Resumen: La cuenca binacional del río Zarumilla es un lugar con gran diversidad de especies y es de gran importancia para la economía de Perú y Ecuador. Estos factores socioeconómicos pueden traer consigo cambios en la distribución de las coberturas del suelo, los cuales afectan de manera negativa los ecosistemas. El presente estudio tiene como objetivo evaluar los cambios de cobertura y uso de suelo durante el periodo 1992-2022 y construir una proyección al año 2040 siguiendo las tendencias de uso actuales y su patrón histórico. Se usaron mapas de cobertura y uso de suelo de los años 1992 y 2022, así como información georreferenciada que fue usada para encontrar los factores condicionantes para el cambio de suelo. El modelo de prospección se basó en la implementación de cadenas de Markov y autómatas celulares usando factores condicionantes para construir una proyección más precisa. Los resultados obtenidos reflejan una pérdida neta de 11378 hectáreas de bosque durante el periodo 1992-2022, ocurriendo el 87.7% de esta pérdida en el lado ecuatoriano de la cuenca. A este ritmo se puede prospectar una pérdida de bosque de 10552 hectáreas para el periodo 2022-2040, de igual manera prevista principalmente en el lado ecuatoriano. El modelo prospectivo también resalta que los principales cambios se producirán en áreas de baja altitud y en proximidad a fuentes hídricas y carreteras. Esta es una pérdida alarmante de bosque nativo, la cual se puede evitar mediante la creación de políticas de desarrollo sostenible y programas de ordenamiento territorial.

Abstract: The binational basin of the Zarumilla River is a place with great species diversity and is particularly important to the economy of Peru and Ecuador. These socioeconomic factors can lead to changes in land cover distribution, which negatively affect ecosystems. The present study aims to evaluate changes in land cover and use during the period 1992-2022 and to generate a projection for the year 2040 following current usage trends and historical patterns. Land cover and use maps from 1992 and 2022 were used, along with georeferenced information that was employed to identify the conditioning factors for land change. The prospection model was based on the implementation of Markov chains and cellular automata using conditioning factors to build a more accurate projection. The results obtained reflect a net loss of 11378 hectares of forest during the period 1992-2022, with 87.7% of this loss occurring on the Ecuadorian side of the watershed. At this rate, a forest loss of 10552 hectares can be projected for the period 2022-2040, similarly projected mainly on the Ecuadorian side. The projection model also highlights that the main changes will occur in low-altitude areas and in proximity to water sources and roads. This is an alarming loss of native forest, which can be avoided through the creation of sustainable development policies and land-use planning programs.

ABSTRACT Wildfires pose a significant threat to natural ecosystems, human lives, and properties worldwide. Developing effective prevention and control strategies is crucial for minimizing the risk of catastrophic wildfire events. In this context, computational modelling has emerged as a valuable tool for simulating and analyzing wildfire behavior, which can aid in identifying potential areas for intervention and prevention measures. This article presents a cellular automata model for simulating forest fires in Sete Cidades National Park, Brazil. The model takes into account the wind direction and speed, vegetation type, and probability of burning in a two-dimensional lattice representing the forest area of interest. The simulation output is analyzed to determine the extent of the damage caused by the fire and to identify critical parameters for wildfire propagation. The results highlight the importance of considering wind direction and speed when developing prevention and control strategies for wildfires. By providing insights into the behavior of wildfires, computational modeling can support decision-making processes and facilitate the development of effective policies to manage the risk of wildfires. ecosystems lives worldwide events context measures Park Brazil type twodimensional two dimensional interest propagation wildfires decisionmaking decision making

