Spatial and Temporal Distribution Characteristics of Stone Age to Warring States Period Sites in Sichuan Province

Abstract

In recent years, significant progress has been achieved in the study of the spatial and temporal distribution of ancient sites and their influencing factors. As an integral component of the pluralistic unity of Chinese civilization, the ancient Shu civilization exhibits unique cultural characteristics and historical significance, rendering the Sichuan region a critical area for exploring the origins and development of Chinese civilization. However, existing research lacks a comprehensive compilation of archaeological sites across the entirety of Sichuan Province, and analyses of spatial and temporal distribution characteristics and their influencing factors often lack multi-scale and multi-dimensional perspectives. This study systematically compiles site data from Sichuan Province and employs GIS spatial analysis methods to examine the distribution characteristics of sites and their relationship with natural geographical factors from a geographical spatial perspective. The findings reveal that site distribution in Sichuan Province exhibits significant clustering, predominantly concentrated near rivers and in higher elevation areas. Factors such as altitude, slope, and proximity to water significantly influence site distribution. In terms of altitude, the elevation of sites’ distribution generally declined from the Stone Age to the Warring States period, dropping below 1000 m during the Shang to Spring and Autumn periods before rising again. Regarding proximity to water, a substantial proportion of sites across all periods are located within a 1 km buffer zone, with approximately 50% during the Stone Age and Warring States period, and up to 70% during the Shang to Spring and Autumn periods, indicating a preference for areas close to water for settlement and production. In terms of slope, most sites across historical periods are located in areas with slopes below 15°, with the highest number of sites during the Shang to Spring and Autumn periods. The evolution of human–environment relationships demonstrates a trend of site concentration shifting from plateau to basin areas from the Stone Age to the Warring States period, reflecting changes in population movement, economic development patterns, and socio-political structures. The research provides new insights into the evolution of human–environment relationships in the region and offers valuable references for related studies.

1. Introduction

Sites refer to physical remains left by ancient human activities, reflecting human activities, human–environment relationships, and historical–cultural landscapes of specific periods [1]. The spatial and temporal distribution of sites is crucial for understanding the origin, development, and evolution of human civilizations, as well as revealing the interactions and influences between ancient human activities and the natural environment [2,3]. Located in the southwestern region of China, Sichuan Province boasts a long history and profound cultural heritage. Represented by the Sanxingdui and Jinsha sites, the ancient Shu civilization exhibits a cultural landscape distinct from that of the Central Plains civilization. The formation and development of the ancient Shu civilization, as well as the ancient sites across Sichuan Province, provide significant case studies for researching the origin and evolution of early civilizations, playing an essential role in revealing the diversity of Chinese civilization. These sites are indispensable for re-evaluating the historical trajectory of the origin and development of Chinese civilization [4].

Currently, numerous scholars have conducted extensive research on the spatial and temporal distribution of archaeological sites. Yang et al. analyzed the spatial distribution characteristics of Peiligang culture settlements in Henan, revealing a clustered distribution pattern during the Peiligang period, which provides insights for re-evaluating and understanding this culture [5]. Květina employed geographic information system (GIS) statistical methods to analyze the structure and spatial distribution patterns of Neolithic sites in the Bylany region, obtaining detailed distribution results within individual sites [6]. Xu et al. conducted a comprehensive investigation of the spatial distribution of Paleolithic sites in the Bose Basin using GIS and remote sensing methods, concluding that site density gradually decreases from the northwest to the southeast [7]. Many scholars have also explored the relationship between rivers and site distribution. Liu et al. studied the spatial distribution and evolution of settlements from the Neolithic to the Bronze Age in the Dalian region, noting that sites were primarily located in river valleys and near water bodies [8]. Gao et al. collected archaeological site data from the Sanhe region to discuss and elaborate on the distribution characteristics of early Chinese civilizations in the middle reaches of the Yellow River, concluding that human settlements were situated on higher ground near rivers [9]. Zhu et al. used kernel density analysis and topographic analysis to identify the distribution characteristics of sites along rivers in Shaanxi Province from the Han to the Sui-Tang periods, as well as their changing trends across different dynasties [10]. Goval et al. conducted archaeological excavations at 26 sites in northern France, revealing that the sites are primarily distributed on hillsides and plains in the region, often located on river terraces or high ground, which facilitated access to water and hunting resources [11]. Oron et al. carried out a systematic geoarchaeological survey in the central and southern Negev Desert of Israel, demonstrating significant regional differences in the spatial distribution of sites. Sites are more densely concentrated in the north and center, while they are sparser in the south [12]. Hamon et al. integrated archaeological site data to analyze population movements and settlement patterns during the Neolithic transition in Western Europe. They found that early sites were predominantly located in coastal plains and estuarine areas, such as the Languedoc region in France, before expanding into inland mountainous and plateau regions [13]. Garcia et al. studied Late Paleolithic sites in the Cantabrian region, noting that most sites are located in caves, with a few in rock shelters, and nearly half of the sites are situated at elevations below 100 m [14].

