Exploring the spatial-temporal arrangements of urban activity space from individual's daily commute: A Geospatial-Agent based Approach Using Empirical Data

Exploring the spatial-temporal arrangements of urban activity space from individual's daily commute: A Geospatial-Agent based Approach Using Empirical Data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Exploring the spatial-temporal arrangements of urban activity space from individual's daily commute: A Geospatial-Agent based Approach Using Empirical Data Mehdi Azari, Sara Moridpour, Mohsen Hatami, Monireh Hosseini This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4835588/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The study explores the significance of individual mobility measures, such as activity space, in understanding how individuals interact with their daily environments. Existing measures often overlook geographical concepts like spatial-temporal arrangements of activity spaces, focusing solely on numerical assessments. To address this gap, a multi-level modeling approach combining Agent-Based Modeling (ABM) and Geographic Information Systems (GIS) is utilized to simulate activity destination selection throughout a workday in Zanjan, Iran. The model integrates individual preferences, built environment characteristics, network attributes, and travel generation data. Real-world data from Emerging Data Sources (EDSs) validate the model's reliability and accuracy. Key findings include: (1) clustering analysis identifying four types of activity destinations at different hourly intervals, (2) a central activity space acting as a hub for activity-based travel with a monocentric distribution pattern, (3) individual preference for destinations with diverse and dense built environments, and (4) a decrease in trip frequency as distance from the main activity space increases, indicating a spatial decay effect on activity-based travels. Activity clusters Destination Choice Travel Pattern Recognition Agent-Based Model GIS Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction The recognition of individual spatial behavior has been a key focus of academic research for many years. Scholars have widely adopted the concept of activity space as a way to assess individual spatial behavior [ 1 – 3 ]. Activity spaces refer to specific areas that individuals visit regularly, representing their spatial footprint [ 4 , 5 ]. Researchers from diverse disciplines have employed various methods and data sets to model and investigate activity spaces. A significant portion of studies has focused on methods to explore the spatial arrangement of activity spaces [ 3 , 6 , 7 ], the fundamental factors influencing the forming of these spaces [ 6 , 8 , 9 ], and their implications for various areas such as transportation planning [ 9 ], travel behaviour change [ 10 ], activity location planning [ 11 ], human health [ 12 ], social exclusion [ 13 ], location services planning [ 14 ], and racial segregation [ 15 , 16 ]. While the concept of activity spaces has been widely utilized by researchers across different fields, some scholars argue that it has evolved into a multi-dimensional concept that remains underexplored [ 3 , 5 ]. The evolution in daily activity patterns has transitioned from being centered around a single location to involving multiple locations, indicating a move towards spatial polygamy [ 2 , 3 ]. This shift suggests that activity spaces are dynamic throughout the day [ 17 , 18 ], with different locations serving various aspects of daily life within specific temporal constraints.[ 8 , 16 , 19 , 20 ]. For example, certain urban areas may attract individuals for specific tasks at particular times, leading to the formation of distinct urban activity spaces during those periods. Conversely, different locations shape alternative activity spaces at different times [ 20 ]. This complexity challenges the traditional view of activity spaces as singular trip destination during peak traffic times. It underscores the significance of factors such as the spatial-temporal evolution of activity spaces and the timing of their utilization [ 15 , 17 ]. This evolution in the concept of activity spaces has significant implications for urban planning, prompting planners to incorporate spatial-temporal dynamics effectively in shaping activity spaces [ 9 , 10 ]. Although previous studies have recognized the location heterogeneity of activity spaces and occasionally incorporated it into activity space modeling [ 3 , 6 , 7 ], there has been limited empirical examinations into the dynamic changes of activity spaces over time [ 15 , 21 , 22 ]. By conducting space-time assessments, we not only gain information on how a specific space is used but also uncover valuable insights about the underlying factors driving human spatial behavior [ 23 ]. Additionally, this process contributes to the building an archive regarding human spatial behavior [ 24 ]. Another aspect that have often overlooked in existing approaches is the inter-relationship between the environmental factors and human spatial behavior in the process of selecting an activity destination [ 23 , 25 ]. Drawing from relevant theories, travel destination choice involves an individual's decision at the origin, as well as the opportunities and limitations posed by the physical environment in the destination [ 11 , 23 , 26 – 28 ]. Accordingly, theories suggest that areas with denser and more diverse land uses closer to travelers, are more attractive for activity-based travel [ 29 ]. Also, some other studies have highlighted the significance of transportation network features in shaping destination choice behaviors [ 11 ]. Conventional travel demand models cannot capture the complex interactions among these factors and there is ongoing debate about how the destination's built-environment and route features impact travel behavior in activity-based and travel demand models [ 25 , 30 ]. They have mainly concentrated on individual’s personal preferences, focusing on basic built-environment characteristics such as household and activity area locations and distances between them [ 30 , 31 ]. However, it’s highly complicated to integrate these into travel demand models to accurately reflect human spatial behavior and include all these factors in a single model. This study aims to fill these gaps through a promising approach combined agent-based models (ABMs) with Geographic Information System (GIS). ABMs represent autonomous decision-making entities called agents that make decisions based on rules [ 32 ]. In a travel demand behavior, an agent symbolizes an individual who decides to travel to an activity area based on their preferences, the output of their own non-linear actions, interpersonal interactions, and the environment [ 20 , 33 ]. In the applied approach, GIS allows for integrating a wide range of geospatial data into ABMs to represent more realistic and detailed simulations and its interactions with the environment [ 32 ]. Moreover, GIS offers a clustering technique to analyse and map the shifts in individual’s activity spaces based on the results of ABM. Accordingly, a Spatial Cluster Detection (SCD) method based on 3D maps was conducted to space-time clustering analysis of activity spaces in an empirical case study in Zanjan, Iran. The SCD method yields a measurable ordinal outcome, i.e., main activity spaces, and second-order activity spaces, providing detailed insights into the spatial-temporal distribution of activity spaces based on travel data. This analysis includes identifying peak visitation times throughout the day, popular activity spaces, and duration of stay, which can offer valuable insights for activity location planning and travel behavior management. The proposed approach goes beyond simply measuring the extent of one's activity space and brings a geographical understanding to pattern-recognition modeling while providing perspective into how urban planning can shape individual spatial behavior. However, this process is notably complex due to the intricate nature of modeling and their parameters and requires careful validation to ensure that the model accurately reflects real-world behaviors [ 23 ]. It should be reminded that making model validation a crucial challenge in ABM construction due to the lack of real-world data [ 28 , 32 ]. Traditionally, researchers often used questionnaires or interviews to validate models, but these methods were laborious and time-consuming. Additionally, these data relied heavily on participants' perceptions [ 34 ]. In order to tackle the validation challenges, an innovative approach involves utilizing precise traffic flow data from emerging data sources (EDSs) was proposed to validate the ABM. Our major contributions can be summarised as follows: Introducing a new activity destination choice model that utilizes Geospatial-ABM to simulate the spatial-temporal arrangements of activity spaces from individual's daily movements. Integrating GIS methods to incorporate destination attractive into ABM, including innovative built-environment features and individual’s personal preferences as well as real travel generation data. Visualizing the spatial-temporal dynamics of activity spaces by SCD and categorizing them into several clusters and sub-clusters. This enables a detailed consideration of geographical patterns in the distribution of activity-based travels. Addressing model validation challenges by using real-world data on the number of visits to activity areas, obtained from sources like EDSs and manual tallying, to validate the effectiveness of the ABM model through the R-square and RMSE procedures. Conducting extensive experiments on a case study in Zanjan, Iran, to demonstrate the effectiveness and superiority of the proposed methodology. The remainder of the paper is structured as follows. The following section provides a concise overview of the case study and details the data collection and preparation process for this research. This is followed by an explanation of our methodological approach, focusing on the ABM and its key characteristics. The paper then delves into the application of the ABM method to the Zanjan case study, showcasing the results pertaining to activity spaces and their temporal variations. Lastly, the discussion and conclusion section encapsulates the main findings and their implications, along with a brief examination of the study's limitations and recommendations for future research endeavors. 2. Data collection and case study 2.1. Study area Zanjan, the provincial capital of Zanjan Province in north-west Iran, serves as the study area for this research. In 1986, Zanjan's population was 215,261, according to the Iranian Statistical Centre. The population increased to 433,475 by 2016, indicating significant growth. However, the city's infrastructural development, including elements like urban facilities and transportation equipment, has not kept pace with this rapid population growth. This mismatch has led to congested narrow streets, particularly in urban cores. Geographically, Zanjan is advantageously located near Tehran, Iran's political center, and the industrial cities of Gazvin and Tabriz. Consequently, it functions as a crucial communication hub within the Iranian urban hierarchy. The city's urban structure is centralized, supporting heavy transportation, especially within certain urban centers. Figure 1 displays the location of Zanjan and the traffic zones the city. 2.2. Data collection and preparation Establishing a realistic simulation requires the extraction of pertinent real-world rules and the identification of significant factors influencing the phenomena being simulated, necessitating a precise database [32]. This section provides an overview of the data collection process. 2.2.1. Travel behaviour survey The travel behavior survey played a pivotal role in our research, providing key insights into both preferred destinations and home location decisions for commuters within Zanjan City. The travel survey data from Zanjan informed basic travel behaviors, serving as samples for agents in simulating multi-location activity space dynamics. This data encompassed a range of factors: travel patterns, modes of transportation, trip destinations, purposes, frequency, duration, and travel time, which bolstered the practicality of our approach. The outcome data from the survey is reflected in the agents' behavior to interact in a wide area, such as a city scale. The survey type was a self-administered questionnaire, distributed among commuters during peak and off-peak hours at different locations in the city. The questionnaire was designed with both open-ended and closed-ended questions. The open-ended questions gathered qualitative information, while the closed-ended questions collected quantitative data. The questionnaire was pre-tested to ensure clarity and comprehensibility. Some important factors in the recruitment of travel behavior surveys are as follows. Representative sample: A representative sample closely mirrors the overall attributes of the population, especially in characteristics pertinent to the study [35]. In this study, two criteria were considered to achieve representativeness: For spatial-based behavior, the survey took the form of a geographically stratified random survey of households. Accordingly, 150 traffic zones (TZs) were clustered into 15 regions compatible with administrative data, based on characteristics such as local proximity, transit accessibility, access to urban services, land uses, and so forth. To represent socio-demographic heterogeneity in the sampling sectors, random sampling was employed. This approach allows all participants in the larger group to have an equal chance of being selected, ensuring the most representative samples and data free of bias [36]. Sample size : Determining the appropriate sample size is a key challenge in travel behavior surveys [36]. Empirical studies recommend basing the sample size on the entire population [37]. This study surveyed a random sample of about 5% of households from each region based on the Zanjan population number (the population is 433,475, and the household number is 131,799) [38]. A total of 6,604 questionnaires were distributed, with 5,942 valid responses received, yielding a response rate of approximately 90%. The questionnaires : Two local students from Zanjan University were chosen for each region and trained to survey travel behavior. It was beneficial to select at least one student from each area, as they were familiar with the environment. The main questions that were asked: The questionnaire consisted of two sections. The first section collected demographic information about the respondents, such as age, gender, and education level. The second section focused on travel characteristics, such as car ownership, travel mode, trip purpose, travel destination, trip frequency, trip duration, and travel time. The survey took approximately 20 minutes to complete. Sampling errors: Another issue associated with survey quality is the degree of data accuracy, which manifests itself in sampling errors. Sampling error is a bias that results from choosing to estimate a sample as representative of a larger population. Sampling errors can be estimated as follows [39]. The outcome of the sampling error estimation demonstrates a statistically significant level with a 95% confidence level. Table 1 presents the respondents' gender, age, and duration of stay in their current location. Among the respondents, 59% were male and 41% were female. Those aged between 35 and 39 years constituted 22% of the total, while individuals over 50 years of age accounted for 12%. The majority of the participants were highly educated individuals who had lived in the area for at least two years, ensuring that they had sufficient experience and knowledge of the area to provide accurate and reliable information. Accordingly, the highest percentages of respondents were those who had lived in the neighborhood for 2–5 years and those with college educations (59% and 42%, respectively). Subsequently, statistical analysis, specifically descriptive statistics, was employed to investigate complex relationships among various personal, geographical, and travel factors collected by the survey. This approach was instrumental in enhancing the understanding of the sample data and in identifying some simple relationships between variables. After this preliminary analysis, the K-means clustering method was implemented to segment individuals based on their destination choice behaviors as indicated by our gathered socio-demographic and travel behavior indicators. This step was crucial for integrating these behaviors into our simulation framework. In our study, K-means clustering was employed to categorize respondents into distinct groups according to their survey responses. Table 1: Socio-demographic characteristics of the respondents. Variables Classification Distribution Gender Male Female 59% 41% Age -30 30-34 35-39 40-44 45-49 +50 16% 14% 22% 19% 17% 12% Duration of stay in the neighbourhood. 2-5 4-7 +7 59% 28% 23% Education level No schooling Primary education Secondary education College education 0% 19% 39% 42% The clustering applied in this context allowed us to recognize underlying factors influencing travel behavior. From this, simple decision rules were derived and incorporated into the simulation platform. This step enabled the training of agents to accurately represent individual travel behavior, particularly in terms of destination choices within the context of activity space [40]. 2.2.2. Built-environment characteristics The influence of the built environment on travel behavior is a key factor that impacts various aspects of individuals' travel behaviors and decision-making processes [27]. This study focuses on examining the three primary land use factors—Density, Diversity, and Distance—that play a significant role in shaping activity-travel behaviors, collectively referred to as the '3Ds measures: Density refers to the concentration of land uses within a specific area. A higher density is typically associated with the close proximity of similar land uses, extensive walkability, and convenient access to similar facilities [27, 29]. Diversity focuses on the variety of land uses present in a given area, with greater diversity correlating to a higher number of destinations and improved access to diverse urban amenities [41]. Distance relates to the travel distance between an origin and a destination, with individuals generally preferring shorter distances to access urban activity areas [41]. In this study, a vector land use map was utilized to represent 3D features in GIS, providing enhanced precision in recording, displaying, and analysing spatial data compared to traditional statistical methods [42]. The land use map encompasses various amenities such as shopping centers, educational institutions, recreational venues, government-banking institutions, and medical facilities. It was initially surveyed and mapped by the Zanjan municipality during the summer of 2019 and revised in the summer of 2020 using OpenStreetMap (OSM) and on-site observations. In this research, the Focal Statistics tool available in ArcGIS Spatial Analyst was employed to compute land use diversity. This tool conducts a neighborhood analysis to generate a new raster, where the value of each cell is determined by the values within a specified surrounding area. The variety function in Focal Statistics calculates the count of unique values in each neighborhood, offering a useful metric for assessing the diversity of land use types [43]. Furthermore, a kernel density estimation technique was utilized to determine the densities of each land use categories. Kernel density approach is a non-parametric method to estimate density from point-based or line-based data, enabling a quantitative evaluation of the distribution of various land use types across the study area [42]. Additionally, a distance function was integrated for the simulation's individual agents, mandating their visits to the nearest TZs. Following the preparation of layers, the Zonal Statistics function (ZSF) was employed to compute the 3D values for each TZ. ZSF is a raster dataset analysis tool that calculates diverse statistics within specified zones, such as TZs [43]. The ZSF output is associated with the corresponding 3D values of all TZs, providing a comprehensive analysis of how the built environment influences travel behavior. 