<p>Urban real-time rainfall-runoff forecasting (URRF) offers an economical and efficient approach to assessing flood risks in urban areas. However, the hydrological processes of urban rainfall are characterized by high nonlinearity and long-term dependencies due to strong uncertainties and significant human influences, making URRF a challenging task in hydrological simulation. Existing methods often fall short in meeting the requirements for real-time response and accuracy. To address these limitations, this paper proposes a novel global-encoder and local-decoder (GL-ED) model. The global encoder extracts global temporal features, while the local decoder focuses on forecasting. A temporal fully connected (TFC) module is introduced within the global encoder to capture the global features of runoff sequences, overcoming the limitation of convolutional operations that primarily focus on local information. Additionally, to tackle the uneven distribution of urban rainfall-runoff data, a novel RD loss function is proposed, combining dynamic time warping (DTW) with RMSE to better guide the training of complex features. The GL-ED model was evaluated using observed urban rainfall events from January 2018 to December 2019 in a 3.52 <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12145_2025_1907_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="32" /> </InlineMediaObject> <EquationSource Format="TEX">$km^2$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mi>k</mi> <msup> <mi>m</mi> <mn>2</mn> </msup> </mrow> </math></EquationSource> </InlineEquation> complex terrain area in Chongqing, China. Experimental results demonstrate that the GL-ED model outperforms conventional deep learning models, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory (LSTM) networks, in terms of NSE, MAE, and RMSE. In ablation experiments, the GL-ED model trained with the RD loss function achieved an average improvement of <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12145_2025_1907_Article_IEq2.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="43" /> </InlineMediaObject> <EquationSource Format="TEX">$2.51\%$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>2.51</mn> <mo>%</mo> </mrow> </math></EquationSource> </InlineEquation> in NSE and <InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12145_2025_1907_Article_IEq3.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="43" /> </InlineMediaObject> <EquationSource Format="TEX">$5.62\%$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>5.62</mn> <mo>%</mo> </mrow> </math></EquationSource> </InlineEquation> in KGE, while reducing RMSE, MAE, and Pbias by <InlineEquation ID="IEq4"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12145_2025_1907_Article_IEq4.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="43" /> </InlineMediaObject> <EquationSource Format="TEX">$6.08\%$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>6.08</mn> <mo>%</mo> </mrow> </math></EquationSource> </InlineEquation>, <InlineEquation ID="IEq5"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12145_2025_1907_Article_IEq5.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="51" /> </InlineMediaObject> <EquationSource Format="TEX">$11.31\%$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>11.31</mn> <mo>%</mo> </mrow> </math></EquationSource> </InlineEquation>, and <InlineEquation ID="IEq6"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="12145_2025_1907_Article_IEq6.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="31" /> </InlineMediaObject> <EquationSource Format="TEX">$93\%$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>93</mn> <mo>%</mo> </mrow> </math></EquationSource> </InlineEquation>, respectively. These findings highlight the model’s capability to provide reliable and accurate real-time rainfall-runoff forecasting, offering significant potential for enhancing urban flood risk management and decision-making processes.</p>

Global to local: a novel encoder-decoder framework for urban real-time rainfall-runoff forecasting

期刊论文
发表日期：

2025年5月17日

查看全文

Yuan Tian,
Wenlong Fu,
Yajing Dong,
Qi Han,
Qian Xiong,
Ruonan Cui,
Zhenyu Ao,
Xinyu Lei

摘要

Urban real-time rainfall-runoff forecasting (URRF) offers an economical and efficient approach to assessing flood risks in urban areas. However, the hydrological processes of urban rainfall are characterized by high nonlinearity and long-term dependencies due to strong uncertainties and significant human influences, making URRF a challenging task in hydrological simulation. Existing methods often fall short in meeting the requirements for real-time response and accuracy. To address these limitations, this paper proposes a novel global-encoder and local-decoder (GL-ED) model. The global encoder extracts global temporal features, while the local decoder focuses on forecasting. A temporal fully connected (TFC) module is introduced within the global encoder to capture the global features of runoff sequences, overcoming the limitation of convolutional operations that primarily focus on local information. Additionally, to tackle the uneven distribution of urban rainfall-runoff data, a novel RD loss function is proposed, combining dynamic time warping (DTW) with RMSE to better guide the training of complex features. The GL-ED model was evaluated using observed urban rainfall events from January 2018 to December 2019 in a 3.52 $km^2$ $k m^{2}$ complex terrain area in Chongqing, China. Experimental results demonstrate that the GL-ED model outperforms conventional deep learning models, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory (LSTM) networks, in terms of NSE, MAE, and RMSE. In ablation experiments, the GL-ED model trained with the RD loss function achieved an average improvement of $2.51\%$ $2.51 %$ in NSE and $5.62\%$ $5.62 %$ in KGE, while reducing RMSE, MAE, and Pbias by $6.08\%$ $6.08 %$ , $11.31\%$ $11.31 %$ , and $93\%$ $93 %$ , respectively. These findings highlight the model’s capability to provide reliable and accurate real-time rainfall-runoff forecasting, offering significant potential for enhancing urban flood risk management and decision-making processes.