Complexity of confluence morphology is not dictated solely by considerations
of alluvial erosion and deposition. The morphology of many confluences can largely
be defined by the rock formations through which the confluent rivers flow. Such
confluences can have quite unique features that lead to ice jamming. For example,
the extensive set of studies concerning ice movement through the upper Niagara
River provide insights into the merging of moving ice in a channel of complex
morphology. The river's Chippawa and Tonawanda Channels merge in a wide
confluence where a large area of shallow flow occurs over a resistant rock bed.
This regularly causes drifting ice to be grounded and form a blockage in the con-
fluence that may extend upstream into the confluent channels (Crissman 1998).
Several further articles describe ice jams formed where a river enters a reservoir
or a lake. Judge et al. (1997), for instance, describe the formation of ice jams in the
headwaters of shallow reservoirs (head ponds) used for hydropower generation
in New Brunswick. They base their description on a numerical model of jam for-
mation and on observations of ice jams formed in the headwaters of hydropower
reservoirs along the Saint John River, New Brunswick. Ice jams also form at the
mouth of the Winnipeg River where it enters Lake Winnipeg in Manitoba*.
Wuebben et al. (1995) provide background information on ice jams that form where
the Aroostook River enters Tinker Reservoir, Maine. Beltaos and Burrell (1991)
describe an instance where a backwater effect from a confluent river downstream
contributes to a recurring jam in a mainstem river. The Restigouche River, New
Brunswick, experiences ice jamming below the river's confluence with the steeper-
sloped Upsalquitch River, which conveys substantial amounts of ice and water
into the Restigouche River. The jam occurs in a reach affected by a backwater ef-
fect attributable to flow from a lower tributary, the Matapedia River.
The most extensive overview of ice jam formation at confluences is the survey
of confluence jam sites reported by Tuthill and Mamone (1997). They cite 44 con-
fluence sites in the United States known to have ice jam problems, and classify the
sites into four groups of confluence type:
Confluences of similar size rivers or channels.
Confluences of different size rivers or channels.
Rivers entering lakes.
Lakes drained by rivers.
These groups are useful for establishing the relative incidence of ice jams. Sub-
sequently an alternative grouping of confluence types is suggested in this report.
The alternative grouping uses terminology coined by fluvial morphologists, and
it simplifies the discussion of ice-jamming mechanisms at confluences.
Of the 44 sites, Tuthill and Mamone identify 15 as involving a river entering a
lake or reservoir (e.g., Aroostook River entering Tinker Dam Reservoir). Of the
remaining confluence sites, several sites involve two rivers in which one river is
subject to a significant backwater condition. One site is the confluence of the Yel-
lowstone River and the Missouri River upstream of Lake Sakakawea. The second
site is the Salmon River merging with the Connecticut River at a reach subject to
tidal slowing of river flow. An objective of the survey was to identify sites for
which some form of structural remediation could be used to mitigate jam forma-
tion. They suggest that eight sites are potentially suitable for this purpose. Of them,
six involve a river entering a lake or a reach of flow slowed by a substantial back-
* R. Carson, ACRES International, Winnipeg, Manitoba, personal communication, 1998.
7
Back to contents page
Previous Page
