Aplicación de un modelo basado en autómatas celulares irregulares para la simulación de escenarios futuros de cambios de uso de suelo urbano
- Autores: Ramón Molinero Parejo
- Directores de la Tesis: Francisco Aguilera Benavente (dir. tes.), Montserrat Gómez Delgado (codir. tes.)
- Lectura: En la Universidad de Alcalá ( España ) en 2023
- Idioma: español
- Tribunal Calificador de la Tesis: Josep Roca Cladera (presid.), Francisco Javier Escobar Martínez (secret.), Marta Gallardo Beltran (voc.)
- Programa de doctorado: Programa de Doctorado en Tecnologías de la Información Geográfica por la Universidad de Alcalá
- Materias:
- Enlaces
- Resumen
- español
La urbanización es uno de los fenómenos más drásticos de transformación del territorio. En las últimas décadas, este fenómeno ha experimentado un aumento vertiginoso. Según indica el informe más reciente de World Urbanization Prospects de Naciones Unidas, se estima que el 68,4% de la población mundial vivirá en zonas urbanas en 2050. Además, se prevé que dicha población se duplique en los países desarrollados y se triplique en los países en vías de desarrollo. Todo ello ha supuesto impactos irreversibles sobre el territorio, afectando enormemente al conjunto de la sociedad en términos de gestión y acceso a recursos, problemas de índole social y económica, contaminación ambiental, etc.
Ante esta situación, se ha observado un creciente interés por el desarrollo y mejora de instrumentos que den soporte a la toma de decisiones y a la gestión de las áreas urbanas. Uno de los instrumentos más empleados para este fin ha sido la planificación de escenarios futuros. Este permite conocer cómo podría afectar la evolución de los usos del suelo urbano a la configuración de los patrones espaciales bajo distintas perspectivas futuras. Con este enfoque, la planificación de escenarios trata de reducir la incertidumbre facilitando la toma de medidas proactivas para minimizar los posibles impactos territoriales.
No obstante, la planificación de escenarios puede verse limitada ante un futuro complejo e incierto si todos los escenarios se mantienen muy próximos a una proyección tendencial. Como ejemplo, el surgimiento de acontecimientos inesperados puede llegar a inhabilitar la utilidad de una planificación lineal basada únicamente en tendencias pasadas. Por dicha razón, y para gestionar de la mejor manera posible los futuros desarrollos urbanos (no) deseados, el pensamiento disruptivo debe formar parte del proceso de previsión, rompiendo así con la linealidad de los acontecimientos actuales para abarcar lo inesperado.
Como parte del proceso de planificación urbana, los modelos de simulación intentan representar el desarrollo futuro de las ciudades para garantizar que puedan desarrollarse de manera eficiente y sostenible. De ellos, los modelos basados en Autómatas Celulares (AC) se encuentran entre los más utilizados como apoyo a la gestión de las áreas urbanas. Estos modelos han experimentado una importante flexibilización, adaptándose a entornos irregulares (parcelario catastral) para ofrecer simulaciones de cambio de uso del suelo urbano a escala local.
En esta línea, son cada vez más los estudios que combinan escenarios narrativos con tareas de modelización de manera participativa, y todo ello con la finalidad de obtener resultados más realistas que contemplen los actuales retos que afronta la planificación urbana. Sin embargo, es difícil que estos modelos consideren por sí solos la amplia gama de factores que intervienen en la evolución futura de las zonas urbanas, especialmente cuando tratan de representar escenarios imaginativos y disruptivos. Ante la situación actual en la que se encuentra la planificación espacial de escenarios urbanos, la presente investigación desarrolla e implementa una metodología que trata de cubrir algunos de los huecos más notables que se observan en este ámbito de estudio.
- English
Urbanisation is one of the most drastic phenomena of territorial transformation. Over the last decades, this phenomenon has been increasing at a dizzying rate. According to the latest World Urbanization Prospects report, it is estimated that 68.4% of the world's population will live in urban areas by 2050. Moreover, this population is expected to double in developed countries and triple in developing countries. This has generated an irreversible impact on the territory, affecting society in terms of management and access to resources, social and economic problems, environmental pollution, etc.
