Spatiotemporal evolution and human-environment relationships of early cultural sites from the Longshan to Xia-Shang periods in Henan Province, China

Abstract

Henan Province, a pivotal region for the origin and development of early Chinese civilization, is home to numerous early cultural sites that serve as vital material carriers documenting early human society. A systematic study of the evolution of these cultural sites in Henan and their relationships with the environment can elucidate the developmental patterns of early state civilizations in the region, reveal the cultural significance of these sites, and provide scientific evidence for their protection and utilization. This study employs GIS spatial analysis techniques to unveil the spatial evolution of cultural site distribution in Henan Province from the Longshan period to the Xia and Shang periods, including characteristics of spatial clustering, distribution density, and directional trends. Furthermore, it examines the relationship between the spatial distribution of cultural sites and environmental factors across different periods. The results indicate: (1) early cultural sites in Henan Province exhibit spatial clustering patterns. During the Longshan period, these sites had the widest distribution and the highest number. By the Xia Dynasty, the number of sites significantly decreased, concentrating in the middle reaches of the Yellow River. By the Shang Dynasty, the distribution expanded northward and southward within Henan Province. (2) Early cultural sites in Henan Province are primarily located in plains, basins, and low mountain and hill areas with elevations below 400 m and gentle slopes. Site selection favored sun-facing slopes with ample sunlight and areas that were convenient for water access and less prone to flooding. (3) Early settlers preferred establishing their communities in areas with relatively complex geographical diversity, offering more possibilities for production and daily life. However, excessive geographical diversity increased the difficulty for early humans to adapt and modify the environment, thus limiting their activities and reducing the number of sites. (4) The results of the geographically weighted regression analysis indicate that there are significant spatial differences and heterogeneity in the influencing factors during the Longshan period and Shang Dynasty, while the factors during the Xia Dynasty did not exhibit statistically significant spatial heterogeneity.

Introduction

The origin of early Chinese civilization resembled a “sky full of stars” pattern, but around 4300 years ago, most regional cultures declined successively, while the Central Plains culture, represented by Henan Province, continued to develop and grow due to its advantageous geographical location, eventually transitioning into a mature expansive monarchical state. Henan Province, a crucial region for the origin and development of early Chinese civilization, is home to numerous early cultural sites dating from the Neolithic to the Xia and Shang periods. Cultural sites served as the settlements of early human societies and are precious immovable cultural heritage. Systematically studying the evolution of these early cultural sites and their human-environment relationships in Henan Province can uncover the developmental patterns of early state civilizations, explore the cultural significance of these sites, and provide a scientific basis for their protection and utilization. Scholars’ research on cultural sites can be categorized into the following types: (1) archeological research, which primarily uses archeological methods to analyze the exchange and interaction of cultural factors among sites within a region, ultimately revealing the social organization and complexity of ancient societies^1,2,3. (2) Research on the spatiotemporal distribution of cultural sites, which explores the distribution characteristics and current preservation status of cultural sites, and proposes protective measures based on their present conditions and risks^4,5. (3) Research on the human-environment relationships of cultural sites, which delves into the spatial distribution of these sites to understand the relationship between human activities and the natural environment in historical periods, thereby providing a scientific basis for the effective management and protection of historical cultural sites^6,7,8,9,10. The human-environment relationship refers to the ongoing interaction of human society, where humans, driven by the need for survival, continuously strengthen their capacity to adapt to the geographical environment, while the geographical environment profoundly influences human activities. (4) Case studies of cultural sites, which include uncovering the cultural relics of individual sites, their spatial forms, and the aspects of ancient societies they reflect^11,12,13. Through analyzing the current status and value of the sites, these studies promote the protection and utilization of the sites from the perspective of local cultural heritage authorities^14,15. (5) Research on the display and utilization of cultural sites, which involves the application of new technologies such as augmented reality and virtual reality in the presentation of cultural sites^16,17,18,19, as well as discussions on restoration and dynamic display measures for archeological sites^20,21. (6) Research on the theoretical framework for the protection of cultural sites, which aims to develop theoretical methods and research frameworks for the protection of cultural sites that are tailored to national conditions^22,23.

Currently, research on cultural sites is abundant, encompassing multiple scales and perspectives, providing numerous approaches and methods for studying the spatiotemporal evolution of cultural sites. Most related studies are conducted from archeological and cultural heritage perspectives, addressing specific issues of cultural disclosure and protection. In terms of the research area and temporal scope, the findings still need to be further enriched, and there is a lack of specialized studies on the transitional stage from the late Neolithic to the formation of early monarchic states. Additionally, scientific quantitative methods for evaluating geographic factors are still lacking. The practical significance of this study lies in understanding the temporal evolution and causes of early cultural sites, providing a research foundation for scholars to interpret the cultural value of such heritage sites, and offering scientific guidance for developing strategies for their protection and utilization. In early human societies, low productivity meant that the geographical environment played a decisive role in the development of civilization. Exploring the evolution of early cultural sites in this region and their human-environment relationship helps to gain a deeper understanding of the patterns of early state civilization’s origin and development. Therefore, in this study, we focused on Henan Province and explored the spatiotemporal evolution of cultural sites from the Longshan period to the Xia and Shang Dynasties, as well as their human-environment relationship. This period witnessed the emergence of early statehood (Longshan period) and the formation and maturation of early states (Xia and Shang Dynasties). During the research, approximately 2500 early cultural sites identified in Henan Province were collected to ensure that the analysis results hold statistical significance. In exploring influencing factors, GIS spatial analysis techniques were used to improve the evaluation methods of hydrological and geographical diversity factors.