2.2.3. Transportation Network To develop a realistic simulation of commuting patterns towards activity spaces, a variety of network attributes were integrated [11, 27]. This dataset included elements such as intersections, traffic lights, pedestrian crossings, speed limits, route widths, and lengths, all visualized on a GIS map. These factors were updated by the Zanjan Traffic Department during May and June 2020. An initial analysis of the network was conducted using ArcGIS 10.6, and the results were saved as supplementary traffic datasets. The amalgamated data led to the development of a Transportation Network Map (TNM). Furthermore, information regarding pedestrian crossings, traffic lights, and intersections was consolidated into a single point map, providing a detailed and accurate depiction of the city's transportation infrastructure. 2.2.4. Synthetic population To determine the population size for a travel model, one can either integrate the full population census of the origin into the model or rely solely on the number of travelers obtained from relevant travel statistics [44]. In our research, the modeled population, consisting of daily Travel Generation Data (TGD) within each traffic zone, was used to feed the simulation. The TGD covered almost the entire traveler population and effectively overcame the sample self-selection issue, which is a challenge in travel demand and activity space literature [28]. However, travel behavior demand for destinations and time dimensions was lacking in TGD, which needed to be enriched by a travel behavior survey (Section 2.2.2. Travel Behavior Survey to Extract Decision Rules). The data, comprising text files and a Traffic Zone Map (TZM), logs raw counts of all daily inbound and outbound journeys for these zones. This data was collated by Tarhe-Haftom Consulting Engineers in collaboration with the traffic department of Zanjan, spanning from March 2 to June 28, 2018. As per this study, Zanjan was segmented into 150 neighborhoods. Certain personal traits were mapped to neighborhood attributes, including a unique code for every Traffic Zone (TZ), the total number of travelers per neighborhood on a typical workday, car ownership, and so forth. 2.2.5. Validation data The validation of the model transfer process is a critical step in determining its predictive capabilities and ensuring its applicability in regions similar to the one it represents [32]. A crucial element for validating any traffic simulation is a reliable count of traffic in origin-destination flows [45]. In this study, a combination of emerging data sources (EDSs) and manual counts was used to gather traffic data across multiple Traffic Zones (TZs). EDSs, which include video vehicle detection and inductive loop counters, utilize data and communication technologies to capture real-time traffic information from various vehicles and transportation networks [46]. However, EDSs in Zanjan did not cover all TZs. Of the 46 EDSs installed throughout the city, only 31 were applicable for the validation process. Therefore, manual counting was also conducted to account for traffic in TZs not covered by EDSs. The selection of TZs for validation was based on several criteria, including the location of EDSs, the spatial distribution of TZs across the city, and the inclusion of diverse types of TZs. Consequently, three major activity spaces (TZs 1, 3, and 8) were selected along with other randomly chosen TZs. Additionally, two peak traffic periods were chosen for data collection—morning and afternoon—due to their high traffic flow and congestion (9-10 am, 11-12 am, 18-19 pm, and 20-21 pm). For comprehensive traffic count data, observations were made on three different days: two weekdays and one holiday. Data collection occurred on April 20, 22, and 25, and the average of these counts was used as the observed data for validation 3. Methodology The agent-based model serves as the foundation of the proposed activity-based model, with the goal of replicating the daily movements of passengers as agents traveling to various activity destinations. The flowchart illustrating the agent-based model can be found in Fig. 2 . The study focused on travel to non-work activity areas and incorporated four types of information into the agent-based model: travel behavior attributes (such as preferred destinations, trip purposes, departure times, and duration of stay), built-environment attributes (including land-use density-diversity and distance from origin to destination), network characteristics (such as intersections, traffic lights, route features), and travel supply data. Each time step in the agent-based model represents an hour from 7:00 to 24:00, excluding data beyond this range due to the infrequency of night trips. The model simulates 17 time steps, equivalent to 17 hours, estimating each agent's next activity destination during every time step. The model framework employs rule-based models to simulate the interaction between spatial environments and human behavior in making trips to activity spaces across different time periods. The study leveraged the capabilities of AnyLogic, a versatile toolkit known for simulating real-world phenomena using discrete events, system dynamics, and agent-based modeling techniques. AnyLogic's compatibility with open-source GIS environments and its ability to schedule individual behaviors within the simulation model using a Java script platform were particularly valuable [ 47 ]. This section outlines the key aspects of the simulation model, drawing inspiration from the approach utilized by Inturri, et al. [ 48 ]. 3.1. Proposed ABM components Environment : The environment in the developed Agent-Based Model (ABM) represents the space where agents engage in activities and interact. It comprises two GIS datasets: the transport network and the traffic zone map. The network consists of a fixed route and optional routes, exported from GIS as the Transportation Network Map (TNM), composed of network nodes and links, stop nodes, and diversion nodes (allowing switching between fixed and optional routes). The TNM includes a variety of attributes such as speed limits, route widths, junctions, one-way routes, traffic lights, and crosswalks. Another key component of the environment is the Traffic Zone Map (TZM), displaying 150 neighborhoods in Zanjan. In addition to tracking daily travels, the TZM also records the number of visits to each traffic zone. Agents : Agents in this model are autonomous and adaptive individuals, optimizing their utility by learning and adapting new behaviours based on experience and training. Each traveler agent aims to find an activity space and reach it. Agents have an overall identical profile (e.g., age, gender, occupation, and residential location) and a daily trip plan including trip origins, destinations, purposes, and departure times. They are grouped into households, sharing characteristics like the available number of vehicles. Trip requests of passengers are stochastically generated according to the demand model in time steps. The status of a traveler agent, regarding commuting, includes four dynamic states: ready for a trip, in travel, at destination, and returning to the beginning or starting another journey from the destination. Agents move from node to node, deciding to accept or reject opportunities at the new node based on specific rules. They are introduced into the simulation by an event and removed either when their visit duration ends or prematurely in cases where the simulated agent population exceeds the expected number. During iterations, agents acquire new information (e.g., traffic conditions) through interactions, which is used to update their perceived utility for subsequent iterations based on a reinforcement learning rule. 3.2. Knowledge learning process Users’ trip requests are generated from an origin (O) zone to a destination (D) zone with a negative exponential distribution, based on a gravitationally distributed probability. Given a set of n zones, \(\:{TR}_{ij}^{t}\) is the probability that a trip with origin \(\:i\) has destination in the zone \(\:j\) , in time \(\:t\) calculated with Eq. 2 , where \(\:{TR}_{i}\) is the generation trip rate of the zone \(\:i\) , proportional to an average trip rate achieved by travel generation data (TGD), and \(\:{P}_{ij}^{t}\) is the probability that a trip with origin \(\:i\) has destination in the zone \(\:j\) in time t calculated with Eq. 3 : $$\:{TR}_{ij}^{t}={TR}_{i}\bullet\:{P}_{ij}^{t}$$ 2 $$\:{P}_{ij}^{t}=\frac{{S}_{ij}^{t}\bullet\:{\left({d}_{ij}\bullet\:{l}_{j}\right)}^{a}\bullet\:{e}^{{\beta\:d}_{ij}}}{{\sum\:}_{k=1}^{n}{S}_{ij}^{t}\bullet\:{\left({d}_{kj}\bullet\:{l}_{j}\right)}^{a}{\bullet\:e}^{{\beta\:d}_{kj}}}$$ 3 \(\:{S}_{ij}^{t}\) represents proportion of individuals who prefer to travel from zone \(\:i\) to zone \(\:j\) in time \(\:t\) , as estimated by the travel behaviour survey results. The distance from \(\:i\) to zone \(\:j\) denoted as \(\:{d}_{ij}\) , is also calculated. Additionally, \(\:{l}_{j}\) is the coefficient of land use impacts including land use density and diversity in traveling individuals to zone \(\:j\) . Land use diversity and density are quantified using the focal statistics function and kernel density, respectively, operating in ArcGIS for each traffic zone. Both of these analytical tools require the ArcGIS Spatial Analyst extension to function [ 49 ]. α and β are the parameters of the decay function: $$\:{f\:\left(d\right)=\:\left(d\bullet\:l\right)}^{a}{e}^{{\beta\:d}_{ij}}$$ 4 3.3. Path learning rule A new setup was implemented to replicate this particular process category, emphasizing adaptable transit services within different system setups. This simulation investigates the interaction between vehicles navigating a flexible road network and users commuting from their origins to their destination TZs. The network configurations are built upon two key components: arcs and nodes. Within this framework, nodes are identified as intersections, traffic lights, and pedestrian crossings. These nodes are fine-tuned using a delay function that allocates specific timings to each node, drawing from information obtained from police operations in Zanjan to ensure a faithful representation of actual traffic patterns The arc \(\:{TZ}_{1}-{TZ}_{2}\) connects origin \(\:{TZ}_{1}\) to destination \(\:{TZ}_{2}\) directly, without any intermediate TZs.. It is characterized by factors such as capacity ( \(\:c\) ), length ( \(\:l\) ), free-flow speed ( \(\:{v}_{f}\) ), flow (q), and additional costs (O) like tolls. The cost of the arc (c) is determined by a function that incorporates these five factors, as described by Zhang, et al. [ 33 ]: $$\:c=g\left(c‚\:l‚\:{v}_{f}‚\:q‚\:o\right)$$ 4 The function \(\:g\left(O\right)\) represents the performance of the arc. In the current model, arc capacities are assumed to be unlimited, resulting in a constant arc cost and no consideration for congestion effects. Travelers learn about the travel cost of links on their route, while TZs gather information on the shortest path from themselves to all other nodes visited by travelers within that TZ. When travelers reach a new TZ, they compare the travel costs from the current TZ to each TZ along their route. Subsequently, both travelers and TZs update their information to reflect the shortest path. The variables used were updated by the Zanjan Traffic Department in May and June 2018. An initial network analysis was performed using ArcGIS 10.6, and the results were saved as additional traffic datasets. 3.4. Pattern recognition and clustering methods In this study, we developed a space-time GIS approach and utilized the K-means clustering technique to facilitate pattern recognition analysis and visualize the dynamics of activity spaces resulting from ABM. The GIS analysis is centered on a space-time GIS framework with temporal dynamic segmentation [ 50 ], and has been implemented in ArcScene, which serves as the 3D viewer within ArcGIS. Referred to as Spatial Cluster Detection (SCD), this approach integrates both spatial and temporal dimensions of activity space shifts. The utilization of the 3D method presents several benefits in uncovering the structure of urban activity spaces. It enables the identification of spatial characteristics such as area, shape, density, as well as spatial relationships like proximity. Additionally, it visualizes the spatial-temporal hierarchy of multi-local activity spaces. Furthermore, the K-means clustering method is efficient in organizing urban areas based on human activity data into measurable ordinal categories, such as main activity areas and secondary activity areas [ 22 ]. This widely-used unsupervised machine learning technique identifies clusters within the data based on similarities and differences among data points [ 51 ]. In our study, we employed the k-means clustering method as implemented in Excel.3.5. Model Calibration and primary implementation Model calibration is crucial for ensuring that simulations accurately reflect real-world scenarios and produce reliable results. During the calibration phase, the model was refined to tune agent behaviours within the ABM and increase its complexity. The first iteration of the model featured homogeneous agents with individual agents uniformly incorporating basic rules such as variations in trip rates for Traffic Zones (TZs), modifications of trip hours, and adjustments to a simplified environment representing only five zones. These modifications were informed by the initial simulation results and the expert knowledge of the research team. During the initial simulation, a significant discrepancy was observed between the actual data and the model's predictions. This gap indicated that certain aspects of the simulation had been overlooked, underscoring the need for further refinements to enhance the simulation's accuracy. Consequently, the model underwent multiple iterations with varied parameters to achieve more realistic outcomes. The refined simulation sets, which demonstrated the highest accuracy, were selected for the final configuration of the model. Subsequent to these adjustments, the model was scaled up to simulate Zanjan, Iran, ensuring that it consistently reflected the behaviours identified during the calibration phase. The model experiments were conducted after the initial settings had been established. These included increasing the number of TZ visits for individual agents and setting a preference for the destination TZ. 4. Results 4.1. Model Validation The model's performance was evaluated by comparing real and simulated data, using statistical methods of R-square. The validation yielded a percentage error of 15.86% and an R-square value of 0.90, indicating a close alignment between the simulated estimates and observed values. Table 2 presents the detailed validation results for the selected TZs and hours. Notably, the validation process revealed that the model is overall consistent with real observation in popular city centers like TZ1, TZ3, and TZ8, while less frequented TZs like TZ50 and TZ85 showed lower accuracy. Table 2 The validation results by R-square Traffic Zones RMSE/Time districts RMSE % R 2 9–10 AM 11–12 AM 18–19 PM 19–20 PM TZ1 8.20 8.93 10.10 9.97 9.30 0.8965 TZ3 10.09 13.21 9.59 8.18 10.27 0.9999 TZ8 8.45 8.21 10.42 10.78 9.46 0.9923 TZ11 15.87 9.57 11.07 16.02 7.74 0.9471 TZ42 36.36 120.00 30.77 0.00 17.05 0.8994 TZ50 40.00 0.00 25.93 16.67 20.65 0.7819 TZ66 36.73 20.69 20.93 17.74 15.21 0.8629 TZ85 5 19.44 1.43 9.33 8.80 0.8376 TZ105 50.98 15.38 18.48 17.61 18.42 0.8933 TZ140 66.67 33.33 33.33 66.67 41.67 0.9048 Average 23.02 16.33 13.44 10.63 15.86 0.9015 Specifically, TZ3 exhibited the highest prediction accuracy with an average percentage error of 9.3%, while TZ4 had the lowest accuracy with a 50% error rate. Additionally, R-square results corroborated the findings from the percentage error method, indicating a strong correlation between the volume of traveler attraction in TZs and the model's accuracy. This pattern aligns with the findings of previous studies by Zhong, et al. [ 52 ] and Apronti, et al. [ 53 ] 4.2. Identifying activity-based travel distribution This section presents the results derived from the integrated models. The practical application of the model was demonstrated through a case study conducted in Zanjan, Iran. The simulation encompassed 150 traffic zones over a span of 17 hours, excluding the time period from 12 AM to 7 AM, resulting in the simulation of 66743 tours. The primary result pertains to the total number of daily travels within the activity space, serving to elucidate the general mobility patterns within the Zanjan and evaluate the predominant activity spaces within the TZs. Figure 2 displays the detailed results of the model regarding a critical mobility indicator, showing the relative impact of each traffic zone on the daily arrival of visitors. The analysis reveals that the distribution of individual activity spaces is predominantly concentrated within a limited number of traffic zones. Notably, TZ 1, situated in the Central Business District (CBD) and constituting a mere 0.14 percent of the city's total area (approximately 10 hectares), attracts around 11 percent of all internal trips. This underscores a monocentric travel distribution pattern, with TZ 1 emerging as the primary activity space. Moreover, a discernible contrast is observed between TZ 1 and its adjacent zones in terms of travel attraction. For instance, TZ 8, serving as a secondary activity space linked to the primary hub, garners nearly one-third of the total trips compared to TZ 1. This disparity underscores the concentrated nature of travel patterns within the CBD. The model's findings further delineate a distinct division in travel attraction among traffic zones. While certain key activity spaces are clustered within the CBD, such as TZ 1, TZ 3, and TZ 8, other activity nodes exhibit a more dispersed distribution, resulting in a relative distribution of trips across various zones. Overall, the observed phenomenon aligns with our expectations and underscores the nuanced dynamics of travel behavior within the Zanjan region. 