Introduction

Intangible cultural heritage (ICH) refers to techniques, performance forms, customs, and other things preserved in a non-material form. It has rich and irreplaceable historical value and is an important source of cultural confidence for a country1,2. Traditional medicine (TMe) represents a significant aspect of Chinese ICH and encompasses a vast repository of knowledge and practices about preventing and treating illnesses over millennia. It is a preferred medical option and has been observed in people’s daily lives, folk festivals, and food habits3. Recently, the TMe has played an active role in preventing and controlling the COVID-19 pandemic in China4, especially in China’s “One Belt, One Road” (OBOR) foreign exchange5 and promoting the Healthy China strategy6. Additionally, the TMe is the essence of the treasure trove of Chinese medicine; it is deeply embedded in the Chinese cultural tradition and irreplaceable in forming cultural identity7.

The predominant focus of global medical development has undergone a gradual transition from disease treatment to disease prevention and health promotion. TMe, which emphasizes holistic perspectives and constitutional studies, has emerged as a significant research domain. In 2014, the World Health Organization (WHO) published the “Traditional Medicine Strategy 2014–2023,” encouraging governments to formulate TMe development strategies and establish comprehensive systems8. In 2018, the WHO incorporated TMe into its medical outline with global influence for the first time9. There has also been growing interest in the development and cultural impact of the TMe internationally10,11,12,13.

In terms of research content, statistical analysis of TMe over the past two decades reveals that the majority of studies have focused on classification14,15,16, protection17,18,19, inheritance20, ethnomedicine21,22,23,24,25,26, and disease prevention and control27,28,29,30,31,32,33,34. Moreover, the significant value of TMe in medicine and healthcare is evidenced by its increasing contribution to international exchange and collaboration35,36,37. Numerous studies have fully embraced the living, inherited, and national aspects of the TMe38,39. However, it is widely recognized that human physiology and disease patterns are influenced by geographical environments, which serve as critical factors in the formation of TMe culture. The transmission of TMe from one generation to the next has been shaped by the distinctive geographical environments of different regions. This process of inheritance, development, promotion and protection, and public service has resulted in the emergence of a distinctive spatial influence. This influence is not only evident in the regional characteristics observed in the spatial dimension40,41 but also manifested in the temporal dimension42. For instance, Nie et al. conducted a preliminary exploration of the spatial distribution pattern of TMe by taking China’s national TMe as the research object43. Nevertheless, their study did not adequately capture the historical evolution over time and the geographical variations in space, which are critical for understanding the inheritance and development of TMe across different regions42,44.

In terms of research methods, the kernel density analysis is a prevalent technique employed to illustrate the spatial distribution of ICH, encompassing TMe. Previous studies investigated the characteristics of the spatial distribution of ICH by type45, conducted retrospective analyses of ICH origins over time46,47,48, and investigated the influencing factors1,49. Notably, this research offers valuable insights into the spatial distribution, origin, and temporal evolution of the ICH48. However, these works only show the spatial distribution of the ICH quantity without demonstrating its regional characteristics and spatial transmission range50. In order to address this limitation, Li and Wang selected China’s national ICH as their research object and employed a deconstruction of the ICH at the provincial, municipal, and county levels through the surface distribution range according to the attributes of the declared regions50. Although this approach, which employs face-shaped symbols to express the distribution and analysis of the ICH, emphasizes spatial transmission characteristics, it does not fully consider the varying scope of influence of community and regional ICH51. Accordingly, from a geographical perspective, a more comprehensive analysis of the factors shaping the spatial and temporal distribution patterns of the TMe, as well as an investigation of its historical development, can contribute to a deeper understanding of the integrity and historical continuity of the TMe43.

In light of this, this study employed the national and provincial TMe in China as its research subject. A spatial expression method of point-area representation was proposed on the basis of the attribute characteristics of the ICH protection units. By employing this method, the geospatial distributions of TMe originating from different periods are visualized, which allows for the qualitative analysis of spatial and temporal distributions. A quantitative analysis is subsequently carried out using GeoDetector to determine the explanatory strength of natural geographic factors, humanistic geographic factors, and historical factors on the TMe based on grid units. It is also used to explore the historical, regional, and ethnic characteristics of the spatial and temporal distribution patterns of the TMe. The findings of this study will aid relevant authorities in gaining a more profound understanding of the distribution and evolution of the TMe from a geographical perspective. This will enhance public recognition and understanding of TMe, thereby fostering collaboration among traditional medicinal practices worldwide. The objective is to establish a community of life for both humanity and nature, characterized by the principles of “life-above-all” and “health-centered.”