Data and methods

Research scope

Temporal scope

The origin and formation of early Chinese state civilizations can be divided into two stages, corresponding to the Longshan period of the late Neolithic and the Xia and Shang periods. Among them, the absolute dates of the Longshan period are approximately 2500-2070 BCE, the Xia Dynasty is around 2070-1600 BCE, and the Shang Dynasty is around 1600-1046 BCE. During the Longshan period, large walled cities or settlements with early urban characteristics began to appear, and many clusters of city sites were discovered in the Henan region. Social stratification, including wealth disparity, had already emerged within these large settlements or cities. This assertion has also been confirmed by archeological findings. For example, in the ancient city-site remains of the Longshan period, large palace foundations were discovered, which were used by the upper aristocracy for residence or political activities. This phenomenon also appears widely in many other sites²⁴. After the social development of the Longshan period, by the Xia Dynasty, the previously fragmented landscape of competing polities during the Longshan period had transformed, with fewer divisions between city sites. Large urban centers became regional hubs, and the Central Plains region, centered on Henan, entered the era of kingdoms, marking the emergence of China’s first monarchical state—the Xia Dynasty. By the Shang Dynasty, the Central Plains remained the focal point, and the Shang Dynasty continuously expanded its territory outward. Due to the geographic characteristics of the Central Plains region, diverse cultures from the origins and formation periods of civilization converged here. The local Longshan culture in Henan had strong cohesion and centripetal force, allowing the formation of Chinese civilization, the emergence and development of early states, and eventually the birth of the Xia and Shang Dynasties during the kingdom era²⁵. Based on this, the temporal scope of this study is confined to the Longshan period through to the Xia and Shang periods.

Spatial scope

Henan Province extends 530 km from north to south and 580 km from east to west. It borders Anhui and Shandong to the east, Hebei and Shanxi to the north, Shaanxi to the west, and Hubei to the south, forming a broad expanse that stretches from north to south and bridges east to west (Fig. 1). The total area of the province is 167,000 square km, accounting for 1.73% of the national land area. The topography of Henan is characterized by higher elevations in the west and lower elevations in the east. The Taihang Mountains, Funiu Mountains, Tongbai Mountains, and Dabie Mountains form a semicircular distribution along the northern, western, and southern borders. The central and eastern regions consist of the Huanghuaihai alluvial plain, while the southwestern region features the Nanyang Basin. The plain and basin areas constitute 55.7% of the total area, while mountainous regions cover 26.6%, and hilly areas account for 17.3%. The formation of the four great ancient civilizations of the world was dependent on large river basins, and the differences in basin environments led to cultural diversity. Henan Province, located in the heart of China, is the primary birthplace of the Yellow River civilization. Against the backdrop of various cultures gradually declining in other regions, Henan’s unique geographical position allowed early states to form and persist here.

Fig. 1: Study area.

a Location of Henan Province (Base map sourced from the National Geographic Information Public Service Platform, review number: GS (2024) 0650, website: https://www.tianditu.gov.cn/). b Topographic Map of Henan Province.

Data

The sites involved in this study are sourced from the list of the Third Cultural Relics Census of Henan Province (source: Henan Provincial Cultural Heritage Administration). This data was collected by the State Council based on the “National Cultural Development Plan Outline during the 11th Five-Year Plan.” The classification of the time periods for the sites in the list was primarily conducted by archeologists using methods such as stratigraphy and typology, combined with the cultural characteristics. This study extracted the Longshan period, Xia Dynasty, and Shang Dynasty cultural sites based on their classification results. In the statistical analysis of cultural sites, some sites have a continuous existence over time, and their cultural remains may span multiple periods. To avoid errors, such cultural sites will be counted multiple times across the classifications of cultural sites from different periods. This study identified a total of 1348 cultural sites from the Longshan period, 225 from the Xia Dynasty, and 1004 from the Shang Dynasty. Previous research indicates that low productivity and inconvenient transportation in early human societies, along with factors such as elevation, slope, aspect, hydrology, and geographical diversity, were dominant constraints on their production and living conditions. Thus, these factors became key considerations for early human settlement location choices^6,7,8,9,10. The DEM elevation data was sourced from Google Maps download software, while slope, aspect, and hydrological data were calculated using the DEM elevation data with ArcGIS Pro software. Geological, soil, and landform data were obtained from publicly available data on the website of the Ministry of Natural Resources of China (source: https://www.mnr.gov.cn/). In this study, the GIS spatial connection function was used to establish links between these natural geographic data and cultural site point data to explore the relationship between the two.

Methods

Nearest neighbor index

By conducting nearest neighbor index analysis on cultural sites, the spatial distribution patterns of cultural sites within the study area are determined, and the proximity values of each cultural site point within the study area are obtained. The classification of spatial distribution types for point features is based on the nearest neighbor index: if the nearest neighbor index is less than 1, the spatial distribution type is clustered; if the nearest neighbor index is equal to 1, the spatial distribution type is random; if the nearest neighbor index is greater than 1, the spatial distribution type is uniform. Here, $\bar{r}1$ represents the average distance between the nearest points, $rE$ represents the average distance in a random distribution model, and D represents point density. Finally, the nearest neighbor index R is calculated^26,27,28. The calculation formula is shown in formula (1):

$$R=\frac{\overline{r}1}{rE}=2\sqrt{Dr1}$$

(1)

Kernel density estimation (KDE)

The application of KDE can reflect the spatial distribution density of cultural sites. The kernel density values of cultural sites indicate their density: the larger the kernel density value, the denser the distribution of cultural sites. KDE is the probability of the density function f at point x, where $k(\frac{x-Xi}{h})$ is the kernel function, h represents a search radius greater than 0, and x − $Xi$ represents the distance (km) between the estimation point x and the event $Xi$^29,30,31. The calculation formula is shown in formula (2):

$$fn(x)=\displaystyle\tfrac{1}{{\rm{nh}}}\mathop{\sum }\limits_{i=1}^{n}k\left(\frac{x-Xi}{h}\right)$$

