Table 1 Description of city green space by city natural characteristics^§

		Indicators of quantity of green space						Indicators of spatial configuration of green space						Urban park metrics
		Greenness (NDVI)	P¹	% Green space area	P	Green space area per capita (m² per 1,000 residents)	P	Patch density (number of patches per ha)	P	Isolation (m)	P	Aggregation	P	Number of parks per capita (number of parks per 1,000 residents)	P	Park area per capita (m² per 1,000 residents)	P
Natural environment
Climate zone²	Arid (N = 79 cities)	0.36 (0.13)	<0.001	30.74 (13.49)	<0.001	107,522 (68,817)	<0.001	49.42 (18.31)	<0.001	37.15 (7.64)	0.0049	0.80 (0.06)	<0.001	0.43 (0.30)	<0.001	2,640 (1,494)	<0.001
	Temperate (N = 125)	0.57 (0.08)		44.87 (8.55)		158,613 (77,921)		37.69 (14.86)		35.61 (4.86)		0.83 (0.04)		0.33 (0.20)		2,308 (1,433)
	Tropical (N = 166)	0.59 (0.08)		42.93 (8.19)		104,578 (51,046)		46.74 (19.07)		34.69 (4.63)		0.82 (0.04)		0.25 (0.19)		1,782 (1,444)
Topography/terrain	Flat (N = 137)	0.50 (0.13)	<0.001	38.42 (11.21)	<0.001	121,135 (64,346)	0.70	47.27 (18.23)	0.014	35.87 (6.19)	0.36	0.34 (0.25)	0.18	0.34 (0.25)	0.19	2,341 (1,493)	0.05
Topography/terrain	Hilly (N = 234)	0.55 (0.13)	<0.001	42.4 (10.76)	<0.001	124,440 (72,654)	0.70	42.53 (17.94)	0.014	35.33 (5.12)	0.36	0.31 (0.22)	0.18	0.31 (0.22)	0.19	2,025 (1,473)	0.05
Coastal	Inland (N = 300)	0.54 (0.12)	0.11	41.58 (10.43)	0.018	126,348 (67,591)	0.06	44.64 (18.32)	0.44	35.06 (5.27)	<0.001	0.82 (0.05)	0.29	0.32 (0.23)	0.87	2,244 (1,527)	0.006
Coastal	Coastal (N = 71)	0.51 (0.16)	0.11	38.19 (13.22)	0.018	110,002 (76,776)	0.06	42.76 (17.56)	0.44	37.51 (6.2)	<0.001	0.83 (0.06)	0.29	0.32 (0.24)	0.87	1,709 (1,222)	0.006

^§The values shown are means (and standard deviations in parentheses) of green space variables for each category of city indicators. Categories of city variables (low/small, medium and high/large) are based on sample-wide tertiles for each variable unless noted otherwise. Detailed definitions of the natural, built and socioeconomic indicators can be found in Extended Data Table 1. Definitions of the green space variables are in Extended Data Table 2.
¹P is the P value from the ANOVA test of differences between groups means. A two-sided post-hoc multi-comparison Tukey’s honestly significant difference (HSD) test was performed on the ANOVA to identify which subgroup means showed statistically significant differences⁶⁵. The P value from Tukey’s HSD accounts for multiple comparisons within the categories of each city indicator and determines whether there is a statistically significant difference in the means of each green space variable across the categories of a city indicator. Grey shading denotes city indicators for whose categories we observe statistically significant differences in the means of green space variables.
²The sample contained one polar city (Puno), which was excluded from the climate stratification.

Search