Methods

Research data

TMe property data

The TMe property data were obtained from the official website of the China Intangible Cultural Heritage Network and the Intangible Cultural Heritage Lists of each province, totaling 1401 pieces of data. The TMe data provide detailed information such as project names, levels, provinces, declared regions, and protective units. The origin of the TMe is determined based on textual details reported in The Encyclopedia of Chinese Intangible Cultural Heritage52 and the China Intangible Cultural Heritage Network. The chronology of China’s history, which is established through archeological research, is divided into eight stages53: the pre-Qin Period (preQ), the Qin and Han Dynasties (QH), the Wei, Jin, Southern and Northern Dynasties (WJSN), the Sui, Tang and Five Dynasties (STF), the Song and Yuan Dynasties (SY), the Ming Dynasty (MD), the Qing Dynasty (QD) and the Modern Period (MP). The initial spatialization of the data was achieved by obtaining the latitude and longitude coordinates of the TMe protection units using the coordinate picker provided by Baidu Map (Fig. 1).

Fig. 1
figure 1

The distribution of TMe in China.

Full size image

Selection of influencing factors and sources

The evolution and enhancement of the TMe have been influenced by the cumulative experiences of Chinese people across successive generations. This phenomenon is distinguished by its historical depth, cultural inheritance, and regional diversity, with various forms of TMe prevalent throughout China. On the basis of the summary and analysis of the existing studies, the influencing factors are classified into three main categories: physical geography, socio-economics, and history and culture. The most common factors are elevation, river system, gross domestic product (GDP), and population size (POP), which are highly prevalent45,54,55. Therefore, river, elevation (DEM), and derived slope are employed as physical geographic factors, while GDP and POP are utilized as socio-economic factors. Since TMe, intangible, traditional villages (TV) with a spatial carrying role are selected as the cultural influence factor, whereas minority characteristic villages (MCV) and minority concentration areas (MCA) are chosen as the ethnic influence factors. Furthermore, three influencing factors are adopted to explore the historical characteristics of the spatial distribution of the TMe: historical capitals (HCap), historical roads (HRoad), and historical population size (HPOP). The HRoad and HPOP are categorized according to the historical period of the origin of the TMe.

To achieve data spatialization, the longitude and latitude coordinates of TV, MCV, and the governments where the HCap are located are obtained using the coordinate picker provided by Baidu Map. Compared with HCap, villages have a relatively smaller spatial area of influence. Concerning the area of ancient Chinese capital cities, a radius of 30 km is set for the influencing range of HCap, whereas a radius of 5 km is fixed for TV and MCV (Table 1).

Table 1 Specific information on influence factors
Full size table

Point-area representation

In the extant literature on the spatial and temporal distribution characteristics of the ICH1,49,56, kernel density analysis is the most commonly employed method for representing the spatial distribution of the ICH43. Additionally, a range of research methods have been utilized to analyze the developmental and evolutionary characteristics of ICH, including spatial autocorrelation56, the nearest neighbor index49, the standard deviation ellipse46, and other geographic information system (GIS) spatial analysis methods. However, a significant limitation of these methods is their reliance on point data, which is inadequate for demonstrating the regional characteristics and spatial spread of the influence of the ICH.

ICH exhibits multiple characteristics owing to its properties of circulation and geospatial distribution50. Similarly, TMe is a ‘living’ culture with a deep impact as people interact. The various spatial levels to which TMe belongs can be categorized from the smallest to the largest levels. These include community-based ICH spaces, regional ICH spaces, and national ICH spaces51. The spatial characteristics of different levels of the ICH are distinct, and simple point vectors are not sufficient to represent the ICH accurately. A point-area representation approach was utilized to accurately illustrate the geospatial distribution of community-based and regional TMe in China, along with the extent of its impact.

The term point-area representation refers to the method of using point symbols to express community-based ICH and area symbols to express regional ICH. In the case of the TMe, protection units, such as factories, companies, hospitals, schools, and associations, exhibited a smaller sphere of influence. This part of the TMe was still represented by the traditional dot notation (Fig. 2), and then its spatially distributed range of influence was extrapolated to all of China through kernel density analysis.