(2)

Standard deviation ellipse

The standard deviation ellipse (SDE) is used to determine the distribution direction of cultural heritage sites. The long axis of the ellipse represents the primary direction of site distribution, the short axis represents the secondary direction, and the tilt angle of the ellipse indicates the distribution trend^32,33,34. The closer the ratio of the long axis to the short axis of the ellipse is to 1, the more the shape of the ellipse approaches a circle, indicating weaker directionality in the distribution of cultural heritage sites. Conversely, a greater disparity indicates stronger directionality. The standard deviation ellipse is shown in Eq. (3), where x_i and y_i are the coordinates of the points, n is the total number, and {$\bar{{\rm{x}}}$, $\bar{y}$} is the mean center of the points. The angle of rotation of the ellipse is calculated as shown in Eq. (4), which $\tilde{{\rm{x}}}{\rm{i}},\tilde{{\rm{y}}}{\rm{i}}$ represents the deviation of the x and y coordinates from the mean center. The standard deviations of the x and y axes are calculated as shown in Eq. (5).

$$\begin{array}{c}SDEX=\sqrt{\frac{\mathop{\sum}\nolimits_{i=1}^{n}{(xi-\overline{x})}^{2}}{n}}\\ SDEy=\sqrt{\frac{\mathop{\sum }\nolimits_{i=1}^{n}{(yi-\overline{y})}^{2}}{n}}\end{array}$$

(3)

$$\begin{array}{c}\tan \theta =\frac{A+B}{C}\\ A=\left(\mathop{\sum }\limits_{i=1}^{n}\tilde{x}{i}^{2}-\mathop{\sum }\limits_{i=1}^{n}\tilde{y}{i}^{2}\right)\\ B=\sqrt{{\left(\mathop{\sum }\limits_{i=1}^{n}\tilde{x}{i}^{2}-\mathop{\sum }\limits_{i=1}^{n}\tilde{y}{i}^{2}\right)}^{2}+4{\left(\mathop{\sum }\limits_{i=1}^{n}\tilde{x}i\tilde{y}i\right)}^{2}}\\ C=2\mathop{\sum }\limits_{i=1}^{n}\tilde{x}i\tilde{y}i\end{array}$$

(4)

$$\begin{array}{l}\sigma x=\sqrt{\frac{\mathop{\sum }\nolimits_{i=1}^{n}{(\tilde{x}i\cos \theta -\tilde{y}i\sin \theta )}^{2}}{n}}\\ \sigma y=\sqrt{\frac{\mathop{\sum }\nolimits_{i=1}^{n}{(\tilde{x}i\sin \theta -\tilde{y}i\cos \theta )}^{2}}{n}}\end{array}$$

(5)

Spatial autocorrelation

Spatial autocorrelation analysis can characterize the spatial clustering features of cultural sites. Among these, global spatial autocorrelation is used to describe the spatial clustering characteristics of attribute values across the entire region, effectively indicating whether cultural sites in Henan Province during different periods exhibit significant clustering. Global spatial autocorrelation is quantified using the Moran’s I index, and its calculation formula is as follows:

$$\begin{array}{c}I=\frac{\sum {w}_{ij}({x}_{i}-x)({x}_{j}-x)}{{s}^{2}\sum \sum {w}_{ij}}\\ {s}^{2}=\frac{\sum {({x}_{i}-x)}^{2}}{n},x=\sum \frac{{x}_{i}}{n}\end{array}$$

(6)

In the formula, n is the number of geographic units; x_i and x_j are the values of units i and j, respectively; x is the mean of all units; and w_ij is the spatial weight matrix element, where w_ij equal to 1 when i and j are spatially adjacent, and w_ij equal to 0 when they are not. When Moran’s I index is greater than 0, the cultural sites in various regions show a positive spatial correlation; if less than 0, it shows a negative spatial correlation; and if equal to 0, it indicates a random distribution of cultural sites.

However, global spatial autocorrelation cannot reflect the specific correlations between individual grids and is often combined with local spatial autocorrelation analysis to further explore the relationships among grids. This study employs hot spot and cold spot analysis to further investigate the spatial autocorrelation characteristics among grids, aiming to analyze the distribution patterns of clustered and dispersed regions. Based on the clustering characteristics of cultural sites, the study sets the grid size to 15 × 15 km². Hot spots are regions with elevated data values, signifying that the observations in these areas are notably higher than in neighboring regions. Cold spots are areas characterized by lower data values, where observed values are substantially lower than in adjacent regions, and its calculation formula is as follows:

$$\begin{array}{c}{G}_{{\rm{i}}}={\sum }_{i}{W}_{ij}{x}_{i}/{\sum }_{j}{x}_{j}\\ Z={G}_{i}-E({G}_{i})/\sqrt{Var({G}_{i})}\end{array}$$

(7)

In the formula, x_i and x_j denote the concentration of cultural sites in grids i and j, respectively; W_ij is the spatial weight function; E(G_i) represents the expected value; and Var(G_i) is the coefficient of variation³⁵.

Geographical factors statistical analysis model

(1) Terrain and hydrological variability model

Henan Province’s terrain slopes from high in the west to low in the east, with varying elevations affecting transportation accessibility, which in turn influences early human settlement choices. Relevant studies indicate that slope gradient affects the scale and intensity of surface material flow and energy conversion, and areas with steep slopes are more prone to landslides and debris flows. Thus, gentler slopes are more suitable for human habitation. The slope is classified into four levels: Level 1 (below 3°, very suitable for human habitation), Level 2 (3° to 6°, moderately suitable for human habitation), Level 3 (6° to 10°, less suitable for human habitation), and Level 4 (above 10°, unsuitable for human habitation)^8,34. Aspect is a critical factor in determining the amount of sunlight an area receives, directly impacting habitation, heating, and agricultural production. Thus, this aspect significantly influences settlement location choices³⁶. This study used a GIS platform to spatially link DEM, slope, and aspect data with cultural site data, statistically analyzing the relationship between topography and cultural sites. Finally, distribution maps and scatter plots were used to determine the impact of topographic factors on the spatial distribution characteristics of cultural sites.

