Spatiotemporal Dataset Development of the Spread and Mortality of the Pneumonic Plague in Northeast China (1910?C1911)

Liu, X. Z.^1,3  Gong, S. S.^1,2*

1. School of Urban and Environmental Sciences, Central China Normal University, Wuhan 430079, China;

2. Center for Sustainable Development Research, Central China Normal University, Wuhan 430079, China;

3. Faculty of Arts and Social Sciences, National University of Singapore, Singapore 119260, Singapore

Abstract: Primarily, the systematic integration of spatiotemporal modeling for epidemic diffusion and mortality geodemographics constitutes a central research focus in historical medical geography with numerous documented epidemic cases as evidence. The 1910?C1911 pneumonic plague outbreak in Northeast China stands as a paradigmatic cautionary case in the history of epidemic control and public health. Therefore, this study formulates a spatiotemporal dataset for the spread and mortality of the plague, using day and county as the temporal and spatial units, respectively. To ensure the authenticity of the dataset, extensive information is extracted from historical documents such as Reports on epidemic affairs in the three northeastern provinces, Report of the international plague conference in Fengtian, Compilation of epidemic disasters in three thousand years of China??s history, and Compilation of historical epidemic records from modern China??s newspapers. This dataset provides comprehensive information, including the geographic distribution of affected areas, the time nodes of the epidemic, and mortality data. The dataset is archived in .shp and .xlsx formats, and consists of 16 data files with data size of 949 KB (Compressed into one file with 612 KB).

Keywords: Pneumonic plague in Northeast China; transmission network; transmission pattern; mortality intensity; mortality differentiation

DOI: https://doi.org/10.3974/geodp.2025.02.04

Dataset Availability Statement:

The dataset supporting this paper was published and is accessible through the Digital Journal of Global Change Data Repository at: https://doi.org/10.3974/geodb.2025.01.06.V1.

1 Introduction

Plague, as a highly lethal infectious disease, has precipitated numerous major epidemics throughout human history, including 3 global pandemics that profoundly shaped human societies. Caused by Yersinia pestis, the disease is mainly characterized by sudden onset, rapid progression, and extremely high mortality. Plague-related general symptoms include high fever, swollen and painful lymph nodes, coughing, chest pain, and distinctive pulmonary inflammation^[1]. Clinically, plague is categorized into 3 types: bubonic, pneumonic, and septicemic. It is designated as a Category A infectious disease under the modern statutory classification system^[2].

The 1910?C1911 pneumonic plague in Northeast China comprised a critical epidemio­logical node of the third global plague pandemic. The outbreak originated in Russia and entered China via Manzhouli^[3], exhibiting a distinctive ??railway-city?? transmission pattern^[4]. Within just a few months, the epidemic caused over 60,000 deaths and induced widespread social panic or even collapse.

Although many scholars have examined this epidemic??with some evaluating its spread and associated mortality data^[4?C9]??systematic construction of spatiotemporal dataset and quantitative analysis remain scarce. This study interprets the topic from the perspective of historical medical geography, utilizing the ??day?? as the temporal unit and the ??county?? as the spatial unit. By employing methods from social network analysis and geographical spatial analysis, the study quantitatively investigates the spatiotemporal transmission network and mortality differentiation of the China??s Northeast plague.

2 Metadata of the Dataset

The metadata of the Spatial-temporal dataset of the spread and mortality of the Northeast China plague during 1910?C1911^[10] is summarized in Table 1. It includes the dataset full name, short name, authors, year of the dataset, temporal resolution, spatial resolution, data format, etc.

3 Data Sources and Methods

3.1 Data Sources

3.1.1 Epidemic Data

The plague in Northeast China is a pivotal public health crisis during the late Qing Dynasty with profound social and medical implications, and it has been extensively recorded in historical sources. Among these sources, the Reports on epidemic affairs in the three northeastern provinces^[12] and the Report of the international plague conference in Fengtian^[13] are considered authoritative, offering specific accounts of outbreak locations, transmission routes, and death tolls. Drawing on these records, the dataset spans the entire duration of the epidemic??from the initial reported case on October 25, 1910, to the last recorded instance on April 29, 1911??effectively representing the full temporal dynamics of the plague??s spread. In addition, the Compilation of epidemic disasters in three thousand years of China??s history^[14] and the Compilation of historical epidemic records from modern China??s newspapers^[15] provide a wealth of supplementary insights into the epidemic. The dataset primarily incorporates epidemic data extracted from these 4 major categories of historical sources.

