Pixels that became de-urbanized in the landuse classification between 1985 and 2000 are not accepted, which imposes a somewhat restrictive condition on the earlier image (1985) that any pixel not urban in 2000 must also be not urban in the earlier images. Since the easiest mistake to make in classification is to mistake a bright, dry field to be urban, all pixels in the earlier image that were not urban in the later one but classified urban in the earlier one are assumed to be in to be in the bare soil/scrub category. This tends to underestimate the impervious surface area and urban cover a bit on the 1985 image, but has no effect on the fractional vegetation cover.
For complete information on the processes and algorithms used see the following: Carlson, T.N. and D.A. J. Ripley, 1997: On the relationship between NDVI, fractional vegetation cover and leaf area index. Remote Sensing of the Environ.,, 62, 241-252.
Gillies. R. R. and T. N. Carlson, 1995: Thermal remote sensing of surface soil water content with partial vegetation cover for incorporation into climate models. J. Appl. Meteor., 34, 745-756.
Hebble, E. E., T. N. Carlson and K. Daniel, 2001: Impervious surface area and residential housing density: a satellite perspective. Geocarto International, 16 13-18.