Performance Evaluation of Machine Learning Models in Predicting Dry and Wet Climatic Phases

Document Type: Regular Article


Department of Civil Engineering, University of Ibadan, Ibadan



Water resource and environmental engineers need accurate information in harnessing water for diverse uses, therefore it is expedient to accurately predict dry and wet climatic phases in order to ensure optimum water resource planning and management. This study examined the applicability of machine learning models for the prediction of extreme dry and wet conditions in Minna, North Central Nigeria. Recorded rainfall, maximum temperature, minimum temperature, relative humidity, wind speed, sunshine hours and estimated potential evapotranspiration were used as predictors in the machine learning models, while hydrological extremes estimated from standardized precipitation index (SPI) served as a response variable. The performance of Support Vector Machine (SVM) based on different kernel types and Artificial Neural Network (ANN) based on different network structures were assessed for the prediction of the different phases of the climate of the study area. The study showed that while normal meteorological conditions occurred for about 74.8% of the study period, 8.9%, 4.6% and 1.9% of this period were moderately wet, severely wet and extremely wet respectively, and 4.7%, 3.4% and 1.7% of the study period were moderately dry, severely dry and extremely dry respectively. Furthermore, SVM based on Radial Basis Kernel with a coefficient of determination of 0.64 outperformed other SVM types and ANN with two hidden layers; with of coefficient of determination 0.68 was found to perform better than ANN with single layers. Generally, ANN was found to have higher accuracy than SVM in predicting dry and wet climatic phases in North Central Nigeria.


Main Subjects

[1]       Wang M, Zhang L, Baddoo TD. Hydrological modeling in a semi-arid region using HEC-HMS. J Water Resour Hydraul Eng 2016;5:105–15.

[2]       Vairavamoorthy K, Gorantiwar SD, Pathirana A. Managing urban water supplies in developing countries – Climate change and water scarcity scenarios. Phys Chem Earth, Parts A/B/C 2008;33:330–9. doi:10.1016/j.pce.2008.02.008.

[3]       Okeke OC, Uzoh OF. Towards achieving sustainable water resources managemt in Nigeria. Glob J Geol Sci 2009;7.

[4]       Akpor OB, Muchie B. Environmental and public health implications of wastewater quality. African J Biotechnol 2011;10:2379–87.

[5]       Gayan L, Upul S. Forecasting Droughts using Artificial Neural Networks. Natl. Symp. Disaster Risk Reduct. Clim. Chang. Adapt., 2009.

[6]       Agbede OA, Aiyelokun OO. Establishment of a Stochastic Model for Sustainable Economic Flood Management in Yewa Sub-Basin, Southwest Nigeria. Civ Eng J 2016;2:646–55.

[7]       ConsortiumDolcine L, Prévil C, Brahm A, Er-Raji A, Qaimi A. Implementation of an Integrated Decision Support System for the Souss-Massa Watershed, Morocco, in UNESCO/IHP, Application of Satellite Remote Sensing to Support Water Resources Management in Africa: Results From the TIGER Initiative, IHP-VII Technical 2010.

[8]       Bhuiyan C. Various drought indices for monitoring drought condition in Aravalli terrain of India. Proc. XXth ISPRS Congr. Istanbul, Turkey, 2004, p. 12–23.

[9]       Tallaksen LM, Van Lanen HAJ. Hydrological drought: processes and estimation methods for streamflow and groundwater. vol. 48. Elsevier; 2004.

[10]     Goddard S, Harms SK, Reichenbach SE, Tadesse T, Waltman WJ. Geospatial decision support for drought risk management. Commun ACM 2003;46. doi:10.1145/602421.602442.

[11]     Ayanshola A, Sule B, Salami A. Evaluation of Willingness to Pay for Reliable and Sustainable household Water Use in Ilorin, Nigeria. Ethiop J Environ Stud Manag 2013;6:754. doi:10.4314/ejesm.v6i6.6S.

[12]     UN-Water Report. World Water Development Report 4. (2012). Managing Water Under Uncertainty and Risk (Vol. 1), Knowledge Base (Vol. 2) and Facing the Challenges (Vol. 3). Paris: UNESCO. 2012.

[13]     McKee TB, Doesken NJ, Kleist J. The relationship of drought frequency and duration to time scales. Proc. 8th Conf. Appl. Climatol., vol. 17, Boston; 1993, p. 179–83.

[14]     Musa JJ. Environmental Appraisal Environmental Appraisal Appraisal of Drainage System Drainage System Drainage System in Minna. Minna, Minna, Niger State Niger State Niger State-Nigeria Niger Niger Int J Appl Biol Res 2012;4:85–93.

[15]     UNDP/NISEPA. Niger state framework for integrated sustainable waste management. Niger State Strategic Waste Management Framework (Unpublished) 2009.

[16]     Ishiaku I, Emigilati MA, Kuta GI, Usman BY, Hassan AB. An Assessment of Alternative Water Source for Domestic Used in Minna Metropolis, Niger State, Nigeria. J Environ Earth Sci 2014;4:19–23.

[17]     Doorenbosm J, Pruitt WO. Crop water requirement, irrigation and drainage. Pap No 1977.

[18]     Jennifer MJ. Faculty Investigator Evaluation of reference evapotranspiration methodologies and afsirs crop water use simulation model (Final Report). Department of Civil and Coastal Engineering University of Florida, Gainesville, Florida 2001.

