Search

Accurate flood runoff and water level predictions are crucial research topics due to their significance for early warning systems, particularly in improving peak flood level forecasts and reducing time lags. This study proposes a novel method, Trend Forecasting Method (TFM), to improve model accuracy and overcome the time lag problem due to data scarcity. The proposed method includes the following steps: (1) select appropriate input factors causing flood events, (2) determine the most suitable AI method as the basis for forecasting models, (3) a forecasting model using a multi-step-ahead approach and a forecasting model with variation in flood depth as input are developed as compared to the selected model in Step 2, and (4) according to the rising limb and falling limb of a flood hydrograph, the maximum and minimum values predicted by the models above are respectively selected as the final outputs. The application to demonstrate the advantages of the proposed method was conducted in the Annan District of Tainan City, Taiwan. Of all the models tested, the Gated Recurrent Unit (GRU) demonstrated superior accuracy in forecasting flood depths, followed by Long Short-Term Memory (LSTM) and Bidirectional LSTM, with the Back Propagation Neural Network falling behind. With a Nash–Sutcliffe efficiency coefficient (NSE) of 0.56 for the next hour's forecast, the GRU model's structure appears particularly fitting for flood depth forecast. However, all four models showed time lag issues. TFM substantially enhanced the GRU model's forecast accuracy, mitigating the time lag. TFM achieved an NSE of 0.82 for forecasting 10-, 20-, 30-, 40-, 50-, and 60-min lead time. The observed flood depths had a 68% probability of consistent rise or fall, validating TFM's underlying hypothesis. Furthermore, including an autoregressive model in TFM reduced the time lag problem. [ABSTRACT FROM AUTHOR]

Since flooding in urban areas is rarely observed using sensors, most researchers use artificial intelligence (AI) models to predict flood hazards based on model simulation data. However, there is still a gap between simulation and real flooding phenomenon due to the limitation of the model. Few studies have reported on the AI model for flood inundation depth forecasting based on observed data. This study presents a novel method integrating long short-term memory (LSTM) with moving average (MA) for flood inundation depth forecasting based on observed data. A flood-prone intersection in Rende District, Tainan, Taiwan, was adopted as the study area. This investigation compared the forecasting performance of the backpropagation neural network (BPNN), recurrent neural network (RNN) and LSTM models. Accumulated rainfall (Ra) and the moving average (MA) method were applied to enhance the LSTM model performance. The model forecast accuracy was evaluated using root mean square error, coefficient of determination (R2) and Nash–Sutcliffe efficiency (NSE). Analytical results indicated that the LSTM had better forecasting ability than the RNN and BPNN, because LSTM had both long-term and short-term memory. Since Ra was an important factor in flooding, adding the Ra to the model input upgraded the LSTM forecasting accuracy for high inundation depths. Because MA reduced the noise of the data, processing the model output using the MA also elevated the forecasting accuracy for high inundation depths. For 3-step-ahead forecasting, the NSE of the model benchmark BPNN was 0.79. Using LSTM, Ra and MA, NSEs gradually increased to 0.83, 0.88 and 0.91, respectively. [ABSTRACT FROM AUTHOR]