Fig. 2
figure 2

Spatial distribution of point TMe.

Full size image

The TMe has a significant influence when it is protected by official institutions such as museums, art galleries, cultural halls, cultural and tourism bureaus, and ICH preservation centers. This part of the TMe is divided into three levels according to the reported region: county-wide, city-wide, and province-wide. The statistical unit used in this study is the county, and the cumulative number of TMe presenting faceted regional characteristics is recorded for each county, along with the cities and provinces to which it belongs. This yielded the distribution of TMe with faceted influence (Fig. 3).

Fig. 3
figure 3

Spatial distribution of faceted TMe.

Full size image

GeoDetector

GeoDetector is a crucial tool in spatial analyses and has been utilized in various research fields, including ecology57, disaster risk58, and ICH43. This statistical method detects the spatial dissimilarity of geographic objects and identifies the underlying factors. It has advantages, as it is independent of covariance between variables and can effectively detect explanatory factors and analyze the interaction between variables59. The factor and interaction detector of GeoDetector helps to assess the impact of various influencing factors on TMe and analyze the reasons behind their spatial distribution of TMe. The geographic detector principle60, as shown in Eq. (1), is applied.

$$q=1-\frac{\mathop{\sum }\nolimits_{h = 1}^{L}{N}_{h}{\sigma }_{h}^{2}}{N{\sigma }^{2}}$$
(1)

where h is the stratification of the variable, N is the number of cells, and σ2 is the variance of the dependent variable Y. The range of q is [0,1], which indicates that the factor explains 100 × q% of the variable. The value of q indicates the influence of the factor and explanatory power of the variable in a directly proportional manner.

Using statistical data as variables for analysis, the majority of extant studies are based on the analysis unit of administrative divisions. To comprehensively examine the factors influencing the spatial and temporal distributions of TMe and adhere to the constraints of GeoDetector data processing, this study employed a grid with a side length of 20 km to encompass China (Fig. 4). The value of each corresponding influencing factor after spatialization and TMe expressed by the point-area representation were extracted and assigned to a grid. The value of TMe obtained from the integrated point-area representation was used as the dependent variable. In contrast, the independent variables were the two major categories of factors, namely, natural geography and human geography, which were both graded using the Natural Breaks (Jenks). Subsequently, GeoDetector was employed to investigate the factors influencing the spatial and temporal distributions of the TMe.

Fig. 4
figure 4

Grid cells used in the study.

Full size image

Results

Quantitative characteristics of the TMe

The spatial and temporal evolution tree of TMe was constructed on the basis of the time of origin, different grades, the province to which it belongs, and the quantity of TMe (Fig. 5). The vertical axis shows the time of origin of the eight dynastic ICH, with the numbers of national and provincial TMe in different provinces displayed on the left and right sides, respectively. A detailed analysis of Fig. 5 reveals significant findings. For example, the total number of TMe originating from the QD is notably greater than that in other periods. The number of provincial TMe is significantly greater than that at the national level. Additionally, the distribution range of provincial TMe is wider, particularly during the period from the SY to the MP.

Fig. 5
figure 5

Spatial and temporal evolution tree of TMe.

Full size image

Statistical charts were constructed using property data from the TMe to identify quantitative differences between TMe from different historical periods (Fig. 6). The trend of changes in the number of TMe at the national and provincial levels is similar for each period, with a clear ‘mountain peak’ development. This indicates that most of the TMe originated during QD, which coincides with the importance placed on the TMe by the Kangxi Emperor61. Before the QD, there was a slight fluctuation in the number of TMe exhibits, with a nadir observed during the WJSN, followed by a ‘stair-step’ increase and a second peak during the MD.

Fig. 6
figure 6

Statistical chart of the number of TMe originating in each historical period.

Full size image

Spatial and temporal distribution characteristics of TMe

Spatial distribution of the TMe

To visually express the spatial distribution characteristics of the TMe, the dissemination range of the point TMe data in Fig. 2 was simulated based on kernel density analysis and counted in counties. The points TMe and faceted TMe (Fig. 3) were subjected to normalization procedures. By utilizing map algebra techniques and adhering to county-level administrative boundaries, the distribution range of the TMe was delineated through point-area representation (Fig. 7).

Fig. 7
figure 7

Spatial distribution of TMe under point-area representation.