4.3. Clustering activity-spaces based on travel volumes The study utilized travel diaries to analyse the spatial dispersion of activity locations. A k-means clustering method was employed to identify key activity destinations. The optimal number of clusters was determined using the elbow method, resulting in the identification of four clusters that explain 94.65% of the variance. The results of the clustering analysis are visualized in Fig. 4 , showcasing the grouping of activity spaces based on daily travels (Fig. 4 a), and sub-clusters of the second cluster (Fig. 4 b). While spatial information was not directly considered in the clustering algorithm, the results provide insights into the geographical distribution of activity spaces. In this context, the horizontal axis signifies the traffic zones, indicating that points in close proximity to each other correspond to nearby traffic zones in reality. The cluster analysis reveals similarities within each cluster and differences among clusters. Cluster 1 represents the main activity hub within the CBD (traffic zone 1), drawing a substantial volume of trips. Statistical analysis revealed notable distinctions between cluster 1 and the remaining clusters in terms of attracted travels. This particular zone, previously highlighted, accounted for 11% of all trip attractions. Cluster 2 consists of second-order activity spaces, with distinct patterns and locations relative to the main activity space; attracting approximately 30% of activity-based travels across 12 traffic zones. As illustrated in Fig. 4 b, this cluster encompasses three sub-clusters characterized by internally consistent activity patterns, predominantly situated on the outskirts of the main activity space. Additionally, two other secondary activity sites are dispersed throughout the city, including a distinct official-marketing area adjacent to the first sub-cluster and newly developed neighborhoods in the suburban regions. The primary distinctions among the three sub-groups are their proximity to the main activity space and the total number of trips attracted to each subgroup. Notably, the trendline depicted in Fig. 4 b indicates a decrease in trip volume to the urban activity center as the distance from the main activity space increases. Cluster 3 primarily includes local activity spaces, encompassing 43 traffic zones and approximately 37% of activity-based trips. As illustrated in Fig. 4 a, the clustering method reveals that the third cluster exhibits a relatively regular spatial distribution pattern across the traffic zones. Cluster 4 comprises entirely local activity spaces distributed across the city, covering 89 traffic zones and approximately 21% of activity-based trips. The spatial distribution of these activity spaces mirrors that of the third activity space, extending across all traffic zones within the city. 4.4. Visualising spatio-temporal shifts of activity spaces Four distinct levels of activity spaces were identified through k-means cluster analysis, as illustrated in Fig. 4 a. The focus of this paper does not extend to analysing and visualizing the spatial-temporal dynamics of cluster 3 and cluster 4, as these clusters do not play a significant role in the city's overall traffic, particularly in activity-based travels. Therefore, this section primarily concentrates on analysing the time slots of the city's main hubs, including the main activity space (cluster 1) and secondary activity spaces (cluster 2). Hourly simulation results from 7:00 to 24:00 for the main activity space are presented in Fig. 5 a, while each sub-cluster of secondary activity spaces is illustrated separately in Fig. 5 b (first sub-cluster), Fig. 5 c (second sub-cluster), and Fig. 5 d (third sub-cluster). The paper additionally incorporates a Spatial Cluster Detection (SCD) technique that employs 3D maps (See Fig. 6 ) to provide a more precise illustration of the congestion impact of activity travels during peak hours, as opposed to the traditional traffic volume functions typically utilized in statistical studies. Each hour is depicted through a unique 3D map, with varying heights indicating the levels of activity-based travel volumes. To emphasize secondary activity areas, the main urban activity zone was excluded from the maps in Fig. 6 b. Analysis of the hourly data generated by the simulation model reveals a notable fluctuation in the volume of trips to the main activity area at different times of the day (see Fig. 5 a and 6 a). The majority of trips occur during the evening peak and the period spanning from the morning peak to the midday peak. As anticipated, the evening rush hour experiences heavy congestion, leading to oversaturation in several time slots. Interestingly, there is a significant decline in activity between 14:00 and 16:00, indicating a lull in activity-based travel to the urban main activity space, reminiscent of the quiet hours observed between midnight and 7:00. It is important to highlight that many activities within this category do not conclude at 21:00. In fact, activities within the traffic zone typically end around 22:00 on average, and they may even extend as late as midnight, which coincides with the closing time of the retail stores and other facilities in the CBD. In Fig. 5 b and 6 b, the first sub-cluster of second-order activity spaces are shown to be located near the main activity space. There is a significant variation in activity-based travel volume across the selected TZs throughout the day. The durations of these trips are generally consistent with those in the main activity space. Particularly, there is a significant increase in traffic numbers during the evening hours, reflecting the pattern observed in the main activity space. The fluctuations among TZs within this sub-cluster exhibit similar variations, especially from midday to midnight. The peak hours of this sub-cluster typically occur from around 18:30 to 20:30, with before noon peak hours ranging from approximately 8:30 to 12:30. Before noon peak times extend across a larger area, indicating a larger variance between the activity spaces. Additionally, as depicted in Fig. 5 b, there is a period of low peak activity from approximately 13:30 to 14:30, showing a high volume of travels returning from the CBD to their origin. Notably, there is a period of weaker activity from 14:30 to 17:30, characterized as blackout hours. Figure 5 c and 6 c depict the second sub-cluster situated at a relatively greater distance from the main activity space. In contrast to the first sub-cluster, the second sub-cluster exhibits distinct space-time paths with unique activity and travel patterns. Within this sub-cluster, the TZs centrality patterns vary significantly across different hours, yet remain relatively stable during midday, particularly during the rush hour from 13:00 to 14:00. Notably, there is a decrease in activity observed between 15:00 and 17:00, suggesting a period of reduced activity akin to a downtime or resting phase in activity-based travels to the sub-cluster activity space. In Fig. 5 d, situated in the inner suburb, similar patterns to those in Fig. 5 c can be observed. However, there is an extension in the sleep time during an additional time slot, particularly from 12:00 to 13:00. The rush hour remains relatively stable in the midday, mirroring the trends seen in the second sub-cluster. Besides, it shows an overall increase in variation compared to the previous sub-clusters. In general, the spatial pattern of activity-based travels indicates that travel volumes are influenced by the distance from the city center. Specifically, travel volumes are higher in cluster 1 and sub-cluster 1, which includes the CBD and surrounding TZs. Conversely, travel volumes are lower in the second sub-cluster, which consists of TZs located between central zones and suburbs, and in the third sub-cluster situated in the inner suburb. In the analysis of temporal patterns, it is observed that cluster 1 and sub-cluster 2 exhibit similar variations in activity travel attraction during evening rush hours. On the other hand, sub-cluster 2 and 3 attract the highest levels of activity travels during midday. However, there is a relative variation in travel patterns during other times of the day. 5. Discussion and Conclusion Motivated by initial discussions from [ 2 , 3 , 5 , 6 , 17 ], our study aimed to investigate the dynamic nature of daily activity spaces with a focus on detailed time dimensions. We introduced a clear conceptual and operational definition of multiple activity spaces, utilizing the flexibility of ABM and a spatial modeling approach from GIS to measure their dynamics and classify them based on individual activity spaces. In addition to exploring the influence of personal preferences on activity-based mobility, we examined how built-environmental characteristics may influence individuals' decisions to travel beyond their immediate neighborhoods. The results obtained from the space–time path based clustering method indicate that activity places may cluster at various hourly intervals within an individual's activity space, with a higher concentration of frequently visited locations. This aligns with previous research that has examined the formation of activity clusters around daily anchor points, showing increased activity intensity in the city's main activity space [ 3 , 6 , 16 , 18 ]. Our analysis demonstrates that the central activity space acts as a focal point for activity-based travel, with a greater number of trips terminating there. This concentration of activities in the CBD influences travel behavior, particularly in terms of activity-based travel patterns. The number of trips decreases as distance from the main activity space increases, indicating a spatial decay effect on travel attraction, which is consistent with some studies [ 7 , 16 ]. The reason behind the formation of this travel behavior pattern is evident in Zanjan, where high-quality public resources such as office buildings and upscale shopping malls are predominantly located in the main activity space. This, along with measures of land-use density and diversity in the built environment, influences intra-urban mobility patterns. The primary activity space and the initial cluster of second-order activity spaces demonstrate similar temporal patterns, while the second and third sub-clusters exhibit distinct temporal patterns. A notable amount of travel to these zones occurs during midday hours, primarily due to time constraints within office buildings. These activity spaces play a vital role in alleviating traffic congestion in the city center and are integral to urban traffic management. These sub-clusters consist of a mix of public services such as schools, commercial facilities, and office buildings. They also offer various benefits associated with urban design features, such as high levels of accessibility, ample parking space, and close proximity to residential neighborhoods. By implementing minor adjustments to the built environment and influencing travel behavior, these areas can function as alternative destinations to redirect a significant portion of travel away from the main activity hub, in particular during evening rush hours. This shift in travel patterns not only improves accessibility and effectively redistributes traffic flow in Zanjan but also contributes to the decentralization of urban activity spaces, transitioning from a monocentric to a more diverse urban layout. This study holds significant implications for the effective management of daily urban traffic and the allocation of resources to underutilized areas in Zanjan. Firstly, our model's findings offer insightful descriptions of the interactions between citizens and the urban built environment. As a result, this information can be utilized by city managers and business leaders to tailor services and locating new infrastructure in response to the evolving profile of people in the activity spaces. Secondly, the analysis of spatio-temporal activity patterns serves as a valuable tool for assessing the utilization of urban spaces, understanding the city's daily movement patterns, evaluating traffic conditions, and assessing the impact of design interventions on changes in travel behavior. This comprehensive approach enables a deeper understanding of urban dynamics and facilitates informed decision-making in urban planning and traffic management initiatives. Our findings support existing theoretical frameworks suggesting that travel destination choices are not solely based on individual decisions but are also influenced by the physical environment at the destination [ 11 , 23 , 26 – 28 ]. One key behavioral motivation for visiting a destination is people's interest in the activities available. When individuals have a common interest in particular type of activity, a destination with diverse and densely packed functions becomes appealing as it caters to a variety of activity demands. Existing literature has confirmed that destinations offering diverse and dense services tend to attract visitors [ 8 , 29 ]. Our findings suggest to policymakers that areas with diverse and densely packed spatial functions, in close proximity to residential neighborhoods, can be effectively leveraged and potentially utilized in initiatives aimed at altering travel behavior and promoting sustainable mobility. Additionally, the incorporation of built-environment attributes into the ABM process represents a significant contribution of our study to the field of travel demand and destination choice modeling research. This emphasizes the importance of integrating GIS and ABM in modeling travel behavior. Furthermore, we utilized real-world data to verify the accuracy of the model in simulating the activity destination choice model in Zanjan; The simulation results were then compared with the actual travel counts recorded during peak hours on various weekdays in selected TZs. Our study emphasizes the significance of incorporating real-world data for model validation, consistent with existing studies in in the field of travel behavior [ 21 , 23 , 45 , 54 ]. We highlight the importance of traffic count data, such as EDSs, in validating urban movements within ABM and traffic simulation, enhancing the realism and reliability of model performance compared to previous studies. This inclusion enhances the realism and reliability of the model's performance. The use of EDSs not only validates the model's accuracy but also serves as a more comprehensive and readily available database for pattern recognition of urban mobility and analyzing travel behavior, particularly in developing countries. The developed model framework was tailored for a medium-sized Iranian city and its travel-to-non-work area, but it holds potential for adaptation to urban transportation modeling in densely populated cities worldwide. Prior to deploying the model in other regions, it is essential to calibrate the prototype model with local time-use data and conduct thorough verification and validation of key system elements such as activity engagement and trip generation. Furthermore, the current model abstracted implementations of agent activities to establish a generalized framework of human spatial behavior based on typical behaviors in activity spaces. Future research will focus on identifying additional spatial behaviors and expanding existing activity types to broaden the model's applicability to various spatial configurations and environments. However, there are limitations associated with the data collection method and the operationalization of urban activity spaces that could be addressed in future studies. Integrating additional data sources like cell phone data using simulation methods can enhance the model's accuracy and aid in validation, although such data may not be readily available in developing countries like Iran due to privacy constraints. Furthermore, the model's activity-based travel structure, comprising sub-models of specific human tasks, could benefit from independent updates by incorporating more advanced approaches, such as integrating informed path-finding or movement strategies at the individual agent level. Declarations Author Contribution M.Azari conceived of the presented idea.M.Hatami and Mr. Hosseini developed the theory and performed the computations.M.Azari verified the analytical methods. S.Moridpour encouraged A.B. to investigate [a specific aspect] andsupervised the findings of this work. All authors discussed the results and contributed to the final manuscript. References Golledge, R.G.: Spatial behavior: A geographic perspective. Guilford Press (1997) Matthews, S.A., Yang, T.-C.: Spatial polygamy and contextual exposures (spaces) promoting activity space approaches in research on place and health. Am. Behav. Sci. 57 (8), 1057–1081 (2013) Hasanzadeh, K., Kyttä, M., Lilius, J., Ramezani, S., Rinne, T.: Centricity and multi-locality of activity spaces: The varying ways young and old adults use neighborhoods and extra-neighborhood spaces in Helsinki Metropolitan Area. Cities. 110 , 103062 (2021) Horton, F.E., Reynolds, D.R.: Effects of urban spatial structure on individual behavior. Econ. Geogr. 47 (1), 36–48 (1971) Li, R., Tong, D.: Constructing human activity spaces: A new approach incorporating complex urban activity-travel. J. Transp. Geogr. 56 , 23–35 (2016) Hasanzadeh, K.: Exploring centricity of activity spaces: From measurement to the identification of personal and environmental factors. Travel Behav. Soc. 14 , 57–65 (2019) van Dijk, J., Krygsman, S.: Analyzing travel behavior by using GPS-Based activity spaces and opportunity indicators. J. Urban Technol. 25 (2), 105–124 (2018) Chen, F., Zhou, S., Lu, J., Zheng, Z.: A behavioral explanation of the activity-space segregation: Individuals’ preference of choosing an activity destination. Environment and Planning B: Urban Analytics and City Science, : p. 23998083241229110. (2024) Cagney, K.A., York Cornwell, E., Goldman, A.W., Cai, L.: Urban mobility and activity space. Ann. Rev. Sociol. 46 , 623–648 (2020) Jones, P.M.: New approaches to understanding travel behaviour: the human activity approach , in Behavioural travel modelling , pp. 55–80. Routledge (2021) Fu, X., Wu, Y., Huang, D., Wu, J.: An activity-based model for transit network design and activity location planning in a three-party game framework. Transp. Res. E. 168 , 102939 (2022) Perchoux, C., Chaix, B., Cummins, S., Kestens, Y.: Conceptualization and measurement of environmental exposure in epidemiology: accounting for activity space related to daily mobility. Health place. 21 , 86–93 (2013) Galster, G., Sharkey, P.: Spatial foundations of inequality: A conceptual model and empirical overview. RSF: The Russell Sage Foundation Journal of the Social Sciences, 3(2): pp. 1–33. (2017) Xu, Y., Belyi, A., Bojic, I., Ratti, C.: Human mobility and socioeconomic status: Analysis of Singapore and Boston, vol. 72, pp. 51–67. Computers, Environment and Urban Systems (2018) Kwan, M.-P.: Beyond space (as we knew it): Toward temporally integrated geographies of segregation, health, and accessibility: Space–time integration in geography and GIScience, vol. 103, pp. 1078–1086. Annals of the Association of American Geographers (2013). 