Time Series Analysis of Solar Energy Production Based on Weather Conditions

Year 2025, Volume: 12 Issue: 4, 1060 - 1077, 31.12.2025

Mehmet Bülüç , Onur Sevli , Lokman Yünlü

https://doi.org/10.54287/gujsa.1797659

Abstract

This study investigates the impact of weather conditions on solar energy production by comparing the performance of different time series forecasting models. Production data obtained from the İkitelli Solar Power Plant in Istanbul, together with simultaneous meteorological variables such as sunshine duration, cloudiness, humidity, and temperature, were analyzed. Three forecasting models—ARIMA, LSTM, and FB-Prophet—were implemented to evaluate their predictive performance. The accuracy of the models was assessed using widely accepted statistical metrics including Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R²). The results show that the ARIMA model achieved the highest accuracy in short-term forecasting with the lowest error rates, demonstrating its effectiveness in handling stationary time series. The LSTM model, a deep learning approach, proved successful in capturing long-term dependencies, offering a robust alternative despite requiring larger datasets for optimal performance. The FB-Prophet model stood out for its ability to account for seasonality and trends but exhibited lower accuracy in short-term fluctuations compared to ARIMA and LSTM. Additionally, the analysis revealed that solar energy production is strongly correlated with weather conditions. In particular, an increase in sunshine duration positively influenced electricity generation, while greater cloud cover led to significant reductions in production levels. These findings highlight the importance of incorporating meteorological data into forecasting models to enhance the accuracy and reliability of renewable energy predictions. Furthermore, the study emphasizes that selecting the appropriate forecasting model according to data characteristics is critical for effective energy management. The outcomes provide methodological insights that may contribute to the optimization of solar energy projects and the integration of renewable energy into sustainable energy strategies.

Keywords

Solar Energy , Time Series Analysis , Renewable Energy Forecasting , Artificial Intelligence , Machine Learning

Ethical Statement

No ethical approval was required for this study.

Supporting Institution

This research received no specific grant from any funding agency.

Project Number

-

Thanks

The authors express their gratitude to the Istanbul Metropolitan Municipality and World Weather Online for providing the data.

