Developing a Context-Sensitive Framework for Evaluating Urban Greenspace Indicators Across Diverse Geographic...

Urban greenspace exposure significantly impacts health within densely built environments. However, commonly used greenspace indicators, such as the Normalized Difference Vegetation Index and Green View Index, often yield divergent estimates of exposure. This reduces comparability across greenspace-health studies and leads to inconsistent findings. While recent research has identified that several factors (e.g., slope, building density) contribute to these discrepancies, these factors often co-occur to form complex geographic contexts. Yet, few studies have systematically examined how these co-occurring environmental factors form geographic contexts and subsequently contribute to these measurement differences. This study fills this gap by designing a context-sensitive diagnostic framework to elucidate these measurement discrepancies. We applied this framework in Hong Kong to evaluate five commonly used greenspace indicators, including Normalized Difference Vegetation Index, Enhanced Vegetation Index, Green View Index, and three-dimensional vegetation metrics like Vegetation Volume Index and Canopy Volume Index, to investigate how these indicators vary across different geographic contexts. The findings reveal three context-dependent patterns of divergence among greenspace indicators: differences in measurement sensitivity between satellite indicators, observation-perspective differences between overhead measures and street-level visibility, and contrasts between two-dimensional and three-dimensional representations of vegetation structure. Drawing on these diagnostic insights from the Hong Kong case study, we developed a practical decision-support protocol that links indicator selection to measurement focus, observation perspective, and geographic context. This empirical application illustrates how the framework can support more context-sensitive interpretation of greenspace exposure indicators and inform indicator selection in high-density urban environments.