5 Zhang, Y., Wang, J., Kan, C.: Temporal variation in activity-space-based segregation: A case study of Beijing using location-based service data. J. Transp. Geogr. 98 , 103239 (2022) Silm, S., Ahas, R., Mooses, V.: Are younger age groups less segregated? Measuring ethnic segregation in activity spaces using mobile phone data. J. Ethnic Migration Stud. 44 (11), 1797–1817 (2018) Zhang, Y., Liu, L., Wang, H.: A new perspective on the temporal pattern of human activities in cities: The case of Shanghai. Cities. 87 , 196–204 (2019) Hafezi, M.H., Daisy, N.S., Millward, H., Liu, L.: Framework for development of the Scheduler for Activities, Locations, and Travel (SALT) model. Transportmetrica A: Transp. Sci. 18 (2), 248–280 (2021) Bekhor, S., Dobler, C., Axhausen, K.W.: Integration of activity-based and agent-based models: case of Tel Aviv. Isr. Transp. Res. record. 2255 (1), 38–47 (2011) Hesam Hafezi, M., Sultana Daisy, N., Millward, H., Liu, L.: Framework for development of the Scheduler for Activities, Locations, and Travel (SALT) model. Transportmetrica A: Transp. Sci. 18 (2), 248–280 (2022) Zhang, S., Yang, Y., Zhen, F., Lobsang, T., Li, Z.: Understanding the travel behaviors and activity patterns of the vulnerable population using smart card data: An activity space-based approach. J. Transp. Geogr. 90 , 102938 (2021) Cheliotis, K.: An agent-based model of public space use. Comput. Environ. Urban Syst. 81 , 101476 (2020) Arentze, T.A., Timmermans, H.J.: A learning-based transportation oriented simulation system. Transp. Res. Part. B: Methodological. 38 (7), 613–633 (2004) Kagho, G.O., Balac, M., Axhausen, K.W.: Agent-Based Models in Transport Planning: Current State, Issues, and Expectations. Procedia Comput. Sci. 170 , 726–732 (2020) Zhong, G., Yin, T., Zhang, J., He, S., Ran, B.: Characteristics analysis for travel behavior of transportation hub passengers using mobile phone data. Transportation. 46 , 1713–1736 (2019) Liu, L., Silva, E.A., Yang, Z.: Similar outcomes, different paths: Tracing the relationship between neighborhood-scale built environment and travel behavior using activity-based modelling. Cities. 110 , 103061 (2021) Bastarianto, F.F., Hancock, T.O., Choudhury, C.F., Manley, E.: Agent-based models in urban transportation: review, challenges, and opportunities. Eur. Transp. Res. Rev. 15 (1), 1–20 (2023) Cervero, R., Kockelman, K.: Travel demand and the 3Ds: Density, diversity, and design. Transp. Res. part. D: Transp. Environ. 2 (3), 199–219 (1997) Gong, S., Dong, X., Wang, K., Lei, B., Jia, Z., Qin, J., Roadknight, C., Liu, Y., Cao, R.: Agent-based modelling with geographically weighted calibration for intra-urban activities simulation using taxi GPS trajectories. Int. J. Appl. Earth Obs. Geoinf. 122 , 103368 (2023) Franco, P., Johnston, R., McCormick, E.: Demand responsive transport: Generation of activity patterns from mobile phone network data to support the operation of new mobility services. Transp. Res. Part. A: Policy Pract. 131 , 244–266 (2020) Brown, D.G., Riolo, R., Robinson, D.T., North, M., Rand, W.: Spatial process and data models: Toward integration of agent-based models and GIS. J. Geogr. Syst. 7 (1), 25–47 (2005) Zhang, L., Levinson, D.: Agent-based approach to travel demand modeling: Exploratory analysis. Transp. Res. Rec. 1898 (1), 28–36 (2004) Gong, S., Cartlidge, J., Bai, R., Yue, Y., Li, Q., Qiu, G.: Geographical temporal huff model. calibration using taxi trajectory data GeoInformatica. 25 , 485–512 (2021) D’Exelle, B.: Representative Sample. In: Encyclopedia of Quality of Life and Well-Being Research, pp. 5511–5513. Springer Netherlands: Dordrecht (2014). A.C. Michalos, Editor Kulpa, T., Szarata, A.: Analysis of household survey sample size in trip modelling process. Transp. Res. Procedia. 14 , 1753–1761 (2016) Egu, O., Bonnel, P.: How comparable are origin-destination matrices estimated from automatic fare collection, origin-destination surveys and household travel survey? An empirical investigation in Lyon. Transp. Res. Part. A: Policy Pract. 138 , 267–282 (2020) Taylor, M.A.: Voluntary travel behavior change programs in Australia: The carrot rather than the stick in travel demand management. Int. J. Sustainable Transp. 1 (3), 173–192 (2007) Duhachek, A., Iacobucci, D.: Alpha's standard error (ASE): an accurate and precise confidence interval estimate. J. Appl. Psychol. 89 (5), 792 (2004) Cote, J.D.R., Diana, M.: Exploring the benefits of a traveller clustering approach based on multimodality attitudes and behaviours. Transp. Res. procedia. 25 , 2552–2565 (2017) Boarnet, M., Crane, R.: The influence of land use on travel behavior: specification and estimation strategies. Transp. Res. Part. A: Policy Pract. 35 (9), 823–845 (2001) Liang, J., Gong, J., Li, W., Ibrahim, A.N.: A visualization-oriented 3D method for efficient computation of urban solar radiation based on 3D–2D surface mapping. Int. J. Geogr. Inf. Sci. 28 (4), 780–798 (2014) Yoshida, T., Tanaka, K.: Land-use diversity index: a new means of detecting diversity at landscape level. Landscape Ecol. Eng. 1 , 201–206 (2005) Ziemke, D., Charlton, B., Hörl, S., Nagel, K.: An efficient approach to create agent-based transport simulation scenarios based on ubiquitous Big Data and a new, aspatial activity-scheduling model. Transp. Res. Procedia. 52 , 613–620 (2021) Zheng, P., Mike, M.: An investigation on the manual traffic count accuracy. Procedia-Social Behav. Sci. 43 , 226–231 (2012) Annur, R., Ponnusamy, V.: Information and communication technology (ICT) for intelligent transportation systems (ITS) , in Employing recent technologies for improved digital governance . IGI Global. pp. 164–194. (2020) Borshchev, A.: The big book of simulation modeling: multimethod modeling with AnyLogic 6. AnyLogic North America (2013) Inturri, G., Le Pira, M., Giuffrida, N., Ignaccolo, M., Pluchino, A., Rapisarda, A., D'Angelo, R.: Multi-agent simulation for planning and designing new shared mobility services. Res. Transp. Econ. 73 , 34–44 (2019) Desktop, E.S.R.I.A.G.I.S.: Release 10, vol. 437, p. 438. Environmental Systems Research Institute, CA (2011). Redlands Shaw, S.-L., Yu, H.: A GIS-based time-geographic approach of studying individual activities and interactions in a hybrid physical–virtual space. J. Transp. Geogr. 17 (2), 141–149 (2009) Rogerson, P.A.: Statistical methods for geography: a student's guide, pp. 1–432. Statistical methods for geography (2019) Zhong, M., Hanson, B.L.: GIS-based travel demand modeling for estimating traffic on low-class roads. Transp. Plann. Technol. 32 (5), 423–439 (2009) Apronti, D.T., Ksaibati, K.: Four-step travel demand model implementation for estimating traffic volumes on rural low-volume roads in Wyoming. Transp. Plann. Technol. 41 (5), 557–571 (2018) Pendyala, R.M., Kitamura, R., Kikuchi, A., Yamamoto, T., Fujii, S.: Florida activity mobility simulator: overview and preliminary validation results. Transp. Res. Rec. 1921 (1), 123–130 (2005) Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4835588","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":335965795,"identity":"6bf90dfe-6994-45e9-a989-a55ed4110071","order_by":0,"name":"Mehdi Azari","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIie3Rv6rCMBTH8V8RdAm4xqW+Qoogyr3QV2kRdHFw1E0R4uQT3Be5Y8sZuhyc46TuLm46aeof3FJHh3ynQ+BDEg7g831hKhOQCaQdg8X+dZq5SP9BfkuyVB+RuCTAsJzr8qOHdYt1bvaguPlHenb5JzRXWUBTF+HNoJeAanKTatNiguQEObuIGXftX6gODrSJNAEGyOcOonbHklxF25JJakm7khhxv0UqS5BboioJjzu9RJGKOFi2FnokIk7nblJwtD1PKQ65cThd9E8YFkQnF3nC9yjsTquBz+fz+dzdAM4iUj9Z/kBrAAAAAElFTkSuQmCC","orcid":"","institution":"University of Maragheh","correspondingAuthor":true,"prefix":"","firstName":"Mehdi","middleName":"","lastName":"Azari","suffix":""},{"id":335965798,"identity":"ddab1aad-24b7-49ce-a8e2-cba78c2cd8cd","order_by":1,"name":"Sara Moridpour","email":"","orcid":"","institution":"RMIT University","correspondingAuthor":false,"prefix":"","firstName":"Sara","middleName":"","lastName":"Moridpour","suffix":""},{"id":335965800,"identity":"08167bda-70d2-4936-aa93-47b9f6094c47","order_by":2,"name":"Mohsen Hatami","email":"","orcid":"","institution":"University of Florida","correspondingAuthor":false,"prefix":"","firstName":"Mohsen","middleName":"","lastName":"Hatami","suffix":""},{"id":335965801,"identity":"a15b9315-c64b-40d9-a1cf-ae82b22de8b9","order_by":3,"name":"Monireh Hosseini","email":"","orcid":"","institution":"University of Zanjan","correspondingAuthor":false,"prefix":"","firstName":"Monireh","middleName":"","lastName":"Hosseini","suffix":""}],"badges":[],"createdAt":"2024-07-31 12:39:48","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4835588/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4835588/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":63626244,"identity":"0b91549c-5312-4fde-bba5-a9b3e7ad8e55","added_by":"auto","created_at":"2024-08-30 09:52:32","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":82131,"visible":true,"origin":"","legend":"\u003cp\u003eLocation of study and Traffic zones of Zanjan. Orange zones were selected to validate model that will be discussed in the section “2.5. Model Validation and Verification”\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/cbf2b39c483e4cdb96b36b01.jpg"},{"id":63626246,"identity":"4e0c0b3b-93d0-4397-9deb-1d91c990e2e2","added_by":"auto","created_at":"2024-08-30 09:52:33","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":132802,"visible":true,"origin":"","legend":"\u003cp\u003eProcess for Identifying Activity Destinations\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/2b85b2fbef2365ad1a7058dd.jpg"},{"id":63627174,"identity":"9b51ad64-d459-488e-ac96-54f1d6315441","added_by":"auto","created_at":"2024-08-30 10:00:33","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":35812,"visible":true,"origin":"","legend":"\u003cp\u003eDetecting Urban Activity Space of Zanjan Using Daily Trip Volumes.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/14505ffe8cbb041901aada33.jpg"},{"id":63626243,"identity":"1d0dcc71-efdf-4a7d-b437-7f58e442984d","added_by":"auto","created_at":"2024-08-30 09:52:32","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":49979,"visible":true,"origin":"","legend":"\u003cp\u003eclustering of activity destinations throughout the day using the k-means method. Figure “a” displays the overall clustering of activity space in four distinct clusters. Figure “b” provides a detailed view of the sub-clusters within the second-order activity space (cluster 2).\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/7a1b2e0f71bd7a554179e74a.jpg"},{"id":63627173,"identity":"dab1ec82-32ca-4062-9803-7358d0638e7e","added_by":"auto","created_at":"2024-08-30 10:00:32","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":81669,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal variations of activity-based trips in a normal day based on simulation results\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/3ea5633565f32818d6ce0c22.jpg"},{"id":63626248,"identity":"1b5a006e-81f2-462d-a26c-e110a13fc0c5","added_by":"auto","created_at":"2024-08-30 09:52:33","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":84873,"visible":true,"origin":"","legend":"\u003cp\u003evisualize the spatial-temporal variations in activity spaces through 3D maps generated from simulation results. The figure classifies activity spaces into three maps: 'a' for the primary activity space, 'b' for secondary activity space sub-cluster 1, and 'c' for secondary activity space sub-clusters 2 and 3. Zones with the highest activity travel attraction are highlighted in red, while green areas represent locations with less prominent travel attractions.\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/205d9d5cd03d5a8921949cb9.jpg"},{"id":70293864,"identity":"6d08704d-5462-4b01-b6be-e379dd3d8404","added_by":"auto","created_at":"2024-12-01 22:46:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1239326,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4835588/v1/637ba448-8616-4bbc-8aec-683481157406.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Exploring the spatial-temporal arrangements of urban activity space from individual's daily commute: A Geospatial-Agent based Approach Using Empirical Data","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe recognition of individual spatial behavior has been a key focus of academic research for many years. Scholars have widely adopted the concept of activity space as a way to assess individual spatial behavior [\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Activity spaces refer to specific areas that individuals visit regularly, representing their spatial footprint [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Researchers from diverse disciplines have employed various methods and data sets to model and investigate activity spaces. A significant portion of studies has focused on methods to explore the spatial arrangement of activity spaces [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], the fundamental factors influencing the forming of these spaces [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], and their implications for various areas such as transportation planning [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], travel behaviour change [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], activity location planning [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], human health [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], social exclusion [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], location services planning [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], and racial segregation [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWhile the concept of activity spaces has been widely utilized by researchers across different fields, some scholars argue that it has evolved into a multi-dimensional concept that remains underexplored [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. The evolution in daily activity patterns has transitioned from being centered around a single location to involving multiple locations, indicating a move towards spatial polygamy [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. This shift suggests that activity spaces are dynamic throughout the day [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], with different locations serving various aspects of daily life within specific temporal constraints.[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. For example, certain urban areas may attract individuals for specific tasks at particular times, leading to the formation of distinct urban activity spaces during those periods. Conversely, different locations shape alternative activity spaces at different times [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. This complexity challenges the traditional view of activity spaces as singular trip destination during peak traffic times. It underscores the significance of factors such as the spatial-temporal evolution of activity spaces and the timing of their utilization [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. This evolution in the concept of activity spaces has significant implications for urban planning, prompting planners to incorporate spatial-temporal dynamics effectively in shaping activity spaces [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAlthough previous studies have recognized the location heterogeneity of activity spaces and occasionally incorporated it into activity space modeling [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], there has been limited empirical examinations into the dynamic changes of activity spaces over time [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. By conducting space-time assessments, we not only gain information on how a specific space is used but also uncover valuable insights about the underlying factors driving human spatial behavior [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Additionally, this process contributes to the building an archive regarding human spatial behavior [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAnother aspect that have often overlooked in existing approaches is the inter-relationship between the environmental factors and human spatial behavior in the process of selecting an activity destination [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Drawing from relevant theories, travel destination choice involves an individual's decision at the origin, as well as the opportunities and limitations posed by the physical environment in the destination [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan additionalcitationids=\"CR27\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Accordingly, theories suggest that areas with denser and more diverse land uses closer to travelers, are more attractive for activity-based travel [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Also, some other studies have highlighted the significance of transportation network features in shaping destination choice behaviors [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Conventional travel demand models cannot capture the complex interactions among these factors and there is ongoing debate about how the destination's built-environment and route features impact travel behavior in activity-based and travel demand models [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. They have mainly concentrated on individual\u0026rsquo;s personal preferences, focusing on basic built-environment characteristics such as household and activity area locations and distances between them [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. However, it\u0026rsquo;s highly complicated to integrate these into travel demand models to accurately reflect human spatial behavior and include all these factors in a single model.\u003c/p\u003e \u003cp\u003eThis study aims to fill these gaps through a promising approach combined agent-based models (ABMs) with Geographic Information System (GIS). ABMs represent autonomous decision-making entities called agents that make decisions based on rules [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. In a travel demand behavior, an agent symbolizes an individual who decides to travel to an activity area based on their preferences, the output of their own non-linear actions, interpersonal interactions, and the environment [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. In the applied approach, GIS allows for integrating a wide range of geospatial data into ABMs to represent more realistic and detailed simulations and its interactions with the environment [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Moreover, GIS offers a clustering technique to analyse and map the shifts in individual\u0026rsquo;s activity spaces based on the results of ABM. Accordingly, a Spatial Cluster Detection (SCD) method based on 3D maps was conducted to space-time clustering analysis of activity spaces in an empirical case study in Zanjan, Iran. The SCD method yields a measurable ordinal outcome, i.e., main activity spaces, and second-order activity spaces, providing detailed insights into the spatial-temporal distribution of activity spaces based on travel data. This analysis includes identifying peak visitation times throughout the day, popular activity spaces, and duration of stay, which can offer valuable insights for activity location planning and travel behavior management.