Laserfiche WebLink
<br />The standard error of estimate, correlation-coefficient, and number of <br /> <br />sites included in each equation (for the 100-year recurrence interval) are <br /> <br />shown in Table 3. For the equations using total drainage area or drainage <br /> <br />area above an elevation level, the standard error of estimate is very <br /> <br />large, and the correlation coefficient is relatively small, indicating <br /> <br />poor relations. The poor relation between the 100-year rainfall flood and <br /> <br />the total drainage area for sites in the South Platte River Basin is shown <br /> <br />in Figure 8a. Regression using drainages above a specified elevation level <br /> <br />are not significant. For the drainage area below a given elevation level, <br /> <br />the standard error of estimate is large until the 8,000-foot (2,438-meter) <br /> <br />level where the standard error of estimate decreases. Similarly, the <br /> <br />correlation coefficient is maximum at this elevation level; therefore, the <br /> <br />drainage area below 8,000 feet (2,438 meters) was selected as the best area <br /> <br />to use to estimate the rainfall flood characteristics in this region. The <br /> <br />improved relation for the 100-year rainfall flood and the drainage area <br /> <br />below 8,000 feet (2,438 meters) for the South Platte River Basin is shown <br /> <br />in Figure 8b. The standard error of estimate improved from 142 to 44 <br /> <br />percent by using the drainage area below 8,000 feet (2,438 meters) rather <br /> <br />than total drainage area in the 100-year regression model. The standard <br /> <br />error of estimate was 207 percent for all 27 stations for the total drain- <br /> <br />age area in the 100-year regression model. An elevation of 7,500 feet <br /> <br />(2,300 meters) may improve the regression results slightly; however, the <br /> <br />1:250,000-scale topographic maps used do not have this contour line so <br /> <br />difficult interpolation would have to be done. <br /> <br />//~ <br /> <br />~ .~""---- <br />