edly used by Korbaylo and Shumilak (1999) in
the calculated equilibrium ice jam thickness. They
modeling the 1986 Churchill River breakup jam.
obtained 46 values of ni from calibrated numeri-
Tuthill and Mamone (1998) used a lower value, ni
cal models of three breakup ice jams. These ranged
= 0.030, when modeling freezeup jams on the
from very smooth (0.020) to very rough (0.150),
middle Mississippi River.
with a mean of 0.066 and standard deviation of
0.023 (Figure 8a). The distribution was found to
Uncertainty in ice roughness
be nearly normal. A Monte-Carlo simulation of
White and Daly (1997) examined the effects of
sample size 10,000, based on the mean of 0.066 and
uncertainties in the estimate of ice roughness on
standard deviation of 0.023, is shown in Figure 8b.
7
6
5
4
3
2
1
0
Breakup Jam "n" Values
a. Distribution of Manning's n values for the underside of the ice from
calibrated numerical models of breakup ice jams at three sites.
1000
800
600
400
200
0
Simulated Breakup Jam "n" Values
b. Distribution of simulated Manning's n values. (From White and Daly 1997.)
Figure 8. Distribution of actual and simulated Manning's n values.
12
Previous Page
