Utilizing Machine Learning and Deep Learning for Precise Intensity-Duration- Frequency (IDF) Curve Predictions

Authors

Sheeraz M. Ameen, Petroleum Technology-Petrochemical Department, Koya Technical Institute, Erbil Polytechnic University, Erbil, Iraq.
Shuokr Qarani Aziz, Department of Civil Engineering, College of Engineering, Salahaddin University-Erbil, Erbil, Kurdistan Region, Iraq
Anwer Hazim Dawood, Department of Geotechnical Engineering, Faculty of Engineering, Koya University, Koya, Kurdistan Region, IraqFollow
Azhin Tahir Sabir, Department of Software Engineering, Faculty of Engineering, Koya University, Koya, Kurdistan Region, Iraq.
Dara Muhammad Hawez, Department of Civil Engineering, University of Raparin, Ranya, Sulaymani, Kurdistan Region, Iraq

How to Cite This Article

Ameen, Sheeraz M.; Aziz, Shuokr Qarani; Dawood, Anwer Hazim; Sabir, Azhin Tahir; and Hawez, Dara Muhammad (2025) "Utilizing Machine Learning and Deep Learning for Precise Intensity-Duration- Frequency (IDF) Curve Predictions," Polytechnic Journal: Vol. 15: Iss. 1, Article 3.
DOI: https://doi.org/10.59341/2707-7799.1848

Document Type

Original Article

Abstract

Intensity-Duration-Frequency (IDF) curves are crucial for the design and management of engineering infrastructure, including storm sewers, retention ponds, dams, and flood mitigation systems. This study adopts a comparative approach to estimate IDF curves using a combination of traditional statistical methods, machine learning techniques, and advanced deep learning models. Rainfall data from Koya City, Iraq (2005–2022), was used, with the 2005–2015 period for training and 2016–2022 for validation. The models evaluated include the Gumbel Distribution, Linear Regression, Support Vector Regression (SVR), and Recurrent Neural Networks (RNN) with Long Short-Term Memory (LSTM), assessed based on three metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R²). Among these, the RNN-LSTM model demonstrated the lowest RMSE (1.44 mm/hr), lowest MAE (0.81 mm/hr), and highest R² (0.99), outperforming the Gumbel Distribution (RMSE: 9.13 mm/hr), Linear Regression (RMSE: 10.76 mm/hr), and SVR (RMSE: 6.19 mm/hr). This establishes RNN-LSTM as the most reliable approach for IDF curve prediction. Leveraging the RNN-LSTM model, rainfall trends for 2023–2043 were forecasted, revealing an expected increase in short-duration, high-intensity rainfall events, heightening flood risks, and emphasizing the need for adaptive stormwater management strategies. The findings underscore the significant potential of deep learning models like RNN-LSTM in enhancing IDF curve predictions and guiding the development of resilient hydraulic infrastructure, particularly in regions like Koya City, where complex topography exacerbates flood challenges during intense rainfall events.

Receive Date

01/10/2024

Revise Date

22/12/2024

Accept Date

12/01/2025

References

[1] Chow V Te, Maidment DR, Mays LW. Applied hydrology. New York City, New York, U.S.: McGraw Hill, Inc; 1988.

[2] Kareem AT, Ahmed AJ, Mohammed SH. Enhancing IDF curve prediction accuracy using deep learning techniques: A case study of the Kurdistan region. Water Resour Manag 2022;36(4):1231e47.

[3] Dawood AH, Mawlood DK. Derivation of the Rainfall Intensity- Duration- Frequency Curve for Erbil city, Using SCS II Method. Zanco J Pure Appl Sci 2023;35(SpA). https:// doi.org/10.21271/zjpas.35.spa.1.

[4] Lionello P, Malanotte-rizzoli P, Boscolo R. MedCLIVAR Mediterranean CLImate VARiability. Earth Environ Sci 2006; (January).

[5] Taylor P. Generalized Gumbel distribution. J Appl Stat 2020; 37(January 2010):171e9. https://doi.org/10.1080/0266476080269 8995.

[6] Griffis VW, Asce M, Stedinger JR, Asce M. Log-Pearson Type 3 Distribution and Its Application in Flood Frequency Analysis I: Distribution Characteristics. J Hydrol Eng 2007; (October):482e91.

[7] Eregno FE, Xu C, Kitterod N. Modeling hydrological impacts of climate change in different climatic zones. Int J Clim Change Strat Manag 2013;5(July):344e65. https://doi.org/ 10.1108/IJCCSM-04-2012-0024.