Full size image

The analysis reveals that the TMe exerts a notable influence in China, with a higher value of the TMe observed in the northwestern region than in the southeastern region. Despite the greater number of TMe in the southeastern region, they are predominantly community-based, exhibiting a fragmented distribution of influence. In contrast, the TMe in the northwestern region is predominantly protected at the municipal and national levels, exhibiting a regional character and a spatially contiguous distribution. One potential explanation for the inverse relationship between the quantity of TMe and the scope of its influence is economic development. In regions with a robust economy, pharmaceutical companies, hospitals, and medical universities are able to assert and safeguard TMe. Conversely, in the Western region, where economic development is a pressing concern, the necessary conditions for cultural development and protection are lacking despite the creation and transmission of TMe through the accumulation of experience and the passing of knowledge from one generation to the next. Consequently, TMe is frequently declared and protected by government agencies.

Spatial distribution of TMe in different historical periods

To examine the evolution of TMe across historical periods, the spatial representation of TMe in each period was realized as a point-area representation. TMe, which originated in the preQ era, was primarily concentrated in Xinjiang, Xizang, Inner Mongolia, and other autonomous ethnic minority regions (Fig. 8a). During the QH period, the origins of the TMe expanded from west to east, becoming more widely distributed across China. This resulted in the formation of clusters in the northern, southern, and central regions of the country (Fig. 8b). During the WJSN era, the origins of the TMe were primarily situated in three distinct centers: central Inner Mongolia, southern Xizang, and the coastal region of southeastern China (Fig. 8c). During the STF period, Xizang became the epicenter of TMe, with the development of Tibetan medicine in Xizang, encompassing practices such as Tibetan bloodletting, medicinal baths, and medicine concoctions (Fig. 8d). The TMe originating in SY presented the greatest spatial aggregation throughout historical development. The highest center of aggregation was observed in Sichuan (Fig. 8e). During the MD and QD, the focus of research into the origin and development of the TMe shifted to the southeastern region. Notably, the TMe originating in the MD had a considerable influence (Fig. 8f). Despite the greater number of TMe originating from the QD, their influence on disseminating these practices was not as pronounced as that of the TMe originating from the MD (Fig. 8g). The distribution of TMe originating in the MP was primarily concentrated in Xizang and the Shaanxi-Shanxi region. Additionally, there are isolated clusters of origin in the vast expanse (Fig. 8h). This indicates that the TMe is still in a state of continuous development.

Fig. 8: Spatial distribution of TMe originating in different historical periods under point-area representation.
figure 8

a TMe originating in preQ. b TMe originating in QH. c TMe originating in WJSN. d TMe originating in STF. e TMe originating in SY. f TMe originating in MD. g TMe originating in QD. h TMe originating in MP.

Full size image

Over time, the TMe, which has been handed down from one age to another and originated in different historical periods, demonstrated a pattern of migration from the northwest to the southeast. This migration occurred in a manner that can be described as a spatial and temporal evolution and distribution pattern of dispersion-concentration-dispersion. The influence of the TMe originating in the STF and SY was relatively concentrated, whereas the influence of the TMe originating in the QH, MD, and QD was relatively scattered. The extensive reach of the TMe that has been preserved to the present day is directly correlated with the duration of the dynasty from which it originated.

Factors affecting the spatial and temporal distributions of the TMe

Results of the factor detector

The explanatory power of different influencing factors on the TMe during different periods revealed a weak relationship between MCV and HRoad and the distribution area of the TMe, and the remaining results were significant at the 1% level (Fig. 9). Generally, the HPOP has the strongest correlation with the TMe, and its ability to explain is significantly greater than that of most other influencing factors in different periods. Moreover, this study covers a relatively wide range of the whole country, and the historical population data used were assessed according to the present administrative divisions, emphasizing the close connection between the TMe and the HPOP. Conversely, MCV have a smaller spatial range and less visible impact, whereas HRoad are primarily governmental with a sparse network and less influence.

Fig. 9
figure 9

Detection results of influencing factors on spatial distribution of TMe (**Indicates significant at the 1% level and *indicates significant at the 5% level. No labeling indicates low level of significance).

