Attempts to map desertification are often non-satisfactory; A key requirement for mapping desertification is that it is defined in a way that leads to objective and practical measurement criteria. Earth-observing instruments carried on satellites (Prince 1999) routinely map land surface variables that respond to desertification, such as albedo, surface temperature, and vegetation cover, all with appropriate spatial resolution and regular global coverage. Unfortunately factors that are not related to desertification also affect these properties, for example AVHRR data have been used to monitor interannual changes of vegetation cover in dry regions (Tucker et al. 1991; Nicholson et al. 1998), but these are caused by rainfall fluctuations, not desertification (Prince 2002).

Persistent reduction in productivity is an expression of desertification (Prince 2002) and it is routinely measured using satellite derived vegetation indices (e.g., normalized difference vegetation index, NDVI). NDVI measures the amount of solar radiation absorbed by the vegetation, from which simple methods can be used to estimate net primary production (NPP Prince 1991). Prince (2002) has suggested that a persistent reduction of NPP below its potential, that does not disappear during wetter periods, could furnish a measurement of desertification that is both practical and based on the underlying mechanism.