地球信息科学学报 ›› 2021, Vol. 23 ›› Issue (2): 246-258.doi: 10.12082/dqxxkx.2021.200362

• 疫情时空分析 • 上一篇    下一篇

政府严控期我国地级市COVID-19疫情的时空集聚、演变及自相关效应研究

巫细波*(), 赖长强, 葛志专   

  1. 广州市社会科学院,广州 510410
  • 收稿日期:2020-07-09 修回日期:2020-09-23 出版日期:2021-02-25 发布日期:2021-04-25
  • 通讯作者: 巫细波 E-mail:wuxibo@gz.gov.cn
  • 基金资助:
    国家社会科学基金项目(20BTJ055);国家自然科学基金青年项目(41801167)

Study on Agglomeration, Evolution and Autocorrelation Effects of Spatio-temporal of COVID-19 Epidemic in Prefecture-level Cities in China during Government's Strict Control Period

WU Xibo*(), LAI Changqiang, GE Zhizhuan   

  1. Guangzhou Academy of Social Sciences, Guangzhou 510410, China
  • Received:2020-07-09 Revised:2020-09-23 Online:2021-02-25 Published:2021-04-25
  • Contact: WU Xibo E-mail:wuxibo@gz.gov.cn
  • Supported by:
    Program of National Social Science Foundation of China(20BTJ055);Youth Program of National Natural Science Foundation of China(41801167)

摘要:

突发性重大公共传染性疫情在地级城市层面政府严格防控时期的时空演变特征能够有效反应我国综合应急防控能力。基于中国2020年1月24日—3月5日312个城市的COVID-19累计确诊数、现有确诊数、治愈数等统计数据,采用ESDA、优化的热点分析、空间马尔科夫链、空间面板数据模型等方法分析了政府严控期COVID-19疫情在312个城市的时空变化特征。研究发现:① 全国COVID-19现存确诊数经历了“快速增长扩散、基本控制、逐渐下降、局部地区完全控制”的变化特征并在2月17日达到峰值,上升期的日均增长率为17.5%,下降期的日均下降率为5.1%,绝大部分城市的疫情变化特征与全国总体情况类似;② 春运期间的人口流动性高是导致疫情快速扩张的主要原因,武汉“封城”之前14 d的百度迁徙强度指数与部分城市的累计确诊数显著相关;③ 优化的热点分析方法识别出疫情热点的空间分布具有固定性且主要分布于以武汉为中心、半径约350 km范围内的36个城市,未识别出具有统计显著性的疫情冷点城市;④ 对各城市现有确诊人数的马尔科夫链转移概率矩阵分析结果显示,各种类型维持现状的概率大于0.85,向下转移的平均概率明显高于向上转移的概率,在不同空间滞后类型的影响下各类型转移概率发生明显变化;⑤ 空间面板数据模型估计结果显示312个城市的现存确诊数具有显著的空间和时间自相关性。本研究从地级市层面多角度分析了政府严控期间COVID-19疫情的时空变化特征,疫情防控重点在于降低其时空自相关效应,为我国当前及未来应对突发性重大公共传染性疫情提供决策参考。

关键词: COVID-19疫情, 政府严控期, 疫情热点, 迁徙规模指数, 时空演变, 时空自相关, 空间马尔科夫链模型

Abstract:

The spatio-temporal evolution of major public infectious epidemics during government's strict control period in prefecture-level city can effectively reflect china's comprehensive emergency prevention and control capabilities. Based on statistical data including number of active cases, total confirmed, deaths of COVID-19 in 312 cities in China from January 24 to March 5, 2020, this paper uses methods including exploratory spatial data analysis, optimized hot spot analysis, spatial Markov chain, spatial panel data model to analyze spatio-temporal evolution characteristics of COVID-19 epidemic in China under government's strict control.The study found that: (1) The number of active cases of COVID-19 in China experienced characteristics of "rapid growth and diffusion, basic control, gradual decline, and complete control in some areas" and reached its peak on February 17, with an average daily growth rate of 17.5% during rising period and an average daily decline rate of 5.1% during falling period, and the epidemic change characteristics of most cities are similar to Nationwide's situation;(2) The high population mobility during Spring Festival transportation period is main reason for rapid expansion of epidemic. The Baidu's migration scale index for the 14 days prior to Wuhan closure was significantly correlated with total confirmed cases of COVID-19 in some cities; (3) The method called optimized hot spot analysis has identified that spatial distribution of hot spots of epidemic is stable and mainly distributed in 36 cities with Wuhan as the center and a radius of about 350 kilometers, while no statistically significant cold spot cities were identified; (4) The results of Markov chain transfer probability matrix analysis of active cased of COVID-19 in 312 cities show that various types are more stable and the probability of maintaining original type is greater than 0.85. The average probability of downward transfer is significantly higher than the probability of upward transfer. The probability of each type of transition changes significantly under the influence of different spatial lag types; (5) The estimation results of the spatial panel data model show that the number of active cases of COVID-19 in cites has spatial-temporal autocorrelation. This paper analyzed spatio-temporal evolution characteristics of COVID-19 epidemic during government's strict control period at prefecture-level city level from multiple perspectives, the focus of COVID-19 prevention and control is to reduce its spatio-temporal autocorrelation effects, this study provides a decision-making reference for government's current and future response to major public infectious epidemics.

Key words: COVID-19 epidemic, government's strict control period, epidemic's hot spot, migration scale index, spatio-temporal evolution, spatio-temporal autocorrelation, spatial markov chain model