Based on the research of the spatial–temporal distribution of sites, many scholars have further explored the factors influencing their distribution or the human–environment relationship. Bi et al. used spatial autocorrelation analysis to study the distribution characteristics of prehistoric settlement sites in the Zhengzhou-Luoyang area and found that there was a significant positive spatial autocorrelation among the sites of different cultural periods, and the site distribution was closely related to geographical environmental factors [15]. Studies have shown that the main factors affecting the spatial–temporal distribution of sites are climate, water sources, elevation, slope, and aspect. Li et al. used GIS tools to analyze the spatial–temporal distribution patterns and influencing factors of ancient sites in China and found that the sites showed an aggregated spatial pattern, mainly concentrated in the Yellow River and Yangtze River basins. They were influenced by multiple factors, including social development, natural environment, geographical elements, and economic–political factors [16]. Xu et al. conducted spatial analysis of Poyang Lake and obtained the distribution of ancient civilization sites and their economic–cultural scales in different periods, exploring the relationship between sites, climate, and environmental changes [17]. Li et al. selected multiple archaeological sites in Guizhou Province from the Paleolithic to the Shang-Zhou periods for spatial–temporal analysis to investigate the relationship between human settlements and the natural environment during these three periods [18]. Gao et al. showed that the location of human settlements in the Sanhe area from the Paleolithic to the Xia, Shang, and Zhou dynasties was mainly determined by topography and hydrology [6]. Hannah et al. used GIS to establish a multi-criteria decision-making analysis model to evaluate the natural factors influencing the spatial distribution of Paleolithic sites in the Wadi Sabra area, including eight parameters, such as elevation, geomorphology, and hydrogeology [19]. Garcia et al. utilized GIS technology to analyze the site selection preferences of Late Paleolithic sites in the Cantabrian region, thereby exploring the factors influencing the spatiotemporal distribution changes of these sites. Variables such as absolute elevation, relative elevation, topography, landscape, and orientation were incorporated into the model to analyze site selection preferences [14]. Neruda et al. conducted a study on the Middle Paleolithic site at Kůlna Cave, revealing the cave’s usage patterns and functional zoning changes across different periods. The cave’s geomorphological features, such as the orientation of its entrance and the internal spatial structure, may have influenced the distribution of sites [20]. Basell et al. analyzed the relationship between site distribution in East Africa during the Middle Stone Age and paleovegetation, volcanic activity, and tectonic movements, exploring how changes in paleovegetation, volcanic activity, and tectonic activities influenced site distribution [21]. Snoeck et al. conducted a multifaceted analysis of human remains from the Neolithic court tomb at Parknabinnia in western Ireland, concluding that strontium isotope values varied across different geological regions, affecting the origins and migration patterns of individuals [22]. Kokkinidou et al. studied multiple basins and plains in western Macedonia, identifying prehistoric sites and recording their types and distribution patterns. Geographic environmental factors, such as rivers and mountain ranges, significantly influenced site distribution [23]. Spencer et al. constructed a logistic regression model to analyze the site dataset from the Mirabello region in Crete, revealing a significant positive correlation between the Topographic Wetness Index (TWI) and settlement intensity. Population growth and social changes also played a role in the distribution changes of sites [24].

In summary, current research on the spatial patterns of archaeological sites primarily focuses on using spatial analysis tools in GIS to reveal the spatial and temporal distribution characteristics of sites. The research scope is mostly concentrated in traditional archaeological hotspots, such as the Yellow River and Yangtze River basins. In contrast, studies on archaeological sites within Sichuan Province are relatively limited, with most research focusing on areas related to the ancient Shu civilization, lacking a systematic investigation of sites across the entire province. Additionally, in terms of research perspectives, there is a lack of comprehensive analysis of the distribution characteristics of sites and their associations with natural geographical factors from a geographical spatial viewpoint. In terms of research methods, existing studies often rely on single methods or data from local regions, lacking a comprehensive revelation of the multi-scale and multi-dimensional spatial distribution patterns. Including the neglected sites in the spatial and temporal distribution research of remains is of great significance for interpreting the more authentic human–environment relationship during the historical periods of the region.

Therefore, this study focuses on Sichuan Province as the research area, systematically compiling comprehensive site data within the region. By employing GIS methods, it analyzes the spatial and temporal distribution characteristics of these sites, exploring their relationships with factors such as elevation, slope, and proximity to water from a geographical spatial perspective. The findings provide new insights into the evolution of human–environment interactions in the region during historical periods, offering valuable references for related research.

2. Data and Methods

2.1. Overview of the Study Area

Sichuan Province is located in the southwestern part of China, with a total area of 486,000 square kilometers. The topography of Sichuan is characterized by higher elevations in the west and lower elevations in the east, with a general tilt from the northwest to the southeast. The province is predominantly mountainous, and its landform can be categorized into four types: plains, tablelands, hills, and mountains. Sichuan has a complex climate due to its varied topography. The Sichuan Basin and its surrounding areas belong to the subtropical humid climate zone, while the southwestern mountainous region of Sichuan is characterized by a subtropical semi-humid climate. The northwestern high-mountain area is a plateau cold climate zone. Human activities in Sichuan can be traced back to more than 2 million years ago during the Paleolithic Age.

The Chengdu Plain, situated in the western part of Sichuan Basin, China, is a piedmont alluvial plain formed by the sedimentation of rivers, such as the Minjiang and Tuojiang. The plain generally slopes from northwest to southeast, with elevations ranging between 400 and 750 m, covering a total area of 18,810 square kilometers, as illustrated in Figure 1. The Chengdu Plain served as the cultural center of Sichuan Province from the Shang Dynasty to the Warring States period and was the core region of the ancient Shu civilization. From the Stone Age to the Warring States period, it remained a stable area for site aggregation, holding significant importance for studying the spatial distribution of sites and human–environment relationships during this era.

