论文部分内容阅读
通过变化基于物理过程分布式参数流域水文模型的单元格大小(Grid size)和时间步长(Time step)研究流域水文过程的尺度依赖性,进而寻求最佳模拟尺度和模拟效果,可以为实现相对准确的流域土地利用变化/植被变化以及气候变异的水文响应预测,制定适应性流域管理措施提供科学基础。本文以基于物理过程的分布式水文模型MIKESHE为工具,以位于西北黄土高原甘肃省天水市吕二沟流域的实测次降水-径流为输入对模型进行校正后,采用多尺度检验的方法探讨分析了单元格及步长变化的水文影响。结果表明:单元格变化对峰值及模拟径流总量有影响。单元格从10m增大到200m,模拟峰值及径流总量均增大,分别变化为1.72～8.56m3/s和1.10～3.68×105m3,这主要源于单元格变化引起流域特征变化,10m和200m单元格对应模拟面积分别较原流域面积变化-0.09%和20.57%,河流链接“river link”分别较原沟系长变化27.23%和5.48%。此外,模型中运动波近似方程有限差分像元△x与单元格水流交换总量之间的关系一定程度也解释了单元格变化对模拟结果的影响。分析认为本研究60～100m为较适宜采用的模拟单元格。步长变化引起单位步长内降水强度的变化,从而影响峰值模拟,但对径流总量无影响。步长减小,模拟峰值增高,0.5h步长和24h步长对应模拟峰值分别为2.38m3/s和1.72m3/s。由于各时段观测记录分辨率不一,研究中步长的选择确定具有很大任意性,使得任一既定采用的模拟步长并不适用于另一观测时段,模拟效果不一。
By changing the grid size and time step of the hydrological model based on the distributed physical parameters of the physical process, the scale dependence of the hydrological process in the basin can be studied to find the best simulation scale and simulation effect, Accurate basin-based land use change / vegetation change and climate change hydrological response prediction, to develop adaptive watershed management measures to provide a scientific basis. In this paper, based on the MIKESHE, a distributed hydrological model based on physical process, the model was calibrated with the measured secondary precipitation-runoff in Lv Ergou catchment located in Tianshui City, Gansu Province, Northwest Loess Plateau. The multi-scale test was used to analyze the model Hydrological effects of changes in cells and steps. The results show that the changes of the cell have an impact on the peak and the total amount of simulated runoff. When the cell size increased from 10m to 200m, the simulated peak and total runoff both increased from 1.72 to 8.56m3 / s and from 1.10 to 3.68 × 105m3, respectively. This was mainly due to the change of cell characteristics caused by the cell changes, The corresponding simulated area of cells changed by -0.09% and 20.57% respectively compared with that of the original drainage area. The river link “river link” changed by 27.23% and 5.48% respectively compared with that of the original channel. In addition, the relationship between the finite difference pixel △ x and the total amount of water exchange in the model is also explained to some extent by the influence of the cell variation on the simulation results. Analysis shows that 60 ~ 100m in this study is the more suitable analog cell. The change of step length causes the variation of precipitation intensity in unit step, thus affecting the peak simulation, but has no effect on the total runoff. The step length decreases and the peak value of simulation increases. The corresponding peak values of 0.5 h step and 24 h step are 2.38 m 3 / s and 1.72 m 3 / s, respectively. Due to the different resolution of observation records at different time intervals, the selection of the step size in the study has a great arbitrariness, so that any given simulated step size is not applicable to another observation period and the simulation results are different.