\u003c/p\u003e \u003cp\u003eThe proposed approach goes beyond simply measuring the extent of one's activity space and brings a geographical understanding to pattern-recognition modeling while providing perspective into how urban planning can shape individual spatial behavior.\u003c/p\u003e \u003cp\u003eHowever, this process is notably complex due to the intricate nature of modeling and their parameters and requires careful validation to ensure that the model accurately reflects real-world behaviors [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. It should be reminded that making model validation a crucial challenge in ABM construction due to the lack of real-world data [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Traditionally, researchers often used questionnaires or interviews to validate models, but these methods were laborious and time-consuming. Additionally, these data relied heavily on participants' perceptions [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. In order to tackle the validation challenges, an innovative approach involves utilizing precise traffic flow data from emerging data sources (EDSs) was proposed to validate the ABM.\u003c/p\u003e \u003cp\u003eOur major contributions can be summarised as follows:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eIntroducing a new activity destination choice model that utilizes Geospatial-ABM to simulate the spatial-temporal arrangements of activity spaces from individual's daily movements.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eIntegrating GIS methods to incorporate destination attractive into ABM, including innovative built-environment features and individual\u0026rsquo;s personal preferences as well as real travel generation data.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eVisualizing the spatial-temporal dynamics of activity spaces by SCD and categorizing them into several clusters and sub-clusters. This enables a detailed consideration of geographical patterns in the distribution of activity-based travels.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eAddressing model validation challenges by using real-world data on the number of visits to activity areas, obtained from sources like EDSs and manual tallying, to validate the effectiveness of the ABM model through the R-square and RMSE procedures.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eConducting extensive experiments on a case study in Zanjan, Iran, to demonstrate the effectiveness and superiority of the proposed methodology.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eThe remainder of the paper is structured as follows. The following section provides a concise overview of the case study and details the data collection and preparation process for this research. This is followed by an explanation of our methodological approach, focusing on the ABM and its key characteristics. The paper then delves into the application of the ABM method to the Zanjan case study, showcasing the results pertaining to activity spaces and their temporal variations. Lastly, the \u003cspan refid=\"Sec20\" class=\"InternalRef\"\u003ediscussion and conclusion\u003c/span\u003e section encapsulates the main findings and their implications, along with a brief examination of the study's limitations and recommendations for future research endeavors.\u003c/p\u003e"},{"header":"2. Data collection and case study","content":"\u003ch2\u003e2.1. Study area\u003c/h2\u003e\n\u003cp\u003eZanjan, the provincial capital of Zanjan Province in north-west Iran, serves as the study area for this research. In 1986, Zanjan\u0026apos;s population was 215,261, according to the Iranian Statistical Centre. The population increased to 433,475 by 2016, indicating significant growth. However, the city\u0026apos;s infrastructural development, including elements like urban facilities and transportation equipment, has not kept pace with this rapid population growth. This mismatch has led to congested narrow streets, particularly in urban cores.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGeographically, Zanjan is advantageously located near Tehran, Iran\u0026apos;s political center, and the industrial cities of Gazvin and Tabriz. Consequently, it functions as a crucial communication hub within the Iranian urban hierarchy. The city\u0026apos;s urban structure is centralized, supporting heavy transportation, especially within certain urban centers.\u0026nbsp;Figure 1\u0026nbsp;displays the location of Zanjan and\u0026nbsp;the traffic zones the city.\u003c/p\u003e\n\u003ch2\u003e2.2. Data collection and preparation\u003c/h2\u003e\n\u003cp\u003eEstablishing a realistic simulation requires the extraction of pertinent real-world rules and the identification of significant factors influencing the phenomena being simulated, necessitating a precise database\u0026nbsp;[32]. This section provides an overview of the data collection process.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e2.2.1. Travel behaviour survey\u0026nbsp;\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eThe travel behavior survey played a pivotal role in our research, providing key insights into both preferred destinations and home location decisions for commuters within Zanjan City. The travel survey data from Zanjan informed basic travel behaviors, serving as samples for agents in simulating multi-location activity space dynamics. This data encompassed a range of factors: travel patterns, modes of transportation, trip destinations, purposes, frequency, duration, and travel time, which bolstered the practicality of our approach. The outcome data from the survey is reflected in the agents\u0026apos; behavior to interact in a wide area, such as a city scale. The survey type was a self-administered questionnaire, distributed among commuters during peak and off-peak hours at different locations in the city. The questionnaire was designed with both open-ended and closed-ended questions. The open-ended questions gathered qualitative information, while the closed-ended questions collected quantitative data. The questionnaire was pre-tested to ensure clarity and comprehensibility.\u003c/p\u003e\n\u003cp\u003eSome important factors in the recruitment of travel behavior surveys are as follows.\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eRepresentative sample:\u0026nbsp;\u003c/strong\u003eA representative sample closely mirrors the overall attributes of the population, especially in characteristics pertinent to the study\u0026nbsp;[35]. In this study, two criteria were considered to achieve representativeness:\u003c/li\u003e\n\u003c/ul\u003e\n\u003col\u003e\n \u003cli\u003eFor spatial-based behavior, the survey took the form of a geographically stratified random survey of households. Accordingly, 150 traffic zones (TZs) were clustered into 15 regions compatible with administrative data, based on characteristics such as local proximity, transit accessibility, access to urban services, land uses, and so forth.\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eTo represent socio-demographic heterogeneity in the sampling sectors, random sampling was employed. This approach allows all participants in the larger group to have an equal chance of being selected, ensuring the most representative samples and data free of bias\u0026nbsp;[36].\u003c/li\u003e\n\u003c/ol\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eSample size\u003c/strong\u003e: Determining the appropriate sample size is a key challenge in travel behavior surveys\u003cspan dir=\"RTL\"\u003e\u0026nbsp;\u003c/span\u003e[36]. Empirical studies recommend basing the sample size on the entire population\u0026nbsp;[37]. This study surveyed a random sample of about 5% of households from each region based on the Zanjan population number (the population is 433,475, and the household number is 131,799)\u0026nbsp;[38]. A total of 6,604 questionnaires were distributed, with 5,942 valid responses received, yielding a response rate of approximately 90%.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eThe\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003equestionnaires\u003c/strong\u003e\u003cstrong\u003e:\u003c/strong\u003e Two local students from Zanjan University were chosen for each region and trained to survey travel behavior. It was beneficial to select at least one student from each area, as they were familiar with the environment.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eThe main questions that were asked:\u003c/strong\u003e The questionnaire consisted of two sections. The first section collected demographic information about the respondents, such as age, gender, and education level. The second section focused on travel characteristics, such as car ownership, travel mode, trip purpose, travel destination, trip frequency, trip duration, and travel time. The survey took approximately 20 minutes to complete.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eSampling errors:\u003c/strong\u003e Another issue associated with survey quality is the degree of data accuracy, which manifests itself in sampling errors. Sampling error is a bias that results from choosing to estimate a sample as representative of a larger population. Sampling errors can be estimated as follows [39].\u0026nbsp;\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"594\" height=\"225\"\u003e\u003c/p\u003e\n\u003cp\u003eThe outcome of the sampling error estimation demonstrates a statistically significant level with a 95% confidence level.\u003c/p\u003e\n\u003cp\u003eTable 1 presents the respondents\u0026apos; gender, age, and duration of stay in their current location. Among the respondents, 59% were male and 41% were female. Those aged between 35 and 39 years constituted 22% of the total, while individuals over 50 years of age accounted for 12%. The majority of the participants were highly educated individuals who had lived in the area for at least two years, ensuring that they had sufficient experience and knowledge of the area to provide accurate and reliable information. Accordingly, the highest percentages of respondents were those who had lived in the neighborhood for 2\u0026ndash;5 years and those with college educations (59% and 42%, respectively).\u003c/p\u003e\n\u003cp\u003eSubsequently, statistical analysis, specifically descriptive statistics, was employed to investigate complex relationships among various personal, geographical, and travel factors collected by the survey. This approach was instrumental in enhancing the understanding of the sample data and in identifying some simple relationships between variables.\u003c/p\u003e\n\u003cp\u003eAfter this preliminary analysis, the K-means clustering method was implemented to segment individuals based on their destination choice behaviors as indicated by our gathered socio-demographic and travel behavior indicators. This step was crucial for integrating these behaviors into our simulation framework. In our study, K-means clustering was employed to categorize respondents into distinct groups according to their survey responses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1:\u003c/strong\u003e Socio-demographic characteristics of the respondents.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"55.73192239858906%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.335097001763668%\" valign=\"top\"\u003e\n \u003cp\u003eClassification\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.932980599647266%\" valign=\"top\"\u003e\n \u003cp\u003eDistribution\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"55.73192239858906%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGender\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.335097001763668%\" valign=\"top\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.932980599647266%\" valign=\"top\"\u003e\n \u003cp\u003e59%\u003c/p\u003e\n \u003cp\u003e41%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"55.73192239858906%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.335097001763668%\" valign=\"top\"\u003e\n \u003cp\u003e-30\u003c/p\u003e\n \u003cp\u003e30-34\u003c/p\u003e\n \u003cp\u003e35-39\u003c/p\u003e\n \u003cp\u003e40-44\u003c/p\u003e\n \u003cp\u003e45-49\u003c/p\u003e\n \u003cp\u003e+50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.932980599647266%\" valign=\"top\"\u003e\n \u003cp\u003e16%\u003c/p\u003e\n \u003cp\u003e14%\u003c/p\u003e\n \u003cp\u003e22%\u003c/p\u003e\n \u003cp\u003e19%\u003c/p\u003e\n \u003cp\u003e17%\u003c/p\u003e\n \u003cp\u003e12%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"55.73192239858906%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDuration of stay in the neighbourhood.\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.335097001763668%\" valign=\"top\"\u003e\n \u003cp\u003e2-5\u003c/p\u003e\n \u003cp\u003e4-7\u003c/p\u003e\n \u003cp\u003e+7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.932980599647266%\" valign=\"top\"\u003e\n \u003cp\u003e59%\u003c/p\u003e\n \u003cp\u003e28%\u003c/p\u003e\n \u003cp\u003e23%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"55.73192239858906%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eEducation level\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.335097001763668%\" valign=\"top\"\u003e\n \u003cp\u003eNo schooling\u003c/p\u003e\n \u003cp\u003ePrimary education\u003c/p\u003e\n \u003cp\u003eSecondary education\u003c/p\u003e\n \u003cp\u003eCollege education\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"13.932980599647266%\" valign=\"top\"\u003e\n \u003cp\u003e0%\u003c/p\u003e\n \u003cp\u003e19%\u003c/p\u003e\n \u003cp\u003e39%\u003c/p\u003e\n \u003cp\u003e42%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe clustering applied in this context allowed us to recognize underlying factors influencing travel behavior. From this, simple decision rules were derived and incorporated into the simulation platform. This step enabled the training of agents to accurately represent individual travel behavior, particularly in terms of destination choices within the context of activity space [40].\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e2.2.2. Built-environment characteristics\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eThe influence of the built environment on travel behavior is a key factor that impacts various aspects of individuals\u0026apos; travel behaviors and decision-making processes [27]. This study focuses on examining the three primary land use factors\u0026mdash;Density, Diversity, and Distance\u0026mdash;that play a significant role in shaping activity-travel behaviors, collectively referred to as the \u0026apos;3Ds measures:\u0026nbsp;\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cstrong\u003eDensity\u003c/strong\u003e refers to the concentration of land uses within a specific area. A higher density is typically associated with the close proximity of similar land uses, extensive walkability, and convenient access to similar facilities [27, 29].\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eDiversity\u003c/strong\u003e focuses on the variety of land uses present in a given area, with greater diversity correlating to a higher number of destinations and improved access to diverse urban amenities [41].\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eDistance\u003c/strong\u003e relates to the travel distance between an origin and a destination, with individuals generally preferring shorter distances to access urban activity areas [41].\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eIn this study, a vector land use map was utilized to represent 3D features in GIS, providing enhanced precision in recording, displaying, and analysing spatial data compared to traditional statistical methods\u0026nbsp;[42]. The land use map encompasses various amenities such as shopping centers, educational institutions, recreational venues, government-banking institutions, and medical facilities. It was initially surveyed and mapped by the Zanjan municipality during the summer of 2019 and revised in the summer of 2020 using OpenStreetMap (OSM) and on-site observations.\u003c/p\u003e\n\u003cp\u003eIn this research, the Focal Statistics tool available in ArcGIS Spatial Analyst was employed to compute land use diversity. This tool conducts a neighborhood analysis to generate a new raster, where the value of each cell is determined by the values within a specified surrounding area. The variety function in Focal Statistics calculates the count of unique values in each neighborhood, offering a useful metric for assessing the diversity of land use types \u0026nbsp; [43]. Furthermore, a kernel density estimation technique was utilized to determine the densities of each land use categories. Kernel density approach is a non-parametric method to estimate density from point-based or line-based data, enabling a quantitative evaluation of the distribution of various land use types across the study area [42]. Additionally, a distance function was integrated for the simulation\u0026apos;s individual agents, mandating their visits to the nearest TZs. Following the preparation of layers, the Zonal Statistics function (ZSF) was employed to compute the 3D values for each TZ. ZSF is a raster dataset analysis tool that calculates diverse statistics within specified zones, such as TZs [43]. The ZSF output is associated with the corresponding 3D values of all TZs, providing a comprehensive analysis of how the built environment influences travel behavior.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e2.2.3. Transportation Network\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eTo develop a realistic simulation of commuting patterns\u0026nbsp;towards\u0026nbsp;activity spaces,\u0026nbsp;a variety of network attributes were integrated\u0026nbsp;[11, 27].\u0026nbsp;This dataset included elements such as intersections, traffic lights,\u0026nbsp;pedestrian crossings, speed limits, route widths, and lengths,\u0026nbsp;all visualized on a GIS map.\u0026nbsp;These factors were updated by the Zanjan Traffic Department during May and June 2020. An initial\u0026nbsp;analysis\u0026nbsp;of the network was\u0026nbsp;conducted\u0026nbsp;using ArcGIS 10.6,\u0026nbsp;and the results were saved as supplementary traffic datasets. The\u0026nbsp;amalgamated\u0026nbsp;data\u0026nbsp;led to the development of a\u0026nbsp;Transportation Network Map (TNM).\u0026nbsp;Furthermore,\u0026nbsp;information regarding pedestrian crossings, traffic lights, and\u0026nbsp;intersections was consolidated into\u0026nbsp;a single point map,\u0026nbsp;providing\u0026nbsp;a detailed and\u0026nbsp;accurate depiction\u0026nbsp;of the city\u0026apos;s transportation\u0026nbsp;infrastructure.