[8] Ewea HA, Elfeki AM, Al-amri NS. Development of intensity e duration e frequency curves for the Kingdom of Saudi Arabia. 5705. Oxfordshire, United Kingdom: Taylor and Francis, Milton Park, Abingdon-on-Thames; 2017. https:// doi.org/10.1080/19475705.2016.1250113.

[9] Yu Q, Tolson BA, Shen H, Han M, Mai J, Lin J. Enhancing LSTM-based streamflow prediction with a spatially distributed approach. Hydrol Earth Syst Sci 2023;1(November): 1e23.

[10] Sidek LM, Basri H, Sammen SS. Machine learning techniques for flood forecasting Uncorrected Proof Machine learning techniques for fl ood forecasting. J Hydroinf 2024; 1(March). https://doi.org/10.2166/hydro.2024.208.

[11] Elsebaie IH. Developing rainfall intensity e duration e frequency relationship for two regions in Saudi Arabia, vol. 1932. Riyadh, Saudi Arabia Kingdom: King Saud University e Engineering Sciences; 2012. p. 131e40. https://doi.org/ 10.1016/j.jksues.2011.06.001.

[12] Le Kim KN, Quoc TN, Thanh PN. Deep Learning Method For Load Forecast In Smart Solar System Based Long ShortTerm Memory. In: 2024 8th international artificial intelligence and Data processing symposium (IDAP), malatya, Turkiye. New York, USA: IEEE, The Institute of Electrical and Electronics Engineers (IEEE); 2024. p. 1e5.

[13] Liu W-B, Bui T-M, Quoc TN, Thi VP, Cho M-Y, Da TN, et al. Failure Classified Method for Diesel Generators Based Long Short-Term Memory Approach. In: 2024 international conference on system science and engineering (ICSSE), hsinchu, Taiwan. New York, USA: IEEE, The Institute of Electrical and Electronics Engineers (IEEE); 2024. p. 1e6.

[14] Da TN, Yimin L, Peng CL, Cho M-Y, Le Kim KN, Thanh PN. Short-term Solar Power Prediction using Long Short-Term Memory in Solar Plant with Deep Learning Machine. In: 2022 6th international conference on green technology and sustainable development (GTSD), Nha Trang city, Vietnam. New York, USA: IEEE, The Institute of Electrical and Electronics Engineers (IEEE); 2022. p. 651e6.

[15] Azi SQ, Mizzouri NS, Ahmed KO, Hamaamin YA, Hawez DM. Intensity-Duration-Frequency (IDF) Modelling for Kurdistan Region Provinces- Intensity-Duration-Frequency (IDF) Modelling for Kurdistan Region ProvincesIraq. Eur Chem Bull 2023;12(10):7177e85. https://doi.org/ 10.48047/ecb/2023.12.10.534.

[16] Aziz SQ, Saleh SM, Muhammad SH, Ismael SO, Ahmed BM. Flood disaster in Erbil city : problems and solutions flood disaster in Erbil city : problems and solutions. Environmental Protection Research, December. Singapore 238875: Universal Wiser Publisher; 2023. https://doi.org/10.37256/epr.3220 232993.

[17] Dawood AH, Mawlood DK. Flood Risk Analysis in Potential Points, Zerin City in Erbil as a Case Study. Iraqi Geol J 2023;54:125e37. https://doi.org/10.46717/igj.56.1E.13ms-2023- 5-23.

[18] Nhat LT, Tachikawa Y, Takara K. Estimation of rainfall intensity duration frequency distribution in the mountainous regions of Vietnam, vol. 49. Kyoto, 606-8501, Japan: Annuals of Disaster Prevention Research Institute, Kyoto University; 2006. p. 73e82.

[19] Eckersten H. Modeling hydrological impacts of climate change using statistical methods. J Hydrol 2016;542:213e26.

[20] Dawood AH, Mawlood DK, Al-ansari N. Flood Modeling on Koya Catchment Area Using Hyfran , Web Map Service , and HEC-RAS Software 2021:107e11. https://doi.org/10.14500/ aro.10824.

[21] Du B, Li W, Tan X, Ye J, Chen W, Sun L. Development of load-temporal model to predict the further mechanical behaviors of tunnel structure under various boundary conditions. Tunn Undergr Space Technol 2021;116:104077.