Laserfiche WebLink
<br />13 <br /> <br />made at or just below bankfull stage best describe the bankfull channel width;" <br />therefore, we analyzed all discharge measurements made at or slightly below <br />the 2-yr recurr~nce interval flood. Inclusion of measurements made at higher <br />discharges in the analysis introduces excessive variability due to large <br />differences in the width of floodplain inundation (Fig 6). <br /> <br />Aerial Photo Interpretation and Analysis <br />Stereo aerial photos taken in 1938, 1952, 1962, 1985, and 1993 (Table <br />2) were analyzed to characterize and quantify channel change in the entire <br />study reach. Map units are defined as areas having a unique combination of <br />four attributes. These attributes include: (1) major feature (flood plain. mid- <br />channel bar, secondary channel, etc.), (2) vegetation density (dense, sparse, <br />unvegetated, etc.), (3) part of the active channel (yes or no--based on breaks in <br />slope and vegetation borders--used to account for different discharges on dates <br />of photography), (4) geomorphic surface (current floodplain with <br />saltcedar/Russian olive, cottonwood terrace, inset willow level) (1993 only). <br />Digital photographic scaling techniques were used to transfer attributes <br />from photos to a common 1 :12,000 scale base map. These data were input into <br />a geographic information system for quantitative analysis. The area of different <br />map units was measured in each photo series and changes over time were <br />calculated. Errors were calculated as the width of the trace line times the scale <br />of the map, and that error was assumed on both sides of the channel. This <br />