Yi Chen

Bio: Yi Chen is an academic researcher from Tongji University. The author has contributed to research in topics: Nonlinear autoregressive exogenous model & Flood myth. The author has co-authored 2 publications. Previous affiliations of Yi Chen include Chinese Ministry of Education.

Using NARX neural network to forecast droughts and floods over Yangtze River Basin

Abstract: Drought and flood events are two extreme climate phenomena which usually bring enormous economic and social loss. For meeting the goal of flood and drought prevention, the nonlinear autoregressive with exogenous input (NARX) neural network is employed to bridge the data gap between the Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow On (GRACE-FO) over Yangtze River Basin (YRB). The precipitation data from NASA Global Precipitation Measurement, temperature data from Global Historical Climatology Network and the Climate Anomaly Monitoring System, and terrestrial water storage anomalies (TWSA) from Global Land Data Assimilation System (GLDAS) are considered as the external inputs. Meanwhile, the performance of NARX models is evaluated for all possible combinations of time delays and neurons in order to find the optimal model structures. Then total storage deficit index (TSDI) is constructed based on TWSA reconstructions to assess drought and flood events over YRB, along with forecasting the extremes during the data gap period. The results show that when the number of time delays and neurons equals one and nine, respectively, the NARX model has an optimal performance with root mean square error (rmse), scaled rmse $$R^{ * }$$ , Nash-Sutcliff Efficiency (NSE) and correlation coefficient r of 1.34 cm, 0.34, 0.95 and 0.94, respectively. As indicated by TSDI and comparisons with previous studies, YRB has switched from drought periods to increased flood risks with a moderate correlation to global warming and El Nino-Southern Oscillation (ENSO). Finally, the most important conclusion that we successfully predict the flood events during the data gap period suggests that NARX neural network is promising for forecasting short-term hydrological extremes over YRB.

The applicability of using NARX neural network to forecast GRACE terrestrial water storage anomalies

Abstract: The Gravity Recovery and Climate Experiment (GRACE) satellite has proven adept at monitoring, characterizing, and predicting hydrological variables. This paper attempted to investigate the applicability of using nonlinear autoregressive with exogenous input (NARX) neural network to forecast GRACE terrestrial water storage anomalies over 11 basins. By using six hydrological indicators as external inputs and forming all possible combinations of these variables, we have found the appropriate external inputs of the optimal NARX model for each basin. In addition, the number of acceptable time delays and the number of hidden neurons are adjusted to train out the optimal NARX model. The results reveal that the NARX models with one time delay perform better than the models with higher than one delay, and the structures of the optimal NARX models vary with basins. The performance of the optimal NARX models of the 11 basins falls into “very good” category whether during training, validating, or testing period. In comparison with GRACE Follow On (GRACE-FO) results, the predictions from the optimal NARX models are satisfactory, with the highest correlation coefficient of 0.97 in the Amazon basin and the lowest correlation coefficient of 0.62 over the Yangtze basin. The results shown here not only bridge the data gap between GRACE and GRACE-FO but also facilitate the applications of using the NARX neural network to predict climate extremes like droughts and flooding.

A Novel Standardized Drought and Flood Potential Index Based on Reconstructed Daily GRACE Data

Abstract: Multiple indicators derived from the Gravity Recovery and Climate Experiment (GRACE) satellite have been used in monitoring floods and droughts. However, these measures are constrained by the relatively short time span (∼20 years) and coarse temporal resolution (1 month) of the GRACE and GRACE Follow-On missions, and the inherent decay mechanism of the land surface system has not been considered. Here we reconstructed the daily GRACE-like terrestrial water storage anomaly (TWSA) in the Yangtze River basin (YRB) during 1961–2015 based on the Institute of Geodesy at Graz University of Technology (ITSG)-Grace2018 solution using the random forest (RF) model. A novel antecedent metric, namely, standardized drought and flood potential index (SDFPI), was developed using reconstructed TWSA, observed precipitation, and modeled evapotranspiration. The potential of SDFPI was evaluated against in situ discharge, VIC simulations, and several widely used indices such as total storage deficit index (TSDI), self-calibrated Palmer drought severity index (sc-PDSI), and multiscale standardized precipitation evapotranspiration index (SPEI). Daily SDFPI was utilized to monitor and characterize short-term severe floods and droughts. The results illustrate a reasonably good accuracy of ITSG-Grace2018 solution when compared with the hydrological model output and regional water balance estimates. The RF model presents satisfactory performances for the TWSA reconstruction, with a correlation coefficient of 0.88 and Nash–Sutcliffe efficiency of 0.76 during the test period 2011–15. Spatiotemporal propagation of the developed SDFPI corresponds well with multiple indices when examined for two typical short-term events, including the 2003 flood and 2013 drought. A total of 22 submonthly exceptional floods and droughts were successfully detected and featured using SDFPI, highlighting its outperformance and capabilities in providing inferences for decision-makers and stakeholders to monitor and mitigate the short-term floods and droughts.

A Novel GRA-NARX Model for Water Level Prediction of Pumping Stations

Abstract: It is necessary but difficult to accurately predict the water levels in front of the pumping stations of an open-channel water transfer project because of the complex interactions among hydraulic structures. In this study, a novel GRA-NARX (gray relation analysis—nonlinear auto-regressive exogenous) model is proposed based on a gray relation analysis (GRA) and nonlinear auto-regressive exogenous (NARX) neural network for 2 h ahead for the prediction of water levels in front of pumping stations, in which an improved algorithm of the NARX neural network is used to obtain the optimal combination of the time delay and the hidden neurons number, and GRA is used to reduce the prediction complexity and improve the prediction accuracy by filtering input factors. Then, the sensitivity to changes of the training algorithm is analyzed, and the prediction performance is compared with that of the NARX and GRA-BP (gray relation analysis back-propagation) models. A case study is performed in the Tundian pumping station of the Miyun project, China, to demonstrate the reliability and accuracy of the proposed model. It is revealed that the GRA-NARX-BR (gray relation analysis—nonlinear auto-regressive exogenous—Bayesian regularization) model has higher accuracy than the model based only on a NARX neural network and the GRA-BP model with a correlation coefficient (R) of 0.9856 and a mean absolute error (MAE) of 0.00984 m. The proposed model is effective in predicting the water levels in front of the pumping stations of a complex open-channel water transfer project.