Estimating urban vegetation cover fraction using Google Earth (R) images

We exploited publicly available satellite-and aircraft-based imagery to estimate urban vegetation cover fraction and land use by class for a semiarid urban area that includes Phoenix, AZ, USA, using low-cost and technologically modest tools. This technique is also used to evaluate two satellite-derived tree cover datasets as well as to compare estimates from the present study with land cover data generated from another study performed using the same study domain. The approach outlined in this article entails the use of Google Earth (R) images that are analyzed either visually or by using a more rigorous visually supervised digital reclassification method. Neither method is automated. Determination of optimal sample size was also an objective of the study. The limitations and advantages associated with these approaches are described.