Derin sinir ağları modeli ile standardize yağış indeksi tahmini

Year 2022, Volume: 11 Issue: 4, 1006 - 1024, 14.10.2022

Levent Latifoğlu

https://doi.org/10.28948/ngumuh.1145279

Abstract

Kuraklık, yaşamı doğrudan etkileyen ve çok çeşitli olumsuz etkileri olan doğal bir afettir. Kuraklığı tahmin etmek üzere farklı kuraklık indeksleri kullanılmaktadır. Bu indekslerden en yaygın olarak kullanılanlardan biri de Standardize Yağış İndeksidir (SYİ). Gerçekleştirilen çalışmada Türkiye’ye ait Rize, Konya ve Şanlıurfa illerinin 3, 6, 9 ve 12 aylık SYİ verileri 1-3 ileri zamanlı olarak tahmin edilmiştir. Tahmin çalışmasını gerçekleştirmek üzere Uzun Kısa Süreli Bellek Ağları (Long Short Term Memory Networks, LSTM) ve Çift Yönlü Uzun Kısa Süreli Bellek Ağlarından (Bidirectional Long Short Term Memory Networks, biLSTM) oluşan Derin Sinir Ağları modelleri geliştirilmiştir. Tahmin performansını değerlendirmek üzere Ortalama Mutlak Hata (Mean Absolute Error, MAE), Ortalama Karesel Hata (Mean Squared Error, MSE), Korelasyon katsayısı (Correlation Coefficient, R) ve Belirlilik katsayısı (Determination Coefficient, R2) parametreleri kullanılmıştır. Elde edilen sonuçlar tahmin parametreleri ve saçılma grafikleri ile değerlendirildiğinde biLSTM içeren derin sinir ağları modelinin performansının oldukça iyi olduğu ve 3 ileri zamanlı tahminde bile yüksek korelasyona sahip sonuçlar elde edilebileceğini göstermiştir.

Keywords

Kuraklık, Standardize yağış indeksi, Derin sinir ağları, LSTM, biLSTM

Standard precipitation index estimation with deep neural network model

Abstract

Drought is a natural disaster that directly affects life and has a wide variety of negative effects. Different drought indices are used to predict drought. One of the most widely used of these indices is the Standardized Precipitation Index (SPI). In this study, the 3, 6, 9 and 12-month SPI data of Rize, Konya and Şanlıurfa provinces of Turkey were estimated 1-3 forward time. Deep Neural Networks models consisting of Long Short Term Memory Networks (LSTM) and Bidirectional Long Short Term Memory Networks (biLSTM) have been developed to perform the prediction study. The Mean Absolute Error (MAE), Mean Squared Error (MSE), Correlation Coefficient, R and Determination Coefficient (R2) parameters were used to evaluate the forecasting performance. When the results obtained are evaluated with the performance parameters and scatter plots, it has been shown that the performance of the deep neural network model with biLSTM is quite good and that high correlation results can be obtained even in 3 forward-time predictions.

