(1) Overview
Context
Hanseong served as the central hub for the origin and development of Baekje, a kingdom that existed on the Korean Peninsula from approximately the 1st century B.C. to A.D. 660. Baekje relocated its capital twice throughout its history; Hanseong functioned as the kingdom’s first capital and political stronghold until its fall to Goguryeo in A.D. 475. Most notably, the region has garnered significant academic attention as it provides relatively clear evidence of the evolutionary process from a small local polity into an ancient state. Geographically, it is situated in the Songpa-gu, one of the most highly urbanized districts within Seoul, the capital of South Korea. Archaeological excavations in this region commenced during the 1980s, spurred by regional development projects in preparation for the 1988 Seoul Olympics, and have continued to the present day.
Pottery recovered from Hanseong has served as the primary chronological indicator for the area. Due to the scarcity of absolute dating references, researchers have predominantly relied on the typological classification of pottery, inferring dates through cross-examinations with Chinese ceramics that have relatively well-established chronologies. However, the lack of distinct morphological changes in the pottery itself has posed significant challenges to research. Under these circumstances, radiocarbon dating results are expected to serve as a highly effective and complementary tool to address these chronological ambiguities.
Although scholarly debate persists regarding the precise spatial extent of Hanseong, it generally refers to the area enclosed by the natural and artificial boundaries formed by the Han River, Samsungdongtoseong, Amsa-dong, and Garak-dong. Centering on the royal fortresses of Pungnaptoseong and Mongchontoseong, numerous ancient tomb complexes and residential sites are distributed throughout the region. During the initial stages of research, investigations primarily focused on major sites with relatively well-preserved archaeological features, including Pungnaptoseong, Mongchontoseong, and the Seokchon-dong ancient tomb complex. Recently, however, the scope of archaeological investigations has expanded to include general residential areas as a result of active urban redevelopment projects. To date, excavations have been conducted in a total of 22 districts; the radiocarbon dating results accumulated from these sites serve as the foundation for constructing a precise chronological dataset, which facilitates multifaceted discussions regarding key academic issues in Hanseong.
Pungnaptoseong and Mongchontoseong were established as a complementary pair, serving distinct purposes while operating in tandem. Pungnaptoseong is a flatland fortress constructed on the alluvial plain along the Han River, serving as the primary administrative center. In contrast, Mongchontoseong is a rear defense fortress built on relatively high ground, strategically utilizing the natural hilly terrain. Meanwhile, the Seokchon-dong Burial Mound District represents the royal and aristocratic cemetery located within the capital area. Together, Pungnaptoseong, Mongchontoseong, the Seokchon-dong Burial Mound District, and surrounding settlements including the sites in Sincheon-dong, Bangi-dong, and Cheonho-dong constitute the capital city of Hanseong Baekje.
Of these, Pungnaptoseong is of particular significance. Pungnaptoseong was designated as a State-designated Heritage in the 1960s when the national heritage designation system was first established; however, this designation was initially limited to the ramparts, leaving the interior of the fortress unprotected. Consequently, during the 1960s and 1970s, the interior was developed into an industrial zone, and as Seoul grew into a high-density metropolis, it further transformed into a dense residential area by the 1990s. During this period of urban expansion, there were numerous instances where archaeological features and artifacts were identified and investigated through rescue excavations during reconstruction processes. Since 2000, institutional frameworks for the restoration and management of Pungnaptoseong have been established, leading the state to begin purchasing private land within the fortress walls. Once a sufficient number of contiguous lots are acquired, a formal archaeological excavation is conducted. Therefore, although Pungnaptoseong is a single fortified site, the specific excavation locations are dispersed across the modern urban landscape. To date, excavations have been carried out at approximately 23 distinct locations, covering about 13% of the total area.
This unique administrative and developmental history explains why various analytical units in this dataset are named after modern residential complexes. For instance, the ‘Mirae-maeul District’ (literally meaning ‘Future Village’) was originally a residential and commercial area. In 2000, development was halted when Baekje cultural layers were identified during a trial excavation conducted by the ‘Mirae-maeul Reconstruction Association,’ and the district was subsequently named after the proposed apartment complex. Similarly, the ‘Gyeongdang District’ within Pungnaptoseong was designated as a State-designated Heritage after Hanseong Baekje artifacts were unearthed during an excavation prior to the reconstruction of the ‘Gyeongdang Row Houses.’ Other sites follow a similar naming convention: the Daejindongsan Row Houses site was named after the housing complex that existed prior to its redevelopment into the present-day Songpa Haemoro Apartments. In the case of Hyundai Riverville, the name refers to the current apartment complex where a rescue excavation was necessitated after artifacts were discovered during the construction process. By compiling data from these geographically separated yet historically connected districts, this dataset provides a comprehensive temporal framework for the entire Hanseong.
