Improved Rainfall Prediction through Nonlinear Autoregressive Network with Exogenous Variables: A Case Study in Andes High Mountain Region

Abstract

Precipitation is the most relevant element in the hydrological cycle and vital for the biosphere. However, when extreme precipitation events occur, the consequences could be devastating for humans (droughts or floods). An accurate prediction of precipitation helps decision-makers to develop adequate mitigation plans. In this study, linear and nonlinear models with lagged predictors and the implementation of a nonlinear autoregressive model with exogenous variables (NARX) network were used to predict monthly rainfall in Labrado and Chirimachay meteorological stations. To define a suitable model, ridge regression, lasso, random forest (RF), support vector machine (SVM), and NARX network were used. Although the results were “unsatisfactory” with the linear models, the specific direct influences of variables such as Niño 1 + 2, Sahel rainfall, hurricane activity, North Pacific Oscillation, and the same delayed rainfall signal were identified. RF and SVM also demonstrated poor performance. However, RF had a better fit than linear models, and SVM has a better fit than RF models. Instead, the NARX model was trained using several architectures to identify an optimal one for the best prediction twelve months ahead. As an overall evaluation, the NARX model showed “good” results for Labrado and “satisfactory” results for Chirimachay. The predictions yielded by NARX models, for the first six months ahead, were entirely accurate. This study highlighted the strengths of NARX networks in the prediction of chaotic and nonlinear signals such as rainfall in regions that obey complex processes. The results would serve to make short-term plans and give support to decision-makers in the management of water resources.

Multi-step-ahead prediction of river flow using NARX neural networks and deep learning LSTM

“…NSE is even better than other metrics, such as the coefficient of determination, a.k.a regression coefficient. An NSE value of above 0.75 is considerably a good fit model, while less than 0.5 indicates unsatisfactory model performance (Pena et al 2020).…”

“…when the model fits the training data, the predictive model developed needs to be evaluated using both training dataset and test dataset. The common metric for regression model accuracy can be found such as in (Pena et al 2020). In this study, the obtained model accuracy is evaluated by calculating RMSE (root mean squared error) value and Nash-Sutcliffe efficiency coefficient (NSE), which are defined, respectively, as follows:…”

Cluster Forecasting of Corruption Using Nonlinear Autoregressive with Exogenous Variables (NARX) – An Artificial Neural Network Analysis

“…The NARX neural network method has been used in various research studies, for example, forecasting heating and cooling electrical loads (Buitrago & Asfour, 2017;Powell et al, 2014), network traffic flows (Alfred, 2015), rainfall (Benevides et al, 2019;Peña et al, 2020), and crop yield and price (Khamis & Abdullah, 2014;Paul & Sinha, 2016). Peña et al (2020) found that NARX provides significantly more accurate results for rainfall predictions compared with nonlinear regression models and the SVM techniques, and Paul and Sinha (2016) determined that NARX outperforms ARIMA time series models in forecasting crop yield. NARX has also been applied in macroeconomic modeling.…”

Cluster Forecasting of Corruption Using Nonlinear Autoregressive Models with Exogenous Variables (NARX)—An Artificial Neural Network Analysis

“…The NARX neural network method has been used in various research studies, for example, forecasting heating and cooling electrical loads [20,21], network traffic flows [22], rainfall [23,24], and crop yield and price [25,26]. Peña et al [24] found that NARX provides significantly more accurate results for rainfall predictions compared with nonlinear regression models and SVM techniques, and Paul and Sinha [25] determined that NARX outperforms ARIMA time series models in forecasting crop yield. NARX has also been applied in macroeconomic modeling.…”