|
Orji, F. N.
Michael Okpara University of Agriculture, Umudike, Abia State, Nigeria.
Ahaneku, I. E.
Michael Okpara University of Agriculture, Umudike, Abia State, Nigeria.
Tom, C. N.
Department of Agricultural and Environmental Engineering, Rivers State University, Port Harcourt, Nigeria.
Awu, J. I.
National Centre for Agricultural Mechanization, Ilorin, Kwara State, Nigeria.
ABSTRACT
The Niger River is the main source of life of large parts of West Africa. Climate change, anthropogenic activities, increasing population possess treats such as flooding, water pollution, increasing river discharge etc. River discharge is a function of the watershed temperature, evaporation, precipitation, infiltration and every other related hydrologic cycle attributes. This study focuses on the historical data such as temperature, precipitation to predict river discharge using Artificial Neural Network (ANN). A ten-year prediction of the river discharge was observed from 2023-2032. Hybrid ANN model and excel spreadsheet forecasting code was employed on the existing discharge from 1992 to 2022. This is to project the historical trend to understand the effect of the fluctuating input parameters on the discharge rate in years to come. The actual discharge rate (over the last 30 years) was compared to the forecasted discharge rate (in the next ten years). The results showed that the performance metrics of the Nonlinear Autoregressive Exogenous Neural Network (NARX-NN) gave Mean Square Error (MSE) of 4.30 x 10-5 and R-squared value of 0.9987. This shows that the neural network models increased with increasing input. Again, the results of the autocorrelation error fall approximately within the confidence limit of 99% (0.99). This implies that the NARX-NN model has the ability to predict accurately river discharge. In conclusion, the predictions were accurate except for climate change fluctuations in the input parameters over the subsequent years.
Keywords: Artificial Neural Network; Anthropogenic Activities; Flooding; Modeling, River Discharge
https://doi.org/10.33922/j.ujet_v10i1_13
|
View: 77 | Download: 3
Published
Tuesday, April 23, 2024
Issue
Vol. 10 No. 1, June 2024
Article Section
GENERAL
The contents of the articles are the sole opinion of the author(s) and not of UJET.
|