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e2.2.4. Synthetic population\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eTo determine the population size for a travel model, one can either integrate the full population census of the origin into the model or rely solely on the number of travelers obtained from relevant travel statistics\u0026nbsp;[44]. In our research, the modeled population, consisting of daily Travel Generation Data (TGD) within each traffic zone, was used to feed the simulation. The TGD covered almost the entire traveler population and effectively overcame the sample self-selection issue, which is a challenge in travel demand and activity space literature\u0026nbsp;[28]. However, travel behavior demand for destinations and time dimensions was lacking in TGD, which needed to be enriched by a travel behavior survey (Section 2.2.2. Travel Behavior Survey to Extract Decision Rules).\u003c/p\u003e\n\u003cp\u003eThe data, comprising text files and a Traffic Zone Map (TZM), logs raw counts of all daily inbound and outbound journeys for these zones. This data was collated by Tarhe-Haftom Consulting Engineers in collaboration with the traffic department of Zanjan, spanning from March 2 to June 28, 2018. As per this study, Zanjan was segmented into 150 neighborhoods. Certain personal traits were mapped to neighborhood attributes, including a unique code for every Traffic Zone (TZ), the total number of travelers per neighborhood on a typical workday, car ownership, and so forth.\u003c/p\u003e\n\u003ch3\u003e\u003cstrong\u003e2.2.5. Validation data\u003c/strong\u003e\u003c/h3\u003e\n\u003cp\u003eThe validation of the model transfer process is a critical step in determining its predictive capabilities and ensuring its applicability in regions similar to the one it represents\u0026nbsp;[32]. A crucial element for validating any traffic simulation is a reliable count of traffic in origin-destination flows\u0026nbsp;[45]. In this study, a combination of\u0026nbsp;emerging data sources (EDSs)\u0026nbsp;and manual counts was used to gather traffic data across multiple Traffic Zones (TZs). EDSs, which include video vehicle detection and inductive loop counters, utilize data and communication technologies to capture real-time traffic information from various vehicles and transportation networks\u0026nbsp;[46].\u003c/p\u003e\n\u003cp\u003eHowever, EDSs in Zanjan did not cover all TZs. Of the 46 EDSs installed throughout the city, only 31 were applicable for the validation process. Therefore, manual counting was also conducted to account for traffic in TZs not covered by EDSs.\u003c/p\u003e\n\u003cp\u003eThe selection of TZs for validation was based on several criteria, including the location of EDSs, the spatial distribution of TZs across the city, and the inclusion of diverse types of TZs. Consequently, three major activity spaces (TZs 1, 3, and 8) were selected along with other randomly chosen TZs. Additionally, two peak traffic periods were chosen for data collection\u0026mdash;morning and afternoon\u0026mdash;due to their high traffic flow and congestion (9-10 am, 11-12 am, 18-19 pm, and 20-21 pm).\u003c/p\u003e\n\u003cp\u003eFor comprehensive traffic count data, observations were made on three different days: two weekdays and one holiday. Data collection occurred on April 20, 22, and 25, and the average of these counts was used as the observed data for validation\u003c/p\u003e"},{"header":"3. Methodology","content":"\u003cp\u003eThe agent-based model serves as the foundation of the proposed activity-based model, with the goal of replicating the daily movements of passengers as agents traveling to various activity destinations. The flowchart illustrating the agent-based model can be found in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The study focused on travel to non-work activity areas and incorporated four types of information into the agent-based model: travel behavior attributes (such as preferred destinations, trip purposes, departure times, and duration of stay), built-environment attributes (including land-use density-diversity and distance from origin to destination), network characteristics (such as intersections, traffic lights, route features), and travel supply data. Each time step in the agent-based model represents an hour from 7:00 to 24:00, excluding data beyond this range due to the infrequency of night trips. The model simulates 17 time steps, equivalent to 17 hours, estimating each agent's next activity destination during every time step. The model framework employs rule-based models to simulate the interaction between spatial environments and human behavior in making trips to activity spaces across different time periods. The study leveraged the capabilities of AnyLogic, a versatile toolkit known for simulating real-world phenomena using discrete events, system dynamics, and agent-based modeling techniques. AnyLogic's compatibility with open-source GIS environments and its ability to schedule individual behaviors within the simulation model using a Java script platform were particularly valuable [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. This section outlines the key aspects of the simulation model, drawing inspiration from the approach utilized by Inturri, et al. [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Proposed ABM components\u003c/h2\u003e \u003cp\u003e \u003cb\u003eEnvironment\u003c/b\u003e: The environment in the developed Agent-Based Model (ABM) represents the space where agents engage in activities and interact. It comprises two GIS datasets: the transport network and the traffic zone map. The network consists of a fixed route and optional routes, exported from GIS as the Transportation Network Map (TNM), composed of network nodes and links, stop nodes, and diversion nodes (allowing switching between fixed and optional routes). The TNM includes a variety of attributes such as speed limits, route widths, junctions, one-way routes, traffic lights, and crosswalks.\u003c/p\u003e \u003cp\u003eAnother key component of the environment is the Traffic Zone Map (TZM), displaying 150 neighborhoods in Zanjan. In addition to tracking daily travels, the TZM also records the number of visits to each traffic zone.\u003c/p\u003e \u003cp\u003e \u003cb\u003eAgents\u003c/b\u003e: Agents in this model are autonomous and adaptive individuals, optimizing their utility by learning and adapting new behaviours based on experience and training. Each traveler agent aims to find an activity space and reach it. Agents have an overall identical profile (e.g., age, gender, occupation, and residential location) and a daily trip plan including trip origins, destinations, purposes, and departure times. They are grouped into households, sharing characteristics like the available number of vehicles. Trip requests of passengers are stochastically generated according to the demand model in time steps. The status of a traveler agent, regarding commuting, includes four dynamic states: ready for a trip, in travel, at destination, and returning to the beginning or starting another journey from the destination. Agents move from node to node, deciding to accept or reject opportunities at the new node based on specific rules. They are introduced into the simulation by an event and removed either when their visit duration ends or prematurely in cases where the simulated agent population exceeds the expected number. During iterations, agents acquire new information (e.g., traffic conditions) through interactions, which is used to update their perceived utility for subsequent iterations based on a reinforcement learning rule.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Knowledge learning process\u003c/h2\u003e \u003cp\u003eUsers\u0026rsquo; trip requests are generated from an origin (O) zone to a destination (D) zone with a negative exponential distribution, based on a gravitationally distributed probability.\u003c/p\u003e \u003cp\u003eGiven a set of n zones, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{TR}_{ij}^{t}\\)\u003c/span\u003e\u003c/span\u003eis the probability that a trip with origin \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e has destination in the zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e, in time \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t\\)\u003c/span\u003e\u003c/span\u003e calculated with Eq.\u0026nbsp;\u003cspan refid=\"Equ2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{TR}_{i}\\)\u003c/span\u003e\u003c/span\u003e is the generation trip rate of the zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e, proportional to an average trip rate achieved by travel generation data (TGD), and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{P}_{ij}^{t}\\)\u003c/span\u003e\u003c/span\u003e is the probability that a trip with origin \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e has destination in the zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e in time t calculated with Eq.\u0026nbsp;\u003cspan refid=\"Equ3\" class=\"InternalRef\"\u003e3\u003c/span\u003e:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{TR}_{ij}^{t}={TR}_{i}\\bullet\\:{P}_{ij}^{t}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:{P}_{ij}^{t}=\\frac{{S}_{ij}^{t}\\bullet\\:{\\left({d}_{ij}\\bullet\\:{l}_{j}\\right)}^{a}\\bullet\\:{e}^{{\\beta\\:d}_{ij}}}{{\\sum\\:}_{k=1}^{n}{S}_{ij}^{t}\\bullet\\:{\\left({d}_{kj}\\bullet\\:{l}_{j}\\right)}^{a}{\\bullet\\:e}^{{\\beta\\:d}_{kj}}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{S}_{ij}^{t}\\)\u003c/span\u003e \u003c/span\u003e represents proportion of individuals who prefer to travel from zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e to zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e in time \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t\\)\u003c/span\u003e\u003c/span\u003e, as estimated by the travel behaviour survey results. The distance from \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e to zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e denoted as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{d}_{ij}\\)\u003c/span\u003e\u003c/span\u003e, is also calculated. Additionally, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{l}_{j}\\)\u003c/span\u003e\u003c/span\u003e is the coefficient of land use impacts including land use density and diversity in traveling individuals to zone \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:j\\)\u003c/span\u003e\u003c/span\u003e. Land use diversity and density are quantified using the focal statistics function and kernel density, respectively, operating in ArcGIS for each traffic zone. Both of these analytical tools require the ArcGIS Spatial Analyst extension to function [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. α and β are the parameters of the decay function:\u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e\n$$\\:{f\\:\\left(d\\right)=\\:\\left(d\\bullet\\:l\\right)}^{a}{e}^{{\\beta\\:d}_{ij}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Path learning rule\u003c/h2\u003e \u003cp\u003eA new setup was implemented to replicate this particular process category, emphasizing adaptable transit services within different system setups. This simulation investigates the interaction between vehicles navigating a flexible road network and users commuting from their origins to their destination TZs. The network configurations are built upon two key components: arcs and nodes. Within this framework, nodes are identified as intersections, traffic lights, and pedestrian crossings. These nodes are fine-tuned using a delay function that allocates specific timings to each node, drawing from information obtained from police operations in Zanjan to ensure a faithful representation of actual traffic patterns\u003c/p\u003e \u003cp\u003eThe arc \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{TZ}_{1}-{TZ}_{2}\\)\u003c/span\u003e\u003c/span\u003e connects origin \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{TZ}_{1}\\)\u003c/span\u003e\u003c/span\u003e to destination \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{TZ}_{2}\\)\u003c/span\u003e\u003c/span\u003e directly, without any intermediate TZs.. It is characterized by factors such as capacity (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:c\\)\u003c/span\u003e\u003c/span\u003e), length (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:l\\)\u003c/span\u003e\u003c/span\u003e), free-flow speed (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{v}_{f}\\)\u003c/span\u003e\u003c/span\u003e), flow (q), and additional costs (O) like tolls. The cost of the arc (c) is determined by a function that incorporates these five factors, as described by Zhang, et al. [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]:\u003cdiv id=\"Equ5\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ5\" name=\"EquationSource\"\u003e\n$$\\:c=g\\left(c\u0026sbquo;\\:l\u0026sbquo;\\:{v}_{f}\u0026sbquo;\\:q\u0026sbquo;\\:o\\right)$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe function \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:g\\left(O\\right)\\)\u003c/span\u003e\u003c/span\u003e represents the performance of the arc. In the current model, arc capacities are assumed to be unlimited, resulting in a constant arc cost and no consideration for congestion effects. Travelers learn about the travel cost of links on their route, while TZs gather information on the shortest path from themselves to all other nodes visited by travelers within that TZ. When travelers reach a new TZ, they compare the travel costs from the current TZ to each TZ along their route. Subsequently, both travelers and TZs update their information to reflect the shortest path.\u003c/p\u003e \u003cp\u003eThe variables used were updated by the Zanjan Traffic Department in May and June 2018. An initial network analysis was performed using ArcGIS 10.6, and the results were saved as additional traffic datasets.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Pattern recognition and clustering methods\u003c/h2\u003e \u003cp\u003eIn this study, we developed a space-time GIS approach and utilized the K-means clustering technique to facilitate pattern recognition analysis and visualize the dynamics of activity spaces resulting from ABM. The GIS analysis is centered on a space-time GIS framework with temporal dynamic segmentation [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e], and has been implemented in ArcScene, which serves as the 3D viewer within ArcGIS. Referred to as Spatial Cluster Detection (SCD), this approach integrates both spatial and temporal dimensions of activity space shifts. The utilization of the 3D method presents several benefits in uncovering the structure of urban activity spaces. It enables the identification of spatial characteristics such as area, shape, density, as well as spatial relationships like proximity. Additionally, it visualizes the spatial-temporal hierarchy of multi-local activity spaces. Furthermore, the K-means clustering method is efficient in organizing urban areas based on human activity data into measurable ordinal categories, such as main activity areas and secondary activity areas [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. This widely-used unsupervised machine learning technique identifies clusters within the data based on similarities and differences among data points [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. In our study, we employed the k-means clustering method as implemented in Excel.3.5. Model Calibration and primary implementation\u003c/p\u003e \u003cp\u003eModel calibration is crucial for ensuring that simulations accurately reflect real-world scenarios and produce reliable results. During the calibration phase, the model was refined to tune agent behaviours within the ABM and increase its complexity. The first iteration of the model featured homogeneous agents with individual agents uniformly incorporating basic rules such as variations in trip rates for Traffic Zones (TZs), modifications of trip hours, and adjustments to a simplified environment representing only five zones. These modifications were informed by the initial simulation results and the expert knowledge of the research team.\u003c/p\u003e \u003cp\u003eDuring the initial simulation, a significant discrepancy was observed between the actual data and the model's predictions. This gap indicated that certain aspects of the simulation had been overlooked, underscoring the need for further refinements to enhance the simulation's accuracy. Consequently, the model underwent multiple iterations with varied parameters to achieve more realistic outcomes. The refined simulation sets, which demonstrated the highest accuracy, were selected for the final configuration of the model. Subsequent to these adjustments, the model was scaled up to simulate Zanjan, Iran, ensuring that it consistently reflected the behaviours identified during the calibration phase.\u003c/p\u003e \u003cp\u003eThe model experiments were conducted after the initial settings had been established. These included increasing the number of TZ visits for individual agents and setting a preference for the destination TZ.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Results","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Model Validation\u003c/h2\u003e \u003cp\u003eThe model's performance was evaluated by comparing real and simulated data, using statistical methods of R-square. The validation yielded a percentage error of 15.86% and an R-square value of 0.90, indicating a close alignment between the simulated estimates and observed values. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e presents the detailed validation results for the selected TZs and hours. Notably, the validation process revealed that the model is overall consistent with real observation in popular city centers like TZ1, TZ3, and TZ8, while less frequented TZs like TZ50 and TZ85 showed lower accuracy.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eThe validation results by R-square\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eTraffic Zones\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eRMSE/Time districts\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eRMSE %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9\u0026ndash;10 AM\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11\u0026ndash;12 AM\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18\u0026ndash;19 PM\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e19\u0026ndash;20 PM\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e9.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e9.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8965\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e13.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e8.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e10.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e9.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9923\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9471\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e120.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e30.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e17.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8994\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e40.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e25.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.7819\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e20.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e20.