Resumen Si bien la identificación de causas y agentes de deforestación en la Amazonia colombiana ha sido abordada en diversos estudios bajo un enfoque histórico relacional, existen pocas investigaciones orientadas a modelar sus relaciones espaciales y la proyección de escenarios futuros a través de la estadística espacial. Como contribución a dicho campo, se realizó un análisis en el cual se aplicó la regresión logística a deforestación ocurrida en el departamento del Guaviare durante el periodo 2005-2020, se buscó identificar y modelar el comportamiento espacial de sus factores explicativos y, con base en ello, proponer tres escenarios de deforestación probable para el área de estudio a 2030. En este último caso se combinó dicho análisis con tres tasas de deforestación distintas para determinar la cantidad de deforestación esperada a 2030 y dos métodos (softy hard) para asignar su localización en el espacio. Mientras que el estado de fragmentación del bosque, la ganadería extensiva, los cultivos de coca y la accesibilidad del área son considerados fuertes determinantes de deforestación, las áreas protegidas y los resguardos ejercen un efecto protector entre leve y nulo. A 2030 se estima la pérdida de entre el 4,4 % y el 8,8 % del área de bosque existente en 2020, afectando principalmente el resguardo Nukak y el Parque Nacional Natural Chiribiquete. Ideas destacadas: En este artículo de investigación se aplica el análisis de regresión logística a la identificación y el modelamiento espacial de factores determinantes de deforestación en el departamento del Guaviare y a la proyección de tres posibles escenarios de deforestación del área a 2030.

Abstract Although the identification of causes and agents of deforestation in the Colombian Amazon has been addressed in various studies under a historical-relational approach, there is little geostatistical research aimed at modeling its spatial relationship with the phenomenon and the projection of future scenarios. As a contribution to this field, a geostatistical analysis of the deforestation that occurred in the Colombian Guaviare during the period 2005-2020 is carried out, seeking to identify and model the spatial behavior of its explanatory factors, and based on this, propose three probable deforestation scenarios for the area to 2030 using logistic regression. In the latter case, this analysis was combined with three different deforestation rates to determine the amount of deforestation expected by 2030 and two methods (soft and hard) to assign its location in space. While the degree of forest fragmentation, extensive cattle ranching, coca crops, as well as the accessibility of the area, are largely determinants of deforestation, protected areas and resguardos exert a slight protective effect and in some cases none. By 2030, is expected the loss of between 4.4 % and 8.8 % of the existing forest area on 2020 with a particular impact on the Resguardo Nukak and the Parque Nacional Natural Chiribiquete. Highlights: Investigation article that applies a logistic regression analysis to the identification and spatial modeling of determinants of deforestation in the department of Guaviare as well as to the projection of three possible deforestation scenarios in the area by 2030.

Resumo Embora a identificação das causas e agentes do desmatamento na Amazônia colombiana tenha sido abordada em vários estudos sob uma abordagem histórica relacional, há poucas pesquisas geoestatísticas voltadas para modelar sua relação espacial com o fenômeno e a projeção de cenários futuros. Como contribuição a este campo, realiza-se uma análise geoestatística, usando regressão logística, do desmatamento ocorrido no departamento de Guaviare no período 2005-2020, buscando identificar e modelar o comportamento espacial de seus fatores explicativos e, com base nisso, propor três cenários prováveis de desmatamento para a área até 2030. Neste último caso, essa análise foi combinada com três diferentes taxas de desmatamento para determinar a quantidade de desmatamento esperada até 2030 e dois métodos (soft e hard) para atribuir sua localização no espaço. Enquanto o estado de fragmentação florestal, a pecuária extensiva, o cultivo de coca e a acessibilidade da área são em grande parte determinantes do desmatamento, as áreas protegidas e reservas exercem um efeito protetor entre pouco e nulo. Até 2030, projeta-se uma perda entre 4,4 % e 8,8 % da área florestal existente em 2020, afetando particularmente a reserva Nukak e o Parque Nacional Natural Chiribiquete. Ideias destacadas: Este artigo de pesquisa aplica a análise de regressão logística para a identificação e modelagem espacial dos determinantes do desmatamento no departamento de Guaviare, bem como para a projeção de três possíveis cenários de desmatamento na área até 2030.

Abstract: The COVID-19 pandemic highlighted the need for new approaches to the urban environment. Among them, physical distancing is intended to minimize contagion and safeguard general welfare. In the present research, a simulated model was designed to identify close contacts during pedestrian traffic in specific urban activities such as a trade fair. This research type was applied and of quasi-experimental design, a parametric simulation model based on intelligent agents was developed at the micro level in urban trade fair scenarios. Initially, theoretical trade fair configurations were compared, considering the shape of the walkable space and the arrangement of trade stands. Then, a manipulation of particular parameters is performed, among them, the probability of making a stall visible and the probability of stopping in case of direct contact or collision. Finally, the results were compared using statistical correlations, the PerMANOVA test, Games-Howell and the graphical analysis of the micro-simulated pedestrian behaviors by the developed model. It is concluded that the model is valid for theoretical pedestrian traffic models for the five simulated urban fairground scenarios.

