Two methods of analysis were evaluated for mapping urban vegetation on high-altitude, color-infrared aerial photographs and Landsat MSS data of Syracuse, NY. The first method consisted of defining the spatial patterns (strata) of urban vegetation occurrence. The second method discriminated woody and herbaceous vegetation classes within defined strata. Emphasis was placed on evaluating the consistency of each method. Results indicate that consistent spatial patterns of urban vegetation strata were not achieved on either of the two data types tested due to the spatial complexity of the urban vegetation. However, for discriminating woody and herbaceous vegetation classes within defined strata, good consistency was noted among the interpreters of the high-altitude aerial photographs. The coarse spatial resolution of the Landsat MSS data resulted in low precision for identifying these two vegetation classes in this highly urbanized area. Where photointerpretation efforts are intended for mapping vegetation within numerous urban areas, the estimation of proportions of vegetation classes within defined strata should be a data analysis procedure more objective and consistently repeatable than is the delineation of vegetation patterns.
