Aerotropolis is an urban development concept characterised by the concentration of land uses fostering economic activities around airports. However, limited empirical evidence exists about the formation of aerotropolises and their effects on the productivity of airports. This study addresses the gaps by examining changes in land use patterns around 75 major airports globally, using data extracted from Google Earth for two time periods: 2004 and 2014. A Two-Step cluster analysis technique was applied to identify natural groupings of the airports in terms of their land use patterns for both periods. Changes in cluster membership were observed between the periods to examine the process of aerotropolis formation. A multivariate multiple linear regression (MMLR) model is estimated to identify factors contributing to aerotropolis formation process. Finally, the effects of the aerotropolis formation process on airport productivity is estimated in a two-stage least square (2SLS) regression model. Cluster analysis results show that airport land side development can be classified into three groups: Island airports (surrounded by largely vacant land), Waterfront airports (adjacent to a large body of water), and Built form airports (well-developed residential, commercial and industrial land uses). Overall, built-up areas increased by 2.1% between 2004 and 2014, and only three airports transitioned from Island to Built form – suggesting a weak evidence of the aerotropolis formation process. Developable land surrounding the airports and city population growth rate contributed significantly to this process, which consequently contributed to the productivity of airports. In conclusion, aerotropolises are a promising concept, yet unreal. It needs to be integrated with city plans to direct development, and thereby, to make it a reality.

Aerotropolis 是一种城市发展概念，其特点是土地用途的集中，促进了机场周围的经济活动。然而，关于航空大都市的形成及其对机场生产力的影响的经验证据有限。本研究使用从 Google 地球中提取的两个时间段的数据（2004 年和 2014 年）检查全球 75 个主要机场周围土地使用模式的变化，从而解决了这些差距。应用了两步聚类分析技术来识别机场的自然分组两个时期的土地利用模式。在各个时期之间观察到集群成员的变化，以检查航空城形成的过程。估计多元多元线性回归 (MMLR) 模型以识别影响航空城形成过程的因素。最后，在两阶段最小二乘 (2SLS) 回归模型中估计了航空城形成过程对机场生产力的影响。聚类分析结果表明，机场陆侧开发可分为三类：岛屿机场（周围有大量空地）、滨水机场（毗邻大片水域）和建成型机场（开发良好的住宅、商业和工业用地）。总体而言，建成区面积在 2004 年至 2014 年间增加了 2.1%，只有三个机场从岛屿形式转变为建成形式——这表明航空城形成过程的证据不足。机场周边的可开发土地和城市人口增长率对这一过程做出了重大贡献，从而促进了机场的生产力。综上所述，航空城是一个很有前途的概念，但并不现实。它需要与城市规划相结合以指导发展，从而使其成为现实。