3.1.2 Population Data

To assess plague mortality intensity and spatial heterogeneity, the study collected demographic data across county-level administrative regions during the epidemic. The sources of population data fall into 3 main categories: First, contemporary historical demography works, notably A history of China??s population^[16], The distribution of China??s population^[17], and A brief history of migration in modern Northeast China^[18]; Second, official census archives of late

Table 1  Metadata summary of the Spatial-temporal dataset of the spread and mortality of the Northeast China plague during 1910?C1911

Qing-early Republican era, such as Agriculture and grain in Northern Manchuria (1909)^[19], Local surveys of counties in Shandong^[20], and 1911 investigation of Heilongjiang, Jilin, and Fengtian provinces near the Chinese Eastern Railway^[21]; Third, local gazetteer demographic data clusters, including those of Heilongjiang^[22], Jilin^[23], Hebei^[24], Shandong^[25], as well as municipal and county gazetteers from Manzhouli^[26], Qing??an^[27], Acheng^[28], Shuangyang^[29], Ningjin^[30], and Tai??an^[31].

3.1.3 Geographic Data

The geospatial data originated from the county-level administrative boundary dataset in the China Historical Geographic Information System (CHGIS V6)[1]. ArcGIS software was applied to convert administrative divisions and associated attributes into vector formats, with population density data sourced separately, enabling systematic spatial analysis workflows.

3.2 Data Preprocessing

3.2.1 Structuring of Historical Textual Sources

Historical epidemic records are largely found in unstructured textual form and must be processed and systematized to support quantitative research. Through thorough examination and validation of textual records, essential information was extracted from the sources, including the timeline of infection in each county, transmission routes, and mortality figures. The extracted information was subsequently formatted into standardized data tables, featuring variables such as county-level administrative units, onset and end dates of outbreaks, origin and destination of transmission, and statistics of deaths??thus creating a complete spatiotemporal dataset of the plague.

3.2.2 Data Cleaning and Missing Value Treatment

(1) Temporal Interpolation Method: Missing epidemic dates were interpolated leveraging adjacent counties?? outbreak chronologies and historical trend documentation. For instance, when specific counties lacked recorded outbreak onset and end dates, these temporal parameters were reconstructed through cross-referencing neighboring counties?? documented outbreaks.

(2) Cross-validation with multiple sources: In cases of inconsistencies and anomalies in reported death tolls or epidemic progression, cross-verification across archival records was conducted. This approach ensured the dataset??s accuracy and credibility through cross-validating data across independent sources.

3.3 Data Generation

3.3.1 Generation of Plague-affected Area Data

Plague-affected areas are defined as the geospatial distribution of the epidemic transmission. The dataset on plague-affected areas comprises 3 key parameters: administrative boundary designations, geographical size, and population metrics. The 1910?C1911 plague devastatingly impacted 130 counties and prefectures across 5 provinces??Heilongjiang, Jilin, Fengtian, Zhili, and Shandong. The administrative boundaries and county seats of these regions were sourced from the China Historical Geographic Information System (CHGIS V6), with boundary data??especially historical national borders??primarily based on The Historical Atlas of China edited by Tan Qixiang^[32]. Meanwhile, geographical sizes were derived using ArcGIS from digitized, vectorized maps, while population data were obtained from historical demographic studies, early census data, and local gazetteer entries.

3.3.2 Generation of Plague Transmission Data

Plague transmission refers to the temporal process of plague transmission and diffusion. The dataset includes the date of first outbreak, termination, epidemiological duration metric, plague network node degree, and plague transmission speed.

(1) Generation of date of first outbreak, termination, and duration

The dataset defines the date of first outbreak (D_f) as the date of the initial plague-associated death in a given county or prefecture, marking the beginning of the epidemic. The date of termination (D_e) denotes the time when the epidemic was completely eradicated, indicating the end of the epidemic incident in every administrative division. The epidemiological duration metric (T_p) is calculated as the temporal interval between disease emergence D_f and fade-out D_e, providing standardized measurement of plague persistence across spatially delineated regions. The calculation formula is as follows:

                                                      T_p =D_e?CD_f                                                      (1)

where, T_p represents the epidemic duration in calendar days (d); greater values signify extended outbreaks, whereas smaller values indicate shorter transmission periods.

