-Parejo, R., Aguilera-Benavente, F., Gómez-Delgado, M., & Shurupov, N. (2023). Combining a land parcel cellular automata (LP-CA) model with participatory approaches in the simulation of disruptive future scenarios of urban land use change. Computers, Environment and Urban Systems, 99, 101895. Nost, E. (2022). Infrastructuring “data-driven” environmental governance in Louisiana’s coastal restoration plan. Environment and Planning E: Nature and Space, 5(1), 104-124. Ocampo-Melgar, A., Vicuña, S., Gironás, J., Varady, R. G., & Scott, C. A. (2016). Scientists, Policymakers, and Stakeholders Plan for Climate Change: A Promising Approach in Chile’s Maipo Basin. Environment: Science and Policy for Sustainable Development, 58(5), 24–37. https://doi.org/10.1080/00139157.2016.1209004 Seto, K. C., Fragkias, M., Güneralp, B., & Reilly, M. K. (2011). A meta-analysis of global urban land expansion. PloS one, 6(8), e23777. Soria-Lara, J. A., Ariza-Álvarez, A., Aguilera-Benavente, F., Cascajo, R., Arce-Ruiz, R. M., López, C., & Gómez-Delgado, M. (2021). Participatory visioning for building disruptive future scenarios for transport and land use planning. Journal of Transport Geography, 90, 102907. Voinov, A., & Bousquet, F. (2010). Modelling with stakeholders. Environmental modelling & software, 25(11), 1268-1281. Voinov, A., Seppelt, R., Reis, S., Nabel, J. E., & Shokravi, S. (2014). Values in socio-environmental modelling: Persuasion for action or excuse for inaction. Environmental Modelling & Software, 53, 207-212. Voinov, A., Jenni, K., Gray, S., Kolagani, N., Glynn, P. D., Bommel, P., ... & Smajgl, A. (2018). Tools and methods in participatory modeling: Selecting the right tool for the job. Environmental Modelling & Software, 109, 232-255. White, R., & Engelen, G. (1993). Cellular automata and fractal urban form: a cellular modelling approach to the evolution of urban land-use patterns. Environment and planning A, 25(8), 1175-1199. Wu, F. (2002). Calibration of stochastic cellular automata: the application to rural-urban land conversions. International journal of geographical information science, 16(8), 795-818.