|
<br />The mild reversal in curvature of the 1921-25
<br />stage-discharge data at the Number 1 Gage occurs at a
<br />stage of 20 ft on this gage. The linear regression fit to the
<br />data above this stage has an R 2 value of 0.982; the
<br />power-law regression fit to the data above this stage
<br />has an R2 value of 0.983, As at the other gages, the
<br />similar goodness-of-fit values for the linear and power-
<br />law regressions suggest that the upper end of the stage-
<br />discharge rating-curve at the Number I Gage can be
<br />approximated either as a line or as a power-law function,
<br />Linear extrapolation of the data suggest that the peak
<br />discharge of the 1921 flood was about 187,000 fi3/s, and
<br />power-law extrapolation of the data suggest that the peak
<br />discharge of the 1921 flood was about 174,000 fi3/s
<br />(fig.12D).
<br />Linear and power-law extrapolations of the stage-
<br />discharge rating curves at these four gages by means of
<br />best-fit regressions lead to an estimated peak discharge of
<br />the June 1921 flood that ranges between 143,.000 and
<br />187,000 ft3/s, This discharge range is somewhat wider
<br />than the 163,000 to 188,000 fi3/s range that was based
<br />on the data trom the Grand Canyon gaging station and
<br />other upstream and downstream gaging stations described
<br />above, but the upper ends of the ranges are nearly equal.
<br />Based on all of the data from the Lees Ferry and other
<br />gaging stations, the best estimate of the peak discharge
<br />of the June 1921 flood at Lees Fen'y is 170,000 fi3/s, at
<br />two significant figures (fig, 13). Given the range in the
<br />estimates of the peak discharge at the Lees Ferry and
<br />other gaging stations, the uncertainty in this value is
<br />approximately 20,000 fi3/s. Our estimate of 170,000:t
<br />20,000 fi3/s is therefore indistinguishable from both the
<br />original USGS unpublished estimate of 174,000 ft3 Is
<br />and the original USGS published value of "about
<br />190,000 fi3/s," This estimate of 170,000:t20,000 fi3/s is
<br />incompatible, however, with the 1939 upward revision of
<br />220,000 fi3/s based on Gatewood and Hunter's 1938
<br />study,
<br />
<br />Estimate of the Peak Discharge of the
<br />1884 Flood at Lees Ferry
<br />
<br />Estimation of the peak discharge of the 1884
<br />flood depends on accurate detennination of maximum
<br />flood stage and appropriate extrapolation of stage-
<br />discharge rating curves to higher elevations. The
<br />original unpublished estimate of the peak discharge
<br />of this flood at Lees FelTY was between 210,000 and
<br />
<br />250,000 ft3/s (G,c. Stevens, unpublished U.S. Geological
<br />Survey memorandum, May 28, 1925), and this estimate
<br />was revised by Gatewood and Hunter in 1938 to
<br />300,000 fi3/s, They made this revision because they had
<br />revised the magnitude of the 192] flood upward; tills
<br />revision resulted in a change in the stage-discharge rating
<br />curve which, in tum, necessitated revision of the 1884
<br />estimate. Reevaluation of stage-discharge data indicate
<br />that the probable range of the peak discharge of this flood
<br />was between 199,000 and 228,000 fi3/s.
<br />
<br />The Original U,S. Geological Survey Estimate otthe
<br />Peak Oischarge of the 1884 Flood
<br />
<br />The original USGS estimate of the peak discharge
<br />of the July 1884 flood was based on extrapolation of
<br />the stage-discharge rating eUrve at the Lees Ferry
<br />Gage, and on extrapolation of crude stage-discharge
<br />relationships developed for the Colorado River at three
<br />sites downstream in Grand Canyon (G,c. Stevens,
<br />unpublished U,S, Geological Survey memorandum,
<br />May 28,1925). These values ranged from 210,000 ft3/s
<br />at River-mile 233,7 (fig, 1A) to 250,000 fi3/s at Lees
<br />Ferry .
<br />For example, Stevens calculated that the
<br />discharge of the July 1884 flood at River-mile 233.7 was
<br />approximately 210,000 ft3/s, based on data collected on
<br />October 8, 1923, during the USGS Birdseye Expedition
<br />(Westwood, 1992). Stevens wrote;
<br />During the survey of the Colorado River in 1923, an
<br />exceptionally high flood mark consisting of a pile of
<br />old drift was noted at Mile 233.7, The following
<br />information has been furnished by Mr, Herman Stabler;
<br />High-water mark, date unknown - - - - _. - - 1298.4'
<br />High-water mark, Sept, 20-23, 1923 - - - - -. 1275, l'
<br />Water level, October8, 1923
<br />(date of survey) - - -- -- - - -- - - - - -- - 1244,9'
<br />The flow atthe Bright Angel station [the Grand
<br />Canyon gaging station] a few days before October 8,1923,
<br />was 13,000-14,000 second-feet [ftl/s]. The flow during
<br />the September flood was 112,000 second-feet. [This flood
<br />largely originated in the Little Colorado River,] Using
<br />these two points and extending as a straight line to
<br />elevation ]298.4' gives 190,000 second-feet. By a curved
<br />extension, from 210,000 to 220,000 may be obtained. Accept
<br />210,000 second-feet forthe present. If the crest of the
<br />September, 1923, flood at Mile 233.7 was lower in
<br />discharge than at Bright Angel [i,e., the Grand Canyon
<br />gaging station] the extension would give a lower figure for
<br />the high-water mark.
<br />
<br />02237
<br />
<br />,
<br />Colorado River Floods at Lees Ferry 29
<br />
|