Long-Term Wind and Solar Resource Variability and Forecasting for Sustainable Energy Planning in the Galapagos Islands
The increasing use of isolated electrical systems based on renewable energy sources (RES) requires a comprehensive understanding and analysis of the variability of renewable energy resources (RER) and, in turn, appropriate forecasting techniques for sound sustainable energy planning. For this reason, this study presents a quantitative assessment of RER variability in response to wind speed and solar irradiance on Santa Cruz Island, Galapagos, Ecuador. This study analyzes the temporality and variability of renewable energy resources (RERs), i.e., wind speed and irradiance, used in renewable energy systems (RES). To this end, four forecasting methods are examined: Autoregressive Integrated Moving Average (ARIMA), Autoregressive Integrated Seasonal Moving Average (SARIMA), regression trees, and Artificial Neural Networks (ANN) trained via backpropagation (BP). These models are evaluated under identical conditions, using standard performance metrics. This research underscores the importance of integrating long-term resource characterization with predictive modeling to support the planning of sustainable power systems in ecologically sensitive islands.
Authors
- Diego Arcos-Avilés (ORCID: https://orcid.org/0000-0002-1511-1614)
- C. P. Bernal (ORCID: https://orcid.org/0000-0002-1340-3234)
- F. Chicaiza
- Rodolfo Gordillo-Orquera (ORCID: https://orcid.org/0000-0002-4866-184X)
- Andrea Godoy
Institutions
- Universidad de las Fuerzas Armadas ESPE (EC)
Publication Details
- Journal
- Sustainability
- Published
- 2026-11-13
- DOI
- https://doi.org/10.3390/su182211576
- Primary Topic
- Energy Load and Power Forecasting
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Universidad de las Fuerzas Armadas ESPE