Previously, several studies have published radiocarbon datasets covering Hanseong [1, 2, 3]. However, because these compilations were primarily intended to examine demographics, they focused almost exclusively on data from dwelling sites. In addition to dwelling sites, various types of features have been identified in Hanseong, and radiocarbon dating has been actively performed on these features as well. This dataset is significant as it integrates and organizes these diverse measurements which had not been previously compiled into a single resource to provide a more comprehensive chronological framework.
Spatial coverage
Projected Coordinate system: EPSG:5186 (Korea 2002/central Belt)
Datum: D Korea Geodetic Datum 2002
Description: Southeastern Seoul
Northern boundary: 552000
Southern boundary: 544000
Eastern boundary: 212000
Western boundary: 208000
Temporal coverage
Proto-Three Kingdoms period-Hanseong period of Baekje (ca. 1st century BC/AD-A.D. 475).
(2) Methods
Steps
This dataset synthesizes the radiocarbon dates identified in Hanseong. Data collection and compilation were carried out during 2025–2026, and the dataset was established through a four-step process.
Step 1: Archaeological excavation reports and academic journals related to Hanseong, published from the 1980s to the present, were collected.
Step 2: Radiocarbon dates recorded in the reports were extracted. In this step, uncalibrated dates were collected, along with metadata such as LabID and sample material.
Step 3: This involved cross-validation and data cleaning, where redundant entries were eliminated and the accuracy of the recorded measurement values was verified.
Step 4: Unique identification codes were assigned to each dating result. Where possible, precise spatial coordinates and archaeological contexts for each sample were documented.
Data collection
This dataset is a systematic compilation of previously published radiocarbon dates rather than a result of new primary sampling. The data collection involved a comprehensive survey of archaeological excavation reports published between the 1980s and the present, focusing on Hanseong (primarily Songpa-gu and parts of Gangdong-gu). To provide a holistic chronological framework for the area, the dataset includes all available radiocarbon dates identified within this spatial boundary. The dataset incorporates certain radiocarbon dates that are not directly associated with Hanseong Baekje. These results were included because they can assist researchers in interpreting complex stratigraphic contexts for example, at the Seokchon-dong ancient tombs, where the severe overlapping of features complicates stratigraphic interpretation, or at the Pungnaptoseong ramparts, where distinguishing between natural sedimentary layers and artificial fill layers used for wall construction can be challenging. All measurement values and metadata such as LabID and sample material were extracted directly from the original documentation to facilitate spatiotemporal analysis.
Data verification and standardization
The methodological intervention in this study was limited to the standardization and verification of existing records. All collected dates underwent a duplication check to identify instances where the same results were reported across different publications. In cases where multiple research institutions conducted excavations at the same site over several decades, unique identification numbers were assigned to prevent the confusion caused by overlapping sub-lot names or numbers. Furthermore accuracy was verified by cross-referencing uncalibrated dates with their respective LabID. As a significant portion of the original reports lacked digital coordinates, sample locations were estimated through georeferencing site layout maps within ArcGIS; where explicit coordinates were available, they were prioritized to minimize positional error.
Constraints
Although archaeological investigations in Hanseong began in the 1980s, systematic radiocarbon dating was only widely implemented after 2000; therefore, since no dating was performed during earlier excavations, such results are non-existent and are thus absent from this dataset. In addition to this, there is a noticeable spatial imbalance in the distribution of the data. Historically, research efforts have been heavily concentrated on the primary sites, the royal fortresses of Pungnaptoseong (n = 114) and Mongchontoseong (n = 68), and the Seokchon-dong ancient tombs (n = 81). As a result, the number of radiocarbon dates for these three sites significantly outweighs those from other areas. While recent urban redevelopment has led to increased investigations in the surrounding districts such as Shincheon-dong (n = 34), Bangi-dong (n = 27), Amsa-dong (n = 3), and Cheonho-dong (n = 2), the data density for regions outside the main fortresses remains comparatively lower. Furthermore, as most original reports do not provide precise digital coordinates, the majority of sample locations were estimated by georeferencing site layout maps within ArcGIS and cross-referencing archaeological contexts. In instances where reports provided explicit coordinates, these values were prioritized and documented as such in the dataset.
