ARIMA models of global solar radiation in the San Jerónimo de Andahuaylas district, Perú

Quispe-Infantes R.R. · Revista de la Academia Colombiana de Ciencias Exactas, Fisicas y Naturales (2025) · DOI: 10.18257/raccefyn.3206

Resumen

The San Jerónimo de Apurimac district experiences high solar radiation, which impacts agricultural, livestock, and mining activities. A reliable forecast would increase production and socioeconomic status. Here, we developed and compared autoregressive moving average (ARIMA) models of global solar radiation (GSR) for the San Jerónimo district using the meteorological data from the stations at the José María Arguedas National University (UNAJMA) and the USA National Aeronautics and Space Administration (NASA). We conducted an applied experimental research with a quantitative approach and a propositional descriptive level, with a sample of 1,037 historical RSG data recorded from January 1, 2022, to November 2, 2024 at the ground station using a light sensor and at the space station through documentary techniques of the data from the geostationary operational environmental satellite (GOES) (POWER, 2024). The following normality tests were applied: normal, Q-Q plot, Dickey-Fuller Test, stationarity tests, Levene and D’Agostino tests, simple autocorrelation (ACF) and partial ACF plots for calculating delays, and the Akaike information criterion (AIC) likelihood for choosing the best model. ARIMA(3,0,2) models were obtained for the NASA station and ARIMA(5,0,4) for the UNAJMA station, with mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean squared error (RMSE) values of 0.290; 5.211% and 0.384, respectively, in the first case; and 0.610; 10.623% and 0.702, in the second. Which allowed forecasting the RSG in the district of San Jerónimo from November 3 to 12, 2024.

Instituciones: Universidad Nacional José María Arguedas