The Chengdu Plain, developed in the Quaternary sedimentary basin at the front of the Longmen Mountains, has seen its fault activities gradually stabilize since the Middle Pleistocene, providing a geological foundation for the development of civilization in the plain area [25,26]. The hydrological system of the Chengdu Plain is primarily dominated by the Minjiang and Tuojiang river systems. The Minjiang River, the main river in the northwest of the plain, is characterized by abundant water flow and a broad valley, carrying a large amount of alluvial and proluvial deposits, which provide essential material for the formation of the plain and the fertility of its soil. The Tuojiang River, a tributary of the Minjiang, flows from the southwest to the southeast through the eastern part of the plain. Together, these two rivers form a complex hydrological network in the Chengdu Plain. The development of the river systems has created multiple terraces and alluvial–proluvial fans, and in lower-lying areas, lakes, marshes, and other wetland landscapes have formed [27,28,29].

The Stone Age (The period described is known as the Stone Age in archaeology, which is the first era used to classify early human history. It encompasses the Paleolithic (Old Stone Age) and the Neolithic (New Stone Age), spanning approximately 2.5 million years ago to around 5000–4000 years ago [30].) corresponds to the early and middle stages of the Pleistocene–Holocene, characterized by a warm and humid climate. The Pleistocene was a critical period for landform formation, with unstable geological structures, often leaving the Chengdu Plain submerged. During the Atlantic period of the Holocene, the Chengdu Plain was likely crisscrossed with waterways and dotted with lakes and marshes, with frequent flooding, which was not conducive to human activities and cultural prosperity [31,32]. The period from the Shang to the Warring States corresponds to the Subboreal period, during which, against a backdrop of global aridity, the multiple rivers within the Chengdu Plain could provide some water replenishment, preventing the climate from becoming too harsh. Although the frequency of flooding decreased, sudden floods still occurred occasionally. From the Shang to the Spring and Autumn periods (from around 1600 BC to 476 BC, approximately 4600 to 2800 years ago), the Minjiang River was well developed, and the Chengdu Plain had a rich variety of vegetation types, with a climate suitable for the growth of various plants. During the Warring States period (from around 475 BCE to 221 BCE, approximately 2490 to 2240 years ago), the construction of the Dujiangyan irrigation system began, addressing the issue of flooding and improving the ancient humans’ ability to build water conservancy facilities. Overall, from the Stone Age to the Warring States period, the humidity in Sichuan Province decreased, and the frequency of flooding reduced, creating more favorable conditions for the rise of ancient cultures [33,34,35].

2.2. Data Acquisition

This study collected data on various period sites in Sichuan Province, including 476 from the Stone Age, 347 from the Shang to Spring and Autumn periods, and 261 from the Warring States period. The site data were compiled from the Atlas of Chinese Cultural Relics: Sichuan Volume, and related archaeological survey reports (

Table 1. Basic information of the heritage sites derived from archaeological survey reports and literature.