Additionally, the Longshan to Xia and Shang periods, approximately 4350–3000 years ago, had significantly different river conditions compared to present-day rivers. Traditional studies often explore the influence of water resources on early human societies by examining the distance between cultural sites and rivers. While this method is reasonable^37,38,39, river course changes introduce considerable uncertainties and errors. Therefore, this study abandoned this approach and did not consider the influence of specific rivers at a micro level. Instead, it attempted to explore the impact of hydrological factors on the spatial distribution of cultural sites from a macro perspective by establishing a hydrology-cultural site model. First, using GIS hydrological analysis tools, surface runoff areas (low-lying, water-collecting regions) were extracted and vectorized. The spatial distribution density of these surface runoff areas was analyzed, and then linked to cultural site data. The number of cultural sites corresponding to each surface runoff density interval was statistically analyzed to explore the influence of hydrological factors on cultural site distribution (Fig. 2)⁸.

Fig. 2

Hydrological-cultural site model.

(2) Geographic diversity-cultural site model

Geographic diversity refers to the richness of geographical features in a region, encompassing topography, soil, geology, and hydrology^40,41,42. Geographic diversity is closely related to human production and life, particularly in agricultural production. Settlers must consider various factors, including soil type, the likelihood of geological disasters, and the convenience of water access. When exploring the impact of a single factor on human settlement, conclusions drawn are often one-sided and have significant limitations, making it difficult to reveal the combined influence of the four major factors: landform, soil, geology, and hydrology on human settlement site selection. Geographic diversity can, to some extent, connect these four geographical factors⁴³. The evaluation of geographic diversity can be achieved through GIS spatial analysis. This study utilizes this technology to construct a geographic diversity evaluation model for Henan Province, and establishes spatial connections with cultural site data to explore the impact of geographic diversity on the distribution characteristics of cultural sites. First, topography, soil, geology, and hydrology are loaded onto the GIS platform and combined into a single dataset. The centroids of the resulting polygons are extracted as point features, and the kernel density of these points is analyzed. The kernel density values represent the geographic diversity index of the region. Finally, the kernel density data is spatially connected with cultural site data, and the number of cultural sites corresponding to each density interval is calculated to explore the impact of geographic diversity on the distribution of cultural sites (Fig. 3).

Fig. 3

Geographic diversity-cultural site model.

Geographically weighted regression (GWR)

The GWR model is a regression model with geographic variation parameters that adhere to the “first law of geography.” It uses spatial relationships as weights in the calculations, enabling the quantification of spatial heterogeneity. The fundamental principle involves using the location of each explanatory variable to measure spatial distances point by point within a certain range, with the weights of explanatory variables changing according to distance⁴⁴. The calculation formula is as follows:

$${y}_{i}={\beta }_{0}({\mu }_{i},{v}_{i})+{\mathop{\sum}\limits_{i=1}^{k}\beta }_{k}({\mu }_{i},{v}_{i}){x}_{ik}+{\varepsilon }_{i}$$

(8)

In the formula, y_i represents the dependent variable; x_ik represents the observation values of the independent variables; (µ_i, v_i) represents the geographic coordinates; β₀(µ_i, v_i) represents the constant term; β_k(µ_i, v_i) represents the regression coefficient; and ε_i represents the random error term.

Research framework

This study utilizes GIS spatial analysis technology to first construct a geospatial database of early cultural sites in Henan Province. Building on this, it combines the nearest neighbor index, KDE, and standard deviation ellipse methods to investigate their spatiotemporal distribution characteristics. Finally, cultural sites are spatially linked with elevation, slope, and aspect to develop hydrology-cultural site and geographic diversity-cultural site models. Using statistical analysis methods, the study explores the impact of the natural geographic environment on the distribution of cultural sites (Fig. 4).

Fig. 4

Research framework.

Results

Spatiotemporal distribution characteristics

Distribution types

According to the nearest neighbor index analysis results, the nearest neighbor index for Longshan period cultural sites is 0.63, for Xia period cultural sites is 0.45, and for Shang period cultural sites is 0.67. All three values are less than 1, indicating that the distribution of cultural sites from the Longshan period to the Shang period is spatially clustered (Table 1).

Table 1 Nearest neighbor index analysis results

Distribution density

According to the results of the KDE analysis, during the Longshan period, cultural sites formed two high-density clusters in the Zhengzhou–Luoyang area, displayed a belt-like distribution in northern Henan, and were primarily patchy in other regions (Fig. 5a). By the Xia period, cultural sites were mainly concentrated in the Zhengzhou–Luoyang area, forming two high-density clusters, with other regions showing a scattered distribution (Fig. 5b). By the Shang period, three high-density clusters had formed in the Zhengzhou–Luoyang and northern Henan areas, with a high-density distribution belt appearing in northern Henan, while eastern and southern Henan mainly exhibited a patchy distribution (Fig. 5c).

Fig. 5: Kernel density distribution of cultural sites in different periods.

a Longshan period. b Xia Dynasty. c Shang Dynasty.

