km2 for the 1991–1992, 1964, and pre-1964 flows, respectively
(Table 3), assuming that the pre-1964 lava could have
resulted from multiple eruptions and could also have been
partially covered by the 1964 flow. Unfortunately, we do not
have any image data acquired before 1964 to make any
further investigation.
After all three lava flows had been defined, a threedimensional
(3D) view of the area was developed to verify
the results. The 3D view was created using ArcScene, a
module of the ArcGis software (ESRI, 2003). The ArcScene
module contains a ‘‘steepest path’’ function, which allowed
us to visualize how well the lava flows from the 1991–1992
eruption follow the topography of Unimak Island as defined
by the NED DEM. The steepest path tool calculates the
direction an object would take if released at a given point on
the surface. The object will take the steepest path downhill
until it reaches the perimeter of the surface model or it
reaches a local topographic minimum. Despite the fact that
the DEM was produced in 1970–1980s and is accurate to
just 60 m horizontally and 15 m vertically, the ‘‘steepest
path’’ confirms that the mapped lava flows conform remarkably
well to the NED topography (Fig. 2).