Table 5. Discharge deficit record

Table 5. Discharge deficit record.

Max

Length

AFDD

Max

Length

AFDD

deficit

of period

at max

deficit

of period

at max

(1000 ft3/s)

(C)

Date

(days)

Date

(days)

in Figure 9. It can be seen that nearly half the

The discharge deficit record can be further ana-

maximum deficits were 1000 ft3/s or less. The

lyzed by grouping the recorded discharge deficits

annual maximum return interval can be found by

into half-month periods, as shown in Table 6. It can

selecting the maximum discharge deficit that oc-

be seen that impact periods do not occur every year

curred during each winter. The return interval is

during each half month. The most likely half

found by assigning each maximum discharge defi-

months are 1631 December (61%), 115 January

(55%) and 1631 January (61%). Again, the return

cit a probability, P, defined by J/(N+1), where J is

interval for each half-month period is found in the

the rank of the deficit and N is the total number of

data points available. The return interval is then 1/

same manner as the annual maximum return in-

(1P). The annual maximum return interval is

terval. The annual maximums occurring in each

shown in Figure 10. It can be seen that the distribu-

half-month period are shown in Figures 11 through

tion is nearly normally distributed, and that there

14. The periods 115 December and 1628 February

is one apparent outlier. This outlier occurred on 22

were not plotted because there were too few data

December 1983 and has a value of 11,500 ft3/s. This

points. We can see that, in large part, the data are

is by far the largest ice-impacted discharge deficit

normally distributed except for the period 1631

recorded during the period of record.

December. The data for this period seem to suggest

12