Index Number	Source of Information	Number of Sites	City and Area of the Sites	Name and Excavation Year of Sites
1	Newly Discovered Paleolithic Sites in the Southern Chengdu Plain	12	Meishan City	Diao Shanzui site, Hehua Village site, Jigong Reservoir site, Jie Xin Chang site, etc. 2021.
2	A Brief Report on the Archaeological Survey of Pre-Qin Period Sites in the Middle Reaches of the Jialing River (Peng’an to Nanchong Section)	9	Nanchong City, 1,650,000 square meters	Ming Jiazui site, Xianba site, Caijia Hao site, etc. 2016–2017.
3	A Brief Report on the Archaeological Survey of the Yanyuan Basin in 2015	4	Liangshan Yi Autonomous Prefecture, 200,000 square kilometers	Dagou Liangzi site, Guan Guan Shan (Daozuo Miao) site, Panjia Ba (Gaoyanzi) site, etc. 2015.
4	On the Hopeful Fields: New Advances in Archaeological Discoveries and Research in the Minjiang Region—Centered on the Fieldwork of 2014	9	Chengdu City, Deyang City, Aba Tibetan and Qiang Autonomous Prefecture	A construction site in Bairren Community, Qingyang District, Chengdu City, Dongyu Village site, Gaoshan City East site, etc.
5	A Brief Report on the Archaeological Survey of Late Neolithic to Shang-Zhou Period Sites along the North Extension of Tianfu Avenue (Deyang Section) in Sichuan	5	Guanghan City, 80,000 square meters	Gao Shi Qiao site, Lan Jia site, Tuan Bo site, etc. 2019.
6	New Achievements in Archaeological Survey, Exploration, and Trial Excavation by the Sichuan Provincial Institute of Cultural Relics and Archaeology in 2006	3	Yibin City, Aba Tibetan and Qiang Autonomous Prefecture, 80 square meters	Dongyi Miao site, Lixian Jianzhai Village site, Songpan Cangping site, 2006.
7	Walking in the Fields: A Glimpse of Archaeological Work in the Minjiang River Basin in 2016	11	Aba Tibetan and Qiang Autonomous Prefecture, Chengdu City, Ganzi Tibetan Autonomous Prefecture, 1600 square meters	Baierwo site, Phase I of the Chengdu Jianong E-commerce Comprehensive Logistics Park, Gaoshan Ancient City site, etc. 2016.
8	Tombs, Culture, and Society of the Pre-Qin Period in the Chengdu Plain	5	Chengdu City, 300,000 square meters	Chongzhou Zizhu City, Dujiangyan Mangcheng, Pixian Ancient City, etc.
9	Overcoming Difficulties Together and Forging Ahead: A Comprehensive Review of Archaeological Work in the Minjiang River Basin in 2020	19	Chengdu City, Meishan City, Liangshan Yi Autonomous Prefecture, Aba Tibetan and Qiang Autonomous Prefecture, 6000 square meters	Bai Ta Community Group 9 site, Bai Ta Community Group 7 site, Da Fen Bao Cemetery, etc. 2019–2010.
10	A Brief Introduction to the Shang and Zhou Archaeological Remains in the Middle and Lower Reaches of the Jialing River	4	Nanchong City, Dazhou City	Langzhong Lanjiaba site, Nanchong Zifu Temple site, Pengcheng site, etc. 2004.
11	New Perspectives on the Study of Regional Archaeological Cultures in the Minjiang River Basin—Insights from the Fieldwork in 2011	9	Chengdu City, 15,000 square meters	The Shang-Zhou site at Fangying Village in Dafeng, the Shang Dynasty site at Wusi Village in Guilin Township the Shang-Zhou site at Shui Guanyin in Xinfan, etc. 2008–2011.
12	Celebrating the Centenary with Abundant Achievements: A Comprehensive Review of Archaeological Work in the Minjiang River Basin in 2021	26	Chengdu City, Guanghan City	Beijiao Village site, Chengguan Grain Station Stone Coffin Burial Ground, Ren Sheng Village site at Sanxingdui in Guanghan, etc. 2013–2021.
13	A Brief Report on the Survey of Paleolithic Remains in Ganzi Tibetan Autonomous Prefecture, Sichuan Province, in 2019	22	Ganzi Tibetan Autonomous Prefecture	Chenjia Mountain site, Dexi Top site, Dong’uo Luo site, etc. 2019.
14	A Brief Report on the Survey of Shang and Zhou Period Sites in the Yazi River Basin, Sichuan Province, from 2011 to 2013	17	Guanghan City, Deyang City, 250,000 square meters	Baqiang Land site, Baimiaozi site, Gan Youfang site, etc. 2013.
15	A Brief Report on the Survey of Pre-Qin Period Sites in the Middle Reaches of the Jialing River (Langzhong to Yilong Section)	10	Nanchong City, 1,300,000 square meters	Bai Sha Dam site, Bao Ben Temple site, Jiang Jia Dam site, etc. 2018.
16	Brief Reports on Archaeological Excavations in Various Districts of Chengdu City	52	Chengdu City, 65,000 square meters	The Shang-Zhou site at the location of the administrative building of the Qing Shui He campus of University of Electronic Science and Technology of China in Chengdu, the location of Luolin Pan, the base of Aiguang Investment Limited Microwave Production, etc. 2002–2018.
17	Brief Reports on Trial Excavations of Sites in Various Districts of Chengdu City	15	Chengdu City, 5000 square meters	The site at Location B line of Huangzhong Village planned road, the site at Chen Jia courtyard in Pixian County, Qinggang Village site, etc. 2002–2012.
18	Excavation Reports of Sites in Various Districts of Chengdu City	12	Chengdu City, 5000 square meters	The ancient site at Jinhe Road in Chengdu, the Songjiahe Dam site in Sanjiaoyan Town, the Shang-Zhou site in Xinhua Village, etc. 2003–2016.
19	Analysis Reports of the Remaining Cultural Sites	3	Chengdu City, Liangshan Yi Autonomous Prefecture, 3000 square meters	Shaping Station site, Tuan Shanbao site, Location C of site No. 5 in Jinniu District at Jinsha Ruins, 2007–2014.
20	Brief Reports on Excavations in Other Regions	34	Guanghan City, Yibin City, 30,000 square meters	Shaxi site in Ya’an, Zhongliang Mountain site, Qingguan Mountain Platform at Sanxingdui site, 2003–2019.
21	Aba Tibetan and Qiang Autonomous Prefecture	16	Chengdu City, Xichang City, Yaan City, 1300 square meters	Liuran Village site, Tangjiapo site, Muqigou site, etc. 2001–2022.
22	Brief Reports on Surveys in Other Regions	13	Aba Tibetan and Qiang Autonomous Prefecture, Meishan City, 15 square meters	Gaopo site, Pujiaoding site in Danba County, Wanghua site, etc. 2000–2014.
23	Other References	26	Chengdu City, Meishan City, Liangshan Yi Autonomous Prefecture, 15,000 square meters	The “Yongjinwan” location at Jinsha Ruins, Raojiadi site, and the “Sports Park” location at Jinsha Ruins, etc. 2001–2021.