Resumen: La pandemia del COVID-19 reveló la necesidad de contar con nuevas aproximaciones para el estudio del entorno urbano. Entre ellas, el distanciamiento físico, el cual tiene como propósito minimizar los contagios y salvaguardar el bienestar general. En la presente investigación, se diseñó un modelo simulado para identificar de contactos cercanos durante el tránsito peatonal en actividades urbanas específicas como la feria comercial. El tipo de investigación fue aplicada y de diseño cuasiexperimental, se desarrolló un modelo de simulación paramétrica basado en agentes inteligentes a nivel micro en escenarios urbanos feriales. Inicialmente, se compararon configuraciones de ferias comerciales teóricas, considerando la forma del espacio caminable y la disposición de puestos comerciales. Luego, se realizó una manipulación de parámetros particulares, entre ellos, la probabilidad de visibilizar un puesto y la probabilidad de detenerse ante un contacto directo o colisión. Finalmente, se contrastaron los resultados usando correlaciones estadísticas, la prueba de PerMANOVA, Games-Howell y el análisis gráfico de las conductas peatonales microsimuladas por el modelo desarrollado. Se concluye que el modelo es válido desde una perspectiva estocástica de estudio del tránsito peatonal para los cinco escenarios urbanos feriales simulados.

Abstract Uncontrolled urban growth in areas with deficient infrastructure and inadequate and/or overloaded services is currently one of the central issues in territorial studies. According to the United Nations estimates, by 2030, 60% of the world’s population will be living in cities and 95% of this expansion will take place in developing countries. This article presents a model based on the use of cellular automata and an artificial neural network, which allows the simulation of urban expansion based on criteria that would define future spatial configurations. The methodological development is applied to the city of Luján (Buenos Aires, Argentina) and is carried out in Geographic Information Systems from the automation of the procedures for the analysis of land use changes available in the MOLUSCE tool (Methods for Land Use Change Evaluation) in QGIS. The results show that the expansion trend to 2030 is 20%, which is equivalent to the incorporation of 6.72 km2 of urban coverage, if no type of intervention is carried out. The relevance of this type of work lies in the fact that the results provide technical-scientific support for planning and management agencies in relation to spatial decision making.

Resumen El crecimiento urbano incontrolado en áreas con deficiencias en la infraestructura y servicios inadecuados y/o sobrecargados, es actualmente uno de los temas centrales en los estudios territoriales. Según estimaciones de la Organización de Naciones Unidas en el 2030, un 60 % de la población mundial vivirá en ciudades y el 95 % de esa expansión se producirá en los países en desarrollo. El artículo presenta un modelo, basado en el uso de autómatas celulares y una red neuronal artificial, que permite simular la expansión urbana a partir de criterios que definirían las configuraciones espaciales futuras. El desarrollo metodológico se aplica a la ciudad de Luján (Buenos Aires, Argentina) y se realiza en Sistemas de Información Geográfica a partir de la automatización de los procedimientos para el análisis de cambios de uso de suelo disponible en la herramienta MOLUSCE (Methods Of Land Use Change Evaluation) en QGIS. Los resultados evidencian que la tendencia de expansión al 2030 es de un 20 %, lo que equivale a la incorporación de 6,72 km2 de cobertura urbana, si no se realizara ningún tipo de intervención. La relevancia de este tipo de trabajo radica en que los resultados brindan apoyo técnico-científico para los organismos de planificación y gestión en relación a la toma de decisiones espaciales.

Abstract For almost 30 years the Lu & Hamilton avalanche model (LH91) has been the paramount tool to study the intermittent nature of solar flares via avalanche models. In this work we used the two-dimensional model of Lu & Hamilton to assess the statistics of the waiting time between avalanches and the possibility forecasting synthetic solar flares. We worked with three different definitions of waiting time between: 𝛥𝑇, 𝛥 𝑇 𝑖𝑖 and 𝛥 𝑇 𝑃 . For the case of the usual definition of the waiting time between avalanches (𝛥𝑇) we found that the statistics can be described statistically through an exponential function. For the other definitions they present power-law statistics with exponents that compare well with solar observations.