(2) Construction of the plague transmission network and generation of node degree

The spatial diffusion of plague can be modeled as a spatial network following a sequential evolution of ??point-line-area??. This study applies Social Network Analysis (SNA) to construct a transmission network model of the epidemic, taking each county as a node (N_i) within the network. In this framework, a source node (N_s) identifies the epidemic focus from which the plague originates and expands outward, while a target node (N_t) specifies the area into which the plague is transmitted. The link between the 2 nodes depicts a transmission path (). During the plague in Northeast China, certain cities served dual roles as both recipients and transmitters of the plague, displaying features of both node types within the network. These cities thus served as critical hubs in the transmission network. To capture the spatial features of this network, the dataset defines 2 key metrics??plague network node degree (C_D(N_s)) and plague network density (D_p)??to quantitatively analyze the transmission patterns and the geographic reach of the epidemic.

The CD(Ns) is employed to measure the functional role of a specific city within the plague transmission network. It quantifies the total number of incoming and outgoing transmission connections for a given node??equivalently, the number of ??edges?? linked to that node. This metric reflects the influence of a city in the spread of the epidemic. The mathematical formula is presented below:

                                              C_D(N_s)=                                              (2)

where, C_D(N_s) denotes the node degree of a source node, and N represents the total number of nodes in the network. If a transmission path exists between epidemic-affected cities s and t, then a_st =1; otherwise, a _st =0. A higher C_D(N_s) value signifies a higher influential role in epidemic propagation, whereas a lower value suggests a reduced involvement in disease diffusion.

Plague network density (D_p) quantifies the degree of interconnectivity among nodes within the epidemic region, capturing the overall spatial cohesion of the epidemic??s transmission pattern. It is calculated as the ratio of the actual number of transmission paths to the theoretical maximum number of possible links. The equation is defined as follows:

                                                   D_p=                                                   (3)

where, D_p represents the plague network density,  denotes the actual number of edges (transmission paths) in the plague transmission network, and N is the number of nodes. The value of D_p ranges from [0,1]. A higher D_p reveals stronger connectivity between cities, suggesting that the epidemic primarily spread through a ??contiguous diffusion?? pattern?? gradually infecting nearby areas. Conversely, when D_p approaches 0, the transmission tends to follow a ??leapfrogging?? pattern, in which the epidemic spreads in a spatially dispersed and non-contiguous mode.

(3) Generation of plague transmission speed

To calculate the plague transmission speed, this study introduces a concept of temporal distance from the initial outbreak, which measures the time lag of the plague outbreak in each county in contrast to the origin point??Manzhouli. This metric reflects the chronological order of the epidemic spread across the infected jurisdictions. The equation is provided hereunder:

                                                      D_d=D_f ?CD₀                                                      (4)

where, D_f represent the date of first outbreak in a given plague-affected area, and D₀ denote the date of first outbreak at the origin of the epidemic (Manzhouli). The time distance from the initial outbreak, denoted as D_d, is measured in days (d). A larger D_d value demonstrates that the epidemic reached the area later.

Plague transmission speed refers to the spatial distance over which the plague spreads per unit of time. The calculation formula is formally expressed as:

                                                                                                     (5)

where, V denotes the plague transmission speed (unit: km/d); n is the number of transmission paths; L_st represents the spatial distance between the source node and the target node (unit: km); and D_st quantifies the temporal interval between the initial outbreak timestamps at 2 geographical nodes (unit: d). If a city demonstrates multiple transmission paths, its average transmission speed is calculated as the ratio of the total path distance to the total temporal difference across all corresponding outbreak onset dates.

3.3.3 Generation of Plague Mortality Data

Plague mortality refers to population-level fatalities directly caused by Yersinia pestis infection during epidemic phases. The plague mortality dataset comprises 5 principal epidemiological metrics: Plague Mortality Count, Per Capita Plague Mortality Rate, Per Area Plague Mortality Rate, Average Daily Plague Mortality Rate, and Composite Plague Mortality Intensity.

(1) Plague Mortality Count (S)

Refers to cumulative deaths directly attributable to plague, as recorded in historical sources.

(2) Per Capita Plague Mortality Rate (D_m)

Defined as the ratio of total plague deaths to the total population (P), standardized per 100,000 individuals. It reflects the burden of mortality relative to population size, allowing for comparative analysis across regions with different population densities.

(3) Per Area Plague Mortality Rate (D_i)

Defined as the number of plague deaths per unit area, expressed as deaths per 10,000 km². It indicates the spatial density of mortality within the epidemic zone.

