Scholarlink

<p>Error correction techniques are often used to improve the accuracy of flood forecasts. The dynamic system response curve (DSRC) is a recently developed error correction method that has been successful in several applications. However, the DSRC method is difficult to apply to multi-site correction problems due to a lack of information. In this study, three DSRC-based multisite error correction methods are developed to address this problem by introducing the temporal, spatial and cyclic proportion coefficients respectively. The proposed methods (T-DSRC, S-DSRC and C-DSRC) can effectively use the available error feedback information to obtain updated rainfall series at multiple sites, thereby improving the accuracy of flood forecasting. The proposed methods were tested in the two catchments, and their performances were evaluated by the Nash-Sutcliffe efficiency coefficient (<i>NSE</i>), relative error of peak flow, deviation of peak time, and relative error of runoff depth. Results showed that compared with the original flood forecasts of the Xinanjiang model, the corrected results obtained by the multi-site error correction methods had higher <i>NSE</i> values. Therefore, the proposed three multi-site error correction methods (T-DSRC, S-DSRC and C-DSRC) had good capabilities in flood correction. Among them, the C-DSRC method had the highest average <i>NSE</i> value and a narrow variation range. Besides, the uncertainty of the proposed correction methods were evaluate by model conditional processor. Results indicated that the C-DSRC method consistently exhibited a larger average relative bandwidth reduction across both catchments while maintaining target containing ratio. These suggested that the C-DSRC method performed better in enhancing the accuracy of real-time flood forecasts compared to the T-DSRC and S-DSRC methods.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhanced flood forecasting models via multi-site rainfall error correction using dynamic system response curve

  • 期刊论文

  • 发表日期：

    2025年8月20日

查看全文
  • Jian Wang,
  • Weimin Bao,
  • Zhangling Xiao,
  • Mingjin Zhang

摘要

Error correction techniques are often used to improve the accuracy of flood forecasts. The dynamic system response curve (DSRC) is a recently developed error correction method that has been successful in several applications. However, the DSRC method is difficult to apply to multi-site correction problems due to a lack of information. In this study, three DSRC-based multisite error correction methods are developed to address this problem by introducing the temporal, spatial and cyclic proportion coefficients respectively. The proposed methods (T-DSRC, S-DSRC and C-DSRC) can effectively use the available error feedback information to obtain updated rainfall series at multiple sites, thereby improving the accuracy of flood forecasting. The proposed methods were tested in the two catchments, and their performances were evaluated by the Nash-Sutcliffe efficiency coefficient (NSE), relative error of peak flow, deviation of peak time, and relative error of runoff depth. Results showed that compared with the original flood forecasts of the Xinanjiang model, the corrected results obtained by the multi-site error correction methods had higher NSE values. Therefore, the proposed three multi-site error correction methods (T-DSRC, S-DSRC and C-DSRC) had good capabilities in flood correction. Among them, the C-DSRC method had the highest average NSE value and a narrow variation range. Besides, the uncertainty of the proposed correction methods were evaluate by model conditional processor. Results indicated that the C-DSRC method consistently exhibited a larger average relative bandwidth reduction across both catchments while maintaining target containing ratio. These suggested that the C-DSRC method performed better in enhancing the accuracy of real-time flood forecasts compared to the T-DSRC and S-DSRC methods.