The Longshan period was a significant transitional phase in history, where the development of productivity facilitated population growth, leading to frequent warfare and intense competition, resulting in substantial changes in settlements and society. During this period, cities had become widespread, and primitive settlement forms gradually differentiated into cities and early villages. The main Longshan sites in the Henan region include Wangchenggang Site, Guchengzhai Site, Pingliangtai Site, Mengzhuang Site, Haojiatai Site, Puchengdian Site, Chaiku Site, Hougang Site, Xubao Site, Xijincheng Site, and Renhezhai Site, among which Wangchenggang Site and Guchengzhai Site exhibited characteristics of early urban centers. In terms of public safety, most villages during this period should have belonged to a certain city, with the responsibility for the public safety of the villages primarily resting with the city⁴⁵. City fragmentation was the main social pattern during the Longshan period; therefore, the kernel density analysis results showed a phenomenon of “multiple areas of concentrated habitation with few dispersed settlements.” By the Xia Dynasty, the method of power succession shifted from a system of abdication to hereditary succession, marking a strengthening of authority, as special public power began to override society. The social structure further integrated, leading to changes in cites fragmentation, and the number of cities significantly decreased. Currently, archeological work has discovered sites such as Xinzai Site, Erlitou Site, Dashigu Site, Dongzhao Site, Wangjinglou Site, Puchengdian Site, and Mengzhuang Site, with Xinzai Site and Erlitou Site serving as the capitals of the early and late Xia Dynasty, respectively. Archeological research confirms that the Xia Dynasty emerged as a widespread royal state, possessing a strong center of power⁴⁶. Therefore, in the kernel density analysis results, compared to the Longshan period, high-density clusters significantly decreased, with the distribution of various sites being relatively dispersed, demonstrating a pattern of expansion from the city center outward. By the Shang Dynasty, building on the foundation of the Xia Dynasty, the Shang Dynasty continually expanded its territory and influence, broadening its realm. Currently, the archeological discoveries of Shang Dynasty cities include the Zhengzhou Shang Dynasty Site, the Yanshi Shang Dynasty Site, the Huanbei Shang Dynasty Site, the Yinxu Site, the Wangjinglou Site, the Fucheng Site, and the Mengzhuang Site. The Shang Dynasty experienced multiple capital relocations, with the main capitals being the Yanshi Shang Dynasty Site, the Zhengzhou Shang Dynasty Site, the Huanbei Shang Dynasty Site, and the Yinxu Site. Therefore, in the kernel density analysis results, high-density clusters appear near each capital, and based on the foundation of the Xia Dynasty, the distribution of sites spreads toward the southern and northern regions of Henan.

Distribution direction

The standard deviation ellipse analysis reveals a west-to-east shift in the overall distribution pattern of cultural sites from the Longshan period to the Xia and Shang periods. During the Longshan period, the standard deviation ellipse covered an area of 194,908 km² (Fig. 6a); by the Xia period, this area had reduced to 81,851 km² (Fig. 6b); and by the Shang period, it had expanded to 211,853 km² (Fig. 6c). This indicates a significant contraction in the distribution range from the Longshan to the Xia periods, followed by a gradual expansion during the Shang period, eventually encompassing the entire province (Fig. 6). The ratio of the major to minor axes of the standard deviation ellipse shows that the directionality of cultural site distribution increased progressively from the Longshan to the Xia and Shang periods (Table 2). The reason is that during the Longshan period, the fragmentation of city sites resulted in a relatively dispersed distribution of cultural sites within Henan Province, leading to a generally weaker directional trend. By the Xia Dynasty, after the formation of early royal states, social structures became more integrated, leading to a contraction in the distribution range of sites and a stronger directional trend compared to the Longshan period. In the Shang Dynasty, building on the foundation of the Xia Dynasty, the Shang Dynasty further developed toward southern and northern Henan, leading to an increased distribution range of sites and an even stronger directional trend compared to the Xia Dynasty.

Fig. 6: Distribution direction of cultural sites in different periods.

a Longshan period. b Xia Dynasty. c Shang Dynasty.

Table 2 Standard deviation ellipse analysis results

Spatial autocorrelation

In this study, based on the distribution characteristics of the site clusters, we use a 15 × 15 km² grid as the basic scale for analysis. The global spatial autocorrelation analysis shows that the global Moran’s I index for the Longshan period cultural sites is 0.51, with a p-value of 0, indicating a statistically significant positive spatial correlation, and the distribution of cultural sites between grids exhibits a spatial clustering trend (Fig. 7a). The global Moran’s I index for the Xia Dynasty cultural sites is 0.38, with a p-value of 0, indicating a statistically significant positive spatial correlation, and the distribution of cultural sites between grids exhibits a spatial clustering trend (Fig. 7b). The global Moran’s I index for Shang Dynasty cultural sites is 0.31, with a p-value of 0, indicating a statistically significant positive spatial correlation, and the distribution of cultural sites between grids exhibits a spatial clustering trend (Fig. 7c).

The local spatial autocorrelation of cultural site distribution between grids is further revealed through cold-spot and hot-spot analysis. The results indicate that high-value areas (hot spots) for cultural sites in each period are concentrated in the Zhengzhou and Luoyang areas. No cold spots are observed in the study area, and local spatial autocorrelation between grids in other regions is not significant, suggesting that the distribution of cultural sites is relatively dispersed (Fig. 7).

Fig. 7: Hot-spot and cold-spot distribution map.

a Longshan period. b Xia Dynasty. c Shang Dynasty.