(4) Average Daily Plague Mortality Rate (D_s)

Calculated by dividing the total number of plague deaths by the duration of the outbreak in days, expressed as deaths per day. It measures the average daily death toll during the plague outbreak.

(5) Composite Plague Mortality Intensity (S_d)

A synthesized index combining the per capita, per area, and daily mortality rates to comprehensively assess the overall intensity of plague mortality. The formula is shown as:

                                                 S_d=                                                  (6)

where, S_d represents the Composite Plague Mortality Intensity, D_i is the Per Area Plague Mortality Rate, D_s is the Average Daily Plague Mortality Rate, and D_m is the Per Capita Plague Mortality Rate.

4 Data Results and Validation

4.1 Dataset Composition

The Spatial-temporal dataset of the spread and mortality of the Northeast China plague during 1910?C1911 covers the period from October 25, 1910, to April 29, 1911, and includes records from 130 counties and prefectures across 5 provinces??Heilongjiang, Jilin, Fengtian, Zhili, and Shandong. The dataset comprises boundary information, administrative seat data, and area statistics for each plague-affected region, along with corresponding data on pneumonic plague area, transmission patterns, and mortality figures. The dataset is archived in .shp and .xlsx formats.

4.2 Data Results

4.2.1 Network Analysis of Plague Transmission

Analysis of the constructed plague transmission network elucidates that the epidemic primarily spread along major railway lines. Pivotal cities located along the Chinese Eastern Railway, South Manchuria Railway, and Peking?CFengtian Railway emerged as core outbreak zones. In addition to terrestrial routes, the plague also spread through maritime routes, extending from Dalian to coastal cities such as Yantai and Qingdao on the Shandong Peninsula. From these coastal hubs, the epidemic further expanded inland into Shandong and Zhili provinces via the Jiaozhou-Jinan Railway.

(1) Plague transmission nodes and paths

Social network analysis demon-  s­trates that in the transmission net­work of the pneumonic plague in Northeast China, major cities such as Harbin, Fengtian (now Shenyang), Changchun, Dalian, Jingshi (now Beijing), Baoding, Yantai, and Jinan constituted the primary transmission nodes (Figure 1). Among the 130 plague-affected counties and prefect­ures, 72.09% (93) were subj­ected to the epidemic through these key nodes. These cities not only functioned as critical transportation hubs, but also formed high-risk areas for epidemic diffusion.

The railway functioned as the primary trans­mission axis of the pn­eumonic plague in North­east China. During the epidemic, 3 major railway-based transmission routes and 1 marit­ime transmission route were formed (Figure 2): the Chinese Eastern-South Manch­uria Railway route, the Peking-Fengtian-Peking-Hankou Rail­way route, the Dalian-Yantai maritime route, and the Tianjin-Pukou-Jiaozhou-Jinan Railway route.

(2) Transmission process, modes, and speed of the plague

Given the time distance from the initial outb­re­ak and the epidemic duration, the pneu­monic plague lasted for 197 days??approx­imately 6.5 months. The epidemic underwent a prolonged transmission process from its initial outbreak to final containment. The dynamic trends in newly affected counties, cumulative affected counties, and recovered counties explicitly delineate distinct spatiotemporal progression patterns of the epidemic. Accordingly, the epidemic??s progression can be divided into 3 phases: the emergence period, the expansion period, and the decline period (Figure 3).

Figure 3  Daily changes in the number of affected counties during the pneumonic plague in Northeast China

The emergence period (October 25 to December 31, 1910) was characterized by minimal and relatively stable increases in the number of newly affected counties. The cumulative count of affected counties rose slowly, with an average of fewer than 1 new county reported per day. During this phase, the epidemic spread at a relatively slow pace and remained geographically confined. The number of actively infected counties fluctuated at a low level, illuminating that the epidemic was still sporadic in nature and exhibited weak transmissibility.

The expansion period (January 1 to February 23, 1911) witnessed a sharp and fluctuant increase in newly affected counties, accompanied by a rapid rise in cumulative case numbers. This phase marked the most intense stage of epidemic spread. The number of actively infected counties surged concurrently, peaking in the mid-to-late stage of this period, uncovering that most counties were in an active outbreak state. Both the speed and spatial scale of transmission expanded significantly.

