【作者】 包红军；

【导师】 李致家；

【作者基本信息】 河海大学 ， 水文学及水资源， 2006， 硕士

【摘要】 沂沭泗流域位于淮河流域东北部,属于半干旱地区。流域内水系复杂,地面起伏大,主要河道源短流急,洪水峰高量大、陡涨陡落,洪水预见期短:现状工程行洪和蓄洪能力较低。流域目前防汛决策系统薄弱,水文站点少,严重影响了洪水预报的精度和调度的准确性、及时性。因此,做好沂沭泗流域的洪水预报调度工作对防洪减灾起着重要的作用。 本文采用基于子流域的三水源新安江模型对沂沭泗流域进行洪水预报。对整个流域进行子流域划分,在每个子流域内分别使用新安江三水源模型进行计算,然后再运用马斯京根法进行汇流演算至出口断面。为了提高预报作业的精度,本文选用了衰减记忆最小二乘法对主要控制站进行实时校正。 沂河临沂以上流域地形起伏较大,水系复杂。临沂站洪水对整个沂沭泗流域的洪水过程有着关键性的决定作用。本文应用Grid-新安江模型、TOPMODEL模型、GTOPMODEL模型于临沂以上流域,以寻求该流域更好的洪水预报模型。 应用结果表明,基于子流域的新安江模型能很好的对整个沂沭泗流域进行洪水模拟,精度好,完全可以应用于实际生产中的洪水预报调度作业;Grid-新安江模型、TOPMODEL模型、GTOPMODEL模型对临沂以上流域的洪水模拟的效果也很明显,其中Grid-新安江模型、GTOPMODEL模型两个模型的模拟效果要略好于TOPMODEL模型,有助于提高临沂以上流域的洪水预报调度作业精度,同时也论证了Grid-新安江模型的合理性。更多还原

【Abstract】 The basin of Yishusi, located in the northeast of the Huaihe River, is a semi-arid region of China. The network of waterways is complicated in the basin. The terrain is much undulant, the main channels are near to the sources and the current is torrential, the flood amount is often large and the flood peak discharge is always on the top, it is very quick for the up-and-down of flood-wind, and the forecasting-time of flood is short. The capacity of the flowing flood and storage flood of existing hydraulic engineering works is very weak. At present, the system of flood prevention and decision-making of the basin is unsubstantial, and the number of hydrologic stations is short, which affects not only the accuracy of flood forecasting, but the veracity and real-time processing of flood scheduling. Therefore it is very important for flood preventing and disaster decreasing to complete the work of flood forecasting and scheduling successfully in the basin of Yishusi.According to the required accuracy of flood forecasting and the cases of the hydrologic station distribution, the basin is divided into several sub-basins.the Three- water Sources Xinanjiang model is carried on the rainfall-runoff calculating in every sub-basin. Flow routing from the sub-basin outlets to the whole basin outlet may be achieved by applying the Muskingum successive reaches model with considering the application of the strobes. In order to improve the accuracy of flood forecasting, the autoregressive model is applied. In this paper the faded-memory least square method is applied on real-time correction and the variation of forgetting factors influencing on the accuracy of flood forecasting of the basin is studied.In upper reaches of the Yishusi basin above the Linyi hydrologic station, the terrain is much undulant and the network of waterways is complicated. The linchpin of flood scheduling is the management of the Linyi flood peak discharge in the basin of Yishusi. In order to find the better flood forecasting model, the Grid-Xinanjiang, TOPMODEL, and GTOPMODEL are applied on the basin.The results show two points. Firstly, the Xinanjiang model can complete flood forecasting and scheduling in the Yishusi basin. Secondly, the Grid-Xinanjiang model, TOPMODEL, GTOPMODEL can do well in Linyi basin. Grid-Xinanjiang model and GTOPMODEL model perform better, which proves the rationality and reliability of Grid-Xinanjiang model.更多还原