Human-environment relationship coupling analysis

Elevation factors

According to the elevation analysis results, it can be observed that the distribution of cultural sites from various periods exhibits a trend of initially increasing and then decreasing with rising elevation, predominantly concentrated within the 200-m elevation zone (Fig. 8). Specifically, the Longshan period cultural sites are most densely distributed within the 0–100 m elevation range, accounting for 39.5% of the total sites from that period. Xia period cultural sites are predominantly found within the 100–200 m elevation range, constituting 62.7% of the total sites. Shang period cultural sites are most concentrated within the 0–100 m elevation range, making up 51.9% of the total sites (Table 3). The results indicate that, within the Henan Province, areas with elevations exceeding 300 m have a significant restrictive effect on the settlement choices of early human societies (Fig. 9). This is because by the end of the Neolithic period, primitive agriculture had advanced, and agricultural production in Henan had formed a diverse cropping system focused on millet and sorghum, encompassing at least four crop types: millet, sorghum, rice, and soybeans. This multifaceted agricultural system was refined in Henan⁴⁷. Furthermore, low-altitude regions are mostly alluvial plains where the fine loess is easy to cultivate and the fertile soil supports agricultural activities⁴⁸. As a result, during the shift from a more primitive gathering and hunting livelihood to an agricultural economy, the loess-rich plains emerged as the primary choice for early ancestors involved in agricultural production. Moreover, during the Longshan to Shang and Zhou periods, the predominant means of travel for common people was still on foot, with the low-altitude plains requiring the least time and physical effort, which was more conducive to permanent settlement.

Fig. 8

Distribution of cultural sites at different elevations.

Table 3 Distribution quantities of cultural sites at different elevations

Fig. 9

Scatter plot of proportional distribution of cultural sites at different elevations

Slope factors

The slope analysis results reveal that during the Longshan period, 98.3% of cultural sites were located within a slope range of 0–3° (Fig. 11a); during the Xia period, this proportion increased to 99.1% (Fig. 11b); and during the Shang period, 98.5% of the sites were within the same slope range (Fig. 11c, Table 4). The results indicate that slope has a significant impact on the settlement choices of early human societies, with areas exceeding a 3° slope exerting a noticeable restrictive effect on settlement locations (Figs. 10 and 11).

Table 4 Distribution quantities of cultural sites at different slopes

Fig. 10

Distribution of cultural sites at different slopes.

Fig. 11: Scatter plot of proportional distribution of cultural sites at different slopes.

a Longshan period. b Xia Dynasty. c Shang Dynasty.

Aspect factors

Research indicates that south-facing slopes are the most suitable for human habitation, southeast-facing slopes are also suitable, while north-facing and northwest-facing slopes are relatively more suitable. Conversely, east-facing, northeast-facing, west-facing, and southwest-facing slopes are considered less suitable for human habitation⁴⁹. According to the slope orientation analysis, cultural sites from various periods are predominantly distributed on sun-facing slopes (Fig. 12). Among these, south-facing slopes account for the largest proportion of sites: 15.3% during the Longshan period, 14.7% during the Xia period, and 13.5% during the Shang period (Table 5). The results suggest that early human societies preferred sun-facing areas for settlement. From the Longshan to the Xia and Shang periods, the proportion of south-facing cultural sites gradually decreased. This trend may be associated with the increased adaptability of early human societies to their environment, reflecting a reduced dependence on south-facing slopes for settlement locations (Fig. 13).

Fig. 12

Distribution of cultural sites in different aspects.

Table 5 Distribution quantities of cultural sites in different aspects

Fig. 13

Radar chart of percentage changes in cultural sites in different aspects.

Hydrological factors

Statistical analysis based on the hydrological-cultural site model reveals a close relationship between the distribution of cultural sites and the distribution of surface runoff (Fig. 14). As the density of surface runoff increases, the number of cultural sites from various periods shows a rising trend, until a certain threshold of runoff density is reached, beyond which the number of cultural sites begins to decline (Table 6). The results indicate that hydrological factors significantly influence the distribution of early cultural sites. The settlement choices of early human societies were closely related to the distribution of water resources within the region, with early inhabitants tending to settle in areas with abundant water resources to facilitate their daily and production needs. However, areas with excessively dense water networks are prone to flooding and other hydrological hazards, which in turn leads to a reduction in the number of cultural sites (Fig. 15).

Fig. 14

Hydrological-cultural site overlay map.

Table 6 Spatial statistical analysis results of hydrological-cultural site

Fig. 15

Scatter plot of proportional distribution of cultural sites with different surface runoff densities.

Geographic diversity factors

The statistical analysis of the geographic diversity-cultural heritage model demonstrates a close relationship between the distribution of cultural sites and geographic diversity (Fig. 16). As the geographic diversity index increases, the number of cultural sites across different periods generally rises, until a certain level of geographic diversity is exceeded, beyond which the number of sites begins to decrease (Table 7). The results indicate that geographic diversity significantly affects the distribution of early cultural sites. Early human settlements were closely related to the geographic diversity of the region, with settlers preferring areas of moderate diversity that offered a wider range of opportunities for their subsistence and lifestyle. However, regions with excessively high geographic diversity present more complex environments, which increase the challenges faced by early settlers in adapting to and modifying their surroundings, thereby constraining their subsistence and lifestyle (Fig. 17).

Fig. 16

Geographic diversity-cultural site overlay map.

Table 7 Spatial statistical analysis results of geographic diversity-cultural site

Fig. 17

Scatter plot of proportional distribution of cultural sites with different geographic diversity indices

Results of GWR analysis

The previous statistical analysis demonstrated the impact of various geographical factors on the temporal and spatial distribution characteristics of cultural sites at the overall scale of Henan Province, but it could not reflect the spatial differences of these influencing factors. This subsection employs the GWR model to compute and further explore the spatial differences of influencing factors, revealing the geographical preferences of early ancestors in site selection during different periods. The dependent variable is the number of cultural site points, while the independent variables include elevation, slope, hydrology, and geographical diversity. The results indicate significant spatial differences in influencing factors during the Longshan period and Shang Dynasty, showing pronounced spatial heterogeneity, while no statistically significant spatial heterogeneity was observed for the influencing factors in the Xia Dynasty (Fig. 18).