Resumen El modelo de avalanchas de Lu & Hamilton (LH91) ha sido, por casi 30 años, una herramienta de gran importancia para modelar la naturaleza intermitente de las erupciones solares. En este trabajo se utiliza el modelo bidimensional de Lu & Hamilton para abordar el problema del comportamiento estadístico del tiempo entre avalanchas y la posibilidad de pronosticar erupciones solares sintéticas. Se trabajó con tres definiciones diferentes de tiempo entre avalanchas: 𝛥𝑇 (el número de iteraciones entre el inicio y el final de una avalancha), 𝛥 𝑇 𝑖𝑖 (número de iteraciones entre el inicio de una avalancha y el inicio de la avalancha siguiente) y 𝛥 𝑇 𝑃 (el número de iteraciones entre dos picos de avalanchas consecutivas. Para el caso de la definición habitual de tiempo entre, avalanchas (𝛥𝑇) se encontró que puede describirse estadísticamente mediante una función exponencial mientras que, para las otras dos definiciones el comportamiento en tipo power law con valores similares a los hallados en observaciones solares.

Abstract We combine urban geography, dynamic system mathematics, geographic information systems (GIS) and computer science to design a general modeling methodology based on Cascade Cellular Automata, which allows the construction of specific models of the expansion of Mexican cities. We submit our methodology to a triple empirical test in contrasting cities. The models consider restrictions on urban expansion (e.g., roads, parks, coasts). We use global and local indicators of goodness of fit between images articulated in a cascade filter, which reduces computational processing. The models showed encouraging adjustments. We illustrate how our method can drive the co-design of urban policies.

Resumen Combinamos geografía urbana, matemáticas de sistemas dinámicos, sistemas de información geográfica (SIG) y ciencias computacionales para diseñar una metodología de modelado de carácter general fundamentada en Autómatas Celulares en Cascada, que permite construir modelos específicos de la expansión de ciudades mexicanas. Sometemos nuestra metodología a una triple prueba empírica en ciudades con características contrastantes. Los modelos consideran restricciones a la expansión urbana (vialidades, parques, etc.). Utilizamos indicadores de bondad de ajuste global y local entre imágenes, articulados en un filtro en cascada, lo que reduce el procesamiento computacional. Los modelos mostraron ajustes alentadores.Ilustramos cómo nuestro método puede impulsar el codiseño de políticas urbanas.

RESUMO Este trabalho propõe estimativas de demanda de água para cenários futuros de uso e ocupação do solo e de verticalização utilizando análises espaciais e modelagem dinâmica com base em autômatos celulares para uma fração urbana do município de Campina Grande (PB). Dados reais de uso e ocupação do solo e de verticalização dos anos de 2011 e 2018 são utilizados para identificar o processo de mudança nesses bairros. Um conjunto de variáveis estáticas e dinâmicas relacionadas é selecionado para subsidiar a parametrização do modelo e simular as mudanças no uso e ocupação do solo e na verticalização, ocorridas no período. Após a validação das simulações para o ano de 2018 com base nos dados observados, novos cenários futuros são propostos para os anos de 2040, 2070 e 2100, identificando, assim, uma tendência de ocupação com algumas características específicas, como a substituição de áreas residenciais de um ou dois pavimentos por novos empreendimentos verticais, seja residencial, empresarial ou de serviços. Dado o aumento de ocupação urbana, as demandas de água capazes de atender à população também aumentam, aumento este que se confirmou neste trabalho. Entre os anos 2018 e 2100, essa tendência de aumento foi estimada em 300%. Uma comparação de cenários futuros considerando usos convencionais e racionais de água (utilizando mecanismos poupadores) também é apresentada.

ABSTRACT This work estimates urban water demands for future land-use scenarios using spatial analysis and dynamic modeling based on cellular automata for an urban area of the city of Campina Grande, Paraíba State, Brazil. Ground truth for two dates (2011 and 2018) is used to identify the process of land-use change in these neighborhoods. Some static and dynamic variables are defined in order to support the model parameterization and to simulate the changes occurred in the period. After the validation of those simulations for 2018, based on the observed data, new future scenarios are proposed for the years 2040, 2070, and 2100, thus identifying a tendency of occupation with some specific characteristics such as the replacement of residential areas new buildings with multiple floors, whether residential, business, or services. Due to the increase in urban occupation, the water demands for attending to the population also increase. Between 2018 and 2100, this upward trend is estimated at 300%. A comparison of future scenarios considering conventional and rational water uses (using saving mechanisms) is also simulated.