Keywords

Drought, Standard precipitation index, Deep neural networks, LSTM, biLSTM

References

• A. Dai, Drought under global warming: a review, Wiley Interdisciplinary Reviews: Climate Change, 2, 45-65, 2011. https://doi.org/10.1002/wcc.81.
• T. D. Delbiso, C. Altare, J. M. Rodriguez-Llanes, S. Doocy, and D. Guha-Sapir, Drought and child mortality: a meta-analysis of small-scale surveys from Ethiopia, Scientific reports, 7(1), 1-8, 2017. https://doi.org/10.1038/s41598-017-02271-5
• S. E. Nicholson, C. J. Tucker, and M. B. Ba, Desertification, drought, and surface vegetation: An example from the West African Sahel, Bulletin of the American Meteorological Society, 79(5), 815-830, 1998.https://doi.org/10.1175/15200477(1998)079<0815:DDASVA>2.0.CO;2
• A. Grainger, S. Smith, V. R. Squires, and E. P. Glenn, Desertification, and climate change: the case for greater convergence, Mitigation and Adaptation Strategies for Global Change, 5(4), 361-377, 2000. https://doi.org/10.3354/cr011051
• C. H. Chung and J. D. Salas, Drought occurrence probabilities and risks of dependent hydrologic processes. Journal of Hydrologic Engineering, 5(3), 259-268, 2000. https://doi.org/10.1061/(ASCE)1084-0699(2000)5:3(259)
• J. H. Stagge, L. M. Tallaksen, C. Y. Xu and H. A. Van Lanen, Standardized precipitation-evapotranspiration index (SPEI): Sensitivity to potential evapotranspiration model and parameters. In Hydrology in a changing world, 363, 367-373, 2014.
• W. M. Alley, The Palmer drought severity index: limitations and assumptions. Journal of Applied Meteorology and Climatology, 23(7), 1100-1109, 1984.https://doi.org/10.1175/15200450(1984)023<1100:TPDSIL>2.0.CO;2
• D. W. Kim, H.R. Byun, ve K. S. Choi, Evaluation, modification, and application of the Effective Drought Index to 200-Year drought climatology of Seoul, Korea. Journal of hydrology, 378(1-2), 1-12, 2009. https://doi.org/10.1016/j.jhydrol.2009.08.021
• G. Tsakiris, D. Pangalou ve H. Vangelis, Regional drought assessment based on the Reconnaissance Drought Index (RDI), Water resources management, 21(5), 821-833, 2007. https://doi.org/10.1007/s11269-006-9105-4
• N. B. Guttman, Comparing the palmer drought index and the standardized precipitation index 1, JAWRA Journal of the American Water Resources Association, 34(1), 113-121, 1998. https://doi.org/10.1111/j.1752-1688.1998.tb05964.x
• M. Svoboda, M. Hayes, and D. Wood, Standardized precipitation index: user guide, 2012.
• S. Sırdaş, Z. Sen, Spatio-temporal drought analysis in the Trakya region, Turkey, Hydrological Sciences Journal, 48(5), 809-820, 2003. https://doi.org/ 10.1623/hysj.48.5.809.51458
• N. B. Guttman, Accepting the standardized precipitation index: a calculation algorithm 1. JAWRA Journal of the American Water Resources Association, 35(2), 311-322, 1999. https://doi.org/10.1111/j.1752-1688.1999.tb03592.x
• A. Ahani, M. Shourian, ve P. Rahimi Rad, Performance assessment of the linear, nonlinear and nonparametric data driven models in river flow forecasting. Water resources management, 32(2), 383-399, 2018. https://doi.org/10.1007/s11269-017-1792-5
• D. P. Solomatine and A. Ostfeld, Data-driven modelling: some past experiences and new approaches, Journal of hydroinformatics, 10(1), 3-22, 2008. https://doi.org/10.2166/hydro.2008.015
• A. Belayneh, J. Adamowski, Standard precipitation index drought forecasting using neural networks, wavelet neural networks, and support vector regression. Applied computational intelligence and soft computing, 6, 6, 2012. https://doi.org/ 10.1155/2012/794061
• H. Citakoglu, Y. Ozeren, ve O. Coskun, Short Time Drought Estimation of Sakarya Basin Station With Wavelet Model‒Adaptive Neuro‒Fuzzy Inference System, IWW'2019: International Conference on Image Processing, Wavelet and Applications, sayfa 190-197, Kocaeli, Türkiye, 18 - 20 Ekim 2019.
• A. Dikshit, B. Pradhan, ve M. Santosh, Artificial neural networks in drought prediction in the 21st century–A scientometric analysis, Applied Soft Computing, 114, 2022. https://doi.org/10.1016/j.asoc.2021.108080
• T. Fischer ve C. Krauss, Deep learning with long short-term memory networks for financial market predictions. European journal of operational research, 270(2), 654-669, 2018. https://doi.org/ 10.1016/j.ejor.2017.11.054
• D. K. Roy, Long short-term memory networks to predict one-step ahead reference evapotranspiration in a subtropical climatic zone, Environmental Processes, 8(2), 911-941, 2021. https://doi.org/10.1007/s40710-021-00512-4
• S. Ghimire, R.C. Deo, N. Raj, ve J. Mi, Deep solar radiation forecasting with convolutional neural network and long short-term memory network algorithms. Applied Energy, 253, 2019. https://doi.org/10.1016/j.apenergy.2019.113541
• NASA, 2020b, National Aeronautics and Space Administration (NASA), Langley Research Center (LaRC), POWER Data Access Viewer, Single Point Data Access (2020) online resource, https://power.larc.nasa.gov/data-access-viewer, Accessed 01 May 2022
• M. R. Al-Kilani, M. Rahbeh, J. Al-Bakri, T. Tadesse, ve C. Knutson, Evaluation of Remotely Sensed Precipitation Estimates from the NASA POWER Project for Drought Detection over Jordan. Earth Systems and Environment, 5(3), 561-573, 2021. https://doi.org/10.1007/s41748-021-00245-2
• Turkish State Meteorological Service, TURKIYE, https://mgm.gov.tr/veridegerlendirme/aylik-normal-yagis-dagilimi.aspx, Accessed 21 September 2022
• T. B. McKee, N. J. Doesken ve J. Kleist, The relationship of drought frequency and duration to time scales, In Proceedings of the 8th Conference on Applied Climatology, 17( 22), 179-183, 1993.
• G. Tsakiris, H. Vangelis, Towards a drought watch system based on spatial SPI. Water resources management, 18(1), 1-12, 2004. https://doi.org/ 10.1023/B:WARM.0000015410.47014.a4
• D.C. Edwards, ve T. B. McKee, Characteristics of 20th Century Drought in the United States at Multiple Times Scales. Atmospheric Science Paper, 634, 1-30, 1997.
• F.K. Sönmez, A. U. Koemuescue, A. Erkan ve E. Turgu, An analysis of spatial and temporal dimension of drought vulnerability in Turkey using the standardized precipitation index, Natural Hazards, 35(2), 243-264, 2005. https://doi.org/10.1007/s11069-004-5704-7
• P. Angelidis, F. Maris, N. Kotsovinos, ve V. Hrissanthou, Computation of drought index SPI with alternative distribution functions. Water resources management, 26(9), 2453-2473, 2012. https://doi.org/10.1007/s11269-012-0026-0
• D. Tigkas, H. Vangelis, G. Tsakiris, DrinC: Ga software for drought analysis based on drought indices. Earth Science Informatics, 8(3), 697-709, 2015. http://dx.doi.org/10.1007/s12145-014-0178-y
• D. Tigkas, H. Vangelis, N. Proutsos, G. Tsakiris, Incorporating aSPI and eRDI in Drought Indices Calculator (DrinC) software for agricultural drought characterisation and monitoring. Hydrology, 9(6), 100, 2022. https://doi.org/10.3390/hydrology9060100
• S. Hochreiter ve J. Schmidhuber, Long short-term memory. Neural computation, 9(8), 1735-1780, 1997. https://doi.org/10.1162/neco.1997.9.8.1735
• S. Zhang, D. Zheng, X. Hu ve M. Yang, Bidirectional long short-term memory networks for relation classification, In Proceedings of the 29th Pacific Asia conference on language, information and computation, 73-78, 2015.