Laserfiche WebLink
<br />001454 <br /> <br />Thus,.even though the station control on a flat-gradient stream <br />remains stable, the stage-discharge relationship is not constant so <br />long as discharge changes rapidly. For any given stage, the discharge <br />is greater if stage is rising, and less if stage is falling. For those <br />gaging stations whose ratings vary in this way during major stream rises," <br />the standard practice of the Geological Survey is to derive daily mean <br />discharge from an average or steady-state rating curve. Published values <br />may be in ",rror for individual days but compensate in large part within <br />the term of the stream rise. Adjustments should be applied with caution. <br />Magnitude of the error can be appraised by a routing comparison with <br />records fI'Qll1 stations up- and down-stream. <br /> <br />At ce~tain gaging stations--especially on the larger streams and in <br />respect to: the earlier records--the gage and measuring section, or the <br />high-water and low-water measuring sections, have been substantial distances <br />apart. In this situation, channel-storage effects such as just described <br />may prevail much of the time that stage is rising or falling, even slowly'. <br />Thus, even.though the stage-discharge relationship were stable at the <br />control, the record of flow might be in error if the station rating were <br />based on measurements made largely on a rising stage, or largely on a <br />falling stage. <br /> <br />Flood flows <br /> <br />Heavy weights and mechanical reels for measuring equipment, necessary <br />for measuring discharge accurately under flood conditions, did not cOllie <br />into gener~ use until about 1935. Earlier flood measurements, and ratingS <br />based on tnem, .are suspect accordingly. Some hydrologists contend that <br />early flood-flow ratings were systematically too large; others, that they <br />were systematically too small. Probably neither contention can be proved <br />to apply universally, so that it seems prudent to consider the errors as <br />random in algebraic sign. Adjustment of early flood flows is not widely <br />practicable, although more reliable ratings are now available for certain <br />relatively stable channels. In hydrologic studies the questioned flow <br />values logically can be given less statistical weight. In flood-frequency <br />studies, it would be wise to identify the early values so that any groupihg <br />at an extreme of the frequency array would be obvious. <br /> <br />.' <br /> <br />I.3 <br /> <br />