基于地理加权随机森林模型的流动人口驱动因子 研究* ——以湖南省县域为例
Decoupling the Driving Factors of Floating Population Using a Geographically Weighted Random Forest Model: A County-Level Case Study in Hu'nan Province
朱 珠
南华大学建筑与设计艺术学院 硕士研究生
郑卫民（通信作者）
南华大学建筑与设计艺术学院 教授，博士，2021000028@usc.edu.cn
李 晟
南华大学建筑与设计艺术学院 副教授，博士
吴 博
南华大学建筑与设计艺术学院 讲师，硕士
周宸旭
南华大学经济与管理法学院 硕士研究生
摘要: 目前，县域人口流动的驱动机制研究多基于传统线性模型，难以捕捉驱动因子的非线性阈值效应和空间异质性。引入机器学 习中的RF和GWRF模型，通过构建“系统阈值—地理权重”双维框架，利用2000—2020年人口普查数据融合多源地理大数 据，以湖南省县域为研究对象，探究经济、社会、环境和区位4类因子对人口县内流动、跨县流入和跨县流出的作用规律。研究 发现：（1）湖南省人口流动呈现梯度分异和多中心演化趋势；（2）湖南省人口流动主要受经济要素限制和生态要素约束；（3） 流动人口的驱动机制存在明显的阈值效应和边际递减规律；（4）流动人口的驱动机制在空间上存在明显的“核心极化—边 缘收缩”特征。研究结果可为中部地区的人口流动管理提供“动态阈值预警—空间靶向施策”的方法论参考。
Abstract: Prevailing studies on the investigation of the driving mechanisms of county-level population movements relying on traditional linear models inadequately capture the nonlinear threshold effects and spatial heterogeneity. This study innovatively integrates machine learning techniques (Random Forest and Geographically Weighted Random Forest) to establish a dual-dimensional "systematic threshold-geographical weighting" analytical framework. Leveraging integrated datasets from population censuses (2000-2020) and multi-source geospatial big data across Hu'nan Province's counties, we systematically investigate how economic, social, environmental, and locational factors govern intra-county mobility, inter-county immigration, and outmigration. Key findings reveal: (1) a gradient differentiation pattern with polycentric evolutionary trends in population redistribution, (2) dual constraints from economic thresholds and ecological carrying capacities, (3) pronounced threshold effects with diminishing marginal impacts across driving factors, and (4) distinct core polarization-edge contraction characteristics in spatial mechanisms. The proposed methodology advances dynamic threshold monitoring and spatially adaptive governance strategies, offering policy insights for managing population mobility in central China's transitional regions.
关键词：流动人口；驱动因子；非线性阈值；空间异质性；地理加权随机森林模型；湖南省县域
Keyword: floating population; driving factors; nonlinear threshold; spatial heterogeneity; geographically weighted random forest (GWRF) model; county-level units in Hu'nan Province
中图分类号：TU984
文献标识码: A
翟振武，金光照. 中国人口负增长：特征、挑战与
应对[J]. 人口研究，2023，47（2）：11-20.
ZHAI Zhenwu, JIN Guangzhao. Negative population growth in China: characteristics, challenges,
and responses[J]. Population Research, 2023, 47(2):
11-20.
刘涛，卓云霞. 中国县级人口变动的空间格局及
影响因素——基于第七次全国人口普查数据的
新探索[J]. 人口研究，2022，46（6）：72-87.
LIU Tao, ZHUO Yunxia. A county-level analysis
of China's population change: insights from the 7th
population census[J]. Population Research, 2022,
46(6): 72-87.
ZELINSKY W. The hypothesis of the mobility
transition[J]. Geographical Review, 1971, 61(2):
219-249.
SKELDON R. Migration and development: a global
perspective[M]. London: Routledge, 2014.
FAN C C. China on the move: migration, the state,
and the household[J]. London: Routledge, 2007.
刘涛，彭荣熙，卓云霞，等. 2000—2020年中国人口
分布格局演变及影响因素[J]. 地理学报，2022，
77（2）：381-394.
LIU Tao, PENG Rongxi, ZHUO Yunxia, et al. China's
changing population distribution and influencing
factors: insights from the 2020 census data[J]. Acta
Geographica Sinica, 2022, 77(2): 381-394.
侯华丽，冯健，张瑞华. 基于日常人口流动的城
市群等级与网络结构——对中国五大城市群的
考察[J]. 城市发展研究，2024，31（9）：91-100.
HOU Huali, FENG Jian, ZHANG Ruihua. Hierarchy
and network structure of urban agglomerations
based on daily population mobility: a study of five
major urban agglomerations in China[J]. Urban
Development Studies, 2024, 31(9): 91-100.
