tude of an anticipated discharge deficit event (water
jams is clear, but in other cases it must be estimated.
placed in storage) based on the severity of the winter
However, the error is small and the ice data sheets are
prior to the current date, forecasted weather, and
extremely valuable in identifying the location of
(once again) an acceptable level of risk. Based on this
initiation points. The location of the upstream edge
estimated discharge deficit and current water levels
of the ice cover is open to interpretation in many
and flows throughout the system, the user could
cases, but very well defined in a few instances. Pro-
determine whether an increase in water release is
gression and regression rates of several ice events can
required, and if so, the amount.
be determined from the ice data sheets.
USGS gaging station records
Normally, discharge in an uncontrolled river in-
ANALYSIS OF ICE AND
creases in the downstream direction, provided con-
sumptive water use is not large in relation to river
To develop a predictive method for determin-
flow. Decreases in discharge in the downstream
ing the required winter releases at Gavins Point
direction are often a result of water going into stor-
Dam, it is necessary to identify ice events that are
age. If a significant amount of water goes into storage
associated with decreases in flow. Two sources of
as a result of ice formation and jamming, discharge
information are available in this case: visual obser-
decreases in the downstream direction so that re-
vations of ice cover presence and USGS gaging
corded discharge is lower at a downstream gaging
station discharge data.
station than at an upstream gaging station. Examina-
tion of stream gaging records during the winter
months can be useful in estimating the occurrence
MRD ice data sheets
Long-term visual ice observations are rare. Fortu-
and location of ice jams.
nately, Missouri River Division (MRD) personnel
To evaluate discharge data in this manner, the
have made careful observations of ice covers on the
Missouri River below Gavins Point Dam was broken
Missouri River. These are shown graphically on the
up into nine reaches (Fig. 2), divided by USGS gaging
MRD ice data sheets, which span from St. Louis, Mo.
stations (Table 1). The reaches are listed in Table 2,
(river mile 0) to Gavins Point Dam (RM 811) for the
along with the ice jam initiation locations indicated
period 1963 to present. A portion of an ice data sheet
on the MRD ice data sheets that occur within each
is shown in Figure 1. These sheets indicate the pres-
reach. The average daily discharge from November
ence or absence of an ice cover and, in some cases, the
through March was plotted for water years 1968
estimated percent coverage of the river by floating
through 1988 for the USGS gaging stations listed in
ice. In this study, an ice cover recorded on the sheets
Table 1. Examples of such plots for the 1968 water
was considered to represent the occurrence of an ice
year are given in Figure 3.
event capable of causing water to go into storage,
The most upstream reach (reach 1) extends from
thereby decreasing flow.
the USGS gage at Sioux City (RM 732.2) to the USGS
Starting and ending dates of ice events are gener-
gage at Yankton (RM 805.8), which is 5.2 miles
ally well defined. The initiation point of some ice
downstream from Gavins Point Dam. There are no
Table 1. Station information for discharge-based reaches.
Kansas City, Missouri
*Continuous daily records