Remote-sensing vegetation indices are widely used as low-cost descriptors of urban vegetation. However, their value as transferable proxies for bird diversity across heterogeneous urban green spaces remains uncertain. Using standardised single-visit summer bird surveys from 34 Chinese cities covering 1363 sites and 225 bird species, we evaluated whether simple site-level normalised difference vegetation index (NDVI) and enhanced vegetation index (EVI) mean and standard deviation (SD) metrics could predict selected observed bird diversity metrics and transfers among cities. Species richness (SR), functional evenness (FEve), and Faith's phylogenetic diversity (PD) were used as representative taxonomic, functional and phylogenetic metrics. Nationwide mixed models, robustness analyses, city-level models, random-slope models, and leave-one-city-out cross-validation were used to evaluate the association strength, explanatory value, city-level consistency, and cross-city transferability. The observed SR and log-transformed total PD showed no stable relationships with the four vegetation metrics. In contrast, FEve showed weak positive associations with the mean NDVI and EVI, with marginal R2 values below 1%. Sensitivity analyses did not materially alter these patterns, and cross-validation indicated limited predictive transferability among cities. Overall, the simple NDVI and EVI metrics tested here are better interpreted as broad greenness descriptors than as reliable stand-alone proxies for the selected observed bird diversity metrics in heterogeneous urban green spaces.
Urban green spaces (UGS) play a central role in mitigating urban heat stress. While urban parks have been extensively studied in this context, other widespread UGS types, such as cemeteries and allotment gardens, have received comparatively less attention. This study analyzed 230 parks, 154 cemeteries, and 512 allotments in the city of Berlin (Germany) using Landsat 8 TM images over six dates to compare the cooling performance of these three urban green spaces. For each UGS, the three cooling capacity indicators, Park Cooling Distance (PCD), Park Cooling Intensity (PCI), and Park Cooling Area (PCA) were determined using buffer analysis and turning point method. Additionally, influencing factors within the sites and their surrounding built-up areas were statistically analyzed. The results show the highest mean PCD of 259 m for parks, compared to 217 m for cemeteries and 187 m for allotments, while the cemeteries demonstrate the highest mean PCI of 2.23 °C, compared to parks (2.04 °C) and allotments (1.08 °C). Vegetation cover and impervious surfaces influence PCI across all UGS, while area size and perimeter further affect PCI, particularly in parks and cemeteries. Within the PCA, vegetation and impervious surfaces remain key determinants of PCI. Cemeteries contribute to mitigating the Urban Heat Island (UHI) effect similarly to parks, highlighting the relevance of considering a broader range of UGS types for urban planners and policymakers. Integrating diverse green spaces can help reduce UHI effects and improve thermal comfort, particularly in densely built areas.
Heat vulnerability indices are widely used to rank urban areas, yet they provide limited information about how heat vulnerability is embedded into urban sustainability or why similar aggregate scores may reflect different component configurations. This study examines whether different sustainability dimensions are associated with heatwave vulnerability and identifies associated urban profiles across 72 Hungarian urban units: the 23 districts of Budapest and the 49 largest cities. It combines long-term gridded climate data for 1971–2024 with recent demographic, socioeconomic, infrastructural, and land-cover indicators. Percentile normalization and equal weighting were used to construct sub-indices, followed by Spearman correlation, group comparison and Ward hierarchical cluster analysis. The results indicate that overall sustainability cannot serve as a direct proxy for heatwave vulnerability; moreover, they point out that economically stronger urban areas may combine high built-environment exposure with greater adaptive resources. Besides, Budapest districts had substantially higher exposure and vulnerability than other cities, while Ward clustering identified four exploratory urban profiles, indicating that comparable aggregate vulnerability levels may correspond to different configurations of climatic, structural and social conditions. The main contribution of the study is to integrate urban sustainability and heatwave vulnerability assessment beyond aggregate rankings by providing an analytical basis for local- and profile-specific adaptation interventions.