Full size image

In addition, TMe is frequently linked to MCA and is more prevalent at lower elevations. Notably, the TMe is rooted primarily in ethnic traditions, as evidenced by the inclusion of many medicines and therapies specific to particular ethnic minorities. Furthermore, flat terrain at lower elevations contributes more to human flourishing, production, and daily activities, resulting in greater TMe availability. The preservation and inheritance of the TMe require human carriers and economic support, making GDP and POP crucial, as they are in the third tier of explanatory power.

The impacts of the influencing factors in different periods were explored individually to gain more profound insight into the spatial distribution characteristics of TMe across various historical periods and to understand the natural, human, and historical factors that contributed to the TMe. The origins of TMe during the preQ are explained mainly by HPOP, DEM, and MCA. The factors have similar explanatory power during this period, indicating a close relationship between the population and the natural environment. The TMe emerged during ethnic origins and is closely associated with the lives of human beings, particularly those of ethnic minorities.

Further, strong spatial connections exist between the TMe originating in the QH and the HPOP, POP, and MCA. Notably, the HPOP had a more significant influence on TMe during this period than other factors, indicating that the number of TMe that originated in the QH increased, which is consistent with human and ethnic traditions as TMe that originated in the preQ. The main difference between the TMe that originated in the preQ and QH is its closer link to modern population size and its better inheritance and protection.

Furthermore, the turbulent history and frequent regime changes in the WJSN period resulted in limited TMe development during this time. Additionally, for the TMe that originated in WJSN, the explanatory strength of each influencing factor is relatively weak, with the main influencing factors being HPOP, POP, and GDP. This further highlights the human nature of the TMe and, to an extent, explains the excavation and attention given to the historical ICH in modern society. Additionally, it is possible that the prevalence of the TMe that originated in this period can be attributed to the influence of TV. This suggests that these villages may hold valuable information regarding TMe from the past, indicating that certain practices have been passed down through generations in these small communities.

The STF period is versatile, including the prosperous Tang era under the rule of Zhenguan, the flourishing of the Kaiyuan and the Yuanhe Zhongxing, as well as frequent dynastic changes and unstable political situations experienced in the Five Dynasties and Ten Kingdoms. The alternation of war and peace promoted ethnic intermingling, resulting in MCA that had the highest explanatory power for the origin of the TMe in the STF, thereby reinforcing the ethnic minority characteristics of the TMe. Additionally, the HPOP and elevation significantly impacted the TMe that originated during this period. The emergence and development of TMe are necessary and dependent on both natural conditions and human involvement, regardless of the political environment.

The TMe, which originated in SY, exhibits a strong ethnic character, as evidenced by the spatial and temporal distribution of the point-area representation (Fig. 8e). During the Yuan Dynasty, previously known as the Great Mongolian State, which was established by Genghis Khan, a significant proportion of ethnic minority populations led to the development of many Mongolian medicines. However, the influence of each factor on the TMe that originated in the SY period is not as evident as that in other periods because of the limited number of available TMe. In the MD period, the HPOP, POP, and GDP had greater explanatory power for the TMe. Moreover, the TMe that originated in the MD period was influenced by various factors, including TV, MCA, HCap, and three physical geographic factors: river, DEM, and slope. These factors have similar explanatory power, indicating that the TMe is more prevalent, widespread, well preserved, and has a pronounced historical human dimension.

The relationship between TMe during the QD and its influencing factors is similar to that of TMe during the MD. However, the factors generally have more substantial explanatory power for the TMe that originated in the QD period. The greater stability of the QD and a greater number of TMe available, most of which originated in the late QD, facilitated timely cultural promotion work because of the shorter inheritance time and lower difficulty of excavation and protection. During the MP, the HPOP significantly influences the TMe more than other factors. However, the historical characteristics of the TMe that originated in the MP are not immediately apparent due to its short inheritance period. The significance of HCap and MCA should not be overlooked, as they contribute to the explanatory power of the TMe during this period. This suggests that the TMe of this period was influenced by medical prescriptions from different dynasties and traditional medical practices by ethnic minorities over a long period.

Results of the interaction detector

GeoDetector’s interaction detector was used to investigate the spatial and temporal distribution patterns of the TMe. The explanatory power of different factors for TMe was enhanced when they interacted with one another (Fig. 10). The distribution of TMe is influenced by multiple factors, indicating the complexity of its formation. In addition, the interaction detector results revealed that the combined effect of POP and HPOP had the greatest explanatory power for the TMe (Fig. 10a). Human invention and inheritance play important roles in TMe, and its human characteristics are the most evident.

