ABSTRACT
Back in the early 1980s, in a seminal article, Welch (1982) outlined the spatial resolution requirements of remotely sensed imagery for classifying urban areas. In terms of spaceborne instruments, the contribution of remote sensing to urban studies was deemed limited at the very least. Simply put, satellite sensors were not capable of generating imagery with a fine enough spatial resolution to identify the majority of urban features. In fact, not only were instruments “at the time” insufficient for accurate urban classification, but Welch (1982) predicted that other finer spatial resolution instruments planned for launch later in the 1980s would not significantly increase classification accuracy. It is only relatively recently, at the start of the third millennium, that the spatial resolution of satellite sensors has increased to the extent that accurate urban classification can be performed, in line with Welch’s (1982) recommendations.