• (1) Longshan period: the elevation correlation coefficient distribution map (Fig. 18a) shows three positively correlated areas in central, southeastern, and northeastern Henan, located in transitional zones from higher to lower elevations. These regions, with favorable natural conditions and transitional elevation characteristics, provided early ancestors with more options for settlement and productive activities, thus showing a positive correlation. In the eastern Henan region, situated within the North China Plain, the distribution of elevation and cultural sites primarily exhibits a negative correlation. This area features flat terrain, resulting in relatively uniform geographical conditions for productive activities, where elevation acts as a constraint on settlement site selection by early ancestors. The slope correlation coefficient distribution map (Fig. 18b) indicates that in the western and central regions of Henan Province, these areas are situated in transition zones between mountainous and hilly terrains. High-slope areas are prone to geological hazards such as landslides and mudslides, where slope acts as a constraint on settlement site selection by early ancestors. In contrast, the eastern Henan region, with its flat terrain, benefits from an appropriate slope that helps settlement points avoid flood impacts, indicating a certain positive role of slope in site selection by early inhabitants. The hydrology correlation coefficient distribution map (Fig. 18c) indicates that in flat terrain areas such as plains and basins, hydrology has a significant promoting effect on the settlement site selection of early ancestors. However, in the central transitional zone of Henan Province, areas with dense water networks may increase the risk of geological disasters, such as floods, thus presenting a negative correlation. The geographical diversity correlation coefficient distribution map (Fig. 18d) shows a clear negative correlation for geographical diversity in most areas of Henan Province. Although the Longshan period witnessed unprecedented advancements in productivity compared to the entire Neolithic era, it still retained a considerable degree of primitivism in contrast to the state of monarchy, where production tools were primarily made of stone. This limited their ability to modify the environment, making it challenging for early ancestors to adapt to complex geographical areas. In areas with complex geographical diversity, the difficulty of modifying and adapting to the environment also correspondingly increases.

• (2) The Xia Dynasty: the results of the geographic-weighted regression analysis indicate that the regression coefficients across regions show little variation, and the geographic-weighted regression results lack significant statistical meaning. This is largely because the Xia Dynasty’s society moved away from the primitivism of the Longshan period and entered the stage of a monarchy. The Luoyang Basin served as the core ruling area of the Xia Dynasty, with cultural sites mainly distributed in the basin, creating a pattern that radiated from this political center outward. In contrast, the distribution of sites in other regions was relatively sparse and limited in number. Consequently, geographic factors did not demonstrate significant spatial heterogeneity across Henan Province. The Xinzai site and the Erlitou site, as early and late capitals of the Xia Dynasty, exhibited a strong absorptive capacity for surrounding site distributions. Thus, political and cultural factors became the dominant influences on settlement site selection by early inhabitants during this period.

• (3) The Shang Dynasty: the elevation correlation coefficient distribution map (Fig. 18i) shows that in most areas of Henan Province, elevation is negatively correlated with Shang cultural sites. In the central region, particularly in the core area where the Shang capital is located, elevation has a positive role; higher ground is beneficial for military defense and flood prevention, reflecting the differences in settlement site preferences between ordinary settlements and those in the royal domain. The slope correlation coefficient distribution map (Fig. 18j) indicates that in most areas of Henan Province, slope acts as a constraint on the site selection of early inhabitants. In eastern Henan, the flat terrain and appropriate slope help settlement points avoid flooding impacts, demonstrating a positive effect on site selection. The hydrological correlation coefficient distribution map (Fig. 18k) reveals that in Henan Province, correlation coefficients are low in the west and high in the east, suggesting that hydrological factors positively influenced the settlement choices of Shang people in the eastern North China Plain. The geographic diversity correlation coefficient distribution map (Fig. 18l) indicates that in most areas of Henan Province, geographic diversity has a positive effect on early inhabitants’ site selection. Compared to the Longshan period, the Shang people exhibited significantly enhanced adaptability to complex geographic environments.

Fig. 18: Spatial distribution of regression coefficients of influencing factors.

a Elevation (Longshan period). b Slope (Longshan period). c Hydrology (Longshan period). d Geo-diversity (Longshan period). e Elevation (Xia Dynasty). f Slope (Xia Dynasty). g Hydrology (Xia Dynasty). h Geo-diversity (Xia Dynasty). i Elevation (Shang Dynasty). j Slope (Shang Dynasty).

Conclusion

This study employs GIS spatial analysis techniques to explore the spatiotemporal distribution characteristics of early cultural sites in Henan Province and the influence of various geographic factors on these characteristics, revealing early human-environment relationships. The main conclusions are as follows:

• (1) The early cultural sites in Henan Province exhibit a spatial aggregation pattern. Among them, the Longshan period sites have the widest distribution and the largest number. By the Xia Dynasty, the number of sites significantly decreased, and their distribution became concentrated in the middle reaches of the Yellow River. By the Shang Dynasty, the distribution range expanded northward and southward within Henan Province.

• (2) Early cultural sites in Henan Province are primarily located in plains, basins, and low hilly areas with elevations below 400 m and gentle slopes. Site selection favored sunny slopes that could receive ample sunlight and areas that facilitated access to water for daily activities and were less prone to flooding.

• (3) Early settlers preferred areas with moderate geographic diversity, which offered more opportunities for their livelihood. However, excessively high geographic diversity indices increased the difficulty of environmental adaptation and transformation, thereby limiting their activities and reducing the number of sites.

• (4) The results of the GWR analysis indicate that there are significant spatial differences and spatial heterogeneity in the influencing factors during the Longshan period and Shang Dynasty, while the factors during the Xia Dynasty did not exhibit statistically significant spatial heterogeneity.