The decline period (February 24 to April 29, 1911) was marked by a rapid drop in new cases, eventually reaching 0, along with a sustained decrease in the number of actively infected counties. The cumulative count of affected counties plateaued, signaling the cessation of geographic spread. During this phase, no new areas were infected, and the situation in previously affected regions gradually improved. The epidemic, overall, entered a stage of containment and resolution.

The pneumonic plague in Northeast China manifested 2 transmission modes: leapfrogging spread and contiguous diffusion. A network density closer to 0 indicates fewer direct links between locations, suggesting that transmission is more likely to occur over long distances. Network analysis shows that the network density (D_p) of the plague transmission network is 0.01??close to 0??confirming that leapfrogging transmission was the predominant pattern of epidemic spread (Figure 4).

The transmission speed of the pneumonic plague in Northeast China exhibited pronounced geographical disparities (Figure 5). The rate at which plague spread across major node cities can be deduced from the first outbreak date isolines: locations farther from the origin experienced longer delays before the disease arrived. Transmission progressed significantly faster along railway routes than in areas distant from rail infrastructure (Figure 5a). Geospatial analytical techniques were used to calculate plague transmission speeds across all 130 impacted counties and prefectures, resulting in a categorized thematic map. The results reveal clear spatial disparities. Among them, Mancheng County in Zhili recorded the fastest transmission, with an average of 43.58 km/d. This was followed by Manzhouli, Tianjin, Jingshi (now Beijing), and Jinan, where rates exceeded 23 km/d. Overall, during the entire 197-day epidemic period, the average transmission speed of the pneumonic plague in Northeast China was 8.09 km/d. During the 54-day expansion period alone, the average speed reached 30 km/d (Figure 5b).

Figure 5  Maps of the transmission speed of the Northeast China pneumonic plague^[32]

4.2.2 Regional Differentiation of Plague Mortality

(1) Provincial-level mortality differentiation

A statistical overview of the pneumonic plague outbreak reveals the following provincial death tolls: Heilongjiang reported 14,812 deaths, Jilin 25,418, Fengtian 6,752, Zhili 1,299, and Shandong 8,006??amounting to a total of 56,287 deaths across the 5 provinces. Adding 4,503 further deaths recorded along the Chinese Eastern Railway (under Russian jurisdiction) and the South Manchuria Railway (under Japanese jurisdiction), the overall death toll from the pneumonic plague in Northeast China reached at least 60,790. In terms of average daily plague mortality, the rate ranked was as follows: Jilin > Heilongjiang > Shandong > Fengtian > Zhili. For the per area plague mortality rate, the order was: Jilin > Shandong > Heilongjiang > Fengtian > Zhili.

Concerning the per capita plague mortality rate, the ranking was: Heilongjiang > Jilin > Shandong > Fengtian > Zhili. Regarding composite plague mortality intensity, Jilin ranked the highest, while Zhili the lowest, with Heilongjiang, Shandong, and Fengtian falling in between. Collectively, Jilin Province was the most severely impacted region during the epidemic^[33].

(2) County-level mortality differentiation

With the natural breaks classification method, the plague mortality indicators for the 130 counties and prefectures were categorized into 5 levels (Figure 6). The results show that the region along the railway corridor from Harbin to Changchun and Fengtian (now Shenyang) experienced the most severe mortality. In this area, the total number of deaths, as well as the per area, per capita, and daily average plague mortality rates, along with the composite mortality intensity, were all noticeably higher than those in other regions.

Figure 6  Maps of the county-level differentiation of plague mortality during the pneumonic plague in Northeast China^[32]

4.3 Data Validation

The collection and extraction of data for the spatiotemporal dataset on the spread and mortality of the pneumonic plague in Northeast China were meticulously carried out throughout the entire process.

The research findings evidence that the dataset is highly validated. Specifically, the dataset documents 60,790 deaths caused by the pneumonic plague in Northeast China, closely matching Wu??s widely cited estimate of ??over 60,000?? deaths^[34]. A further validation procedure involves measuring the distance from each of the 130 affected locations to the nearest railway line and analyzing its correlation with the same mortality indicators. The findings demonstrated a statistically significant inverse relationship: plague mortality increased with proximity to railways and declined with increasing distance. This quantitative pattern supports the historical account by Minister Shi, Zhaoji, who stated that ??all the towns through which the disease spread lay along railway lines??^[35].