). For sites without coordinates in the reports, latitude and longitude were obtained using the Baidu Coordinate Picker and spatially located using ArcMap 10.8. The study comprehensively covers and efficiently integrates information from archaeological reports across multiple periods, including all publicly available data published after the Atlas of Chinese Cultural Relics: Sichuan Volume. The DEM data and geomorphological data were obtained from the NASA Earthdata (https://www.earthdata.nasa.gov, accessed on 30 November 2013), specifically the GDEM V3 30 M resolution elevation dataset. Elevation and slope data for the study area were extracted from the DEM. The distribution of sites in Sichuan Province is shown in Figure 2, with blue, red, and yellow representing Stone Age, Shang to Spring and Autumn periods, and Warring States period sites, respectively. The map was created using the standard topographic map of Sichuan Province issued by the Sichuan Bureau of Surveying and Mapping. In addition, the number of site points in each geomorphological type was statistically analyzed, and the results are presented in

Table 2. Information on the number of heritage sites in each geomorphological type in Sichuan Province.

Geomorphological Type	Number of Sites	Typical Site Names
Plain	272	Zhengjiaba site, Baodun Ancient City site, etc.
Terrace	34	Tanguan Mountain site, Feihu Village Warring States Cemetery, etc.
Hilly	73	Gaojiayuanzi site, Gaoshan Ancient City site, etc.
Low Mountain	40	Baishaba site, Jiangjiaba site, etc.
Middle mountain	414	Guzhai Village site, Buwa Stone Coffin Burial, etc.
High Mountain	140	Liujiazhai site, Danidong site, etc.

.

2.3. Analysis Methods

2.3.1. Kernel Density Analysis

Kernel density estimation is a non-parametric statistical method used to estimate the probability density function of a random variable. Kernel density analysis is commonly employed to study the spatial distribution patterns of point data, revealing the clustering characteristics and density variations of these points. The method involves placing a kernel function around each point data and then superimposing all the kernel functions to obtain the density distribution across the entire area. The formula is as follows:

$$\widehat{f}\left(x\right)={\displaystyle \frac{1}{nh}}{{\displaystyle \sum }}_{i=1}^{n}K\left({\displaystyle \frac{x-X_i}{h}}\right)$$

(1)

In the formula, $\widehat{f}\left(x\right)$ represents the kernel density estimate at location $x$, which in this study refers to a site, n is the number of point data, i.e., the number of archaeological sites, and h is the bandwidth. In this study, the bandwidths used were 80 km for the Neolithic period, 120 km for the Shang to Spring and Autumn periods, and 60 km for the Warring States period. $K$ is the kernel function, and $X_i$ is the location of the point data sorted by $i$, which means the location of the site sorted by $i$.

2.3.2. Buffer Analysis

Buffer analysis is a method of spatial data analysis that automatically establishes polygonal buffer zones of a certain width around point, line, and polygon features in a database, thereby extending the spatial data in the horizontal direction. A buffer zone represents the range of influence or service area of a geographical spatial object in terms of scale. It is a dependent variable that changes with the form of the element being studied. From a mathematical perspective, a buffer zone is the neighborhood obtained for a given spatial object or set, and the size of the neighborhood is determined by the radius of the neighborhood or the conditions for establishing the buffer zone. Therefore, for a given object A, its vector buffer can be defined as:

$$P=\left\{x|d\left(x,A\right)\le r\right\}$$

(2)

In the formula, $A$ is a given spatial object or collection, $P$ represents the buffer range of object $A$, and $d$ generally refers to the Euclidean distance, although other types of distances can also be used. In this study, the Euclidean distance is employed. Additionally, $r$ is the radius of the neighborhood or the condition for establishing the buffer zone. In this study, 1 km and 3 km buffer zones around rivers were created to investigate the relationship between the distribution of archaeological sites and their distance from water sources.

2.3.3. Slope Analysis

Slope indicates the degree of inclination of the terrain at a given point. Numerically, it is equal to the angle between the normal vector $\overrightarrow{n}$ of the differential unit of the terrain surface at that point and the z-axis, that is:

$$\mathrm{slope}=\arccos \left({\displaystyle \frac{\overrightarrow{z}·\overrightarrow{n}}{\left|\overrightarrow{z}\right|·\left|\overrightarrow{n}\right|}}\right)$$

(3)

When extracting slope specifically, a simplified decomposition formula is commonly used. The complete mathematical representation is as follows:

$$\mathrm{slope}=arctan\sqrt{f_x^2+f_y^2}$$

(4)

In the formula, $f_x$ is the rate of elevation change in the X direction, and $f_y$ is the rate of elevation change in the Y direction. In this study, the slope was extracted and analyzed using DEM data with the ArcMap software 10.8.

3. Results and Discussion

3.1. Spatial and Temporal Distribution Characteristics of Sites

This study utilized the kernel density analysis method to investigate the distribution characteristics of archaeological sites from the Paleolithic and Neolithic periods, as well as the Shang to Spring and Autumn periods and the Warring States period in Sichuan Province. The results are presented in Figure 3.

During the Stone Age, the spatial distribution of sites generally aligned in an EN–WS orientation. The city of Chengdu and its surrounding areas were the most densely distributed regions for site points. This period is represented by sites such as the Baodun Ancient City, Pixian Ancient City, and Chongzhou Zizhu Ancient City, which formed the most prominent kernel density hotspots. This indicates that this region was likely a major center of human habitation during the Stone Age. These sites are typical representatives of the Baodun culture (approximately 4800 years ago), the oldest archaeological culture traceable in the Chengdu Plain. Most of these sites were settlements where ancient humans gathered, and they were gradually excavated in the 1990s. Similarly, a significant cluster of sites was observed in the southern part of the Aba Tibetan and Qiang Autonomous Prefecture, suggesting that this area may have been an important location for human activities. In the southern part of the northwestern Sichuan Plateau, counties such as Barkam and Jinchuan in the Aba Tibetan and Qiang Autonomous Prefecture were areas where various ethnic groups settled. Representative sites in this region include the Shenxianbao site and the Shiguangdong site, which are historical and cultural landscapes reflecting the convergence and integration of the Yellow River Basin culture, Yangtze River Basin culture, Central Plains culture, and ethnic cultures. In contrast, other regions, such as eastern, southern, and southwestern Sichuan, had more dispersed sites with lower densities, which could be attributed to the natural environmental conditions or patterns of human activity at the time.

