Laserfiche WebLink
<br />001455 <br /> <br />Some specific sources of error in flood flows are worthy of note. <br />Flood peak~ C9Jlll1l0nly exceed critical velocity and :flow is non:"lroninar; <br />in this sit.uation, :flow deteI'1llinations by slope-area procedures would <br />be too large. During the higher floods, water may have occupied one or <br />more channels that by-pass the measuring section. Especially in a flood <br />whose stag~ exceeded that of antecedent measurements, by-pass :flow may <br />have escap+d notice, been neither measured nor estimated, and been <br />omitted frbm station rating and published flow record. Subsequent <br />measurements may have recognized and included the by-pass :flow and <br />established a basis for adjusting the antecedent record. Or, post-flood <br />surveys may disclose the by-pass channels, determine their cross-section, <br />and establish an approximate auxiliary rating by slope-area methods. <br />Some such channels that were active years ago have been closed by dikes. <br />Others have been enlarged in capacity, or new channels have been opened <br />for plU"poses of :flood relief. In aIV of these circumstances, intensive <br />field surveYs and historical investigation may be required before early <br />:flow recor,;ls can be adjusted appropriately. $0 far as such surveys and <br />investigations by the Geological Survey have disclosed competent infor- <br />mation, rerlsed determinations of flood :flows have been published in <br />Water-8upp:j.y Papers. <br /> <br />For aEy flood that exceeds the previously established station rating, <br />the record* of flow or the computed volume involve obvious uncertainties.' <br />Adjustment'may not be practicable but the records can be tested for <br />consistency by "routing", correlation with data on rainfall, or post- <br />flood field surveys to establish slope-area parameters. The record for a <br />single station can be tested by separating the data for the cross_section' <br />into area llnd' velocity components. If it is known that stream..bed profile <br />remained stable but the computed velocity component is so large as to imply <br />erosion orbed materials, the values of :flow may be too large. On the otl1.er <br />hand, if t~e stream bed 'scoured, the values of :flow may be too smalJ.. <br /> <br />Scour and fill of stream beds during floods is especially severe if <br />the stream; gradient is steep and the floods are flashy. Alinement of the <br />thalweg may change greatly. Instances are known of scour during the <br />initial flt;lOd rise, then fill, then scour on the waning phase of the flood <br />and long into a subsequent period of ordinary flow. For stations on such' <br />streams, t~e pre-flood and post-flood ratings may differ little or much. <br />Published telues of flow may be substantially in error for certain days, <br />and may or;may not compensate within the flood record as a whole.. Co_only <br />there is n~ basis for rational adjustment. <br /> <br /> <br />1.4 <br />