邵云通，吴晓. 中国城际人口流动网络的多尺度
空间差异——基于腾讯迁徙大数据的分析[J].
地理科学，2025，45（1）：189-201.
SHAO Yuntong, WU Xiao. Multi-scale spatial
differences of intercity population flow network
in China: based on Tencent migration big data[J].
Geographical Science, 2025, 45(1): 189-201.
金丹，周婧楠，许尊. 京津冀城市群流动人口
空间格局及影响因素研究[J]. 城市发展研究，
2024，31（12）：34-44.
JIN Dan, ZHOU Jingnan, XU Zun. Study on the
spatial pattern and influencing factors of floating
population in the Beijing-Tianjin-Hebei urban
agglomeration: analysis based on the sixth national
census and the seventh national census[J]. Urban
Development Studies, 2024, 31(12): 34-44.
王桂新. 中国省际人口迁移变化特征——基于
第七次全国人口普查数据的分析[J]. 中国人口
科学，2022（3）：2-16.
WANG Guixin. Research on characteristics of
China's inter-provincial migration: based on the data
of China's seventh population census[J]. Chinese
Journal of Population Science, 2022(3): 2-16.
曹广忠，刘涛. 中国城镇化地区贡献的内陆化演
变与解释——基于1982—2008年省区数据的分
析[J]. 地理学报，2011，66（12）：1631-1643.
CAO Guangzhong, LIU Tao. Rising role of
inland regions in China's urbanization in the 21st
century: the new trend and its explanation[J]. Acta
Geographica Sinica, 2011, 66(12): 1631-1643.
林李月，朱宇，柯文前. 城镇化中后期中国人口
迁移流动形式的转变及政策应对[J]. 地理科学
进展，2020，39（12）：2054-2067.
LIN Liyue, ZHU Yu, KE Wenqian. Changes in
the form of population migration and mobility in
China and corresponding policy responses at the
late-intermediate stage of urbanization[J]. Progress
in Geography, 2020, 39(12): 2054-2067.
吴瑞君，薛琪薪，罗志华. 我国人口迁移和城镇
化格局的转折性变化：2000—2020年[J]. 上海行
政学院学报，2022，23（1）：74-86.
WU Ruijun, XUE Qixin, LUO Zhihua. The
turning change of the spatial patterns of population migration and urbanization in China from
2000 to 2020[J]. The Journal of Shanghai Administration Institute, 2022, 23(1): 74-86.
刘艳军，汤爽爽，吴康，等. 经济地理学视角下中
国人口研究热点与展望[J]. 经济地理，2021，
41（10）：97-105.
LIU Yanjun, TANG Shuangshuang, WU Kang, et
al. Hotspots and prospects of population research
in China from the perspective of economic geography[J]. Economic Geography, 2021, 41(10): 97-105.
罗小龙，曹姝君，顾宗倪. 回流城镇化：中部地区
城镇化开启新路径[J]. 地理科学，2020，40（5）：
685-690.
LUO Xiaolong, CAO Shujun, GU Zongni. The
returning-migrant-induced urbanization: the
new path of urbanization in middle China[J].
Geographical Science, 2020, 40(5): 685-690.
王新贤，高向东. 中国流动人口分布演变及其对
城镇化的影响——基于省际、省内流动的对比
分析[J]. 地理科学，2019，39（12）：1866-1874.
WANG Xinxian, GAO Xiangdong. The evolution
of China's floating population and its impact on
urbanization: a comparative analysis based on interand intra-provincial perspectives[J]. Geographical
Science, 2019, 39(12): 1866-1874.
柯文前，肖宝玉，林李月，等. 中国省际城乡流动
人口空间格局演变及与区域经济发展的关系[J].
地理学报，2023，78（8）：2041-2057.
KE Wenqian, XIAO Baoyu, LIN Liyue, et al.
Interprovincial urban and rural floating population
evolution of China and its relationship with regional
economic development[J]. Acta Geographica
Sinica, 2023, 78(8): 2041-2057.
赖建波，朱军，郭煜坤，等. 中原城市群人口流动
空间格局与网络结构韧性分析[J]. 地理与地理
信息科学，2023，39（2）：55-63.
LAI Jianbo, ZHU Jun, GUO Yukun, et al. Spatial
pattern of population flow and the resilience of
network structure of central plains urban agglomeration[J]. Geography and Geo-Information Science,
2023, 39(2): 55-63.
曹广忠，陈思创，刘涛. 中国五大城市群人口流
入的空间模式及变动趋势[J]. 地理学报，2021，
76（6）：1334-1349.
CAO Guangzhong, CHEN Sichuang, LIU Tao.
Changing spatial patterns of internal migration to
five major urban agglomerations in China[J]. Acta
Geographica Sinica, 2021, 76(6): 1334-1349.
