Günlük İklim Tahmini için İstatistiksel ve Derin Öğrenme Modellerinin Karşılaştırmalı Analizi: Delhi Veri Kümesi Üzerine Bir Vaka Çalışması

Öz

İklim değişikliğinin etkilerinin giderek arttığı günümüzde, meteorolojik parametrelerin doğru ve güvenilir bir şekilde tahmin edilmesi, tarımsal planlamadan afet yönetimine kadar pek çok alanda kritik öneme sahiptir. Bu çalışma, 2013-2017 yıllarını kapsayan Delhi iklim verilerini kullanarak ortalama sıcaklık, nem, rüzgâr hızı ve ortalama basıncın günlük tahmini için yedi tahmin modelinin (iki temel istatistiksel model (ARIMA, SARIMA) ve beş derin öğrenme modeli (LSTM, CNN, MLP, CNN-BiLSTM-Attention ve Wavelet-CNN-LSTM)) sistematik bir karşılaştırmasını sunmaktadır. Tüm modellerin hiperparametreleri Rastgele Arama yöntemiyle optimize edilmiş ve istatistiksel güvenilirliği sağlamak için farklı rastgele tohumlarla 10 bağımsız çalışma yapılmıştır. Model performansı RMSE, MAE, MAPE, MASE ve R² metrikleri kullanılarak değerlendirilirken, performans farklılıklarının istatistiksel anlamlılığını değerlendirmek için Nemenyi post-hoc analizi ile Friedman testi kullanılmıştır. Sonuçlar, Wavelet-CNN-LSTM modelinin genel sıralamada en iyi performansı gösterdiğini, ardından sırasıyla LSTM, MLP ve CNN modellerinin geldiğini ve bunların hepsinin ARIMA ve SARIMA modellerinden önemli ölçüde daha iyi performans sergilediğini göstermektedir. Sıcaklık (R² = 0,922) ve basınç (R² = 0,9) için yüksek tahmin doğruluğu elde edilirken, rüzgâr hızı, doğası gereği stokastik yapısı nedeniyle tüm modeller için en zorlu değişken olma özelliğini korumaktadır (R² < 0,2). Tahmin edilebilirlik analizi, düşük otokorelasyon ve yüksek değişkenlik ile karakterize edilen rüzgâr hızının kaotik davranışının, günlük çözünürlükte deterministik tahmin üzerinde temel sınırlamalar getirdiğini doğrulamaktadır.

Anahtar Kelimeler

• İklim tahmini
• derin öğrenme
• rastgele arama optimizasyonu
• Wavelet-CNN-LSTM
• zaman serisi tahmini.

Comparative Analysis of Statistical and Deep Learning Models for Daily Climate Forecasting: A Case Study on the Delhi Dataset

Öz

As the impacts of climate change continue to intensify, accurate forecasting of meteorological parameters is of critical importance for many applications ranging from agricultural planning to disaster management. This study presents a systematic comparison of seven forecasting models – two baseline statistical (ARIMA, SARIMA) and five deep learning (LSTM, CNN, MLP, CNN-BiLSTM-Attention, and Wavelet-CNN-LSTM) – for daily prediction of mean temperature, humidity, wind speed, and mean pressure using Delhi climate data covering 2013–2017. Hyperparameters for all models are optimized via Random Search, and 10 independent runs with different random seeds are conducted to ensure statistical reliability. Model performance is evaluated using RMSE, MAE, MAPE, MASE, and R² metrics, while the Friedman test with Nemenyi post-hoc analysis is employed to assess statistical significance of performance differences. Results indicate that the Wavelet-CNN-LSTM model achieves the best overall ranking, followed by LSTM, MLP, and CNN, all of which performed significantly better than the ARIMA and SARIMA models. High prediction accuracy is obtained for temperature (R² = 0.922) and pressure (R² = 0.9), whereas wind speed remains the most challenging variable for all models (R² < 0.2) due to its inherently stochastic nature. Predictability analysis confirms that the chaotic behavior of wind speed, characterized by low autocorrelation and high variability, imposes fundamental limits on deterministic forecasting at daily resolution.