During the Shang to Spring and Autumn periods, the distribution of site points became more concentrated. Notable examples include the Sanxingdui site in Guanghan, a core site of the Sanxingdui culture (approximately 4000 years ago), as well as the Jinsha site and Shierqiao site, core sites of the Shierqiao culture (approximately 3200 years ago). These sites formed a distinct cluster around the areas of Chengdu and Mianyang, indicating that this region likely served as a central hub of human activity during the Shang to Spring and Autumn periods. Additionally, Ya’an and Meishan also exhibited a considerable number of site points, and although their density was lower than that of the core areas, they still showed a notable degree of clustering, suggesting a supplementary role in human settlement. In contrast, the eastern and southern parts of the Sichuan Basin, as well as western Sichuan, had fewer site points and did not form clear high-density cores, possibly due to topographical constraints, indicating that these areas were less significant centers of human activity during this period.

During the Warring States period, the distribution of site points displayed a multi-center clustering pattern. A distinct high-density core emerged in the southern part of the Liangshan Yi Autonomous Prefecture, characterized by a dense concentration of site points. This may be attributed to the development of transportation networks, ethnic migration, and intercultural exchanges at the time. Representative sites near Liangshan Prefecture include the Muxigou site and the Wangjiatian site, among others. These sites are closely associated with the Dashi Tomb culture (approximately 3000 years ago) distributed in Liangshan Prefecture. The area near Chengdu still exhibited a significant high-density clustering of sites, indicating that by the Warring States period, the Chengdu region may have developed relatively mature settlements. However, during this time, new types of burial sites emerged, such as earthen pit tombs and boat-coffin tombs. Examples include the Warring States earthen pit tombs in the Shiren Community on the western outskirts of Chengdu and the boat-coffin tombs on Chengdu’s Commercial Street. The diversification of archaeological site types provides more comprehensive evidence for the study of Shu culture. Conversely, the eastern and western regions of Sichuan had relatively sparse site distributions without distinct high-density cores, implying that these areas were not primary centers of human activity during this period. This could be due to geographical limitations, such as the unsuitability of high-altitude regions for habitation, as well as constraints on transportation and social development at the time.

In kernel density calculations, varying bandwidths significantly influenced the results. Considering the substantial differences in the spatial distribution of archaeological site points across different historical periods in Sichuan Province, distinct bandwidths were employed for each period’s kernel density analysis. Specifically, this study selected bandwidths of 80,000 m for the Stone Age, 12,000 m for the Shang to Spring and Autumn periods, and 60,000 m for the Warring States period. These choices were guided by the unique distribution characteristics and density of site points in each era, ensuring that the results accurately reflected both localized clustering features and overall distribution trends. Overall, from the Stone Age to the Warring States period, the spatial distribution of site points in Sichuan Province transitioned from widespread dispersion to relative concentration. The continuous presence of site distributions from the Stone Age through the Shang, Zhou, Spring and Autumn, and Warring States periods indicates that these regions maintained their significance as areas of human activity over an extended period. Notably, the Chengdu Plain consistently served as a core cluster of archaeological sites. Additionally, the emergence of new high-density site areas in southern Sichuan during the Warring States period may be closely linked to factors such as the natural environment, transportation routes, cultural expansion, and economic development.

As shown in Figure 4, the spatial distribution of archaeological sites in Sichuan Province from the Stone Age to the Warring States period demonstrated a significant regional evolution trend. During the Stone Age, sites were extensively dispersed throughout the province, with Aba Prefecture (186 sites), Chengdu (78 sites), Liangshan Prefecture (52 sites), and Ganzi Prefecture (31 sites) hosting the highest concentrations. The selection of settlement locations during this period was likely influenced by natural environmental factors, such as water sources, climate, and resource availability. In the Shang to Spring and Autumn periods, the distribution of sites experienced a marked shift, with the Chengdu Plain emerging as a prominent cultural center. The number of sites in Chengdu increased dramatically to 157, while Deyang (39 sites) and Nanchong (34 sites) also became areas with relatively dense site distributions. Conversely, the number of sites in highland regions, like Aba, Ganzi, and Liangshan Prefectures, significantly decreased. This transformation can be attributed to agricultural development, advancements in social organization, and the establishment of centralized governance, leading to a concentration of human activity in the Chengdu Plain. During the Warring States period, the distribution pattern of sites underwent another adjustment. The number of sites in Chengdu sharply declined to 40, while the number of sites in Liangshan Prefecture surged to 136, forming a new high-density archaeological site area. This change may have been driven by social upheaval, population migration, and economic restructuring.

Overall, from the Stone Age to the Warring States period, the spatial distribution of archaeological sites in Sichuan Province exhibited a trend of shifting concentration from the highlands to the basin. This evolutionary process reflects changes in population movements, economic development patterns, and the transformation of social and political structures.

