In designing this model, we chose to focus on
the potential of a crown-®re initiating from a surface
®re or of a ®re spreading through the canopy of a
stand, rather than on the effects of an actual ®re. At
the landscape scale, individual stand risk represents
only a portion of the ®re risk affecting each stand
(Turner and Romme, 1994). The effects of neighboring
stands and their ®re risk need to be incorporated.
Having calculated risk for each stand on the landscape,
we created another set of queries to address
the effect of neighboring stands. In particular, we
were interested in those stands that are upwind and/
or downslope from a focus stand because of the
implications for ®re movement across the landscape.
We designed queries to identify downslope neighbors
based on average elevation for each stand. A wind
direction variable was created allowing the user to
set the direction of the prevailing wind during severe
®re weather (e.g., N, NE, E, SE,...). A query then
identi®es the adjacent polygons that are upwind
from each stand. The risk from these adjacent upwind
and downslope stands is then calculated. The area of
stands in each risk category and their proportion
across the landscape are summed and displayed in a
graph as output. For this discussion, slope risk is