Introduction

Urban traffic in contemporary metropolises represents a multidimensional, dynamic, and nonlinear challenge with significant economic, environmental, and social consequences. Tehran metropolis, with a population exceeding 9 million within the city boundary possesses one of the most complex traffic systems. Traffic conditions in this city are shaped by a confluence of factors, including the concentration of administrative, economic, and educational functions within the urban core; population flows from surrounding cities toward Tehran; and regulatory measures such as the Traffic Scheme and the Odd-Even Vehicle Restriction Plan. Despite this complexity, a comprehensive understanding of the spatio-temporal patterns of traffic congestion in Tehran - particularly through integrated approaches that simultaneously address both spatial and temporal dimensions - remains a notable gap in the research literature.

The primary objective of this study is to identify, analyze, and model the spatio-temporal patterns of traffic congestion in Tehran metropolis using the novel Emerging Hot Spot Analysis (EHSA) approach, built upon a Space-Time Cube framework. The research seeks to answer three core questions: What structural form do the spatio-temporal patterns of traffic congestion in Tehran take? Does this distribution exhibit statistically significant spatial autocorrelation? And how are high-density and low-density traffic clusters distributed across space and time?

Data and Methodology

The research database was assembled through systematic web crawling and real-time online recording of traffic data across Tehran's road network during the period from April 22 to April 29, 2026. A total of 259 data collection rounds were conducted at approximately 30-minute intervals, yielding a cumulative dataset of 7,945,396 traffic records covering an average of 30,667 road segments per round. The daily mean record count was 984,000 with a standard deviation of 325,000. Data were stored in a GeoPackage database format and subsequently converted to Geodatabase format for processing within the ArcGIS Pro environment.

The research methodology was structured across five sequential stages. In the first stage, raw online traffic data were extracted and organized into a structured format. The second stage was dedicated to data cleaning, quality control, and imputation of missing values. In the third stage, linear road network data were converted to point data through centroid extraction for each road segment. The fourth stage involved the construction of the Space-Time Cube. The fifth and central stage of the research executed three levels of spatial analysis in sequence: global spatial autocorrelation analysis using Moran's I index; High/Low Clustering analysis; and Emerging Hot Spot Analysis (EHSA).

Resulsts

Descriptive Structure of Traffic Distribution: Descriptive analysis revealed that 68.93% of road segments operated under low-flow conditions, 22.59% under near-free-flow conditions, 5.33% under heavy traffic, 1.30% under server traffic, and 1.85% were close during the study period. While the high proportion of free-flow conditions may initially suggest a favorable overall traffic situation, these figures must be interpreted in light of the spatial concentration patterns of traffic load. The bulk of the traffic burden is concentrated on a limited number of arterial corridors - a pattern consistent with what is known in transport planning literature as the Pareto principle in traffic networks, whereby approximately 20% of roads bear 80% of the total traffic load.

Spatial Autocorrelation: The Moran's I analysis yielded a z-score of 212.41, substantially exceeding the critical threshold of 2.58, thereby conclusively establishing the presence of statistically significant positive spatial autocorrelation at the 99% confidence level. This finding demonstrates that the spatial distribution of traffic congestion in Tehran does not follow a random pattern, but instead exhibits a strong clustering structure. The positive z-score obtained from the High/Low Clustering analysis further confirmed that this clustering is primarily driven by the co-location of high traffic values — that is, congested road segments tend to aggregate spatially rather than disperse, forming dense traffic hotspot clusters. This pattern is theoretically interpretable within the frameworks of the center-periphery theory and the monocentric city model.

EHSA Patterns: The most substantive findings of the study emerged from the EHSA, which identified eight distinct spatio-temporal pattern classes organized into hot spot and cold spot categories.

Among hot spot patterns, the Persistent Hot Spot pattern (4.08%) was identified across most of District 10, the eastern portion of District 9, and the western portion of District 11. This area, characterized by dense and aging urban fabric, a high-intensity mix of residential, commercial, and administrative land uses, convergence of major highways, and a large floating population, functioned as the most critical and consistently congested traffic node throughout the entire study period. The Intensifying Hot Spot pattern (1.3%) was identified in the western extremity of Districts 21 and 22, where traffic intensity exhibited a continuous upward trend — the most alarming finding of the study, indicating that this zone is undergoing a transition toward a future traffic crisis, directly linked to rapid residential development projects such as Chitgar Township and the Persian Gulf Martyrs Lake complex. The Oscillating Hot Spot (7.61%) and Sporadic Hot Spot (3.64%) patterns together account for more than 11% of spatial units and form a concentric ring around the central persistent hot spot zone, representing transitional areas between the congested core and the more fluid periphery. The Diminishing Hot Spot pattern (0.1%) was identified in neighborhoods south of Mehrabad Airport.

Among cold spot patterns, the Sporadic Cold Spot (13.34%) was the most spatially extensive, predominantly encompassing Districts 22 and 20 along with parts of Districts 4, 15, and 18. The Oscillating Cold Spot (11.58%) covered peripheral zones in Districts 5, 2, 4, and 19 through 22. The Diminishing Cold Spot (3.64%) covered nearly the entirety of District 21, indicating a gradual deterioration of free-flow conditions in this area. The Persistent Cold Spot (2.32%) and Intensifying Cold Spot (2.98%) patterns were identified along the urban fringe in areas including the eastern part of District 4, the Sharif neighborhood in western Tehran, and the Resalat Township in the south.

Overall Spatial Structure Identified

The synthesis of EHSA patterns reveals a four-ring concentric spatial structure in Tehran's traffic geography: a persistent critical core at the center; a transitional ring of oscillating and sporadic hot spots in the immediate surroundings; a middle zone of sporadic and oscillating cold spots; and, finally, the urban fringe characterized by persistent and intensifying cold spot patterns. This structure follows a center-periphery model in which traffic congestion intensity diminishes progressively as distance from the urban core increases.

Conclusions and Implications

The integration of real-time web-crawled data with the EHSA approach within the ArcGIS Pro environment provides an efficient, scalable, and reproducible analytical framework for continuous urban traffic monitoring. The principal planning and management implications of this research can be summarized across three axes. First, immediate policy prioritization on Tehran's urban core is warranted, through travel demand management, public transport enhancement, and land use redistribution. Second, proactive intervention in Districts 21 and 22 is urgently needed to prevent their transformation into new critical congestion nodes. Third, distinct and time-sensitive traffic management strategies must be designed for the transitional ring zones, which exhibit high sensitivity to hourly and daily fluctuations. The alignment of this study's findings with results from comparable research in major international metropolises validates the methodological applicability of EHSA as a standard analytical instrument for urban traffic analysis.