Fig. 10: Interaction results of influencing factors on the spatial distribution of TMe.
figure 10

a TMe originating in all of eight historical period. b TMe originating in preQ. c TMe originating in QH. d TMe originating in WJSN. e TMe originating in STF. f TMe originating in SY. g TMe originating in MD. h TMe originating in QD. i TMe originating in MP.

Full size image

The explanatory results of the interaction detector in the form of visualization graphs revealed distinct bands in the preQ, STF, and SY (Fig. 10b, e, and f). The interaction of HPOP, POP, and MCA with other factors has greater explanatory power and plays a dominant role in explaining the TMe that originated in the preQ, STF, and SY. In all three periods, especially in the STF and SY, the TMe had more pronounced human and ethnic characteristics, mainly due to strong ethnic intermingling.

In the modern era, the preservation of TMe from the QH, MD, and QD has improved due to secondary explanatory factors such as GDP and POP (Fig. 10c, g, and h). It is important to mention that TMe from the QH and QD have a plain nature, whereas TMe from the MD has a more distinctive traditional village character. There is inconsistency in the results with varying reasons from the interplay of various factors in explaining the TMe that originates from the WJSN and the MP (Fig. 10d, i). The WJSN period, marked by frequent wars and brief periods of stability, constrained the number of TMe, with a significant proportion of them hailing from small villages. As the history of TMe from the MP is relatively brief, factors only demonstrate high explanatory power when interacting with the HPOP.

When examining the spatial and temporal distribution patterns of the TMe (Fig. 8), it becomes evident that the TMe originating from the STF and the SY exhibits a notably concentrated distribution. The spatial arrangement of the TMe is influenced primarily by HPOP, MCA, and DEM, with a particularly pronounced ethnic dimension. Conversely, in the MD and QD periods, the spatial distribution pattern of TMe origination was influenced primarily by GDP and POP. In general, the widespread distribution of TMe is influenced by socio-economic factors, whereas factors with strong historical roots can better explain the concentrated influence of TMe.

Discussion

As is widely recognized in the extant research, the majority of studies on the spatial distribution and influencing factors of the TMe concentrate on the ICH as a whole1,49,56, and the point distribution is widely used. However, this method cannot accurately reflect the spatial extent of the ICH, as the ICH is characterized in geospatial space by a faceted distribution50. Specifically, for TMe, which requires special medical skills, some TMe are transmitted only within specific communities, such as families and villages. Hence, this study advocates the adoption of a point-area representation to show the scope of influence according to the type of TMe protection unit (Fig. 11a).

Fig. 11: Comparison of the spatial distribution of TMe under different representation methods.
figure 11

a Based on the point-area representation. b Based on the kernel density analysis method.

Full size image

A comparison of the traditional approach with the one proposed in this paper reveals that the traditional point-kernel density analysis merely shows the spatial distribution of the number of TMe (Fig. 11b). In contrast, the point-area representation introduced in this study is specifically designed to convey the spatial extent of the influence of the TMe. This new approach not only retains the centers of TMe aggregation—such as Beijing, Shaanxi-Shanxi, Sichuan, and Shanghai—as identified by the traditional point-kernel density method, but also effectively captures the extensive reach of the TMe’s influence in the northwestern region, a detail that the traditional method fails to adequately represent. As a result, this method provides a more accurate depiction of the quantitative distribution of TMe at the microlevel of county-level administrative divisions, thereby offering a more comprehensive understanding of the spatial distribution of TMe. The analysis of influencing factors via GeoDetector also suggests that the point-area representation for spatial expression of TMe is a viable approach, with most findings being significant at the 1% level. Furthermore, this method can be extended to investigate other types of ICH. Nevertheless, whether TMe exerts a small-scale influence by point radiation or a large-scale impact by faceted distribution remains open and needs to be addressed.

In the context of economic globalization and cultural diversity, the transformation of the global medical model, the change of the spectrum of diseases, and the growing demand for healthcare have led to a consensus among governments and people around the world regarding the development of TMe. Since 2006, the Chinese government has announced a total of 182 representative national-level ICH projects related to TMe in five batches62. Using these data, Nie et al. utilized coordinate picking to achieve data spatialization43, employed the kernel density to visualize the spatial distribution characteristics of TMe points, and used GeoDetector to analyze the factors influencing the spatial distribution pattern of the TMe. In addition to summarizing the spatial distribution characteristics of the national TMe, the authors also identified the significant impact of socio-economic factors on its spatial distribution, thereby contributing to the existing body of research on this topic.

