Laserfiche WebLink
Review Findings — Modeled Zone A Streams <br />WSE reviewed each HEC -RAS model and proposed delineation focusing on areas where there <br />is either existing development or future development likely. WSE did not examine areas where <br />the floodplains are confined by natural topographic relief and therefore remain narrow and have <br />little potential for future development. General findings common to many of the modeled Zone <br />A delineations are described below. Eight sites have specific issues worthy of a more detailed <br />description. These are presented in Figures 1 to B. <br />General Comments <br />Hydrology — StreamStats appears to have been used to determine the magnitude of the <br />100 -year flood for each stream. While Streem5tats may be a useful and simple tool to <br />estimate flows for ungauged sates, it is often inaccurate within Kittitas County. WSE has <br />found that annual instantaneous peak discharges obtained from StreamStats for streams <br />In this area are frequently under -predicted. This is based on our personal history of <br />conducting hydraulic investigations within the County. In one recent case it under- <br />predicted the 100 -year event by approximately 300 percent. The problem occurs <br />because the regression equations used in StreemStats are based upon a relatively small <br />annual instantaneous peak flow dataset obtained from a wide range of gage sites -- from <br />large rivers Including the Yakima and Teanaway to very small streams. <br />An example of the problem described above occurs within subreach 4 of the Yakima <br />River. The proposed modeled Zone A delineation for this reach is based upon a <br />StreamStats 100 -year discharge estimate of 20,700 cls. WSE recently completed a <br />detailed hydrologic investigation for a reach of the Yakima River immediately upstream <br />from subreach 4 and the resulting 100 -year discharge was 35,300 cfs, a value <br />approximately 70% larger than the StreamStats estimate (WSE, 2015). Based upon our <br />study, 20,700 ds is approximately a 10 -year event, not a 100 -year flood (WSE, 2015). <br />In a separate issue, there is a specific problem with the 100 -year discharge estimate for <br />Quilomene Creek obtained from StreamStats. The drainage basin area used to estimate <br />the flow from the program includes the watershed of Brushy Creek, an area which does <br />not contribute flow to Quiiomene Creek. <br />2. Cross Section Orientation — Cross sections must be oriented perpendicular to the <br />direction of flow in a 1 D model. In many models, the ends of some cross sections are <br />bent so sharply up -valley that these portions of the cross sections are not perpendicular <br />to flow. In these cases the cross sections erroneously tie into high ground creating a <br />fictitious narrow flow corridor and floodplain. Examples of this problem are specifically <br />ldentlfled in Figures 1, 2, 3, 4, 5, and 7. <br />3. Main Channel Location — The main channel is incorrectly located along a number of <br />reaches and therefore, the floodplain delineations are incorrect. Obvious mistakes <br />Include reaches of Big Creek, Crystal Creek, and Taneum Creek (Figures 1, 2, and 6). It <br />can be difficult to locate the main channel of a small stream using topography or aerial <br />photography. When uncertainty exists, the location of the main channel should be field <br />verified. <br />2 <br />