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e17.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e15.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8629\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e19.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e9.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8376\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e50.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e15.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e18.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e17.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e18.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.8933\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZ140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e66.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e33.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e33.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e66.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e41.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9048\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e16.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e15.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eSpecifically, TZ3 exhibited the highest prediction accuracy with an average percentage error of 9.3%, while TZ4 had the lowest accuracy with a 50% error rate. Additionally, R-square results corroborated the findings from the percentage error method, indicating a strong correlation between the volume of traveler attraction in TZs and the model's accuracy. This pattern aligns with the findings of previous studies by Zhong, et al. [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e] and Apronti, et al. [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Identifying activity-based travel distribution\u003c/h2\u003e \u003cp\u003eThis section presents the results derived from the integrated models. The practical application of the model was demonstrated through a case study conducted in Zanjan, Iran. The simulation encompassed 150 traffic zones over a span of 17 hours, excluding the time period from 12 AM to 7 AM, resulting in the simulation of 66743 tours.\u003c/p\u003e \u003cp\u003eThe primary result pertains to the total number of daily travels within the activity space, serving to elucidate the general mobility patterns within the Zanjan and evaluate the predominant activity spaces within the TZs. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e displays the detailed results of the model regarding a critical mobility indicator, showing the relative impact of each traffic zone on the daily arrival of visitors. The analysis reveals that the distribution of individual activity spaces is predominantly concentrated within a limited number of traffic zones. Notably, TZ 1, situated in the Central Business District (CBD) and constituting a mere 0.14 percent of the city's total area (approximately 10 hectares), attracts around 11 percent of all internal trips. This underscores a monocentric travel distribution pattern, with TZ 1 emerging as the primary activity space. Moreover, a discernible contrast is observed between TZ 1 and its adjacent zones in terms of travel attraction. For instance, TZ 8, serving as a secondary activity space linked to the primary hub, garners nearly one-third of the total trips compared to TZ 1. This disparity underscores the concentrated nature of travel patterns within the CBD.\u003c/p\u003e \u003cp\u003eThe model's findings further delineate a distinct division in travel attraction among traffic zones. While certain key activity spaces are clustered within the CBD, such as TZ 1, TZ 3, and TZ 8, other activity nodes exhibit a more dispersed distribution, resulting in a relative distribution of trips across various zones.\u003c/p\u003e \u003cp\u003eOverall, the observed phenomenon aligns with our expectations and underscores the nuanced dynamics of travel behavior within the Zanjan region.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Clustering activity-spaces based on travel volumes\u003c/h2\u003e \u003cp\u003eThe study utilized travel diaries to analyse the spatial dispersion of activity locations. A k-means clustering method was employed to identify key activity destinations. The optimal number of clusters was determined using the elbow method, resulting in the identification of four clusters that explain 94.65% of the variance. The results of the clustering analysis are visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, showcasing the grouping of activity spaces based on daily travels (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea), and sub-clusters of the second cluster (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). While spatial information was not directly considered in the clustering algorithm, the results provide insights into the geographical distribution of activity spaces. In this context, the horizontal axis signifies the traffic zones, indicating that points in close proximity to each other correspond to nearby traffic zones in reality.\u003c/p\u003e \u003cp\u003eThe cluster analysis reveals similarities within each cluster and differences among clusters. Cluster 1 represents the main activity hub within the CBD (traffic zone 1), drawing a substantial volume of trips. Statistical analysis revealed notable distinctions between cluster 1 and the remaining clusters in terms of attracted travels. This particular zone, previously highlighted, accounted for 11% of all trip attractions. Cluster 2 consists of second-order activity spaces, with distinct patterns and locations relative to the main activity space; attracting approximately 30% of activity-based travels across 12 traffic zones. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, this cluster encompasses three sub-clusters characterized by internally consistent activity patterns, predominantly situated on the outskirts of the main activity space. Additionally, two other secondary activity sites are dispersed throughout the city, including a distinct official-marketing area adjacent to the first sub-cluster and newly developed neighborhoods in the suburban regions. The primary distinctions among the three sub-groups are their proximity to the main activity space and the total number of trips attracted to each subgroup. Notably, the trendline depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb indicates a decrease in trip volume to the urban activity center as the distance from the main activity space increases. Cluster 3 primarily includes local activity spaces, encompassing 43 traffic zones and approximately 37% of activity-based trips. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, the clustering method reveals that the third cluster exhibits a relatively regular spatial distribution pattern across the traffic zones. Cluster 4 comprises entirely local activity spaces distributed across the city, covering 89 traffic zones and approximately 21% of activity-based trips. The spatial distribution of these activity spaces mirrors that of the third activity space, extending across all traffic zones within the city.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Visualising spatio-temporal shifts of activity spaces\u003c/h2\u003e \u003cp\u003eFour distinct levels of activity spaces were identified through k-means cluster analysis, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea. The focus of this paper does not extend to analysing and visualizing the spatial-temporal dynamics of cluster 3 and cluster 4, as these clusters do not play a significant role in the city's overall traffic, particularly in activity-based travels. Therefore, this section primarily concentrates on analysing the time slots of the city's main hubs, including the main activity space (cluster 1) and secondary activity spaces (cluster 2). Hourly simulation results from 7:00 to 24:00 for the main activity space are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, while each sub-cluster of secondary activity spaces is illustrated separately in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb (first sub-cluster), Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec (second sub-cluster), and Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed (third sub-cluster).\u003c/p\u003e \u003cp\u003eThe paper additionally incorporates a Spatial Cluster Detection (SCD) technique that employs 3D maps (See Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e) to provide a more precise illustration of the congestion impact of activity travels during peak hours, as opposed to the traditional traffic volume functions typically utilized in statistical studies. Each hour is depicted through a unique 3D map, with varying heights indicating the levels of activity-based travel volumes. To emphasize secondary activity areas, the main urban activity zone was excluded from the maps in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb.\u003c/p\u003e \u003cp\u003eAnalysis of the hourly data generated by the simulation model reveals a notable fluctuation in the volume of trips to the main activity area at different times of the day (see Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). The majority of trips occur during the evening peak and the period spanning from the morning peak to the midday peak. As anticipated, the evening rush hour experiences heavy congestion, leading to oversaturation in several time slots. Interestingly, there is a significant decline in activity between 14:00 and 16:00, indicating a lull in activity-based travel to the urban main activity space, reminiscent of the quiet hours observed between midnight and 7:00. It is important to highlight that many activities within this category do not conclude at 21:00. In fact, activities within the traffic zone typically end around 22:00 on average, and they may even extend as late as midnight, which coincides with the closing time of the retail stores and other facilities in the CBD.\u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb, the first sub-cluster of second-order activity spaces are shown to be located near the main activity space. There is a significant variation in activity-based travel volume across the selected TZs throughout the day. The durations of these trips are generally consistent with those in the main activity space. Particularly, there is a significant increase in traffic numbers during the evening hours, reflecting the pattern observed in the main activity space. The fluctuations among TZs within this sub-cluster exhibit similar variations, especially from midday to midnight. The peak hours of this sub-cluster typically occur from around 18:30 to 20:30, with before noon peak hours ranging from approximately 8:30 to 12:30. Before noon peak times extend across a larger area, indicating a larger variance between the activity spaces. Additionally, as depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, there is a period of low peak activity from approximately 13:30 to 14:30, showing a high volume of travels returning from the CBD to their origin. Notably, there is a period of weaker activity from 14:30 to 17:30, characterized as blackout hours.\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec depict the second sub-cluster situated at a relatively greater distance from the main activity space. In contrast to the first sub-cluster, the second sub-cluster exhibits distinct space-time paths with unique activity and travel patterns. Within this sub-cluster, the TZs centrality patterns vary significantly across different hours, yet remain relatively stable during midday, particularly during the rush hour from 13:00 to 14:00. Notably, there is a decrease in activity observed between 15:00 and 17:00, suggesting a period of reduced activity akin to a downtime or resting phase in activity-based travels to the sub-cluster activity space.\u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed, situated in the inner suburb, similar patterns to those in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec can be observed. However, there is an extension in the sleep time during an additional time slot, particularly from 12:00 to 13:00. The rush hour remains relatively stable in the midday, mirroring the trends seen in the second sub-cluster. Besides, it shows an overall increase in variation compared to the previous sub-clusters.\u003c/p\u003e \u003cp\u003eIn general, the spatial pattern of activity-based travels indicates that travel volumes are influenced by the distance from the city center. Specifically, travel volumes are higher in cluster 1 and sub-cluster 1, which includes the CBD and surrounding TZs. Conversely, travel volumes are lower in the second sub-cluster, which consists of TZs located between central zones and suburbs, and in the third sub-cluster situated in the inner suburb. In the analysis of temporal patterns, it is observed that cluster 1 and sub-cluster 2 exhibit similar variations in activity travel attraction during evening rush hours. On the other hand, sub-cluster 2 and 3 attract the highest levels of activity travels during midday. However, there is a relative variation in travel patterns during other times of the day.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Discussion and Conclusion","content":"\u003cp\u003eMotivated by initial discussions from [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], our study aimed to investigate the dynamic nature of daily activity spaces with a focus on detailed time dimensions. We introduced a clear conceptual and operational definition of multiple activity spaces, utilizing the flexibility of ABM and a spatial modeling approach from GIS to measure their dynamics and classify them based on individual activity spaces. In addition to exploring the influence of personal preferences on activity-based mobility, we examined how built-environmental characteristics may influence individuals' decisions to travel beyond their immediate neighborhoods.\u003c/p\u003e \u003cp\u003eThe results obtained from the space\u0026ndash;time path based clustering method indicate that activity places may cluster at various hourly intervals within an individual's activity space, with a higher concentration of frequently visited locations. This aligns with previous research that has examined the formation of activity clusters around daily anchor points, showing increased activity intensity in the city's main activity space [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Our analysis demonstrates that the central activity space acts as a focal point for activity-based travel, with a greater number of trips terminating there. This concentration of activities in the CBD influences travel behavior, particularly in terms of activity-based travel patterns. The number of trips decreases as distance from the main activity space increases, indicating a spatial decay effect on travel attraction, which is consistent with some studies [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The reason behind the formation of this travel behavior pattern is evident in Zanjan, where high-quality public resources such as office buildings and upscale shopping malls are predominantly located in the main activity space. This, along with measures of land-use density and diversity in the built environment, influences intra-urban mobility patterns.\u003c/p\u003e \u003cp\u003eThe primary activity space and the initial cluster of second-order activity spaces demonstrate similar temporal patterns, while the second and third sub-clusters exhibit distinct temporal patterns. A notable amount of travel to these zones occurs during midday hours, primarily due to time constraints within office buildings. These activity spaces play a vital role in alleviating traffic congestion in the city center and are integral to urban traffic management. These sub-clusters consist of a mix of public services such as schools, commercial facilities, and office buildings. They also offer various benefits associated with urban design features, such as high levels of accessibility, ample parking space, and close proximity to residential neighborhoods. By implementing minor adjustments to the built environment and influencing travel behavior, these areas can function as alternative destinations to redirect a significant portion of travel away from the main activity hub, in particular during evening rush hours. This shift in travel patterns not only improves accessibility and effectively redistributes traffic flow in Zanjan but also contributes to the decentralization of urban activity spaces, transitioning from a monocentric to a more diverse urban layout.\u003c/p\u003e \u003cp\u003eThis study holds significant implications for the effective management of daily urban traffic and the allocation of resources to underutilized areas in Zanjan. Firstly, our model's findings offer insightful descriptions of the interactions between citizens and the urban built environment. As a result, this information can be utilized by city managers and business leaders to tailor services and locating new infrastructure in response to the evolving profile of people in the activity spaces.\u003c/p\u003e \u003cp\u003eSecondly, the analysis of spatio-temporal activity patterns serves as a valuable tool for assessing the utilization of urban spaces, understanding the city's daily movement patterns, evaluating traffic conditions, and assessing the impact of design interventions on changes in travel behavior. This comprehensive approach enables a deeper understanding of urban dynamics and facilitates informed decision-making in urban planning and traffic management initiatives.\u003c/p\u003e \u003cp\u003eOur findings support existing theoretical frameworks suggesting that travel destination choices are not solely based on individual decisions but are also influenced by the physical environment at the destination [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan additionalcitationids=\"CR27\" citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. One key behavioral motivation for visiting a destination is people's interest in the activities available. When individuals have a common interest in particular type of activity, a destination with diverse and densely packed functions becomes appealing as it caters to a variety of activity demands. Existing literature has confirmed that destinations offering diverse and dense services tend to attract visitors [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Our findings suggest to policymakers that areas with diverse and densely packed spatial functions, in close proximity to residential neighborhoods, can be effectively leveraged and potentially utilized in initiatives aimed at altering travel behavior and promoting sustainable mobility. Additionally, the incorporation of built-environment attributes into the ABM process represents a significant contribution of our study to the field of travel demand and destination choice modeling research. This emphasizes the importance of integrating GIS and ABM in modeling travel behavior.