This study differs from the study conducted by Nie et al. in three principal respects. First, to gain further insight into the development and evolution of the TMe, this study employs a comprehensive dataset comprising 182 items of the national TMe in China and 1219 items of the TMe at the provincial level, for a total of 1401 items. This more detailed data allows for a more effective exploration of the spatial distribution characteristics of the TMe. Moreover, our research traces and analyzes the origin of the TMe, which is not addressed in Nie’s research, thereby enriching the historical analysis of the TMe.

Second, regarding the methodology, a data spatialization method called point-area representation is introduced. This method leverages the unique features of the TMe. Specifically, point-like symbols represent community-based TMe, and face-shaped symbols denote regional TMe. This approach builds on traditional kernel density analysis to accurately depict the spatial distribution range of the TMe.

Third, this study endeavors to delve into the underlying reasons behind the spatial and temporal distribution patterns of TMe from a geographical perspective. In selecting the pertinent influencing factors, we place significant emphasis on historical factors, encompassing HPOP, HRoad, and HCap. The empirical evidence garnered from this research substantiates the assertions made by scholars, including Cheng and Xu, who argue that the conceptualization, evolution, and dissemination of the ICH are closely linked to historical trajectories and developments63,64.

This study primarily addresses the spatial and temporal evolution of the TMe in China. When examining TMe at the regional level, such as a province, relying solely on national and provincial data to elucidate its distribution characteristics is insufficient. Accordingly, future research will focus on the integration of more comprehensive data at the municipal and county levels to facilitate a more nuanced examination of the spatial and temporal distribution patterns of ICH in smaller geographical areas.

Two major aspects, i.e., physical geography and human geography, are important factors that influence the spatial distribution of the ICH. In the case of human geography, it is important to note that several studies have incorporated the significant impact of ethnic minorities into their analyses. However, they have considered only the number or proportion of ethnic minority populations as an indicator1,49,60,65. With respect to the selection of cultural factors, various studies have identified one or more cultural bearers, including national key cultural relic protection units, national historically and culturally famous cities, towns and villages, museums, cultural centers, and TV45,66. From a geographical perspective, the aforementioned analysis employs statistical data while disregarding the spatial attributes of the research object. Accordingly, this study primarily investigates the factors influencing the spatial and temporal distribution patterns of the TMe from three perspectives: historical and cultural context, social and economic factors, and the geographical environment. This approach is justified by the evident historical, cultural, and regional characteristics of the TMe that have shaped its evolution and dissemination. Furthermore, in this study, the GDP and POP are based on existing spatialized data, where the HPOP is derived from the results of books examined and vectorized by administrative division, and the vectorization of MCA is included in the reference. Furthermore, vector data are employed to delineate the spatial distribution ranges of MCV, TV, and HCap. All types of influencing factors possess spatial attributes, which facilitated the investigation of the factors that contributed to the formation of the distribution pattern of TMe with a faceted distribution.

Importantly, the literature lacks a discussion of influencing factors by historical period. Our analysis revealed that the TMe originating in different historical periods exhibited diverse distribution characteristics. Consequently, the spatial and temporal distributions of the TMe cannot be fully explained without analyzing the historical causes. In our study, two historically characteristic influencing factors, viz. the historical population and roads, are classified based on historical periods and analyzed separately for TMe in different periods. The results of the GeoDetector analysis indicate that the HPOP significantly impacts TMe more than the other influencing factors, whether they are considered in an overall context or by period. Given such findings, one of the most effective ways to increase the depth of research is to synthesize current research and examine a broader range of historical factors to gain a more holistic comprehension of the diversity of factors at play.

Furthermore, the optimization of influencing factors in a quantitative manner illustrates an important avenue for enhancing the study. For instance, is it necessary to spatially distribute historical population data by weight according to the topography of each province when it is examined by provinces and regions? Another representative example involves finding a more scientific and reasonable scheme for defining the scope of influence of TV, MCV, and HCap. The above questions may facilitate improvements in the analysis of the TMe in the future.