3.2. Relationship Between Spatial and Temporal Distribution of Sites and Environmental Factors

3.2.1. Elevation

The elevation range of the study area spans from 107 m to 7473 m. In ArcMap, site points from different periods were overlaid onto a digital elevation model (DEM) map. The number and percentage of sites within various elevation ranges were calculated and visualized in statistical charts (Figure 5 and Figure 6). During the Stone Age, the average elevation of site distribution was 1866.5 m. Approximately 20% of the sites were located below 500 m, followed by 18% within the 500–1000 m range and 14% within the 3000–3500 m range. Apart from these ranges and a small cluster (3%) within 4000–4500 m, site distribution across other elevation intervals was relatively uniform. In the Shang to Spring and Autumn periods, the average elevation of site distribution significantly decreased to 977.8 m. Human settlements were predominantly concentrated in areas below 1000 m, accounting for 76% of the total sites, with only a few scattered at higher elevations. This lower elevation distribution was mainly attributed to the concentration of sites in the Chengdu Plain. During the Warring States period, the average elevation of site distribution increased again to 1561.5 m. The most densely distributed sites were within the 1500–2500 m range, accounting for 48% of the total, while only a few sites were found above 3000 m. Overall, the analysis revealed a general downward trend in site distribution elevation from the Stone Age to the Warring States period. After dropping below 1000 m during the Shang to Spring and Autumn periods, the average elevation of site distribution rose again, which may be related to environmental changes in the Chengdu Plain.

3.2.2. Proximity to Water Sources

From the Stone Age to the Warring States period, limited agricultural techniques and flood control measures made proximity to water a critical factor in human settlement and burial site selection. Using hydrological data in ArcMap, buffer zones of 1 km and 3 km around water bodies were created and overlaid with site locations from different periods to calculate their distances to water sources. The results are summarized in

Table 3. Statistical table of distances from water sources for heritage sites across different periods.

Proximity to Water Sources		<1 km	1–3 km	>3 km
Stone Age period	Number	241	154	81
	Percent	51%	32%	17%
Shang to Spring and Autumn periods	Number	243	84	20
	Percent	70%	24%	6%
Warring States period	Number	128	99	34
	Percent	49%	38%	13%

. During the Stone Age, 51% of the sites were located within 1 km of a water source, while 32% were situated between 1 and 3 km away. The frequent flooding of rivers in the Chengdu Plain during this period may explain why some settlements were positioned more than 3 km from water sources. The Shang to Spring and Autumn periods exhibited the most pronounced preference for proximity to water, with 70% of the sites located within 1 km of a water source and only 6% situated beyond 3 km. During the Warring States period, 49% of the sites were within 1 km of water, while 38% were between 1 and 3 km away. Climate fluctuations during this period, including both droughts and floods, likely influenced site selection strategies. Overall, the proportion of sites located within the 1 km buffer zone remained consistently high across all periods, peaking at 70% during the Shang to Spring and Autumn periods and remaining around 50% during the Stone Age and Warring States periods. This pattern underscores the strong preference of ancient populations for settling near water sources to support daily life and agricultural activities, highlighting a distinct tendency toward water proximity.

3.2.3. Slope

Using DEM data, the slope of Sichuan Province was extracted and classified according to the national Technical Regulations for Land Use Status Survey [36]. The site locations from different periods were overlaid with the slope classification to generate distribution maps (Figure 7). During the Stone Age, sites predominantly occurred in areas with slopes of 6–15° and >25°, while regions with slopes below 2° had relatively few sites. A notable concentration of sites appeared within Aba Tibetan and Qiang Autonomous Prefecture, characterized by generally steep terrain. In the Shang to Spring and Autumn periods, sites were mainly distributed in areas with slopes of 2–6° and 6–15° within the Chengdu Plain, whereas in other slope ranges, they exhibited a more even distribution. During the Warring States period, the overall distribution of sites became more balanced, with the highest concentration observed in areas with slopes of 6–15°. According to the distribution maps and statistical results, the majority of sites across all historical periods were located in areas with slopes below 15°. The number of sites in this slope range peaked during the Shang to Spring and Autumn periods, mirroring the pattern observed in the relationship between site distribution and elevation.

3.3. Discussion

The spatial distribution of archaeological sites is closely influenced by factors such as elevation, proximity to water sources, slope, and the climatic and environmental conditions of different historical periods. Therefore, analyzing the spatial and temporal patterns of site distribution is of significant value in understanding historical human–environment interactions.

This study covered a longtime span from the Stone Age to the Warring States period, encompassing multiple cultural development stages and various types of sites across the entire province. In contrast, most other studies focused on a single period, with relatively shorter timeframes, often limited to specific phases of the Paleolithic or Neolithic eras. For example, the study of Kůlna Cave [20] is confined to cave sites, while research in the Cantabrian region [19] primarily examines cave and rock shelter sites along specific coastal areas. Thus, there are differences in the scope and types of research subjects. This study was based on a comprehensive dataset compiled from historical atlases, archaeological survey reports, and relevant literature, including 476 sites from the Stone Age, 347 from the Shang to Spring and Autumn periods, and 261 from the Warring States period. For sites lacking explicitly recorded coordinates, we employed a coordinate retrieval system to determine their locations based on place names and geographical descriptions, thereby ensuring the accuracy and completeness of the dataset. This approach enabled a more systematic and representative analysis of site distribution across different historical periods.