\u003c/p\u003e \u003cp\u003eFurthermore, we utilized real-world data to verify the accuracy of the model in simulating the activity destination choice model in Zanjan; The simulation results were then compared with the actual travel counts recorded during peak hours on various weekdays in selected TZs. Our study emphasizes the significance of incorporating real-world data for model validation, consistent with existing studies in in the field of travel behavior [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. We highlight the importance of traffic count data, such as EDSs, in validating urban movements within ABM and traffic simulation, enhancing the realism and reliability of model performance compared to previous studies. This inclusion enhances the realism and reliability of the model's performance. The use of EDSs not only validates the model's accuracy but also serves as a more comprehensive and readily available database for pattern recognition of urban mobility and analyzing travel behavior, particularly in developing countries.\u003c/p\u003e \u003cp\u003eThe developed model framework was tailored for a medium-sized Iranian city and its travel-to-non-work area, but it holds potential for adaptation to urban transportation modeling in densely populated cities worldwide. Prior to deploying the model in other regions, it is essential to calibrate the prototype model with local time-use data and conduct thorough verification and validation of key system elements such as activity engagement and trip generation.\u003c/p\u003e \u003cp\u003eFurthermore, the current model abstracted implementations of agent activities to establish a generalized framework of human spatial behavior based on typical behaviors in activity spaces. Future research will focus on identifying additional spatial behaviors and expanding existing activity types to broaden the model's applicability to various spatial configurations and environments.\u003c/p\u003e \u003cp\u003eHowever, there are limitations associated with the data collection method and the operationalization of urban activity spaces that could be addressed in future studies. Integrating additional data sources like cell phone data using simulation methods can enhance the model's accuracy and aid in validation, although such data may not be readily available in developing countries like Iran due to privacy constraints. Furthermore, the model's activity-based travel structure, comprising sub-models of specific human tasks, could benefit from independent updates by incorporating more advanced approaches, such as integrating informed path-finding or movement strategies at the individual agent level.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eM.Azari conceived of the presented idea.M.Hatami and Mr. Hosseini developed the theory and performed the computations.M.Azari verified the analytical methods. S.Moridpour encouraged A.B. to investigate [a specific aspect] andsupervised the findings of this work. All authors discussed the results and contributed to the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGolledge, R.G.: Spatial behavior: A geographic perspective. Guilford Press (1997)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatthews, S.A., Yang, T.-C.: Spatial polygamy and contextual exposures (spaces) promoting activity space approaches in research on place and health. Am. Behav. Sci. \u003cb\u003e57\u003c/b\u003e(8), 1057\u0026ndash;1081 (2013)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHasanzadeh, K., Kytt\u0026auml;, M., Lilius, J., Ramezani, S., Rinne, T.: Centricity and multi-locality of activity spaces: The varying ways young and old adults use neighborhoods and extra-neighborhood spaces in Helsinki Metropolitan Area. Cities. \u003cb\u003e110\u003c/b\u003e, 103062 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorton, F.E., Reynolds, D.R.: Effects of urban spatial structure on individual behavior. Econ. Geogr. \u003cb\u003e47\u003c/b\u003e(1), 36\u0026ndash;48 (1971)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, R., Tong, D.: Constructing human activity spaces: A new approach incorporating complex urban activity-travel. J. Transp. Geogr. \u003cb\u003e56\u003c/b\u003e, 23\u0026ndash;35 (2016)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHasanzadeh, K.: Exploring centricity of activity spaces: From measurement to the identification of personal and environmental factors. Travel Behav. Soc. \u003cb\u003e14\u003c/b\u003e, 57\u0026ndash;65 (2019)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Dijk, J., Krygsman, S.: Analyzing travel behavior by using GPS-Based activity spaces and opportunity indicators. J. Urban Technol. \u003cb\u003e25\u003c/b\u003e(2), 105\u0026ndash;124 (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, F., Zhou, S., Lu, J., Zheng, Z.: \u003cem\u003eA behavioral explanation of the activity-space segregation: Individuals\u0026rsquo; preference of choosing an activity destination.\u003c/em\u003e Environment and Planning B: Urban Analytics and City Science, : p. 23998083241229110. (2024)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCagney, K.A., York Cornwell, E., Goldman, A.W., Cai, L.: Urban mobility and activity space. Ann. Rev. Sociol. \u003cb\u003e46\u003c/b\u003e, 623\u0026ndash;648 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJones, P.M.: \u003cem\u003eNew approaches to understanding travel behaviour: the human activity approach\u003c/em\u003e, in \u003cem\u003eBehavioural travel modelling\u003c/em\u003e, pp. 55\u0026ndash;80. Routledge (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFu, X., Wu, Y., Huang, D., Wu, J.: An activity-based model for transit network design and activity location planning in a three-party game framework. Transp. Res. E. \u003cb\u003e168\u003c/b\u003e, 102939 (2022)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerchoux, C., Chaix, B., Cummins, S., Kestens, Y.: Conceptualization and measurement of environmental exposure in epidemiology: accounting for activity space related to daily mobility. Health place. \u003cb\u003e21\u003c/b\u003e, 86\u0026ndash;93 (2013)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGalster, G., Sharkey, P.: \u003cem\u003eSpatial foundations of inequality: A conceptual model and empirical overview.\u003c/em\u003e RSF: The Russell Sage Foundation Journal of the Social Sciences, 3(2): pp. 1\u0026ndash;33. (2017)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, Y., Belyi, A., Bojic, I., Ratti, C.: Human mobility and socioeconomic status: Analysis of Singapore and Boston, vol. 72, pp. 51\u0026ndash;67. Computers, Environment and Urban Systems (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKwan, M.-P.: Beyond space (as we knew it): Toward temporally integrated geographies of segregation, health, and accessibility: Space\u0026ndash;time integration in geography and GIScience, vol. 103, pp. 1078\u0026ndash;1086. Annals of the Association of American Geographers (2013). 5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, Y., Wang, J., Kan, C.: Temporal variation in activity-space-based segregation: A case study of Beijing using location-based service data. J. Transp. Geogr. \u003cb\u003e98\u003c/b\u003e, 103239 (2022)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilm, S., Ahas, R., Mooses, V.: Are younger age groups less segregated? Measuring ethnic segregation in activity spaces using mobile phone data. J. Ethnic Migration Stud. \u003cb\u003e44\u003c/b\u003e(11), 1797\u0026ndash;1817 (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, Y., Liu, L., Wang, H.: A new perspective on the temporal pattern of human activities in cities: The case of Shanghai. Cities. \u003cb\u003e87\u003c/b\u003e, 196\u0026ndash;204 (2019)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHafezi, M.H., Daisy, N.S., Millward, H., Liu, L.: Framework for development of the Scheduler for Activities, Locations, and Travel (SALT) model. Transportmetrica A: Transp. Sci. \u003cb\u003e18\u003c/b\u003e(2), 248\u0026ndash;280 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBekhor, S., Dobler, C., Axhausen, K.W.: Integration of activity-based and agent-based models: case of Tel Aviv. Isr. Transp. Res. record. \u003cb\u003e2255\u003c/b\u003e(1), 38\u0026ndash;47 (2011)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHesam Hafezi, M., Sultana Daisy, N., Millward, H., Liu, L.: Framework for development of the Scheduler for Activities, Locations, and Travel (SALT) model. Transportmetrica A: Transp. Sci. \u003cb\u003e18\u003c/b\u003e(2), 248\u0026ndash;280 (2022)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, S., Yang, Y., Zhen, F., Lobsang, T., Li, Z.: Understanding the travel behaviors and activity patterns of the vulnerable population using smart card data: An activity space-based approach. J. Transp. Geogr. \u003cb\u003e90\u003c/b\u003e, 102938 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheliotis, K.: An agent-based model of public space use. Comput. Environ. Urban Syst. \u003cb\u003e81\u003c/b\u003e, 101476 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArentze, T.A., Timmermans, H.J.: A learning-based transportation oriented simulation system. Transp. Res. Part. B: Methodological. \u003cb\u003e38\u003c/b\u003e(7), 613\u0026ndash;633 (2004)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKagho, G.O., Balac, M., Axhausen, K.W.: Agent-Based Models in Transport Planning: Current State, Issues, and Expectations. Procedia Comput. Sci. \u003cb\u003e170\u003c/b\u003e, 726\u0026ndash;732 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhong, G., Yin, T., Zhang, J., He, S., Ran, B.: Characteristics analysis for travel behavior of transportation hub passengers using mobile phone data. Transportation. \u003cb\u003e46\u003c/b\u003e, 1713\u0026ndash;1736 (2019)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, L., Silva, E.A., Yang, Z.: Similar outcomes, different paths: Tracing the relationship between neighborhood-scale built environment and travel behavior using activity-based modelling. Cities. \u003cb\u003e110\u003c/b\u003e, 103061 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBastarianto, F.F., Hancock, T.O., Choudhury, C.F., Manley, E.: Agent-based models in urban transportation: review, challenges, and opportunities. Eur. Transp. Res. Rev. \u003cb\u003e15\u003c/b\u003e(1), 1\u0026ndash;20 (2023)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCervero, R., Kockelman, K.: Travel demand and the 3Ds: Density, diversity, and design. Transp. Res. part. D: Transp. Environ. \u003cb\u003e2\u003c/b\u003e(3), 199\u0026ndash;219 (1997)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGong, S., Dong, X., Wang, K., Lei, B., Jia, Z., Qin, J., Roadknight, C., Liu, Y., Cao, R.: Agent-based modelling with geographically weighted calibration for intra-urban activities simulation using taxi GPS trajectories. Int. J. Appl. Earth Obs. Geoinf. \u003cb\u003e122\u003c/b\u003e, 103368 (2023)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFranco, P., Johnston, R., McCormick, E.: Demand responsive transport: Generation of activity patterns from mobile phone network data to support the operation of new mobility services. Transp. Res. Part. A: Policy Pract. \u003cb\u003e131\u003c/b\u003e, 244\u0026ndash;266 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown, D.G., Riolo, R., Robinson, D.T., North, M., Rand, W.: Spatial process and data models: Toward integration of agent-based models and GIS. J. Geogr. Syst. \u003cb\u003e7\u003c/b\u003e(1), 25\u0026ndash;47 (2005)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, L., Levinson, D.: Agent-based approach to travel demand modeling: Exploratory analysis. Transp. Res. Rec. \u003cb\u003e1898\u003c/b\u003e(1), 28\u0026ndash;36 (2004)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGong, S., Cartlidge, J., Bai, R., Yue, Y., Li, Q., Qiu, G.: Geographical temporal huff model. calibration using taxi trajectory data GeoInformatica. \u003cb\u003e25\u003c/b\u003e, 485\u0026ndash;512 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eD\u0026rsquo;Exelle, B.: Representative Sample. In: Encyclopedia of Quality of Life and Well-Being Research, pp. 5511\u0026ndash;5513. Springer Netherlands: Dordrecht (2014). A.C. Michalos, Editor\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKulpa, T., Szarata, A.: Analysis of household survey sample size in trip modelling process. Transp. Res. Procedia. \u003cb\u003e14\u003c/b\u003e, 1753\u0026ndash;1761 (2016)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEgu, O., Bonnel, P.: How comparable are origin-destination matrices estimated from automatic fare collection, origin-destination surveys and household travel survey? An empirical investigation in Lyon. Transp. Res. Part. A: Policy Pract. \u003cb\u003e138\u003c/b\u003e, 267\u0026ndash;282 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor, M.A.: Voluntary travel behavior change programs in Australia: The carrot rather than the stick in travel demand management. Int. J. Sustainable Transp. \u003cb\u003e1\u003c/b\u003e(3), 173\u0026ndash;192 (2007)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuhachek, A., Iacobucci, D.: Alpha's standard error (ASE): an accurate and precise confidence interval estimate. J. Appl. Psychol. \u003cb\u003e89\u003c/b\u003e(5), 792 (2004)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCote, J.D.R., Diana, M.: Exploring the benefits of a traveller clustering approach based on multimodality attitudes and behaviours. Transp. Res. procedia. \u003cb\u003e25\u003c/b\u003e, 2552\u0026ndash;2565 (2017)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoarnet, M., Crane, R.: The influence of land use on travel behavior: specification and estimation strategies. Transp. Res. Part. A: Policy Pract. \u003cb\u003e35\u003c/b\u003e(9), 823\u0026ndash;845 (2001)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang, J., Gong, J., Li, W., Ibrahim, A.N.: A visualization-oriented 3D method for efficient computation of urban solar radiation based on 3D\u0026ndash;2D surface mapping. Int. J. Geogr. Inf. Sci. \u003cb\u003e28\u003c/b\u003e(4), 780\u0026ndash;798 (2014)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoshida, T., Tanaka, K.: Land-use diversity index: a new means of detecting diversity at landscape level. Landscape Ecol. Eng. \u003cb\u003e1\u003c/b\u003e, 201\u0026ndash;206 (2005)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiemke, D., Charlton, B., H\u0026ouml;rl, S., Nagel, K.: An efficient approach to create agent-based transport simulation scenarios based on ubiquitous Big Data and a new, aspatial activity-scheduling model. Transp. Res. Procedia. \u003cb\u003e52\u003c/b\u003e, 613\u0026ndash;620 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng, P., Mike, M.: An investigation on the manual traffic count accuracy. Procedia-Social Behav. Sci. \u003cb\u003e43\u003c/b\u003e, 226\u0026ndash;231 (2012)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnnur, R., Ponnusamy, V.: \u003cem\u003eInformation and communication technology (ICT) for intelligent transportation systems (ITS)\u003c/em\u003e, in \u003cem\u003eEmploying recent technologies for improved digital governance\u003c/em\u003e. IGI Global. pp. 164\u0026ndash;194. (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorshchev, A.: The big book of simulation modeling: multimethod modeling with AnyLogic 6. AnyLogic North America (2013)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInturri, G., Le Pira, M., Giuffrida, N., Ignaccolo, M., Pluchino, A., Rapisarda, A., D'Angelo, R.: Multi-agent simulation for planning and designing new shared mobility services. Res. Transp. Econ. \u003cb\u003e73\u003c/b\u003e, 34\u0026ndash;44 (2019)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesktop, E.S.R.I.A.G.I.S.: Release 10, vol. 437, p. 438. Environmental Systems Research Institute, CA (2011). Redlands\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShaw, S.-L., Yu, H.: A GIS-based time-geographic approach of studying individual activities and interactions in a hybrid physical\u0026ndash;virtual space. J. Transp. Geogr. \u003cb\u003e17\u003c/b\u003e(2), 141\u0026ndash;149 (2009)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogerson, P.A.: Statistical methods for geography: a student's guide, pp. 1\u0026ndash;432. Statistical methods for geography (2019)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhong, M., Hanson, B.L.: GIS-based travel demand modeling for estimating traffic on low-class roads. Transp. Plann. Technol. \u003cb\u003e32\u003c/b\u003e(5), 423\u0026ndash;439 (2009)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eApronti, D.T., Ksaibati, K.: Four-step travel demand model implementation for estimating traffic volumes on rural low-volume roads in Wyoming. Transp. Plann. Technol. \u003cb\u003e41\u003c/b\u003e(5), 557\u0026ndash;571 (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePendyala, R.M., Kitamura, R., Kikuchi, A., Yamamoto, T., Fujii, S.: Florida activity mobility simulator: overview and preliminary validation results. Transp. Res. Rec. \u003cb\u003e1921\u003c/b\u003e(1), 123\u0026ndash;130 (2005)\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Activity clusters, Destination Choice, Travel Pattern Recognition, Agent-Based Model, GIS","lastPublishedDoi":"10.21203/rs.3.rs-4835588/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4835588/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe study explores the significance of individual mobility measures, such as activity space, in understanding how individuals interact with their daily environments. Existing measures often overlook geographical concepts like spatial-temporal arrangements of activity spaces, focusing solely on numerical assessments. To address this gap, a multi-level modeling approach combining Agent-Based Modeling (ABM) and Geographic Information Systems (GIS) is utilized to simulate activity destination selection throughout a workday in Zanjan, Iran. The model integrates individual preferences, built environment characteristics, network attributes, and travel generation data. Real-world data from Emerging Data Sources (EDSs) validate the model's reliability and accuracy. Key findings include: (1) clustering analysis identifying four types of activity destinations at different hourly intervals, (2) a central activity space acting as a hub for activity-based travel with a monocentric distribution pattern, (3) individual preference for destinations with diverse and dense built environments, and (4) a decrease in trip frequency as distance from the main activity space increases, indicating a spatial decay effect on activity-based travels.\u003c/p\u003e","manuscriptTitle":"Exploring the spatial-temporal arrangements of urban activity space from individual's daily commute: A Geospatial-Agent based Approach Using Empirical Data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-08-30 09:52:28","doi":"10.21203/rs.3.rs-4835588/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"4ac29785-be47-40b2-82b4-bcf763894b9e","owner":[],"postedDate":"August 30th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-12-01T22:38:19+00:00","versionOfRecord":[],"versionCreatedAt":"2024-08-30 09:52:28","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4835588","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4835588","identity":"rs-4835588","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}