[19]     Janaki B. Hydrological Modeling of a Typical Ungauged Basin Of Odisha 2013.

[20]     Saracoglu Ö. An Artificial Neural Network Approach for the Prediction of Absorption Measurements of an Evanescent Field Fiber Sensor. Sensors 2008;8:1585–94. doi:10.3390/s8031585.

[21]     Zahraie B, Nasseri M. Basin scale meteorological drought forecasting using support vector machine (SVM). Int. Conf. drought Manag. Strateg. Arid semi Arid Reg. Muscat, Oman, December, 2011.

[22]     Chen S-T. Mining Informative Hydrologic Data by Using Support Vector. Entropy 2015;17:1023–41. doi:10.3390/e17031023.

[23]     Granata F, Gargano R, de Marinis G. Support Vector Regression for Rainfall-Runoff Modeling in Urban Drainage: A Comparison with the EPA’s Storm Water Management Model. Water 2016;8:69. doi:10.3390/w8030069.

[24]     Du J, Liu Y, Yu Y, Yan W. A Prediction of Precipitation Data Based on Support Vector Machine and Particle Swarm Optimization (PSO-SVM) Algorithms. Algorithms 2017;10:57. doi:10.3390/a10020057.

[25]     Sahai AK, Soman MK, Satyan V. All India summer monsoon rainfall prediction using an artificial neural network. Clim Dyn 2000;16:291–302. doi:10.1007/s003820050328.

[26]     Toth E, Brath A, Montanari A. Comparison of short-term rainfall prediction models for real-time flood forecasting. J Hydrol 2000;239:132–47. doi:10.1016/S0022-1694(00)00344-9.

[27]     Somvanshi VK, Pandey OP, Agrawal PK, Kalanker N V, Prakash MR, Chand R. Modeling and prediction of rainfall using artificial neural network and ARIMA techniques. J Ind Geophys Union 2006;10:141–51.

[28]     Karamouz M, Rasouli K, Nazif S. Development of a Hybrid Index for Drought Prediction: Case Study. J Hydrol Eng 2009;14:617–27. doi:10.1061/(ASCE)HE.1943-5584.0000022.

[29]     Khalili. Rainfall prediction using artificial neural networks. Thesis of Master degree of Irrigation and drainage engineering. Faculty of Agriculture, Ferdowsi university of Mashhad (In Persian).2008. (6) (PDF) Daily Rainfall Forecasting for Mashhad Synoptic Stati. 2009.

[30]     Geetha G, Selvaraj RS. Prediction of monthly rainfall in Chennai using back propagation neural network model. Int J Eng Sci Technol 2011;3.

[31]     Belayneh A, Adamowski J, Khalil B. Short-term SPI drought forecasting in the Awash River Basin in Ethiopia using wavelet transforms and machine learning methods. Sustain Water Resour Manag 2016;2:87–101. doi:10.1007/s40899-015-0040-5.

[32]     Lu K, Wang L. A Novel Nonlinear Combination Model Based on Support Vector Machine for Rainfall Prediction. 2011 Fourth Int. Jt. Conf. Comput. Sci. Optim., IEEE; 2011, p. 1343–6. doi:10.1109/CSO.2011.50.

[33]     Nayak MA, Ghosh S. Prediction of extreme rainfall event using weather pattern recognition and support vector machine classifier. Theor Appl Climatol 2013;114:583–603. doi:10.1007/s00704-013-0867-3.

[34]     Ortiz-García EG, Salcedo-Sanz S, Casanova-Mateo C. Accurate precipitation prediction with support vector classifiers: A study including novel predictive variables and observational data. Atmos Res 2014;139:128–36. doi:10.1016/j.atmosres.2014.01.012.

[35]     Sánchez-Monedero J, Salcedo-Sanz S, Gutiérrez PA, Casanova-Mateo C, Hervás-Martínez C. Simultaneous modelling of rainfall occurrence and amount using a hierarchical nominal–ordinal support vector classifier. Eng Appl Artif Intell 2014;34:199–207. doi:10.1016/j.engappai.2014.05.016.

[36]     Sehad M, Lazri M, Ameur S. Novel SVM-based technique to improve rainfall estimation over the Mediterranean region (north of Algeria) using the multispectral MSG SEVIRI imagery. Adv Sp Res 2017;59:1381–94. doi:10.1016/j.asr.2016.11.042.

[37]     Young C-C, Liu W-C, Wu M-C. A physically based and machine learning hybrid approach for accurate rainfall-runoff modeling during extreme typhoon events. Appl Soft Comput 2017;53:205–16. doi:10.1016/j.asoc.2016.12.052.

[38]     Wu CL, Chau KW. Rainfall–runoff modeling using artificial neural network coupled with singular spectrum analysis. J Hydrol 2011;399:394–409. doi:10.1016/j.jhydrol.2011.01.017.

[39]     Deka PC, Prahlada R. Discrete wavelet neural network approach in significant wave height forecasting for multistep lead time. Ocean Eng 2012;43:32–42. doi:10.1016/j.oceaneng.2012.01.017.

[40]     Oluwatobi A, Gbenga O, Oluwafunbi F. An Artificial Intelligence Based Drought Predictions In Part Of The Tropics. J Urban Environ Eng 2017;11:165–73.