Compared to previous studies, this research incorporated a more extensive and up-to-date dataset, providing a valuable supplement to existing findings. Prior research has documented 151 sites in Sichuan from the Paleolithic to the Bronze Age, revealing an evolutionary shift from a multi-core distribution pattern to a single-core pattern centered on the Chengdu Plain, with varying degrees of environmental influence across different periods [2]. In contrast, this study classified sites according to historical periods and employed distinct analytical methodologies, yielding more detailed insights. During the Stone Age, site distribution exhibited a dispersed, multi-core pattern. By the Shang to Spring and Autumn periods, settlement patterns became more concentrated, forming a prominent core in the Chengdu Plain. In the Warring States period, a new high-density site cluster emerged in southern Sichuan, particularly in Liangshan Yi Autonomous Prefecture. While the overall trends in site distribution aligned with previous findings, differences in data volume and kernel density bandwidth selection have resulted in distinct spatial distribution patterns for each period.

Despite its contributions, this study has certain limitations. In terms of data, while significant efforts were made to ensure the precise location of sites, some records with insufficient location details had to be excluded. Additionally, the study primarily explored the relationship between site distribution and historical climatic conditions from a qualitative perspective. Future research could integrate historical climate data for a more rigorous quantitative analysis, further elucidating the interplay between environmental changes and human settlement patterns over time.

4. Conclusions

• This study systematically compiled site data across the entire region of Sichuan Province and employed GIS spatial analysis methods to explore the distribution characteristics of sites and their relationship with natural geographical factors from the geographical spatial perspective. By addressing the comprehensiveness of the data and utilizing multi-dimensional GIS analytical approaches, we expanded the findings presented in other studies within this field. The research findings analyzed the spatial distribution characteristics of site distributions in Sichuan Province from perspectives such as clustering features and water adjacency. Furthermore, we investigated the reasons influencing the distribution of remains based on factors including altitude, slope, and distance from water sources.
• The kernel density values of site distribution in Sichuan Province from the Stone Age to the Warring States period initially increased and then decreased, indicating a clustering phenomenon in site distribution, with the highest density occurring during the Shang to Spring and Autumn periods. This reflects a trend from dispersion to concentration. The Chengdu Plain remained the most central area of site aggregation throughout this period. From the Warring States period, a new high-density site cluster emerged in southern Sichuan, centered around the Liangshan Yi Autonomous Prefecture. During the Stone Age, sites were widely distributed across the province. However, by the Shang to Spring and Autumn periods, the Chengdu Plain became the cultural center, and the number of sites in Chengdu increased significantly. The flat terrain and fertile soil of the Chengdu Plain, combined with advancements in irrigation technology, led to a substantial increase in agricultural output. This attracted large populations to settle in the area, resulting in a higher number of sites and a more concentrated distribution. During the Warring States period, the distribution pattern of sites in Sichuan Province underwent another adjustment. The number of sites in Chengdu decreased, while the number of sites in Liangshan Prefecture surged, forming a new high-density site cluster.
• The study considered the influence of elevation, distance to water, and slope on site distribution. The results indicated that from the Stone Age to the Warring States period, the elevation of site distribution in the study area first decreased and then increased. During the Shang to Spring and Autumn periods, the average elevation of sites was 1866.5 m, which dropped to 977.8 m during the Shang to Spring and Autumn periods. During the Shang to Spring and Autumn periods, sites were primarily concentrated in the low-elevation areas of the Chengdu Plain, a distribution pattern likely closely related to environmental changes in the region. The Chengdu Plain during this period had a suitable climate and abundant water resources, making low-elevation areas more conducive to agricultural production and daily life, which explains the high concentration of sites there. Additionally, due to limitations in farming techniques and flood control technology, human settlements exhibited a significant preference for proximity to water. Approximately 50% of sites during the Stone Age and Warring States periods were located within 1 km of water sources, while this proportion rose to 70% during the Shang to Spring and Autumn periods. This may be because, during the Stone Age, the threat of river flooding was greater, leading people to settle farther from water. By the Warring States period, climate fluctuations caused alternating droughts and floods, forcing people to balance the convenience of accessing water with the need to reduce flood risks when choosing settlement locations. Furthermore, sites in the study area were mainly distributed in areas with slopes below 15°. During the Shang to Spring and Autumn periods, the highest concentration of sites was found in areas with slopes between 2° and 15°, reflecting a similar relationship to elevation. This demonstrates the preference and adaptation of ancient humans for terrain slope when selecting settlement locations.
• During the Stone Age to the Warring States periods, Sichuan Province experienced a generally warm and humid climate, with a gradual decline in humidity over time and a reduction in flood frequency. In the Stone Age, frequent flooding of the Minjiang River posed threats to habitats, leading to higher elevations and steeper slopes for settlement sites. From the Shang to the Spring and Autumn periods, the climate in the Chengdu Plain was most favorable, resulting in lower elevations and concentrated distribution of sites in this area. By the Warring States period, settlements were located farther from rivers, likely to mitigate flood risks, a trend also linked to advancements in water conservancy projects, like the Dujiangyan. Overall, the climate in Sichuan transitioned from warm and humid to drier during this period, coinciding with improvements in ancient human agricultural practices and water management, which were closely related to the distribution of settlement sites.