Laserfiche WebLink
<br />16 <br /> <br />taken prior to 1914 by USGS personnel were matched in spring 1997 (Fig. 9): <br />The new saltcedar-covered floodplain is clearly visible, as is a lower elevation <br />willow-covered surface. Cottonwoods that once were near the channel are now <br />greater than 25 m away. It is not possible to determine the timing or rate of <br />bank accretion at the old cableway because no measurements were made at <br />the site between 1930 and 1997. <br />The rate and style of channel narrowing were measured at the present <br />cableway by analyzing changes in the bankfull channel width between 1930 <br />and 1993. Figure 10a summarizes these findings. Regression lines in Figure <br />10a were computed for time intervals selected by visual assessment of the <br />data. These groupings were analyzed for statistical significance using an <br />. analysis of covariance that accounts for the uneven distribution of data points <br />over time. Channel widths for the three time periods are significantly different <br />at the a = 0.15 level (p < 0.01). To ensure that the identified changes in <br />channel width were not a product of systematic variations in discharge at the <br />time of measurement, river discharge at the time of each measurement also <br />was analyzed using the same statistical technique. Discharge values for the <br />three time intervals are statistically equivalent; thus, the channel width changes <br />are real and are not an artifact of any measurement bias. Changes in mean <br />section velocity and mean depth for each of the three time periods are also <br />significant at the a = 0.15 level (p l::= 0.1). <br />