(3) Dataset description
This dataset is provided in .csv and .xlsx formats. The data are structured according to the columns listed in Table 1. In cases where data are unavailable, the respective fields have been left blank.
Table 1
List of data columns included in the database.
| FIELD | DESCRIPTION |
|---|---|
| Easting | The projected X coordinate (in meters) of the sample location in EPSG:5186. Where explicit coordinates were provided in excavation reports, those values were adopted; otherwise, positions were estimated through georeferencing excavation maps in ArcGIS. |
| Northing | The projected Y coordinate (in meters) of the sample location in EPSG:5186. Where explicit coordinates were provided in excavation reports, those values were adopted; otherwise, positions were estimated through georeferencing excavation maps in ArcGIS. |
| Code | The unique number of newly assigned analysis sample |
| SiteName | The name of the archaeological site where the sample was collected. In cases where excavations were conducted across multiple distinct areas within a single site, the specific names of the area are also included. |
| FeatureName | The specific name of the archaeological feature from which the sample was collected. |
| Context | The archaeological context of the sample, detailing the stratigraphic and functional relationship between the sample and its associated feature or layer. |
| SiteType | The classification of the archaeological feature from which the sample was retrieved (e.g., dwelling, pit, rampart, or tomb). |
| SampleID | The identification number or designation assigned to the sample as recorded in the original archaeological excavation report. |
| Lab | The name of the institution or laboratory where the radiocarbon dating analysis was performed. |
| LabID | The unique identification name or code assigned to the sample by the laboratory that performed the radiocarbon dating analysis. |
| Material | The physical substance or material composition of the sample selected for radiocarbon dating analysis (e.g., charcoal, wood, bone). |
| BP | The measured radiocarbon age of the sample, expressed in years Before Present (BP) prior to any atmospheric calibration. |
| Error | The measurement uncertainty or standard deviation associated with the uncalibrated radiocarbon age, as reported by the laboratory. |
| δ13C | The δ13C values of the analyzed samples, as reported by the laboratory. |
| References | A citation of the source from which the data has been drawn |
Object name
Data/ – compilation files
AMS_Master_en.csv
AMS_Master_ko.csv
Graph/ – calculated analysis graph
Figure 1. Distribution of radiocarbon dating sampling locations in Hanseong region.tif
Figure 1
Distribution of radiocarbon dating sampling locations in Hanseong (n = 329).
Figure 2. Locations of radiocarbon dating samples within Pungnaptoseong.tif
Figure 2
Locations of radiocarbon dating samples within Pungnaptoseong (n = 114).
Figure 3. Number of samples(Hanseong).tiff
Figure 3
Number of radiocarbon dates by site in Hanseong.
Figure 4. Summed Probability Distributions (SPD) of radiocarbon dates by site in Hanseong.tiff
Figure 4
Summed Probability Distributions (SPD) of radiocarbon dates by site in Hanseong.
SPD(Hanseong).tiff
SPD(Mongchontoseong).tiff
SPD(Outside of PNTS&MCTS).tiff
SPD(Pungnaptoseong).tiff
SPD(Tombs).tiff
Script/
barplot.r
SPD.r
Reference.txt
README.md
Data type
Secondary data and processed data
Format names and versions
*.csv, *.r, *.tif, *.tiff
Creation dates
01/09/2025 – 02/01/2026
Dataset Creators
Ju Chanhyeok, Park Hyeonghoo
Language
English, Korean
License
CC-BY 4.0
Repository location
Github Repository: https://github.com/ChanToRe/Hanseong_C14
Publication date
09/01/2026
(4) Reuse potential
Research on Hanseong Baekje has garnered significant attention, spanning diverse perspectives across the fields of history and archaeology. This scholarly interest is reflected in the extensive volume of academic literature, which includes over 350 journal articles and approximately 160 academic dissertations [4, 5, 6, 7, 8, 9, 10]. This radiocarbon database is a crucial resource for advancing these discussions, particularly regarding royal fortresses like Pungnaptoseong and Mongchontoseong. Research on these ramparts focuses on initial construction and expansion [11, 12, 13, 14], yet limited artifacts within structures hinder traditional typological chronology. Radiocarbon analysis, including wiggle-matching, offers a vital alternative for resolving debates over 3rd- versus 4th-century origins.