Anahtar Kelimeler

• Climate forecasting
• deep learning
• random search optimization
• Wavelet-CNN-LSTM
• time series prediction

Kaynakça

1. Hayati M, Mohebi Z. Application of Artificial Neural Networks for Temperature Forecasting. International Journal of Electrical and Computer Engineering 2007; 1: 654-658.
2. Dombaycı ÖA, Gölcü M. Daily means ambient temperature prediction using artificial neural network method: A case study of Turkey. Renewable Energy 2009; 34(4): 1158-1161. https://doi.org/10.1016/j.renene.2008.07.007
3. Hochreiter S, Schmidhuber J. Long Short-Term Memory. Neural Comput. 1997; 9(8): 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
4. Karevan Z, Suykens JAK. Transductive LSTM for time-series prediction: An application to weather forecasting. Neural Networks 2020; 125: 1-9. https://doi.org/10.1016/j.neunet.2019.12.030
5. Hou J, Wang Y, Zhou J, Tian Q. Prediction of hourly air temperature based on CNN–LSTM. Geomatics Nat. Hazards Risk 2022; 13(1): 1962-1986. https://doi.org/10.1080/19475705.2022.2102942
6. Guo Q, He Z, Wang Z. Monthly climate prediction using deep convolutional neural network and long short-term memory. Sci Rep 2024; 14, 17748. https://doi.org/10.1038/s41598-024-68906-6
7. Shrivastava VK, Shrivastava A, Sharma N, Mohanty SN, Pattanaik CR. Deep learning model for temperature prediction: A case study in New Delhi. Journal of Forecasting 2023; 42(6): 1445-1460. https://doi.org/10.1002/for.2966
8. Elshewey AM, Shams MY, Elhady AM, Shohieb SM, Abdelhamid AA, Ibrahim A, Tarek Z. A Novel WD-SARIMAX Model for Temperature Forecasting Using Daily Delhi Climate Dataset. Sustainability 2023; 15(1), 757. https://doi.org/10.3390/su15010757

9. Hanoon MS, Ahmed AN, Zaini N, Razzaq A, Kumar P, Sherif M, Sefelnasr A, El-Shafie A. Developing machine learning algorithms for meteorological temperature and humidity forecasting at Terengganu state in Malaysia. Sci Rep 2021; 11, 18935. https://doi.org/10.1038/s41598-021-96872-w
10. Bilgili M, Sahin B. Prediction of Long-term Monthly Temperature and Rainfall in Turkey. Energy Sources Part A 2009; 32(1): 60-71. https://doi.org/10.1080/15567030802467522
11. Hanifi S, Cammarono A, Zare-Behtash H. Advanced hyperparameter optimization of deep learning models for wind power prediction. Renewable Energy 2024; 221, 119700. https://doi.org/10.1016/j.renene.2023.119700
12. Wu Z, Luo G, Yang Z, Guo Y, Li K, Xue Y. A comprehensive review on deep learning approaches in wind forecasting applications. CAAI Trans. Intell. Technol. 2022; 7(2): 129-143. https://doi.org/10.1049/cit2.12076
13. Liu Y, Guan L, Hou C, Han H, Liu Z, Sun Y, Zheng M. Wind Power Short-Term Prediction Based on LSTM and Discrete Wavelet Transform. Applied Sciences 2019; 9(6), 1108. https://doi.org/10.3390/app9061108
14. Elshewey AM, Jamjoom MM, Alkhammash EH. An enhanced CNN with ResNet50 and LSTM deep learning forecasting model for climate change decision making. Sci Rep 2025; 15, 14372. https://doi.org/10.1038/s41598-025-97401-9
15. Daily Climate Time Series Dataset, Kaggle, dataset. [Online]. Available: https://www.kaggle.com/datasets/abbasi1214/daily-climate-time-series-dataset. Accessed: Dec. 15, 2025.
16. Box GEP, Jenkins GM, Reinsel GC, Ljung GM. Time Series Analysis: Forecasting and Control. 5th ed. Hoboken, NJ: John Wiley & Sons, 2015.
17. Lecun Y, Bottou L, Bengio Y, Haffner P. Gradient-based learning applied to document recognition. In Proceedings of the IEEE 1998; 86(11): 2278-2324. https://doi.org/10.1109/5.726791
18. Rumelhart D, Hinton G, Williams, R. Learning representations by back-propagating errors. Nature 1986; 323: 533-536. https://doi.org/10.1038/323533a0
19. Schuster M, Paliwal KK. Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing 1997; 45(11): 2673-2681. https://doi.org/10.1109/78.650093
20. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser L, Polosukhin I. Attention is all you need. In: 31st Conference on Neural Information Processing Systems (NIPS 2017); 2017; Long Beach, CA, USA.
21. Renaud O, Starck J-L, Murtagh F. Wavelet-based combined signal filtering and prediction. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics) 2005; 35(6): 1241-1251. https://doi.org/10.1109/TSMCB.2005.850182
22. Hyndman RJ, Koehler AB. Another look at measures of forecast accuracy. International Journal of Forecasting 2006; 22(4): 679-688. https://doi.org/10.1016/j.ijforecast.2006.03.001