References

• Akkaya, M., Gültekin, A., Sabancı, K., Balcı, S., & Sağlam, H. (2020). Analysis and applicability of Mersin region wind speed data with artificial neural networks. Nevşehir Bilim ve Teknoloji Dergisi, 9(1), 39–51. https://doi.org/10.17100/nevbiltek.691120
• Adjiski, V., Kaplan, G., & Mijalkovski, S. (2023). Assessment of the solar energy potential of rooftops using LiDAR datasets and GIS-based approach. International Journal of Engineering and Geosciences, 8(2), 188–199. https://doi.org/10.26833/ijeg.1112274
• Atalay, B. A., & Zor, K. (2025). XGBoost (Aşırı gradyan artırımlı karar ağaçları) ile hidroelektrik enerji tahmini. Çukurova Üniversitesi Mühendislik Fakültesi Dergisi, 40(1), 205–218. https://doi.org/10.21605/cukurovaumfd.1666062
• Ateş, F., & Şenol, R. (2023). Solar energy prediction using regression methods. Yekarum, 8(2), 94–104.
• Balcı, F., & Oralhan, Z. (2020). LSTM ile EEG tabanlı kimliklendirme sistemi tasarımı. Avrupa Bilim ve Teknoloji Dergisi, Ejosat Özel Sayı 2020 (HORA), 135–141. https://doi.org/10.31590/ejosat.779526
• Balti, H., Abbes, A. B., Mellouli, N., Sang, Y., Farah, I. R., Lamolle, M., & Zhu, Y. (2021, July 4-5). Big data-based architecture for drought forecasting using LSTM, ARIMA and Prophet: Case study of the Jiangsu Province, China. In: 2021 International Congress of Advanced Technology and Engineering (ICOTEN), Taiz, Yemen. https://doi.org/10.1109/ICOTEN52080.2021.9493513
• Başaran Caner, C., & Engindeniz, S. (2020). Türkiye’de pamuk üretiminin ARIMA modeli ile tahmini. Tarım Ekonomisi Dergisi, 26(1), 63–70. https://doi.org/10.24181/tarekoder.681079
• Berus, Y., & Yakut, Y. B. (2024). Derin öğrenme (1D-CNN, RNN, LSTM, BiLSTM) ile enerji tüketim tahmini: Diyarbakır AVM örneği. Dicle University Journal of Engineering, 15(2), 311–322. https://doi.org/10.24012/dumf.1415055
• Çolak, M. B., & Özhan, E. (2025). Renewable energy forecasting in Turkey: Analytical approaches. Journal of Intelligent Systems: Theory and Applications, 8(1), 25–34. https://doi.org/10.38016/jista.1447980
• Demirtop, A., & Sevli, O. (2024). Wind speed prediction using LSTM and ARIMA time series analysis models: A case study of Gelibolu. Turkish Journal of Engineering, 8(3), 524–536. https://doi.org/10.31127/tuje.1431629
• Ismail, A. A., Wood, T., & Bravo, H. C. (2018). Improving long-horizon forecasts with expectation-biased LSTM networks. https://doi.org/10.48550/arXiv.1804.06776
• Kartal, B., & Alkan, R. M. (2025). Solar energy based model for decarbonization: a case study in Istanbul. International Journal of Engineering and Geosciences, 11(1), 64–77. https://doi.org/10.26833/ijeg.1648588
• Mohamed, H., Hassan, A., & Elhag, A. (2025). A five-year study using Sentinel-5P data observing seasonal dynamics and long-term trends of atmospheric pollutants. International Journal of Engineering and Geosciences, 10(2), 262–271. https://doi.org/10.26833/ijeg.1587122
• Özşahin, E., Girgin, K., & Köse, T. (2025). Topographical features affecting the distribution of wind power plants (WPPs) in Türkiye. International Journal of Engineering and Geosciences, 10(3), 339–351.
• Özsoy, K., & Acar, E. (2017). Isparta-Senirkent ilçesinde güneş enerjisi potansiyeli üzerine bir araştırma. Teknik Bilimler Dergisi, 7(1), 29–37.
• Özer, O. O., & İlkdoğan, U. (2013). Box–Jenkins modeli yardımıyla dünya pamuk fiyatının tahmini. Tekirdağ Ziraat Fakültesi Dergisi, 10(2), 13–20.
• Öztürk, Ö. F., & Pashaei, E. (2021). Konuşmalardaki duygunun evrişimsel LSTM modeli ile tespiti. Dicle University Journal of Engineering, 12(4), 581–589. https://doi.org/10.24012/dumf.1001914
• Polat, N., & Memduhoğlu, A. (2024). An unmanned aerial vehicle-based investigation of roof patch suitability for solar panel installation. International Journal of Engineering and Geosciences, 9(2), 281–291. https://doi.org/10.26833/ijeg.1424400
• Suryawan, I. T., Putra, I. N., Meliana, P. M., & Sudipa, I. I. (2024). Performance comparison of ARIMA, LSTM, and Prophet methods in sales forecasting. Sinkron: Jurnal dan Penelitian Teknik Informatika, 8(4), 2410–2421. https://doi.org/10.33395/sinkron.v8i4.14057
• Taştan, S. (2015). Genetik algoritmalar ile ARIMA modellerinin belirlenmesi. Cumhuriyet Üniversitesi İktisadi ve İdari Bilimler Dergisi, 16(2), 161–171.
• Tamay, M., & Türker, G. F. (2024). Machine learning based energy forecasting for photovoltaic solar plants. Yekarum, 9(2), 128–146.
• Umut, Y. (2023). Yapay zeka ve makine öğrenimi araçları ile elektrik tüketimi tahmini MSc Thesis, İzmir Katip Çelebi Üniversitesi.
• Vargas-Forero, V. M., Manotas-Duque, D. F., & Trujillo, L. (2025). Comparative study of forecasting methods to predict the energy demand for the market of Colombia. International Journal of Energy Economics and Policy, 15(1), 65–76. https://doi.org/10.32479/ijeep.17528
• Yıldız, M., & Atış, E. (2019). Türkiye organik incir ihraç fiyatının ARMA yöntemi ile tahmini. Tarım Ekonomisi Dergisi, 25(2), 141–147. https://doi.org/10.24181/tarekoder.583972
• Yolcan, O. O., & Köse, R. (2020). Türkiye’nin güneş enerjisi durumu ve güneş enerjisi santrali kurulumunda önemli parametreler. Kırklareli University Journal of Engineering and Science, 6(2), 196–215. https://doi.org/10.34186/klujes.793471
• Yurttabir, A., & Şen, İ. K. (2021). Prophet model in financial performance forecast: Implementation in manufacturing sector. Journal of Economics, Finance and Accounting, 8(4), 160–166. https://doi.org/10.17261/Pressacademia.2021.1470
• Zhang, B., Song, C., Li, Y., & Jiang, X. (2022). Spatiotemporal prediction of O₃ concentration based on the KNN-Prophet-LSTM model. Heliyon, 8(11), e11670. https://doi.org/10.1016/j.heliyon.2022.e11670