Figure 6: LUCAS output maps of Little Tennessee River Basin.
LUCAS has recently been used [14] to assess the influence of land ownership on land-cover change in the Little Tennessee River Basin of western North Carolina. Transition probabilities for 100 year simulations of land-cover change have been generated [16] using historical data from two periods: 1975--1986, and 1986--1991. As discussed in Section 2, these probabilities can reflect transitions between discrete land-cover types with respect to certain socioeconomic and ecological variables. In this particular simulation, the land-cover types are
while the physical landscape attributes (independent variables) are
Transition probability matrices (TPM's) are then generated
for each time period and initial land-cover for both
public and private land ownership. Hence, 
TPM's are generated by the Socioeconomic Model Module (see Figure
1) in LUCAS. Figures 6(a) through 6(d)
show the LUCAS output maps (only one of five replicates shown) from 100-year
simulations (beginning
with the year 1991) of land-cover change for the Little
Tennessee River Basin. In these figures, green denotes forest land,
yellow denotes grassy/brushy cover, red denotes unvegetated
area, and blue denotes water. Scenarios of land-cover change
based on Historical Transition Probabilities (HTP's) [16] for
public and private lands are listed in Table 5.
Table 5: Scenarios of land-cover change for Little Tennessee
River Basin according to historical transition probabilities.
As discussed in [14], the importance of shifts in land-cover-change dynamics is illustrated by the differences between Figures 6(a) and 6(d). The greatest change in the landscape was observed when the empirically observed 1975-1986 transition probabilities, which were greater than the 1986-1991 transition probabilities, were applied across lands under both ownership types (Scenario 4, Figure 6(d)). Forest cover exhibited the greatest decline and fragmentation when transition probabilities for the period 1975-1986 are applied to both public and private lands. The least change in the landscape is observed when the 1986-1991 rates of transition were applied across lands under both ownership types (Scenario 1, Figure 6(a)). In this scenario, landscape patterns remained relatively stable through time, as indicated by the proportion of forest (see Table 7 in the Appendix for agreements in proportion of land cover between simulated and actual landscapes).