Geological Survey Professional Paper 1547 Sedimentology, Behavior, and Hazards of Debris Flows at Mount Rainier Washington

HISTORICAL FLOODS COMPARED WITH DEBRIS FLOWS

Glacial-outburst debris flows and some smaller examples of both cohesive and noncohesive lahars are all likely to be less destructive than some historical floods have been. The largest floods of record were caused by intense precipitation on snow during prolonged warm periods, and they are described for comparison with the smallest, most frequent volcanic flows. Historical floods in the Nisqually River have been analyzed by Nelson (1987); peak annual discharges in the Puyallup River at Puyallup from 1915 to 1986 were compiled by Prych (1987, table 2).

Probably the largest post-settlement flood occurred in 1867 (described by Summers, 1978, p. 235). In early December, much of a heavy snowpack on Mount Rainier was melted by four days of warm rain, causing a major flood in at least the Cowlitz River system. The city of Monticello at the mouth of the Cowlitz was completely destroyed on December 17. Upstream flooding was not reported, because settlement there had not begun. However, several other large floods between 1886 and 1911 probably inundated the entire valley bottom of the upper Cowlitz River. The valley downstream from Packwood, which is as much as 3 km wide, was apparently inundated in March 1907 and possibly again in 1909 (Packwood History Committee, 1954; Superintendent's Reports, Mount Rainier National Park, 1907 and 1910; U.S. Geological Survey stream-gaging records). Subsequent high flows occurred on the Cowlitz River at Packwood in 1933, 1959, and 1977, but cannot be directly compared. Inundation of the Cowlitz valley to a depth of approximately 2 m is described in several undated early accounts. The Nisqually River drainage was also flooded early in 1910, when flood waters from a drainage to the south overflowed into that river (Bretz, 1913, p. 27). In general, historical flood inundation has been similar in depth to that by the most recent lahar-runout flows (table 4).

¹Numbers used for comparison with table 6.
²Transformation types:
    A, Direct, progressive transformation of wave front to hyperconcentrated flow.
    B, Deposition of successive flow fronts; bypassed by dilute tail of hyperconcentrated flow.
    C. Dewatering of course deposits to yield secondary debris flow.
³Risk is much lower than in the three described planning cases.
⁴Site-specific hazard-zone mapping based on techniques described in the text.

Historical flood data for the Nisqually River near and downstream from Longmire (Nelson, 1987) show that floods having recurrence intervals of 25 to 500 yr generally have smaller areas of inundation is than do lahars with similar recurrence intervals. This is especially true as recurrence intervals reach and exceed 100 years, because the more frequent volcanic and glacial-outburst debris flows attenuate rapidly at the base of the volcano (fig. 13), whereas rainfall floods amplify downstream with increased tributary inflow. A 500-yr flood will locally affect flood plains outside the active channel (Nelson, 1987, pls. 1 and 2), whereas a 500-yr volcaniclastic flow, like the National or even the Tahoma Lahar, could be catastrophic at a location like Longmire.

While dating the younger noncohesive lahars and their runout phases, we also dated some flood deposits and groups of such deposits (table 4). Some were probably local; others were the deposits of floods affecting all drainages of the mountain. Yet others may have been the distal flood waves evolved from upstream lahar-mount flows. In assessing risk, no presumption of a debris flow is made from fluvial sediment unless a direct correlation is possible. The distal streamflow deposits of the National Lahar (fig. 11C) are an example of such a correlation.