Assessing Spatial Equity and Potential Green Gentrification in Urban Farming Using GIS-based Machine Learning: A Case Study of Taipei’s Garden City Program

Abstract

This study explores the spatial equity and socio-spatial implications of urban agriculture, focusing on the distribution of community gardens under Taipei’s Garden City Program (TGCP) and their relationship to neighborhood-level income growth. By integrating unsupervised machine learning (DBSCAN) with spa­tial statistical methods (Moran’s I and LISA), the analysis reveals that garden service density is highly concentrated in central districts, while peripheral areas exhibit significant service gaps. At the citywide scale, no significant correlation is found between garden density and income growth. This finding aligns with existing research suggesting that community gardens—being low-cost, small-scale, and decentralized—reflect the characteristics of the “just green enough” approach and are less likely to trigger conventional forms of green gentrification. However, several local areas do exhibit potential spatial coupling between garden presence and income change, highlighting the need for future policies to address distributional equity and neighborhood transition risks in urban agriculture planning.