• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Flood Forecasting and Water Resources Assessment for IAHS(International Association of Hydrological Sciences)-PUB >Modeling Streamflow Processes with Univariate Long-memory ARFIMA Model
【24h】

Modeling Streamflow Processes with Univariate Long-memory ARFIMA Model

机译：使用单变量长内存ARFIMA模型对流过程进行建模

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The applicability of long-memory fractionally integrated autoregressive moving average model, i.e., ARFIMA model, in the modeling of daily streamflow processes is investigated in the present study. The ARFIMA mode, autoregressine moving average (ARMA) model and artificial neural network (ANN) model are applied to three daily streamflow series of the headwaters of three rivers in high cold mountain areas. The result shows that all the univariate time series models work very well for forecasting short-term daily average discharges of the three rivers, especially for two of them. The ARFIMA model built on the basis of deseasonalized daily streamflow series outperforms the ARMA model in two out of three cases, meanwhile it outperforms or is comparable to the ANN model in two cases. The ARFIMA model is therefore recommended as an alternative to the conventional ARMA model for modeling univariate daily streamflow processes.
机译：本研究研究了长记忆分数积分自回归移动平均模型（即ARFIMA模型）在日常流量过程建模中的适用性。将ARFIMA模式，自回归移动平均（ARMA）模型和人工神经网络（ANN）模型应用于高寒山区三河源头的三个日流量序列。结果表明，所有的单变量时间序列模型都可以很好地预测三条河流（尤其是其中两条）的短期日均流量。在每日减少流量的基础上建立的ARFIMA模型，在三分之二的情况下胜过ARMA模型，而在两种情况下，其表现优于或可与ANN模型媲美。因此，建议使用ARFIMA模型作为常规ARMA模型的替代模型，以对单变量每日流量过程进行建模。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号