改革开放以来，随着户籍制度改革和城镇化的推行，人口大规模跨区域流动已经成为我国重要的社会现象。青藏高原地区因其独特的生态环境和重要的战略地位，人口流动越来越频繁，流动人口规模的扩大不仅深刻影响着青藏高原的社会环境等多方面的发展，同时也带来流动人口分布不均衡、城市发展不均衡的严峻问题。本文基于2000年、2010年和2020年3期人口普查数据，采用基尼系数和空间自相关方法，分析2000—2020年间青藏高原各区县人口流动的空间分布格局。通过最小二乘法线性回归模型和时空地理加权回归模型来探究人口流动的动力机制，为青藏高原地区间城市的发展、人口流动引导提出政策建议。研究表明：

（1）2000—2020年间，青藏高原的流出人口规模持续增加，但增长的速度逐年放缓，在空间分布上呈现显著的区域异质性，流出人口规模较大的区县在空间上呈现集聚分布，随着时间的推移，东部地区和东南部地区呈现出了更为明显的流出趋势。与此同时，2000—2020年间，青藏高原流入人口规模也呈现出持续上升的趋势，增长的速度也逐年放缓。在空间分布上，流入人口空间分布格局和流出人口相似，流入人口的空间分布格局也出现了明显的区域异质性，并呈现出集群分布和点状分布并存的空间分布格局。

（2）从县域尺度采用净流动率来衡量青藏高原流动人口的方向，并划分四种地域类型，得出2000—2020年青藏高原流动人口的四种地域类型在空间上均表现出明显的差异性：主要净流入区空间分布从南北向东西两侧扩展，显示出明显的空间变异性、一般净流入区在青藏高原的空间分布格局是从全域广泛分布逐渐向部分区县集中、主要净流出区空间分布从离散模式转变为带状分布、一般净流出区从局部分布转变为全域覆盖。

（3）利用基尼系数和空间自相关方法探究青藏高原不同区县流动人口的空间格局演变。基尼系数表明，20年来流动人口的分布由原来的“相对均匀”状态逐渐转向“不均匀”状态。全局空间自相关结果表明，流动人口的Moran's I值均处于接近0的较低水平，这表明在这20年间，青藏高原的流动人口分布状态基本保持稳定，主要呈现出随机分布的特点。

（4）构建时空地理加权回归模型对青藏高原县域流动人口的影响因素进行检验，研究发现：人均GDP，地方一般预算收入对人口流动具有显著的正向影响；城镇等级对人口流动的影响力越来越明显，在2000年和2010年，正向影响和负向影响共存，但到2020年，全域156个区县全为正向影响区，这意味着城镇等级的提升越来越能成功吸引人口流入；到省会或首府的距离对人口流动的影响在2000年主要为正负影响共存，随着时间的推移，正向影响的区县数量不断增加，逐渐转变为以正向影响为主。海拔和降水在青藏高原某些县域以高值正向影响为主，但某些县域以高值负向为主，在空间上呈现二元分布的态势。

Since the reform and opening up, accompanied by the reform of the household registration system and the advancement of urbanization, large-scale cross-regional population movement has emerged as a significant social phenomenon in China. In the Qinghai-Tibet Plateau region, due to its distinct ecological environment and crucial strategic status, population mobility has become increasingly frequent. The expansion of population movement not only profoundly affects the social environment and various aspects of development in the Qinghai-Tibet Plateau, but also poses stark challenges related to the uneven distribution of the mobile population and imbalanced urban growth. This study, utilizing population census data from the years 2000, 2010, and 2020, employs the Gini coefficient and spatial autocorrelation methods to analyze the spatial distribution pattern of population movement among the counties of the Qinghai-Tibet Plateau from 2000 to 2020. By applying ordinary least squares linear regression models and spatiotemporal geographically weighted regression models, it investigates the driving mechanisms behind population mobility and offers policy suggestions for urban development and the steering of population movements within the Qinghai-Tibet Plateau region. The research shows:

Firstly, during the period from 2000 to 2020, the outflow of population from the Qinghai-Tibet Plateau continued to grow, although the pace of growth slowed down year by year. The spatial distribution of this outward migration displayed significant regional heterogeneity, with densely populated districts showing a clustered distribution. As time progressed, the eastern and southeastern regions exhibited a more pronounced trend of population exodus. Meanwhile, the scale of the inflow population to the Qinghai-Tibet Plateau also showed a continuous upward trend during 2000-2020, with the rate of growth similarly slowing over the years. In terms of spatial distribution, the pattern of incoming population mirrored that of the outgoing population, characterized by significant regional heterogeneity and a spatial pattern where cluster distributions and scattered distributions coexisted.

Secondly, Using net migration rates at the county level to measure the direction of population mobility on the Qinghai-Tibet Plateau and to delineate four types of geographic areas, it was found that from 2000 to 2020, the spatial distribution of these four types of areas on the Qinghai-Tibet Plateau exhibited clear differences: main net inflow areas showed spatial expansion from north to south towards the east and west sides, indicating significant spatial variability; general net inflow areas' spatial distribution shifted from being widely spread across the plateau to a concentration in certain counties; main net outflow areas underwent a notable change, with their spatial distribution transitioning from a dispersed to a zonal pattern; and general net outflow areas evolved from a localized distribution to covering the plateau in its entirety.

Thirdly, Utilizing the Gini coefficient and spatial autocorrelation methods to examine the evolution of the spatial pattern of the migrating population across different counties in the Qinghai-Tibet Plateau. The Gini coefficient results indicate that over the past 20 years, the distribution of the migrating population has shifted from a state of "relative uniformity" to more "unevenness." The global spatial autocorrelation outcome, revealed by Moran's I index being close to 0, signifies that during these two decades, the distribution of the migrating population on the Qinghai-Tibet Plateau has remained essentially stable, characterized by a predominantly random distribution.

Lastly, by constructing a spatio-temporal geographically weighted regression model, an examination was carried out on the factors influencing population mobility in the counties of the Qinghai-Tibet Plateau. The study found that per capita GDP and local general budget income have a significant positive impact on population movement; urban status is increasingly affecting population mobility. In the years 2000 and 2010, both positive and negative effects co-existed, but by 2020, all 156 districts of the region showed a positive influence, indicating that the improvement of urban status is increasingly successful in attracting population inflow. The distance to the provincial capital or regional capital had a mixed positive and negative impact on population mobility in 2000, and as time progressed, the number of districts with positive impacts continued to grow, gradually shifting to a predominantly positive influence. Altitude and precipitation exert a high-value positive impact in certain counties of the Qinghai-Tibet Plateau, but in other counties, they predominantly have a high-value negative influence, presenting a bipolar distribution in the spatial dimension.