A fusion-based neural network methodology for monthly reservoir inflow prediction using MODIS products

Abstract

A model fusion approach was developed based on five artificial neural networks (ANNs) and MODIS products. Static and dynamic ANNsthe multi-layer perceptron (MLP) with one and two hidden layers, general regression neural network (GRNN), radial basis function (RBF) and nonlinear autoregressive network with exogenous inputs (NARX)were used to predict the monthly reservoir inflow in Mollasadra Dam, Fars Province, Iran. Leaf area index and snow cover from MODIS, and rainfall and runoff data were used to identify eight different combinations to train the models. Statistical error indices and the Borda count method were used to verify and rank the identified combinations. The best results for individual ANNs were combined with MODIS products in a fusion model. The results show that using MODIS products increased the accuracy of predictions, with the MLP with two hidden layers giving the best performance. Also, the fusion model was found to be superior to the best individual ANNs.

Physically based vs. data-driven models for streamflow and reservoir volume prediction at a data-scarce semi-arid basin

“…There are only a few studies in the literature regarding reservoirs, and in those studies the reservoir inlet or outlet flow was predicted rather than reservoir volumes. Ghazali et al ( 2018 ) studied model performance for simulating monthly reservoir inflow and the R 2 results ranged from 0.73 to 0.90. Yang et al ( 2019 ) also examined the reservoir inflow prediction and they found the NSE result as 0.85.…”

“…NARX has been successfully used for modeling nonlinear systems (Wunsch et al 2018) with its capability to store information in memory much longer than other RNNs (Lin et al 1996), which leads to faster convergence and better generalization (Lin et al 1998). Most researchers used NARX for predicting groundwater levels (Guzman et al 2017(Guzman et al , 2019; Javadinejad et al 2020;Nunno and Granata 2020;Wunsch et al 2018), reservoir inflows (Ghazali et al 2018;Yang et al 2019), streamflow (Nunno et al 2021), water temperatures (Kwak et al 2017), and floods (Chang et al 2014(Chang et al , 2022Nanda et al 2016;Rjeily et al 2017). The number of studies that focused on reservoir volume and streamflow prediction is comparatively lower (Ghazali et al 2018;Nunno et al 2021;Yang et al 2019).…”

“…Most researchers used NARX for predicting groundwater levels (Guzman et al 2017(Guzman et al , 2019; Javadinejad et al 2020;Nunno and Granata 2020;Wunsch et al 2018), reservoir inflows (Ghazali et al 2018;Yang et al 2019), streamflow (Nunno et al 2021), water temperatures (Kwak et al 2017), and floods (Chang et al 2014(Chang et al , 2022Nanda et al 2016;Rjeily et al 2017). The number of studies that focused on reservoir volume and streamflow prediction is comparatively lower (Ghazali et al 2018;Nunno et al 2021;Yang et al 2019). To the best of our knowledge, no studies used NARX for reservoir volume prediction.…”