ABSTRACT The urban dynamics analysed by geographic information systems contributes to the identification of local specificities of space occupation and transformation. This study analyses the spatial growth based on cellular automata for the proposition of future scenarios of the expansion of north-northwest area of Campinas- sp , Brazil. The modelling was performed by sleuth software, which is based on the Clarke Urban Growth Model, and requires information on land use, slope, road network, hill shade, excluded areas and urban areas to simulate urban growth up to 2040. The studied area showed a growth trend with constant dynamics of expansion and densification from existing nucleus of buildings and along the road network. The limiting factors of the expansion were the large unbuilt institutional areas, preservation areas and rural properties. Therefore, the urban growth model simulated by sleuth was adequate for the analysis of physical-environmental and infrastructure factors that determine the urban dynamics.

RESUMO A dinâmica urbana analisada por Sistema de Informação Geográfica contribui para identificação de especificidades locais da ocupação e transformação do espaço. O presente estudo analisou o crescimento espacial baseado em autômato celular para a proposição de cenários futuros da região norte-noroeste de Campinas– sp , Brasil. A modelagem foi realizada pelo softwaresleuth , que tem por princípio o Modelo de Crescimento Urbano de Clarke, e requer informações sobre uso do solo, declividade, malha viária, relevo sombreado, áreas de restrição e manchas urbanas para simular o crescimento até 2040. A área analisada apresentou tendência de crescimento com dinâmica constante de expansão e adensamento a partir dos núcleos urbanos existentes e no entorno da malha viária. Os fatores limitantes da expansão foram as grandes áreas institucionais não edificadas, as áreas de preservação e propriedades rurais. Portanto, o modelo de crescimento simulado pelo sleuth se apresentou adequado para a análise dos fatores físico-ambientais e de infraestrutura determinantes da dinâmica urbana.

RESUMEN La flota de los países en vías de desarrollo está compuesta por motocicletas y automóviles. Esta condición heterogénea en el tráfico presenta ventajas y desventajas, lo que resulta en puntos de vista conflictivos (por ejemplo, los motociclistas que disfrutan de su mayor movilidad mientras que los conductores de automóviles se resienten con su velocidad disminuida). En este artículo corroboramos la idea de que la evaluación del tráfico depende de la métrica escogida (por ejemplo, flujo de vehículos, consumo de combustible, costos mensuales) y del punto de vista (conductor, motociclista y responsables de formular políticas). Para ello, estudiamos una condición de tráfico mixto, considerando que el rendimiento del vehículo se ve afectado por tres escalas: motor, vehículo y tráfico. Modelamos el motor usando correlaciones empíricas de potencia y eficiencia energética, la del vehículo a través de un equilibrio de fuerzas propulsoras y resistivas, y la del tráfico mediante un modelo de autómata celular. Simulamos 189 condiciones de tráfico y evaluamos el flujo de vehículos, el consumo de energía promedio, la emisión total de CO2 de la vía y los costos mensuales. También discutimos los resultados desde el punto de vista del conductor, el motociclista y la sociedad. Llegamos a la conclusión de que la condición óptima depende tanto de la elección de la métrica como del punto de vista, además de que no es apropiado usar los resultados del tráfico homogéneo para analizar condiciones de tráfico heterogéneo, incluso si ambos escenarios presentan el mismo flujo total de vehículos.

