7628 - Laserfiche WebLink

Home Browse Search

'- ,.~ The aver2ge size of all backwaters ranged from 285 m2 at 5,260 cfs to 527 m at 1,101 cfs. Linear Regression Linear regressions of flow vs. area were analyzed for flows from 1,101 to 2,42j cfs for the Island Park, Jensen, and Ouray sites. The 5,260 cfs flow was deleted because it was far outside the range of 1,101 to 2,423 cfs and is not considered a realistic flow scenario for the summer season. The 1,430 cfs flow was also deleted at the Sand Wash site only, as this flow had not reached the site when the photography was acquired. Regression of backwater flow vs. area indicates two different relationships occurred. Significant, negative correlations occurred at the Island Park and Jensen sites (P = 0.05). Figures 4 and 5 show the regression lines at the Island Park (r = -0.87) and Jensen (r = -0.91) sites, respectively. These regressions indicate that as flow decreased backwater area increased. Figures 6 and 7 show the regression lines at the Ouray (r = +0.29) and Sand Wash (r = -0.39) sites, respectively. There was no correlation between flow and area at these two sites. For all four sites combined (with the 5,260 and 1,430 cfs flows deleted), a significant (P = .05) negative correlation existed between flow and backwater area. Figure 8 shows the regression line for the combined upper site totals (r = -0.91). Comparability of relationships within the four upper sites was analyzed. The hypothesis that relationships were equal was rejected (P = 0.10), which is sufficient to question that the correlations were equal. Furthermore, when the Island Park and Jensen sites were paired and the Ouray and Sand Wash sites were paired, it is apparent that the correlations between flow and backwater area were more alike than expected by chance. The pooled correlation for the Island Park and Jensen pair was -0.88 (P <0.01), and +0.17 (P >0.05) for the Ouray and Sand Wash pair. The statistical difference between the two pairs was highly significant (P <0.02), indicating the relationship between flow and backwater area for the Island Park and Jensen pair was statistically different from the Ouray and Sand Wash pair. Consequently, backwater area can reliably be predicted from flow (between 2,423 and 1,101 cfs) for the Island Park and Jensen sites, but cannot be predicted for Ouray and Sand Wash. Regression analyses of flow vs. backwater number indicate there was no significant relationship at any site (P = .05). However, at a lower significance level (P = .10), both the Island Park site and the combined four upper sites had significant correlation coefficients.