|
<br />The final step in the construction of the 1921-86
<br />continuous record of instantaneous discharge was quality
<br />control; daily mean discharges were computed /i'om this
<br />continuous time series of instantaneous discharge and
<br />compared with the daily mean discharges previously
<br />published by the USGS. All periods during which the
<br />computed daily mean discharges deviated from the
<br />published daily mean discharges by I percent or more for
<br />more than several days were examined for errors, During
<br />this process, the major digitizing errors were found and
<br />corrected, In addition, errors in the published daily mean
<br />discharge data were also identified. The most cornmon
<br />type of error found was a transcription or copying error,
<br />where either stage or discharge was incOlrectly written or
<br />copied either on the graph or in the final records (fig. 19).
<br />The record for days with substantial ice effects could not
<br />always be corrected because the published daily mean
<br />discharges on these days were actually based on the
<br />record from the Grand Canyon gaging station. Therefore,
<br />some disagreement still exists between the instantaneous
<br />discharges in the continuous record and the published
<br />daily mean discharges on days of extreme ice effect.
<br />Appendix E contains a complete listing of the causes of all
<br />disagreements greater than or equal to 5 percent between
<br />the daily mean discharges computed from the continuous
<br />record of instantaneous discharge and the published daily
<br />mean discharges.
<br />Average disagreement between the daily mean
<br />discharges computed from the continuous record of
<br />instantaneous discharge and the published daily mean
<br />discharges was found to be only +0.055 percent for the
<br />entire period from water year 1921 through water year
<br />1986 (fig. 20). In fig. 20, percent disagreement is defined
<br />as
<br />
<br />100(Qcomp- Qpub)
<br />Qpub
<br />where Qcomp = the daily mean discharge computed from
<br />the continuous record of instantaneous discharge and
<br />Qpub = the daily mean discharge previously published by
<br />the USGS. All days with disagreements greater than 5
<br />percent or less than -5 percent are either (1) days affected
<br />by ice, or (2) days in error in the published record.
<br />Disagreement increases slightly after the closure of Glen
<br />Canyon Dam because the increased variability in
<br />
<br />discharge within each day made it more difficult for the
<br />USGS to compute daily mean discharges prior to the
<br />availability of more advanced computers,
<br />
<br />The Continuous Record of Instantaneous
<br />Discharge for Water Years 1921-2000
<br />
<br />(1)
<br />
<br />Following the computation and quality control of
<br />the continuous record of instantaneous discharge for water
<br />years 1921-86, these data were combined with 30- and
<br />IS-minute stage and discharge data collected after
<br />October I, 1986, to form a continuous record of the
<br />instantaneous discharge of the Colorado River at Lees
<br />Ferry from May 8, 1921, through September 30, 2000
<br />(fig. 21A), Also included in fig. 21A is the peak discharge
<br />of the 1884 flood, the largest flood between] 884 and
<br />1921 at Lees Ferry. Stage and discharge of the Colorado
<br />River at Lees Ferry were recorded at discrete 30-minute
<br />intervals between October I, 1986, and May 3], 1998, and
<br />have been recorded at discrete IS-minute intervals since
<br />June I, 1998. These post-October I, 1986, data were
<br />previously and are currently available from the Arizona
<br />District of the USGS, The resolution of the May 8, ]921,
<br />through January ]8, ]923, part of the continuous record
<br />(the period before installation of the Lees Ferry Gage with
<br />its continuous stage recorder) is twice daily; the resolution
<br />of the January ]9, 1923, through September 30, ]986, part
<br />of the continuous record is approximately] 5-30 minutes.
<br />The resolution of the October I, ] 986, through May 3].
<br />1998, part of the continuous record is 30 minutes; the
<br />resolution of the June I, 1998, through September 30,
<br />2000, part of the continuous record is ]5 minutes. The
<br />entire record can be either requested from the USGS
<br />Grand Canyon Monitoring and Research Center,
<br />Flagstaff, Arizona, or obtained electronically at
<br />http://www,gcllU'c,gov.
<br />The daily range in the discharge of the
<br />Colorado River at Lees Feny between May 8, ] 921, and
<br />September 30, 2000, (fig, 21B) was computed from the
<br />data in fig, 2]A, The largest daily range in discharge of
<br />68,100 ft3/s occurred on September ]3,1927, during the
<br />rising limb of the September 13, 1927, 125,000 ft3/s flood.
<br />Most of the water in this flood originated within the San
<br />Juan River Basin.
<br />
<br />0223~
<br />
<br />Computation of the Continuous Record of Instantaneous Discharge at lees Ferry for Water Years 1921-2000 39
<br />
|