ABSTRACT The fleet of developing countries consists of motorcycles and cars. This heterogeneous traffic condition has its advantages and disadvantages, which results in conflicting points of view (e.g., motorcyclists enjoying a higher mobility while car drivers resent their decreased speed). In this paper, we corroborated the notion that traffic evaluation depends on the chosen metric (e.g., vehicle flow, fuel consumption, monthly costs) and the point of view (driver, rider, and policy makers). To this effect, we studied a mixed traffic condition, considering that the vehicle performance is affected by three scales: engine, vehicle, and traffic. We modeled the engine using empirical correlations of power and energy efficiency, the vehicle based on a balance of propulsive and resistive forces, and traffic with a cellular automata model. We simulated 189 traffic conditions and evaluated vehicle flow, average energy consumption, total CO2 emission of the road, and monthly costs. We also discussed the results from the point of view of the driver, rider, and society. We concluded that the optimal condition depends both on the choice of metric and point of view, and that is not appropriate to use results from homogeneous traffic to analyze heterogeneous traffic conditions, even if both scenarios present the same total vehicle flow.

ABSTRACT: Urbanization advances vertiginously and its impacts are visible beyond metropolitan spaces. In this context, urban sprawl simulation methodologies become relevant to understand and add to future urban growth scenarios. In this work, the case of the city of Temuco in the La Araucanía Region is analyzed, using the following tools: Markov chains; Cellular Automaton; Multicriteria-Multi-objective Assessment; and the determination of land usage/cover using Geographic Information Systems (GIS). Based on the above, the urban scenario was determined for 2049 from the spatial patterns of the area under study, since 1985. The model shows a trend to a peripheral and diffuse sprawl towards the north of the city and a strong development towards the western sector in the Labranza neighborhood. That is to say, the first sector is projected as a possible node of expansion, and the second, as one of consolidation. It is concluded that both zones require pertinent updated urban planning instruments, which the city has not had until now.

RESUMEN: La urbanización avanza de manera vertiginosa y sus impactos son visibles más allá de los espacios metropolitanos. En ese contexto, las metodologías de simulación del crecimiento urbano adquieren relevancia para comprender (y aportar a) futuros escenarios de crecimiento urbano. En concreto, aquí se analiza el caso de la ciudad de Temuco en la Región de La Araucanía empleando las siguientes herramientas: Cadenas de Markov, Autómata Celular, Evaluación Multicriterio-Multiobjetivo, como también de la determinación de usos/coberturas de suelo mediante Sistemas de Información Geográfica (SIG). Así, el trabajo determina el escenario urbano al año 2049 a partir de los patrones espaciales del área de estudio desde el año 1985. El modelo muestra la tendencia al crecimiento periférico y difuso hacia el norte de la ciudad y un fuerte desarrollo hacia el sector oeste en el barrio de Labranza, es decir, el primer sector se proyecta como un foco de posible expansión y el segundo, como uno de consolidación. Se concluye que ambas zonas requieren de instrumentos pertinentes y actualizados de planificación urbana, con los que la ciudad no ha contado hasta ahora.

Abstract: The present work seeks to model the simulation of the landscape of the city of Curitiba, through the use of cellular automata (CA) algorithms, together with the Geographic Information Systems (GIS) and Remote Sensing tools. Four models were elaborated, having as input data classified images from the years 2006, 2009, 2011 and 2014; and different time intervals between the initial and final landscape of the models. Geographic data of the city were also used, as well as the current legislation of the municipality, and such data contributed to the robustness of the modelling. Two validation tests were applied to verify the adequacy of the simulated models concerning the observed reality. The validation performed by the Multiple Resolutions Adjustment method indicated that the model elaborated with data from 2009 and 2011 reached the highest similarity index, being equal to 0.88. Thus, it was possible to carry out geosimulations and these indicated that, as long as the trends observed in the past are maintained, the urban expansion of the municipality will occur at the expense of vegetation.

Resumo Previsões do comportamento das mudanças que ocorrem no uso do solo são indispensáveis para um planejamento urbano adequado, indicando áreas mais propensas à ocupação humana. Dessa forma, o estudo tem por objetivo prever a expansão urbana da cidade de Fortaleza, localizada no nordeste do Brasil, utilizando um modelo dinâmico que combina autômatos celulares com as Cadeias de Markov (CA-Markov). Ele foi desenvolvido por meio de três etapas: aquisição e processamento dos dados; espacialização dos dados; e, modelagem dinâmica espacial. A matriz de transição markoviana forneceu o percentual de mudanças de uma classe para a outra ao algoritmo de simulação. Dados reais de entrada de uso do solo dos anos de 2009 e 2017 mostraram uma transição de 26,50% de áreas vegetadas para áreas urbanizadas. A validação do modelo apresentou excelentes resultados (índice de similaridade fuzzy superior a 0.8) para realizar simulações futuras (2020 até 2025). Segundo as previsões, em 2025, dos 297,10 km² de área urbanizada e vegetada, 82,43% serão de área urbanizada e apenas 17,57% serão de área vegetada. O modelo de simulação pode ser alimentado por mudanças que aconteçam dentro do período previsto, sendo incluídas na simulação como variáveis dinâmicas, obtendo uma simulação compatível com essas mudanças.