5 Discussion and Conclusion

From the perspective of historical medical geography, this study collected and processed historical documents related to the 1910?C1911 pneumonic plague in Northeast China. By adopting social network analysis and geospatial-temporal analysis, it effectively constructed a comprehensive spatiotemporal dataset on the transmission and mortality of the epidemic. The data analysis yielded the following key conclusions:

(1) The pneumonic plague in Northeast China affected 130 counties and prefectures across 5 provinces: Heilongjiang, Jilin, Fengtian, Zhili, and Shandong. Major transmission nodes included Harbin, Fengtian (now Shenyang), Changchun, Dalian, Jingshi (now Beijing), Baoding, Yantai, and Jinan.

(2) The epidemic persisted for 197 days, progressing through 3 phases: the emergence period (October 25 to December 31, 1910), the expansion period (January 1 to February 23, 1911), and the decline period (February 24 to April 29, 1911).

(3) Transportation lines played a critical role in the spread of the epidemic, forming 3 major railway-based transmission routes and one maritime transmission route.

(4) The plague predominantly expanded through a leapfrogging transmission pattern?? characterized by rapid, long-distance jumps??combined with contiguous diffusion, which propagated more gradually over shorter distances. The interplay of these 2 modes resulted in pronounced distance decay effects and heavy reliance on transportation networks in shaping the epidemic??s spatial dynamics. During the expansion period, the average transmission speed reached approximately 30 km/d.

(5) The epidemic caused at least 60,790 recorded fatalities, with mortality severity showing prominent spatial clustering along principle railway corridors. Across all metrics??cumulative deaths, mortality per unit area, per capita mortality, daily average mortality rates, and composite mortality intensity??the ??railway pull effect?? (i.e., epidemic amplification driven by transport connectivity) emerged as a defining feature: rail hubs and their adjacent zones consistently functioned as disease hotspots. The region between Harbin, Changchun, and Fengtian (now Shenyang) experienced the highest death tolls.

By leveraging social network analysis and geospatial analysis, this study quantitatively examined and visually presented the spatiotemporal dynamics of the pneumonic plague in Northeast China. Compared with conventional descriptive historiography, this data-driven approach demonstrated enhanced scientific rigor through its systematic integration of spatiotemporal analytics. The multidimensional dataset enables not only quantifiable reconstruction of the epidemic??s diffusion patterns across temporal and spatial dimensions but also establishes a reproducible analytical framework??with theoretical, methodological, and evidentiary implications??for modeling historical and contemporary disease transmission dynamics in China and globally.

Author Contributions

Gong, S. S. conceived and designed the dataset framework, developed the core algorithms, supervised the processing of key data, and contributed to the review and revision of the data paper. Liu, X. Z. spearheaded for the collection and processing of data regarding plague-affected areas, transmission progression, and mortality data, optimized the algorithm logic, carried out the practical development of the dataset, drafted and revised the data paper.

Conflicts of Interest

The authors declare no conflicts of interest.

References

[1]        Li, L. J. Infectious Diseases [M]. Beijing: Higher Education Press, 2004: 153.

[2]        Jin, W. T., Pan, J. Diagnosis, treatment and prevention of plague [J]. Chinese Clinical Medicine, 2019, 26(6): 803?C806.

[3]        Du, L. H. Plague prevention and traffic blockade in Northeast China during the late Qing Dynasty [J]. Historical Research, 2014, 61(2): 73?C90+190.

[4]        Cao, S. J., Li, Y. S. Plague: War and Peace??China??s Environmental Conditions and Social Change: 1230?C1960 [M]. Jinan: Shandong Pictorial Press, 2006: 224.

[5]        Gu, Y. Q. Modern Epidemic Prevention in China (1840?C1937) [M]. Beijing: Guangming Daily Press, 2012.

[6]        Wang, Y. A study of the 1910?C1911 plague in Northeast China and its prevention and control [D]. Suzhou: Soochow University, 2005.

[7]        Chen, Y. The Northeast pneumonic plague in late Qing and its historical implications [J]. Shandong Social Sciences, 2020(5): 74?C79.

[8]        Jiao, R. M. The 1910?C1911 northeast plague and responses by the court and people [J]. Modern History Studies, 2006, 28(3): 106?C124.

[9]        Shi, J. Japan??s observation and response to the late Qing Northeast Plague [J]. Journal of Chongqing University (Social Science Edition), 2022(1): 202?C220.

