Study on the response analysis of LID hydrological process to rainfall pattern based on framework for dynamic simulation of urban floods.

An in-depth analysis of the urban flood disaster level in response to different rainfall characteristics and Low Impact Development (LID) measures is of significant importance for addressing unfavorable management conditions and implementing effective flood control measures. This study proposes a dynamic urban flood simulation framework based on the Storm Water Management Model (SWMM) and Geographic Information System (GIS) spatial analysis, incorporating an active inundation seed search algorithm. The framework is calibrated and validated using nine historical urban flood events. Subsequently, the impact of rainfall patterns on urban inundation under LID measures is analyzed based on the dynamic urban flood simulation framework. The results show that the urban flood simulation framework exhibits good applicability, with Nash-Sutcliffe Efficiency (NSE) values of 0.825 and 0.763 during the calibration and validation periods, respectively. The extent of inundation shows little variation for rainfall events with a return period greater than 20 years, and the location of flooding is minimally affected by rainfall patterns. LID measures have a decreasing effect on urban inundation control as the return period of rainfall increases, and there are variations in hydrological responses to different rainfall patterns under the same return period. For single-peak rainfall events with the same return period, the control rates of inundation volume, flow, and infiltration decrease as the rainfall peak coefficient increases, indicating a weakening effect of LID measures on flood control with increasing rainfall peak coefficient. Under the same return period conditions, LID measures exhibit the best runoff control effect for uniform rainfall, while their effectiveness is lower for double-peak rainfall events and single-peak rainfall events with an r = 0.75 coefficient. The findings of this study provide a theoretical basis for urban flood warning and management of Low Impact Development measures.

Study on long short-term memory based on vector direction of flood process for flood forecasting.

Accurate flood forecasting is crucial for flood prevention and mitigation, safeguarding the lives and properties of residents, as well as the rational use of water resources. The study proposes a model of long and short-term memory (LSTM) combined with the vector direction (VD) of the flood process. The Jingle and Lushi basins were selected as the research objects, and the model was trained and validated using 50 and 49 measured flood rainfall-runoff data in a 7:3 division ratio, respectively. The results indicate that the VD-LSTM model has more advantages than the LSTM model, with increased NSE, and reduced RMSE and bias to varying degrees. The flow simulation results of VD-LSTM better match the observed flow hydrographs, improving the underestimation of peak flows and the lag issue of the model. Under the same task and dataset, with the same hyperparameter settings, VD-LSTM can more quickly reduce the loss function value and achieve a better fit compared to LSTM. The proposed VD-LSTM model couples the vectorization process of flood runoff with the LSTM neural network, which contributes to the model better exploring the change characteristics of rising and receding water in flood runoff processes, reducing the training gradient error of input-output data for the LSTM model, and more effectively simulating flood process.

Application of a hybrid algorithm of LSTM and Transformer based on random search optimization for improving rainfall-runoff simulation.

Flood forecasting using traditional physical hydrology models requires consideration of multiple complex physical processes including the spatio-temporal distribution of rainfall, the spatial heterogeneity of watershed sub-surface characteristics, and runoff generation and routing behaviours. Data-driven models offer novel solutions to these challenges, though they are hindered by difficulties in hyperparameter selection and a decline in prediction stability as the lead time extends. This study introduces a hybrid model, the RS-LSTM-Transformer, which combines Random Search (RS), Long Short-Term Memory networks (LSTM), and the Transformer architecture. Applied to the typical Jingle watershed in the middle reaches of the Yellow River, this model utilises rainfall and runoff data from basin sites to simulate flood processes, and its outcomes are compared against those from RS-LSTM, RS-Transformer, RS-BP, and RS-MLP models. It was evaluated against RS-LSTM, RS-Transformer, RS-BP, and RS-MLP models using the Nash-Sutcliffe Efficiency Coefficient (NSE), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Bias percentage as metrics. At a 1-h lead time during calibration and validation, the RS-LSTM-Transformer model achieved NSE, RMSE, MAE, and Bias values of 0.970, 14.001m3/s, 5.304m3/s, 0.501% and 0.953, 14.124m3/s, 6.365m3/s, 0.523%, respectively. These results demonstrate the model's superior simulation capabilities and robustness, providing more accurate peak flow forecasts as the lead time increases. The study highlights the RS-LSTM-Transformer model's potential in flood forecasting and the advantages of integrating various data-driven approaches for innovative modelling.

Prediction of water level at Huayuankou station based on rating curve.

The construction of large reservoirs has modified the process of water and sediment transport downstream, resulting in changes in the morphology of the river cross-section. Changes in water and sand transport and cross-sectional morphology are reflected in the rating curve at the cross-section. This study analyzed the variations in the rating curve at the Huayuankou (HYK) section and their influencing factors, and conducted water level predictions based on this relationship. The findings revealed that while the annual mean water level has shown a declining tendency over the past 20 years, the annual mean discharge has shown a constant pattern. The rating curve at this stretch narrowed from a rope-loop type curve in its natural condition to a more stable single curve as a result of the construction of the dam upstream of the HYK section. The effect of pre-flood section morphology and the water-sediment process on the scattering degree of the rating curve is inverse; increasing roughness and hydraulic radius decreases scattering degree, while increasing sand content and sand transport rate increases scattering degree. Using the measured data from 2020 as an example, the feasibility of predicting cross-sectional water levels using the rating curve was verified. The prediction results were accurate when the flow was between 1000 and 2800 m3/s; However, when the flow was between 2800 and 4000 m3/s, the forecast results were typically slightly lower than the measured values. Overall, the method demonstrates good predictive accuracy. Insight from the method can be used to predict water levels to better inform decision making about water resources management, and flood emergency response in the lower Yellow River.