Discussion

This study uses GIS spatial analysis techniques, including kernel density analysis, nearest neighbor index, standard deviation ellipse, and spatial autocorrelation analysis, to explore the spatiotemporal distribution characteristics of early cultural sites in Henan Province. It also analyzes the impact of altitude, slope, aspect, hydrology, and geographical diversity on these distribution characteristics. The results show that the geographical factors involved in the study significantly influence the site selection of early human settlements, and there is significant spatial heterogeneity in the impact of these factors.

(1) Comparison with existing research

In previous studies, GIS spatial analysis techniques have been widely applied. Elevation, slope, aspect, and hydrology are typically considered primary influencing factors^7,8,10. This study is consistent with the conclusions of these studies in these aspects; however, these are merely the effects of single factors on the sites. Geographical factors in nature do not exist in isolation; they usually have a combined impact. Therefore, we introduced the concept of geographical diversity. Due to the low productivity of early human societies, agricultural activities, gathering, and hunting were crucial for survival. Regions with more complex geographical diversity might have provided more opportunities for early humans to survive. Therefore, this study considers geographical diversity as a factor influencing the location of early human settlements. This hypothesis is supported by the constructed geographical diversity-cultural heritage model, which compensates for the previous studies’ insufficient consideration of this factor. The results indicate that regions with moderate geographical diversity are more likely to be more suitable for early human settlement.

The selection of research samples in this study differs from previous studies. Previous studies typically prioritize cultural heritage sites protected at the national level, assuming these sites have higher research and preservation value. However, in investigating the spatiotemporal distribution of early cultural heritage sites to reveal early human-environment relationships, we believe sample selection should not be based on preservation value. Each cultural heritage site is a location and activity area for early human settlements, and in this respect, they all contribute significantly to the study of early human-environment relationships. Therefore, this study collected almost all discovered early cultural heritage sites in Henan Province (from the Longshan to Xia and Shang periods) and loaded them into the GIS platform for spatial statistical analysis. This approach avoids errors due to sample omission, ensuring the statistical significance of the research results.

(2) Limitations of this study and prospects for future research

However, this study has some limitations. First, the number of early cultural heritage sites in Henan Province is vast, and some newly discovered sites may not yet be publicly available, potentially leading to omissions in the sample database. Additionally, this study adopts a macro perspective on cultural heritage sites and does not conduct targeted research on specific site cases, indicating a direction for future research.

Further research is needed in the following areas: a. Combining archeological methods to conduct in-depth examinations of the formation environments of site cases from a micro perspective. b. Summarizing the correlations between the distribution of early cultural heritage sites and the environment to identify general patterns in site distribution characteristics. These findings can be applied to the construction of archeological predictive models, providing practical value in archeological excavation and survey.

In terms of theoretical and practical significance, this research not only enriches our understanding of the spatial distribution characteristics of early human societies but also provides valuable references and guidance for several key fields.

(3) The theory and practice of this study

Through GIS spatial analysis techniques, we have gained insights into the temporal and spatial distribution characteristics of early cultural sites in Henan Province and their relationships with various geographical factors. These findings offer a new perspective on the field of cultural heritage protection. By understanding how early inhabitants chose their settlements based on natural conditions such as topography and hydrology, we can more scientifically assess the vulnerability and protection needs of cultural heritage. This enables us to formulate more effective conservation measures, ensuring that these precious historical legacies are properly preserved and transmitted. The specific applications of the results of this study can be developed in the following aspects:

a. Provides important data support for building archeological prediction models. By analyzing the distribution characteristics of early cultural sites, this study identifies several key environmental variables that significantly influence the distribution of sites. These variables can serve as important inputs for prediction models and provide a basis for improving model accuracy. For example, this study shows that early cultural sites tend to be distributed in areas with abundant water sources and flat terrain, which provides a clear direction for feature selection in machine learning models. The environmental preference analysis results obtained in this study can be used to generate prediction maps of potential site distributions. These maps can assist archeologists in prioritizing field surveys in high-suitability areas, thus significantly improving survey efficiency.

b. Provides important insights for urban planning. Understanding the principles and strategies behind ancient cultural site selection can help us better comprehend the historical context and regional characteristics of urban development, providing useful references for modern urban planning and construction. Additionally, by drawing on the wisdom and experience of ancient inhabitants in the site selection process, we can more effectively address challenges in contemporary urban planning and create more livable and harmonious urban environments.

c. It has significant implications for cultural heritage protection. By clarifying the role of environmental factors in site selection, this study can provide scientific evidence for site preservation. For example, in future conservation planning, emphasis could be placed on site areas vulnerable to climate change or natural disasters, thereby enhancing the focus and effectiveness of preservation efforts. In combination with environmental features such as the hydrophilic nature of cultural sites, flood barriers, protective walls, or diversion structures can be set up for sites vulnerable to flooding, safeguarding the sites from flood damage. Additionally, by relying on natural water bodies such as rivers and lakes, hydrophilic sites in the region can be linked together to form a cultural heritage corridor, creating an integrated whole. In corridor planning, the role of water bodies as links in ancient transportation, economic, and cultural exchange should be emphasized to highlight the evolutionary context of regional culture. Creating a “site + ecology + culture” composite tourism product can enhance the vitality of cultural heritage tourism development, compensating for the lack of visual appeal of cultural sites.

In summary, this study utilizes GIS spatial analysis techniques to offer new perspectives and evidence for understanding the spatial distribution of early human societies while providing important theoretical and practical support for cultural heritage protection, archeological prediction, and urban planning. These findings contribute to our understanding and preservation of human cultural heritage and guide sustainable urban development. In the future, we will continue to deepen research in this field to achieve more breakthroughs and progress, contributing further wisdom and strength to the sustainable development of human society.