Furthermore, the dataset facilitates examining spatial transformations. Urbanization has limited excavations to fragmentary areas, creating complex stratigraphy that challenges existing chronologies [3, 15]. In this context, Summed Probability Distribution (SPD) analysis facilitates the comparison of temporal activity patterns across Pungnaptoseong, Mongchontoseong, their internal and external residential sites, and the Seokchon-dong and Bangi-dong burial mound districts. The relative fluctuations inferred from the SPD curves allow for identifying the timing of rampart construction, the expansion of residential spaces, and the intensification of funerary activities. This approach, in turn, helps elucidate the temporal relationships between the concentration of activities in specific areas, urban growth, and the establishment of defensive systems.
Additionally, GIS analysis integrating radiocarbon dates with spatial coordinates can be employed to directly reconstruct the spatiotemporal organizational changes of human activity within the capital. As this dataset is confined to specific excavated districts and is not uniformly distributed across Hanseong, the analytical scope is necessarily restricted to the investigated areas. Consequently, rather than estimating activity intensity, this dataset allows for the visualization and comparative analysis of the spatial distribution patterns of locations where activity has been confirmed by period. By progressively visualizing these locations—including the fortress interior, rampart sections, and external cemeteries and residential zones—based on the distribution of radiocarbon dates, it is anticipated that the processes of spatial expansion, as well as the differentiation and reorganization of functional areas within the capital, can be identified.
Moreover, given that the Pungnaptoseong rampart is a large-scale monument with clearly defined structural stratigraphy and units, Bayesian modeling of the radiocarbon dates obtained from each layer can more definitively reconstruct the construction phases and the temporal span of each stage. This enables a quantitative verification of whether the rampart was formed through a single construction event or via successive stages of expansion and reinforcement. Ultimately, by comparing the timing of construction and the developmental sequences of Pungnaptoseong and Mongchontoseong, it is expected that the temporal structure of the formation of Hanseong Baekje capital system and the process of political centralization can be more precisely elucidated.
Consequently, by utilizing this dataset, researchers can achieve a holistic view of the region’s evolution. Integrated studies using these methods are already active [10, 14, 15], and this database provides the foundation for further synthesis.
Since the 2000s, radiocarbon dating has been actively conducted in South Korea. While the majority of these investigations are rescue excavations, the sheer density of archaeological surveys relative to the land area has led to a vast accumulation of data, making it highly feasible to utilize radiocarbon measurements. However, to date, cases where structured datasets have been shared with the academic community are extremely rare. This is largely due to the absence of a systematic sharing platform and a lack of consensus among researchers regarding the importance of data sharing [16]. As previously mentioned, archaeological activity in the Korean Peninsula is exceptionally high, with over a thousand excavations conducted annually, a volume that far exceeds the capacity of any single researcher to compile. Consequently, more proactive data sharing is essential, and there is a critical need for collaborative frameworks that allow multiple researchers to manage and utilize this information. In this regard, the present study is significant as it raises the necessity for such a paradigm shift within the field of Korean archaeology.
A dataset organized in such a standardized manner can be further expanded and utilized in diverse contexts. For instance, a comparative analysis with data from the capitals of Goguryeo, Silla, and Gaya, which coexisted on the Korean Peninsula during the same period would allow for an examination of the distinct characteristics and standardized aspects of central cities at that time. Even if the Goguryeo region is excluded due to its location in North Korea and the resulting lack of accessibility, a vast amount of data has been accumulated for the Gyeongju region, the capital of Silla, necessitating an integrative analysis. Since such diachronic analyses fundamentally require datasets structured in a standardized format, it is essential to compile various archaeological data in the same manner as the dataset produced in this study. In this regard, the present dataset can serve as a comprehensive guideline for future data-driven research.
As excavations continue such as recent investigations of Pungnaptoseong’s west and east ramparts yielding over a hundred new samples, the database will receive periodic updates. Continuous management is essential to support long-term integrated research, ensuring the database remains a living resource for understanding Hanseong Baekje.
Acknowledgements
The authors express their deepest gratitude to all the researchers who have dedicated themselves to the archaeological investigations in Hanseong.
Competing Interests
The authors have no competing interests to declare.
Author Contributions
All authors contributed equally to this paper. Ju C. performed data collection and wrote portions of the manuscript; Dzon S. conceptualized the study and wrote portions of the manuscript; and Park H. was responsible for creating figures and collecting coordinate information for the samples.