Floods in small mountainous watersheds cover a wide spectrum of flow. They can range from clear water flows and hyperconcentrated flows to debris floods and debris flows, and calculation of the peak discharge is crucial for predicting and mitigating such hazards. To determine the optimal approach for discharge estimation, this study compared water flow monitoring hydrographs to investigate the performance of five hydrological models that incorporate different runoff yields and influx calculation methods. Two of the models performed well in simulating the peak discharge, peak time, and total flow volume of the water flood. The ratio (γ) of the monitored debris flood discharge (Q_d) to the simulated water flow discharge (Q_w) was investigated. Qualitatively, γ initially increased with Q_w but then decreased when Q_w exceeded a certain threshold, which corresponded to rainfall of 95 and 120 mm in a 6‐ and 24‐h event with a normal distribution of precipitation, respectively. The decrease might be attributable to a threshold of sediment availability being reached, beyond which increased flow rate is not matched by increased sediment input in the large watershed. Uncertainty of hydrological calculation was evaluated by dividing the catchment into sub‐basins and adopting different rainfall time steps as input. The efficiency of using a distributed simulation exhibited marginal improvement potential compared with a lumped simulation. Conversely, the rainfall time step input significantly affected the simulation results by delaying the peak time and decreasing the peak discharge. This research demonstrates the applicability of a discharge estimation method that combines a hydrological water flow simulation and an estimation of γ. The results were verified on the basis of monitored flow densities and videos obtained in two watersheds with areas of 2.34 and 32.4 km².

中文翻译：

---

估算小流域的水流量和泥石流排放量

山区流域的洪水泛滥。它们的范围从清澈的水流和高浓度的水流到泥石流和泥石流，计算峰值排放量对于预测和减轻此类危害至关重要。为了确定最佳的估算排放量的方法，本研究比较了水流监测水位图，以研究结合了不同径流量和流量计算方法的五个水文模型的性能。其中两个模型在模拟洪峰流量，洪峰时间和总流量方面表现出色。监测的泥石流排放量（Q _d）与模拟水流排放量（Q _w）之比（γ））进行了调查。定性地，γ最初随着Q _w增大，但是当Q _w时减小。超过某个阈值，分别对应于6小时和24小时事件中95 mm和120 mm的降雨，且降水呈正态分布。减少可能归因于达到沉积物可利用量的阈值，超过该阈值，流量增加与大型流域中增加的沉积物输入不匹配。通过将流域划分为子流域并采用不同的降雨时间步长作为输入来评估水文计算的不确定性。与集总仿真相比，使用分布式仿真的效率显示出了边际改进潜力。相反，降雨时间步长输入会延迟峰值时间并减少峰值流量，从而极大地影响了仿真结果。γ。根据监测到的流量密度和在面积分别为2.34和32.4 km ^2的两个流域中获得的视频对结果进行了验证。