Population mobility is a critical driver of population redistribution and settlement patterns, influenced by socioeconomic factors like wage imbalances and living conditions. It impacts economic activities, social norms, energy consumption, and disease transmission. Past research often relied on static data, hindering the dynamic capture of rapid mobility and urban development. The advent of GPS and LBS technologies has enabled the analysis of large-scale population behavior. Existing studies on population mobility during the COVID-19 pandemic broadly fall into three categories: (1) correlation between epidemic cases and population movements; (2) using population movement data to predict pandemic infection trends; and (3) analysis of the epidemic's impact on population movements. While extensive research exists for the first two categories, studies on the long-term impact of the pandemic on population migration are limited. This study investigates the long-term effects of the COVID-19 outbreak on large-scale population movements in China, examining changes in the scale of population migration, population return, and the population flow network and city network status before and after the epidemic. The study utilizes Baidu migration big data to construct the Population Migration Scale Index (PMSI) and the Recovery of Population Return Scale Index (PRSI) to analyze spatiotemporal characteristics and the recovery of population activity. Social network analysis is employed to assess changes in the city network status.

The literature extensively documents the impact of population mobility on various aspects of society. Studies have explored the correlation between population movement and disease spread, utilizing mobility data to predict pandemic trends, and analyzing the pandemic's impact on mobility patterns. While the first two areas have been extensively researched, understanding the long-term impacts of such events on population distribution and mobility remains understudied. Existing research largely focuses on the immediate consequences of pandemic outbreaks, with limited analysis of the delayed and long-term effects. This gap in understanding necessitates a focused investigation into the enduring influence of major public health events on the dynamics of population migration and the resilience of urban systems. The study addresses this gap by specifically examining the long-term effects of COVID-19 on population movement and city network structure in China, a country that implemented stringent movement restrictions early in the pandemic, providing a unique case study for understanding these long-term impacts.

This research utilized Baidu Migration Big Data, a platform tracking the movement of LBS service users, providing indices of migration scale for 369 Chinese cities over 40 days during the Spring Festival from 2019 to 2023. The data was pre-processed to focus on the Spring Festival travel rush, a period of massive population movement. Two key indicators were developed: the Population Migration Scale Index (PMSI), which measures the overall activity of population migration in a city, and the Population Return Scale Index (PRSI), which reflects the net population inflow after the holiday, representing the resumption of work and production. These indices were calculated daily for each city and compared to 2019 data to assess recovery levels. Social Network Analysis (SNA) was employed using a 369x369 inter-urban population migration matrix with population migration (PM) between cities as weights. Five network indicators – Degree, Weighted Degree, Density, Clustering Coefficient, and Characteristic Path Length – were used to quantitatively analyze the network structure characteristics. Natural Breaks (Jenks) classification in ArcGIS 10.5 was used for spatial analysis, categorizing cities based on PMSI and PRSI values. Anselin Local Moran's I was used for spatial autocorrelation analysis to identify clusters and outliers of PMSI recovery indices. Community detection in Gephi was used to analyze the evolution of city network communities based on population mobility networks, identifying shifts in urban agglomerations over time.

The study revealed significant findings on the impact of COVID-19 on population migration in China: **Temporal Analysis of Population Migration Scale:** The Spring Festival period was divided into three phases: Homecoming, Holiday, and Return. In 2020, the initial impact was delayed, but the overall migration scale declined significantly in 2021. Recovery began in 2022 and accelerated in 2023, exceeding 2019 levels. **Spatial Differentiation of Population Migration Scale:** The impact varied across regions, with less active migration cities showing minimal effect. Megacities were less affected. By 2023, most cities recovered to or exceeded pre-pandemic PMSI levels, indicating resilience and minimal residual effects. Spatial analysis showed recovery varied geographically; some areas recovered faster than others. **Changes in Population Return Scale:** Analysis at the provincial and city levels showed that COVID-19's impact was mainly short-term. By 2023, most provinces and cities had largely recovered their population return scale, although some shifts in provincial employment opportunities were observed. Spatial analysis revealed a 'west-to-east, south-to-north' recovery trend. **Changes in City Network and Community Evolution:** Network analysis showed a decrease in Average Degree, Density, and Average Clustering Coefficient from 2020 to 2023, while Average Weighted Degree and Average Path Length increased. This reflects higher direct connectivity during the pandemic, with people opting for direct travel to minimize risk. Community detection showed considerable stability in some regions and significant shifts in others. Southern regions exhibited more dynamic community structure changes than northern regions, with 119 cities changing communities (32.25%). Community boundaries largely aligned with provincial borders, highlighting the influence of administrative divisions on population movements.

The findings demonstrate the resilience of China's population migration system. While COVID-19 caused substantial short-term disruptions, the long-term impacts proved relatively weak, with most cities recovering to pre-pandemic levels. Regional differences were prominent, reflecting variations in economic development, transportation infrastructure, and pandemic control measures. The observation of a 'west-to-east, south-to-north' recovery pattern suggests that more developed, easily accessible regions experienced quicker recovery. The evolution of urban communities highlighted the interplay between geographical proximity and administrative boundaries in shaping population flow patterns. The study's implications extend to pandemic preparedness, urban planning, and the effectiveness of movement restriction policies. The use of spatiotemporal big data offers valuable insights into the impact of such events on population dynamics.

This study provides a comprehensive analysis of the long-term impacts of COVID-19 on China's intercity population migration. The resilience of the population migration system was evident, despite substantial short-term disruptions. Regional disparities in recovery were observed, and urban community structures exhibited notable shifts, particularly in southern China. The findings have crucial implications for policy-making and future research, emphasizing the value of spatiotemporal big data in analyzing the effects of major public health emergencies. Future work could explore the underlying drivers of migration changes in more detail, incorporating variables such as traffic conditions, settlement policies, and economic development.

The study's limitations include potential biases in Baidu Migration Data, which may not represent the entire population. The 2019 data might not accurately reflect typical population movement due to policy changes. Further research is needed to explore the reasons behind observed changes in migration patterns by building models to analyze the impact of a wider range of factors on migration.