Laserfiche WebLink
<br />001243 <br /> <br />-3- <br /> <br />Special emphasis is warranted in regard to adequacy of sampling for <br />records of chemical quality. Inherently, a single analysis shows only <br />the chemical constituents at a single point within the body of water <br />sampled.' Also, unless it represents a composite of samples taken over a <br />period of time, it shows only the constituents momentarily at that <br />sampling point. All too infrequently is a water body sampled at enough <br />points or at enough times to represent all the potential variability <br />wi thin that body. All too rew ohemioal-qual:!. ty reoorde derine point and <br />'time of sampling in relation to features of the envirornnent. Thus, an <br />incautious hydrologist easily can be misled into undefendable chemical <br />and geochemical conolusions. The sole defense is to "screen" all data <br />for consistency, search out a fUll explanation of any seemingly abnormal <br />data, and draw no conclusion by extrapolation. <br /> <br />The difficulties just outlined are compounded in regard to ground- <br />water quality. Here, available data generally are' for samples withdraw <br />from wells that may range widely in depth and that, when pumped, may <br />yield a blend of two or more waters of unlike chemical character, from <br />distinct aquifers. The possible ramifications go far beyond the scope <br />of this manual. Suffice it to point out that (l) a full explanation of <br />all variations in chemical quality may require comprehensive infOI'lllation <br />on depth and casing records of wells, types and settings of pumps, and <br />regimen of withdrawals for some indeterminate period prior to sampling) <br />but (2) rarely is such ancillary infoI'lllation reported as part of the <br />chemical-quality record. <br /> <br />BIAS IN Rl!X)ORDS OF LONG TERM <br /> <br />Some hydrologic records have been published over a long term of <br />years under a single station name when in actuality the station has been <br />relocated one or more times during the term. In this situation, it is <br />unwise to assume that the "same water" has been measured, and that the <br />composite record is not "biased" by enviromnental differences among the <br />several locations. For streamflow,records in Water-Supply Papers of the <br />Geological Survey, the latest station description should be scanned to <br />identity possible relevant changes in location. All such records through <br />October 1950 have been compiled and summarized in Water-Supply Papers <br />l30l through 1319 and l372; these reports include the history of gage <br />changes, if aqy, at stations discontinued prior to the cut-off date of <br />the compilation. For climatologic stations, changes in location are <br />listed in the Weather Bureau's "Substation Histories" (Key to Meteoro- <br />logical Records Documentation No. l.l). Each change of location should <br />be treated as though a new 'and independent record had been started, <br />unless the several partial records are shown to be consistent one with <br />another, by double-mass plotting against records for one or more adjacent <br />statiol1S that are in the same environment and that have not been relocated. <br /> <br />