Assessing Urban Forest Canopy Cover Using Airborne or Satellite Imagery

Article Figures & Data

Table 1.

Accuracy assessment of cover classification for Syracuse, New York, using high-resolution digital color infrared imagery.^*

Class name	Reference					Row total	User’s accuracy (%)	Standard error (%)
Class name	Tree/shrub	Grass/herbaceous	Bare soil	Water	Impervious surface	Row total	User’s accuracy (%)	Standard error (%)
Classified
Tree/shrub	25.0	3.75	0.0	0.0	0.25	29	86.2	3.2
Grass/herb	3.5	13.0	0.75	0.0	1.5	18.75	69.3	5.4
Bare soil	0.0	0.0	0.5	0.0	0.75	1.25	40.0	24.5
Water	0.0	0.0	0.0	6.5	0.0	6.5	100.0	0.0
Impervious	2.25	4.5	1.0	0	36.75	44.5	82.6	2.8
Column total	30.75	21.25	2.25	6.5	39.25	100
Producer’s accuracy (%)	81.3	61.2	22.2	100.0	93.6		Overall 81.75
Standard error (%)	3.5	5.3	14.7	0.0	2.0			Overall 1.9

↵*Myeong et al. (2001). Error matrix entries are based on percent of 400 ground-truth points.

Table 2.

Urban forest canopy cover comparison for Syracuse, New York.^{^z}

Method	Imagery	Year	Percent canopy cover	Standard error or accuracy	Source^{^y}
Photograph interpretation	Photographic prints	1984	24%		A
	Emerge 2 ft (60 cm) (digital)	1999	25.7%	SE: 2.5%	F
	NYS Orthos 1 ft (30 cm) (digital)	2003	24.7%	SE: 2.5%	F
Supervised classification	Emerge 2 ft (60 cm)	1999	26.6%	Acc: 82%	D
Subpixel–Cubist (specific)	Landsat 30 m	ca. 2001	26.6%	MAE: 11.7%	F
NLCD subpixel (regional)	Landsat 30 m	ca. 2001	12.7%		E
USFS Forest Density	AVHRR 1.1 km	1991	21.6%		B
USFS RPA (water removed)	AVHRR 1.1 km	1991	23.0%		C
UFORE ground sample		1999	24.4%	SE: 1.97%	F
		2001	23.1%	SE: 1.82%	F
		2004	21.4%	SE: 1.79%	F

↵^zResults in shaded rows were generated using digital classification techniques.
↵^yA = Rowntree (1984); B = Zhu (1994); C = Dwyer et al. (2000); D = Myeong et al. (2001); E = USGS National Land Cover Database (2001); Homer et al. (2007); F = USDA Forest Service, Northern Research Station, Syracuse, NY, unpublished data.
NLCD = National Land Cover Database; USFS RPA = U.S. Forest Service’s Resources Planning Act; UFORE = Urban Forest Effects; AVHRR = Advanced Very-High Resolution Radiometer; MAE = mean absolute error; SE = standard error.