Abstract Forecasts of the behaviour of changes that occur in land use are indispensable for adequate urban planning, indicating areas more prone to human occupation. Thus, the study aims to predict the urban expansion of the city of Fortaleza, located in northeastern Brazil, using a dynamic model that combines cellular automata with the Markov Chains (CA-Markov). It was developed through three steps: data acquisition and processing; spatialization of data; and, dynamic spatial modelling. The Markovian transition matrix provided the percentage of changes from one class to another to the simulation algorithm. Actual land use input data for the years 2009 and 2017 showed a transition of 26.50% from vegetated to urbanized areas. The validation of the model showed excellent results (fuzzy similarity index greater than 0.8) to perform future simulations (2020 to 2025). According to forecasts, in 2025, of the 297.10 km² of urbanized and vegetated area, 82.43% will be urbanized, and only 17.57% will be vegetated. The simulation model can be fed by changes that happen within the predicted period, included in the simulation as dynamic variables, obtaining a simulation compatible with these changes.

ABSTRACT Background: The spread of forest fire is a complex natural phenomenon. The cellular automata(CA) is a common model of forest fire spread, which fails to combine the unique combustion properties of forest fire spreading, resulting in inaccurate simulation results. In order to improve the accuracy of forest fire spread simulation, the Rothermel forest fire rate formula is simplified and combined with CA to form a multi-dimensional cellular automata(MD-CA) model of forest fire spread with different combustion properties in each cell. The formulas of the spread rate of forest fire in eight directions are obtained through the training datasets, and the testing datasets are used to compare the simulation results of CA model, MD-CA model and the actual fire spread. Results: It is concluded that the simulated areas and perimeters of the MD-CA are closer to the actual forest fire spread, the area error rate is 9.42% - 15.63%, and the perimeter error is 4.21% - 8.99%, The errors are less than CA model. Conclusion: The MD-CA model based on Rothermel has the strong ability to simulate the spread of fire in the laboratory, however, how to eliminate the errors over time is the task of the next stage.

Resumen La Tarea de Clasificación de Densidad (TCD) es un problema bien conocido, donde el objetivo principal es construir un autómata celular cuya regla local dé lugar a una coordinación global emergente. Describimos los métodos utilizados para identificar nuevos autómatas celulares que resuelven este problema. Nuestro enfoque identifica tanto la vecindad como su regla estocástica utilizando un conjunto de datos de configuraciones iniciales que cubre de manera predefinida el rango completo de densidades en TCD. Comparamos nuestros resultados con algunos modelos disponibles actualmente en el campo. En algunos casos, nuestros modelos muestran un mejor rendimiento que la mejor solución informada en la literatura, con una eficacia de 0.842 para conjuntos de datos con distribución uniforme alrededor de la densidad crítica. Las pruebas se llevaron a cabo en conjuntos de datos de diversos tamaños de malla y condiciones de muestreo. Finalmente, mediante una prueba estadística no paramétrica demostramos que no hay diferencias significativas entre nuestros autómatas celulares identificados y el modelo más conocido.

Abstract Density Classification Task (DCT) is a well-known problem, where the main goal is to build a cellular automaton whose local rule gives rise to emergent global coordination. We describe the methods used to identify new cellular automata that solve this problem. Our approach identifies both the neighborhood and its stochastic rule using a dataset of initial configurations that covers in a predefined way the full range of densities in DCT. We compare our results with some models currently available in the field. In some cases, our models show better performance than the best solution reported in the literature, with efficacy of 0.842 for datasets with uniform distribution around the critical density. Tests were carried out in datasets of diverse lattice sizes and sampling conditions. Finally, by a statistical non-parametric test, we demonstrate that there are no significant differences between our identified cellular automata and the best-known model.