In brief, this study analyses TMe from a national perspective and thereby establishes the feasibility of this analysis. Subsequent studies may define the study areas as a cultural belt, a city cluster, or a smaller area, such as a province, a city, or even a county characterized by deep historical and cultural heritage. This offers a focused discussion on the characteristics of spatial and temporal distributions. Concurrently, combining the multidisciplinary knowledge and the attributes of the TMe is recommended to facilitate an in-depth discussion of the factors influencing the spatial and temporal distributions of the TMe. Moreover, it is anticipated that the aforementioned improvements will benefit the study of TMe and other types of ICH.

In recent years, generative artificial intelligence (AI) and large language models, with ChatGPT serving as a notable representative, have led to a profound information technology revolution. In this transformation, the activation and application of AI technology in the domain of cultural heritage has garnered widespread global attention. This trend is exemplified by the launch of several projects in 2023 by the U.S. National Science Foundation67 and the European Commission68, which aim to utilize AI technologies for the conservation of cultural heritage, further underscoring the prominence and promise of this field.

Currently, AI technology is widely applied in multiple fields of ICH research. The restoration of the Dunhuang murals was achieved through the utilization of a deep learning model by professional restorers69. Technological tools such as text recognition, semantic classification, and speech recognition are employed for the digital recording and organization of ICH, thereby significantly enriching the content of ICH databases70,71,72. Various constructed knowledge graphs of ICH have been shown to intuitively demonstrate the connections among ICH elements71,73,74. Furthermore, Shao’s analysis of the spatial distribution characteristics and formation mechanisms of ICH has been based on explainable machine learning and coupled coordination models, providing a new method for the preservation of ICH75. Notably, the integration of GIS with minimum cumulative resistance (MCR)76, along with the application of virtual reality (VR) technology77 and natural language processing technology73, has further enhanced the comprehensibility and accessibility of ICH. This development has resulted in the presentation of ICH to the public in a more vivid and intuitive manner. By analyzing existing social media data and applying deep learning algorithms to analyze user preferences78, the precise delivery of ICH content has been accomplished, thereby effectively reshaping urban characteristics and enhancing urban cultural charm79.

Despite the noteworthy advancements in integrating AI technology with ICH, numerous challenges persist. The complexity and widespread dissemination of ICH necessitate enhanced recognition algorithms capable of processing intricate cultural information. Concurrently, the capacity of AI to deeply understand the connotation of ICH is limited, which may result in inaccuracies in its representation. Moreover, urgent clarification and regulation are required for issues related to data privacy, intellectual property ownership, and definitions.

In this study, the spatial and temporal evolution characteristics of the TMe are analyzed. The factors for forming spatial and temporal distribution patterns of TMe based on grid units are elucidated through GeoDetecter. Overall, the present study concludes the following:

  1. (1)

    A spatial and temporal evolution tree of the TMe is constructed, taking the time of origin, the different grades, and the quantity of the TMe as the main content. This tree provides a visual representation of spatial, temporal, and quantitative three-dimensional information, offering a novel approach to the multidimensional expression of the ICH.

  2. (2)

    A spatial expression method based on the point-area representation is proposed to study the TMe as an object. The method enables the expression of the spreading range of TMe in geospatial space, with counties serving as the research unit. The spatial distributions of TMe during different periods are classified and discussed. The method facilitates a more reasonable representation of the influence range of the ICH.

  3. (3)

    The spatial and temporal distribution patterns of TMe are explored at the grid scale, with a focus on the influences of natural geographic factors and human geographic factors, including socio-economic, cultural, ethnic, and historical factors. The results of the study illustrate that socio-economic factors have a relatively high degree of explanatory power for the spatial and temporal distributions of the TMe. Specifically, HPOP is a vital explanatory factor, followed by MCA and DEM, whereas GDP and POP also play influential roles.

In conclusion, the spatial expression method named point-area representation, as proposed in this paper, proves to be an effective tool for illustrating the spatial distribution and dissemination extent of cultural phenomena. The spatial and temporal distribution patterns derived from this analysis provide valuable insights that can inform macro-level decision-making processes aimed at optimizing resource allocation for the protection and promotion of ICH and other cultural phenomena. By carefully selecting influencing factors that align with the intrinsic attributes of the research subject, this study ensures a more focused analytical approach, thereby yielding more meaningful and impactful findings. Furthermore, the visualization of multidimensional information through the construction of a spatial and temporal evolution tree enhances our comprehensive understanding of the multifaceted characteristics of the ICH and other cultural phenomena.