Short-Term Streamflow Forecasting Using the Feature-Enhanced Regression Model

Abstract

Reservoir inflow forecasting is extremely important for the management of a reservoir. In practice, accurate forecasting depends on the feature learning performance. To better address this issue, this paper proposed a feature-enhanced regression model (FER), which combined stack autoencoder (SAE) with long short-term memory (LSTM). This model had two constituents: (1) The SAE was constructed to learn a representation as close as possible to the original inputs. Through deep learning, the enhanced feature could be captured sufficiently. (2) The LSTM was established to simulate the mapping between the enhanced features and the outputs. Under recursive modeling, the patterns of correlation in the short term and dependence in the long term were considered comprehensively. To estimate the performance of the FER model, two historical daily discharge series were investigated, i.e., the Yangtze River in China and the Sava Dolinka River in Slovenia. The proposed model was compared with other machine-learning methods (i.e., the LSTM, SAE-based neural network, and traditional neural network). The results demonstrated that the proposed FER model yields the best forecasting performance in terms of six evaluation criteria. The proposed model integrates the deep learning and recursive modeling, and thus being beneficial to exploring complex features in the reservoir inflow forecasting. Moreover, for smaller catchments with significant torrential characteristics, more data are needed (e.g., at least 20 years) to effectively train the model and to obtain accurate flood-forecasting results.

Impact of the Mean Daily Air Temperature Calculation on the Rainfall-Runoff Modelling

“…The results show that performance criteria used in the process of model calibration and the length of the warm-up period had a larger effect on the model results than selection of the equation used for calculating the mean daily air temperature (Section 3.2). Similar results would be expected in the case of shortening or extending the length of the calibration period [40]. However, some studies have indicated that shorter periods could be sufficient for efficient model calibration [41,42].…”

[Retracted] Forecast of the Employment Situation of College Graduates Based on the LSTM Neural Network

“…Yang and Wang [ 15 ] provided the employment situation of graduates of civil aviation aircraft; due to its complex and irregular curve shape, it is laborious for the existing single model to realize the ideal forecast effect. To this end, Xu et al [ 16 ] conceived a hybrid method based on SSA and SVR. The method first selects the fault feature components from the original data and then models and forecasts them separately, and obtains better experimental results than a single model.…”

Exploring Options for Flood Risk Management with Special Focus on Retention Reservoirs

“…Moreover, for the Ljubljanica river, such water retention is even increasing the peak discharge, which can be attributed to the karst characteristics of the Ljubljanica river catchment [53,55,69] and outflow from the reservoirs that shifts the peak discharge for a few hours as shown in Section 2.4. More specifically, several natural karst poljes (e.g., lake Cerknica or Planinsko polje) are located in the Ljubljanica river catchment that act as extensive natural retention reservoirs leading to a lagged (i.e., slower) response of the Ljubljanica river catchment compared to more torrential rivers such as Sava or Savinja [70]. A similar conclusion can be seen if one would construct some dry retention reservoirs in the Gradaščica river catchment (Tables 6 and 7).…”