In view of this situation, there has been a growing interest in the development and improvement of instruments to support decision-making and management of urban areas. One of the most widely used instruments for this purpose has been future scenario planning. This provides insight into how the evolution of urban land uses might affect the configuration of spatial patterns under different future scenarios. With this approach, scenario planning seeks to reduce uncertainty by facilitating proactive measures to minimise potential spatial impacts.
However, scenario planning may be limited in the face of a complex and uncertain future if all scenarios remain very close to a trend projection. As an example, the occurrence of unexpected events can render linear planning based solely on past trends unhelpful. For this reason, and in order to best manage (un) desired future urban developments, disruptive thinking must be part of the visioning process, thus breaking with the linearity of current events to embrace the unexpected.
As part of the urban planning process, simulation models attempt to represent the future development of cities to ensure that they can develop efficiently and sustainably. Of these, Cellular Automata (CA) models are among the most widely used to support the management of urban areas. These models have undergone significant flexibilization, adapting to irregular environments (cadastral parcels) to provide simulations of urban land use change at a local scale.
In this respect, an increasing number of studies are combining narrative scenarios with participatory modelling tasks, all with the aim of obtaining more realistic results that address the current challenges facing urban planning. However, it is difficult for these models alone to consider the wide range of factors involved in the future evolution of urban areas, especially when they attempt to represent imaginative, disruptive scenarios. Given the current situation of urban scenario planning, the present research develops and implements a methodology that attempts to fill some of the most notable gaps in this area of study.
Firstly, a study based on the design and mapping of disruptive scenarios is presented through a participatory workshop in which experts from various fields related to urban planning and transport collaborated together. The results derived from this workshop were analysed using a statistical method called Geographically Weighted Logistic Regression (GWLR) in order to determine the main factors that explain the location of urban land uses in the different disruptive scenarios.
Subsequently, the results of the previous analysis were used to calibrate a new model developed based on vector CA, called Land Parcel - Cellular Automata (LP-CA). This model simulates different imaginative and disruptive future scenarios reproducing urban dynamics of growth, change, and loss of land use. At the same time, a partial validation method was applied to observe the robustness of the model with respect to the influence of the factors in the simulations.
Finally, an innovative methodology designed to assess the different disruptive scenarios was applied. It employs multi-scale spatial metrics based on the use of moving windows applied at the parcel level that allow characterising the diversity and type of urban expansion. The developed methodology was applied to a sector of the Henares Corridor (Spain), which was used as an experimental territorial laboratory. The overall results have demonstrated the usefulness of the integration of disruptive scenarios in spatial planning to show contrasts between different scenarios, highlighting the usefulness of the visions and the participatory workshop in the spatial representation of the amount and direction of growth of urban uses and the organisation of the transport network. In a complementary way, the statistical analysis by means of RLGP allowed a relevant adjustment of the suitability parameter in CA-based models, which favoured a calibration more adapted to each scenario. Concerning the progress of the LP-CA model, the simulations successfully reproduced disruptive urban dynamics (in addition to growth, transformation of uses and abandonment). The spatial patterns generated were in line with the narrative scenarios described. Additionally, the sensitivity analysis verified the balanced incidence of all factors in the simulations generated by the LP-CA model.
Finally, the scenario evaluation allowed for a more detailed characterisation and comparison of the territorial implications of each scenario in terms of diversity and type of urban expansion.
In conclusion, the information provided by this research provides new tools and improves some of the existing methods within spatial scenario planning. In particular, it offers a novel methodology capable of generating simulations of growth and urban land use change for disruptive future scenarios. The results facilitate the observation of the spatial propagation of uncertainty associated with future events through the patterns that shape new land uses. Ultimately, it seeks to extract complex information from different approaches to future urban evolution, presenting it in a simple way so that it can be used by decision-makers.
- español
