method (also used by Betts et al., 2003 and Schiefer and Gilbert, 2007)
stable control areas (CAs) were identified outside the landslide body
(stratified selection) where 16 CAs (i.e. ~1 600 points, corresponding
to a density of ~100 points per area) were selected randomly in these
stable areas. For the “1977–1969” differential DEM, the number of CAs
was lowered to 14 (~1 400 points) due to its reduced area. The
number of CAs (i.e. points) required was determined from the desired
computation accuracy of the statistical parameters (i.e. 1–2 cm in this
case). Then ME, SD, and RMSE from the “false” elevation differences at
these points were computed. The obtained results are only slightly
different from those described above (see Tables 3 and 4). However,
the ME values of −0.1 and +0.1 m, respectively, calculated for the
“1998–1988” and “1988–1977” differential DEMs have been found to
be significant by the t-test (p