Laserfiche WebLink
<br />2 <br /> <br />Basin characteristics were investigated to learn how they may explain <br />the perviously described variations in seasonal and in chance distribution <br />of monthly base flow. By regression analyses on a digital computer, 27 <br />possible basin characteristics were investigated in over 1,000 different <br />combinations to develop equations defining base flow. Six characteristics <br />were found significant at the 9S-percent confidence level, or higher. The <br />equations have the following form: <br /> <br />log Q = a + bl log L + b2 log W + b3 log qm + b4 log Rsw <br /> <br />+ bS v + b6 Rlm <br /> <br />where Q is base flow for annual or seasonal periods or for 20-, 50-, and 80- <br />percent probability in any month; a and b are coefficients whose values are <br />tabulated for 39 different categories of Q; L is stream length; W is basin <br />width; qm is mapped mean total streamflow; Rsw is mapped ratio of sum- <br />mer to winter base flow; v is mapped variability index; and Rlm is mapped <br />ratio of 1 percent flow to mean flow. Residual errors of combinations of <br />this general equation were mapped by digital computer and show no re- <br />gional trends. The average standard error of estimate of mean annual base <br />flow was 34 percent. <br /> <br />Although the foregoing equation will provide preliminary estimates <br />of base flow at any site, tests at 16 gaging stations indicated that if time is <br />available for the collection of field data, there is a 62-percent chance that <br />base flow may be estimated more closely at an ungaged site by correlating <br />a selection of base-flow measurements at the site with concurrent base <br />flow at an established nearby gaging station. <br />