• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Hydrological Research Letters >Strategy to automatically calibrate parameters of a hydrological model: a multi-step optimization scheme and its application to the Xinanjiang model
【24h】

Strategy to automatically calibrate parameters of a hydrological model: a multi-step optimization scheme and its application to the Xinanjiang model

机译：水文模型参数自动校正策略：多步优化方案及其在新安江模型中的应用

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Parameter calibration is fundamental for the implementation and operation of a hydrological model. Automatic calibration techniques have been widely studied. However, even the most modern optimization schemes cannot always help us to obtain an optimal parameter set due to high dimensionality of the parameter space and complex interactions between parameters. The main purpose of this study was to test our strategy for automatic parameter calibration: lowering the dimensionality. Our modified Xinanjiang model was selected for study. It consists of 15 parameters controlling data adjustment and representing hydrological processes. Morris’ global sensitivity analysis technique was used to get better understanding about the structure of the parameter space. Parameters were found to have significantly different sensitivities at yearly, monthly and daily temporal scales. Also strong interactions between the parameters were observed at all three scales. A multi-step optimization scheme was designed and tested based on these observations. In this scheme, the 15 parameters are divided into three groups and optimized group by group at the time scale they are most sensitive to by using the SCEM-UA algorithm, a global optimization algorithm. The newly developed scheme is shown to be very efficient and robust.
机译：参数校准是水文模型实施和运行的基础。自动校准技术已被广泛研究。但是，由于参数空间的高维性和参数之间的复杂交互作用，即使是最现代的优化方案也无法始终帮助我们获得最佳参数集。这项研究的主要目的是测试我们的自动参数校准策略：降低尺寸。我们选择了改进的新安江模型进行研究。它由15个控制数据调整并代表水文过程的参数组成。莫里斯（Morris）的全局灵敏度分析技术用于更好地了解参数空间的结构。发现参数在每年，每月和每天的时间尺度上具有显着不同的敏感性。在所有三个尺度上也观察到参数之间的强相互作用。基于这些观察结果，设计并测试了多步骤优化方案。在此方案中，将15个参数分为三个组，并按时标对每个组进行最优化，这是使用全局优化算法SCEM-UA算法最敏感的。新开发的方案被证明是非常有效和健壮的。

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