李聪，宗会明，肖磊. 中国典型人口流出地区人
口流动格局——以川渝地区为例[J]. 热带地理，
2021，41（3）：516-527.
LI Cong, ZONG Huiming, XIAO Lei. Spatial
pattern of population flow in China's typical outflow
areas: a case study of the Sichuan-Chongqing
area[J]. Tropical Geography, 2021, 41(3): 516-527.
SHEN J. Modelling interregional migration in
China in 2005–2010: the roles of regional attributes
and spatial interaction effects in modelling error[J].
Population, Space and Place, 2017, 23(3): e2014.
GU H. Understanding the migration of highly and
less-educated labourers in post-reform China[J].
Applied Geography, 2021, 137: 102605.
STARK O. The migration of labor[M]. Hoboken:
Blackwell Publishing, 1991.
HARRIS J R, TODARO M P. Migration, unemployment and development: a two-sector analysis[J]. The
American Economic Review, 1970, 60(1): 126-142.
KRUGMAN P. Increasing returns and economic
geography[J]. Journal of Political Economy, 1991,
99(3): 483-499.
BLACK R, ADGER W N, ARNELL N W, et al.
The effect of environmental change on human
migration[J]. Global Environmental Change, 2011,
21: S3-S11.
MCLEMAN R. International migration and climate
adaptation in an era of hardening borders[J]. Nature
Climate Change, 2019, 9(12): 911-918.
王婧，刘奔腾，李裕瑞. 京津冀人口时空变化特
征及其影响因素[J]. 地理研究，2018，37（9）：
1802-1817.
WANG Jing, LIU Benteng, LI Yurui. Spatialtemporal characteristics and influencing factors
of population distribution and floating changes
in Beijing-Tianjin-Hebei region[J]. Geographical
Research, 2018, 37(9): 1802-1817.
王婧，李裕瑞. 中国县域城镇化发展格局及其影
响因素——基于2000和2010年全国人口普查分
县数据[J]. 地理学报，2016，71（4）：621-636.
WANG Jing, LI Yurui. Spatial pattern and influencing factors of urbanization development in
China at county level: a quantitative analysis based
on 2000 and 2010 census data[J]. Acta Geographica
Sinica, 2016, 71(4): 621-636.
田盼盼，朱宇，林李月，等. 省际与省内流动人口
空间分布及其影响因素的差异——以福建省为
例[J]. 人口学刊，2015，37（6）：56-67.
TIAN Panpan, ZHU Yu, LIN Liyue, et al. Differences in the spatial distribution and its determinants between inter- and intra-provincial floating
population: the case of Fujian Province[J]. Population Journal, 2015, 37(6): 56-67.
王超，阚瑷珂，曾业隆，等. 基于随机森林模型的
西藏人口分布格局及影响因素[J]. 地理学报，
2019，74（4）：664-680.
WANG Chao, KAN Aike, ZENG Yelong, et al.
Population distribution pattern and influencing
factors in Tibet based on random forest model[J].
Acta Geographica Sinica, 2019, 74(4): 664-680.
GEORGANOS S, GRIPPA T, NIANG GADIAGA
A, et al. Geographical random forests: a spatial
extension of the random forest algorithm to
address spatial heterogeneity in remote sensing and
population modelling[J]. Geocarto International,
2021, 36(2): 121-136.
邓楚雄，李民，宾津佑. 湖南省人口分布格局时
空变化特征及主要影响因素分析[J]. 经济地理，
2017，37（12）：41-48.
DENG Chuxiong, LI Min, BIN Jinyou. Spatialtemporal variation characteristics and main influence
factors analysis of population distribution pattern in
Hu'nan Province[J]. Economic Geography, 2017,
37(12): 41-48.
段成荣，盛丹阳，吕利丹，等. 迁徙中国视野下我
国县域人口流动状况与发展挑战[J]. 西北人口，
2022，43（6）：1-13.
DUAN Chengrong, SHENG Danyang, LYU Lidan,
et al. Study on migration of China's county and
its challenges from the perspective of migrating
China[J]. Northwest Population Journal, 2022,
43(6): 1-13.
曾永明，钟子康，刘厚莲. 中国县域流动人口分
布样态及其演变——基于柯西分布的模拟及预
测[J]. 中国人口科学，2024，38（1）：113-128.
ZENG Yongming, ZHONG Zikang, LIU
Houlian. The pattern and dynamics of countylevel population flow in China: simulation and
projection using the Cauchy Distribution[J].
Chinese Journal of Population Science, 2024,
38(1): 113-128.
BREIMAN L. Random forests[J]. Machine
Learning, 2001, 45: 5-32.
GRÖMPING U. Variable importance assessment
in regression: linear regression versus random
forest[J]. The American Statistician, 2009, 63(4):
308-319.