[10]     Liu, X. Z., Gong, S. S. Spatial-temporal dataset of the spread and mortality of the Northeast China Plague during 1910?C1911 [J/DB/OL]. Journal of Global Change Data Repository, 2025. https://doi.org/10.3974/ geodb.2025.01.06.V1.

[11]     GCdataPR Editorial Office. GCdataPR data sharing policy [OL]. https://doi.org/10.3974/dp.policy.2014.05 (Updated 2017).

[12]     Fengtian Anti-Epidemic Bureau. Epidemic Reports of the Three Eastern Provinces [M]. Tianjin: Tianjin Ancient Books Press, 2010: 1?C692.

[13]     Editorial Committee of the International Plague Conference. Report of the International Plague Conference in Fengtian [M]. Trans. Zhang, S. Z. Beijing: Central Compilation Press, 2010: 1?C501.

[14]     Gong, S. S. Compilation of Historical Plague Materials in China (Qing Dynasty Volume) [M]. Jinan: Qilu Press, 2019: 505?C1314.

[15]     Gong, S. S. Compilation of Epidemic Reports in Modern China??s Newspapers (Vol.1) [M]. Jinan: Shandong Pictorial Press, 2023: 1?C659.

[16]     Ge, J. X. History of China??s Population, Vol. 5 [M]. Shanghai: Fudan University Press, 2002: 1?C1343.

[17]     Hu, H. Y. Distribution of China??s population [J]. Acta Geographica Sinica, 1935, 2(2): 33?C74.

[18]     Wu, X. Y. Brief history of modern Northeast migration [J]. Northeast Journal, 1941, 1(2): 2.

[19]     Boloban. Agriculture and grain in Northern Manchuria [R]. Trans. Heilongjiang Library. Harbin: Far East Railway Bureau, 1909.

[20]     Lin, X. Z. Local surveys of counties in Shandong [R]. Jinan: Education Department of Shandong Governor??s Office, 1919.

[21]     Boloban. 1911 Survey of Heilongjiang, Jilin and Fengtian along the Chinese Eastern Railway [R]. Trans. Heilongjiang Library. Harbin: Far East Railway Bureau, 1912.

[22]     Heilongjiang Provincial Gazetteer Committee. Heilongjiang Gazetteer, Vol. 57: Population [M]. Harbin: Heilongjiang People??s Press, 1996: 99?C127.

[23]     Jilin Provincial Gazetteer Committee. Jilin Gazetteer, Vol. 5: Population [M]. Changchun: Jilin People??s Press, 1992: 1?C481.

[24]     Hebei Provincial Gazetteer Committee. Hebei Gazetteer, Vol. 12: Population [M]. Shijiazhuang: Hebei People??s Press, 1991: 19?C52.

[25]     Shandong Provincial Gazetteer Committee. Shandong Gazetteer: Population [M]. Jinan: Qilu Press, 1994: 154?C186.

[26]     Manzhouli Gazetteer Committee. Manzhouli Gazetteer [M]. Hohhot: Inner Mongolia People??s Press, 1998: 53?C71.

[27]     Qing??an County Gazetteer Office. Qing??an County Gazetteer [M]. Harbin: Heilongjiang People??s Press, 1995: 424?C432.

[28]     Acheng County Gazetteer Office. Acheng County Gazetteer [M]. Harbin: Heilongjiang People??s Press, 1988: 115?C138.

[29]     Shuangyang County Gazetteer Committee. Shuangyang County Gazetteer [M]. Changchun: Jilin Literature and History Press, 1992: 120?C132.

[30]     Ningjin County Gazetteer Committee. Ningjin County Gazetteer [M]. Jinan: Qilu Press, 1992: 103?C118.

[31]     Tai??an Gazetteer Committee. Tai??an Gazetteer [M]. Jinan: Qilu Press, 1996: 85?C97.

[32]     Tan, Q. X. Historical Atlas of China, Vol. 8 [M]. Beijing: China Map Publishing House, 1996: 5?C6.

[33]     Zhao, E. C. Draft History of Qing, Vol. 449: Biography of Xi Liang [M]. Beijing: Zhonghua Book Company, 2020: 12534.

[34]     Wu, L. D. Plague Fighter: The Autobiography of a Modern Chinese Physician (Vol. 1) [M]. Trans. Cheng, G. S., Ma, X. B.; Ed. Wang, L. F. Changsha: Hunan Education Press, 2011.

[35]     Shengjing Times. Speech by Shi Zhaoji [N]. Shengjing Times, 1911-04-05(2).