Modelled Stage Elevation Shoreline 878 m3s

Using widely-available software intended for modeling rivers, a new one-dimensional hydraulic model was developed for the Colorado River through Grand Canyon from Lees Ferry to Diamond Creek. Solving one-dimensional equations of energy and continuity, the model predicts stage for a known steady-state discharge at specific locations, or cross sections, along the river corridor. This model uses 2,680 cross sections built with high-resolution digital topography of ground locations away from the river flowing at a discharge of 227 m3/s; synthetic bathymetry was created for topography submerged below the 227 m3/s water surface. The synthetic bathymetry was created by adjusting the water depth at each cross section up or down until the model’s predicted water-surface elevation closely matched a known water surface. This approach is unorthodox and offers a technique to construct one-dimensional hydraulic models of bedrock-controlled rivers where bathymetric data have not been collected. An analysis of this modeling approach shows that while effective in enabling a useful model, the synthetic bathymetry can differ from the actual bathymetry. The known water-surface profile was measured using elevation data collected in 2000 and 2002, and the model can simulate discharges up to 5,900 m3/s. In addition to the hydraulic model, GIS-based techniques were used to estimate virtual shorelines and construct inundation maps.
The error of the hydraulic model in predicting stage is within 0.4 m for discharges less than 1,300 m3/s. Between 1,300-2,500 m3/s, the model accuracy is about 1.0 m, and for discharges between 2,500-5,900 m3/s, the model accuracy is on the order of 1.5 m.
In the absence of large floods on the flow-regulated Colorado River in Grand Canyon, the new hydraulic model and the accompanying inundation maps are a useful resource for researchers interested in water depths, shorelines, and stage-discharge curves for flows within the river corridor with 2002 topographic conditions.

本研究基于广泛可用的通用河流建模软件，针对科罗拉多大峡谷段（利斯费里至钻石溪区间）的科罗拉多河，开发了一款全新的一维水力学模型。该模型通过求解一维能量方程与连续方程，可沿河道走廊内的特定点位（即河道断面），针对已知稳态流量预测水位。该模型共包含2680个河道断面：河道裸露区域的地形采用高分辨率数字地形数据构建，该数据对应流量为227 m³/s时的裸露河床；而水位低于227 m³/s对应水面以下的淹没地形，则通过合成测深数据（synthetic bathymetry）生成。合成测深数据的生成方式为：逐一调整各断面的水深，直至模型预测的水面高程与实测水面高程高度吻合。该建模方法属于非常规手段，为缺乏测深数据的基岩控制型河道提供了一种一维水力学模型构建方案。对该建模方法的分析表明，尽管其可有效构建实用的水力学模型，但合成测深数据与实际测深数据可能存在偏差。研究所用的已知水面剖面基于2000年与2002年采集的高程数据测得，该模型可模拟的最大流量可达5900 m³/s。除水力学模型外，本研究还基于地理信息系统（GIS）技术，估算了虚拟岸线并构建了淹没范围图。
该水力学模型的水位预测误差表现如下：当流量低于1300 m³/s时，误差不超过0.4 m；当流量处于1300~2500 m³/s区间时，预测精度约为1.0 m；当流量处于2500~5900 m³/s区间时，预测精度约为1.5 m。
鉴于科罗拉多大峡谷段的科罗拉多河为受调控径流河道，并无大型洪水事件发生，因此这款全新的水力学模型及配套的淹没范围图，可为关注2002年地形条件下河道内水流的水深、岸线及水位-流量曲线的研究人员提供实用的参考资源。