Data Pre-Processing Effects on the Artificial Neural Network Performance to Predict Monthly Rainfall (Case Study: Abadeh County)

Document Type : Research Article


1 Department of Water Engineering, Faculty of Agriculture, Fasa University, Iran

2 MSC Student of Irrigation and Drainage, Department of Water Engineering, Faculty of Agriculture, Fasa University, Iran

3 Department of Information Technology, Faculty of Engineering, Fasa University, Fasa, Iran


Since many time series are not normal, it is required to normalize data by transformation functions prior to any analysis and modeling. In this study, the next month rainfall of Abadeh County station was predicted using the average monthly rainfall, minimum and maximum temperatures and minimum and maximum humidity as inputs of MLP network, both normally and raw, at period 1976 to 2013. After scrutiny the existence or nonexistence of missing and outlier data, meteorological data were normalized using three normalization methods: minimum-maximum, rank normalization and z- score. After obtaining the best network structure using try and error for each method, the minimum-maximum method with a three-layer network structure and 13 number of hidden layers of neurons chose as the best method with R=0.92 and MSE=0.13 compared to other methods. Also comparing the performance of the network in using raw and Pre-Processed data showed that Pre-Processing the data improved greatly network performance. The results of the sensitivity analysis showed the maximum sensitivity of model to remove maximum humidity parameter, and the second the maximum temperature had the greatest impact on precipitation forecast. Also comparing the performance of the network with the different numbers of inputs indicated that network with two inputs including minimum temperature and minimum humidity had good results (MSE = 0.13) compare with five inputs.


Main Subjects


    1. Lookzadeh S. Evaluation of some methods for reconstruction of rainfall data in Alborz region, MSc, thesis, Tehran University,2004; P. 96. [In Persian].
    2. Kohzadi N, Boyd M, Kaastra I, Kermanshahi B, Scuse D. Neural networks for forecasting: an introduction. Canadian Journal of Agricultural Economics. 1995; 43: 463-474.
    3. Naghdi R, Shayannejhad M, Sadatinejad S.J. Comparison of different methods for estimating of monthly discharge missing data in Grand Karoon River Basin. Journal of Watershed Management Research. 2010; 1(1): 59-73. [In Persian].
    4. Mohammadi Takami M. The methods of data processing and pattern recognition. Faculty of electrical engineering. Khaje Nasireddin Toosi University. 2005.[In Persian].
    5. Nazeri Tahrudi M, Khalili K, Abbaszade Afshar M, Nazeri Tahrudi Z. Compared to the normal mechanism becomes the normal monthly rainfall data from different regions of Iran. Journal of Water and soil. 2014; 28 (2): 365- 372. [In Persian].
    6. Adab H, Fallah Ghalhari Gh, Mirzabayati R. Evaluation of interpolation methods of Kriging and linear regression based on the DEM to mapping annual rainfall in Khorasan Razavi province. Geomatics Conference. Tehran. National Cartographic Center, Iran. 2008.[In Persian].
    7. Hamidi R, Emamgholizade S. Stochastic modeling of Maroon River annual discharge using ARMA model. The first National Conference of Applied Research of Iran Water Resources. Tehran. 2009. [In Persian].
    8. Shafiei M, Ghahraman B, Ansari H, Sharifi M. B. Stochastic simulation of drought severity based on Palmer Index. Journal of Water and Irrigation Management. 2011; 1 (1).1-13. [In Persian].
    9. Saghafian B, Razmkhah H, Ghermezcheshmeh B. Investigation of regional variations in annual rainfall by geostatistical methods, case study: Fars Province. Journal of Water Resources Engineering.2011; 4 (9): 29-38. [In Persian].
    10. Ahmadi F, Dinpajouh Y, Fakherifard A, Khalili K. Modeling of river discharge using time series linear models (case study: Barandoozchay River). The first national conference of Strategies to achieve sustainable development in agriculture, natural resources and the environment. Tehran. 2012.[In Persian].
    11. Nikmanesh M, Taleb Bidokhti N. Comparison of wavelet theory and time series in modeling of monthly rainfall of Saadatshahr and Arsenjan in Fars Province. Journal of Natural Geography.2012; 5 (16): 1-10. [In Persian].
    12. Nazeri Tahrudi M, Khalili K, Ahmadi F, Nazeri Tahrudi Z. Modeling of temperature using periodic ARMA model (case study: Kerman Synoptic Station). Conference of applied researches in science and engineering. Tehran. 2012.[In Persian].
    13. Nazeri Tahrudi M, Ahmadi F, Khalili K, Nazeri Tahrudi Z. Application of SAMS software to modeling of future climate of Kordestan Province (case study: Sanandaj Synoptic Station). The first conference of semi-arid regions hydrology. Sanandaj. 2013.[In Persian].
    14. Nazeri Tahrudi M, Khalili K, Nazeri Tahrudi Z, Shahnazi M. Evaluation of ARIMA and PARMA models in modeling and forecasting maximum wind speed (case study: Bandar Abbas Synoptic Station). National conference of applied researches in science and engineering. Takestan. 2013.[In Persian].
    15. Azadi S, Sepaskhah A. R. Annual precipitation forecast for west, southwest, and south provinces of Iran using artificial neural networks. Theoretical and Applied Climatology. 2012; 109(1): 175.
    16. Leahy P, Kiely G, Corcoran G. Structural optimisation and input selection of an artificial neural network for river level prediction. J Hydrol,. 2008; 355:192–201.
    17. Tabari H, Marofi S, Sabziparvar A.A. Estimation of daily pan evaporation using artificial neural network and multivariate non-linear regression. Irrigation Science. 2010; 28: 399-406.
    18. Kim IS, Son JS, Park CE, Kim IJ, Kim HH. An investigation into an intelligent system for predicting bead geometry in Arc welding process. Int. J. of Materials Processing Technology. 2005; 159: 113–118.
    19. Nasri M, Modarrs R, Dastoorani MT. Validation of ANN model of rainfall- runoff relationship in Zaynderood Dam Watershed. Journal of Watershed Researches.2010; 88: 17-26. [In Persian].
    20. Hagan MT, Menhaj M. Training feed-forward networks with the Marquardt algorithm. IEEE Transactions on Neural Networks. 1994; 5(6): 989-993.
    21. Aksoy S, Haralick M. Feature Normalization and Likelihood-based Similarity Measures for Image Retrieval. Intelligent Systems Laboratory, Department of Electrical Engineering, University of Washington, Seattle, WA 98195-2500, U.S.A. 2000.
    22. Nawi NM, Atomi WH, Rehman MZ. The Effect of Data Pre-Processing on Optimized Training of Artificial Neural Networks. Procedia Technology. 2003; 11: 32 – 39.
    23. Willmotte CJ. Some comments on the evaluation of model performance. American Meteorological Society,1982; 63:1309-1313.
    24. Haliban AH, Darand M. Forecasting rainfall using ANN. Journal of Applied Researches of Geographic Sciences. 2012; 12 (26): 47-63. [In Persian].
    25. Amiri MJ, Abedi-Koupai J, Eslamian SS, Mousavi SF, Hasheminejad H. Modeling Pb(II) adsorption from aqueous solution by ostrich bone ash using adaptive neural-based fuzzy inference system, J. Environ. Sci. Health, Part A 48. 2013; 543–558.
    26. Erfanian M, Ansari H, Alizade A. Forecasting monthly rainfall and average temperature using remote link templates with ANN (case study: Mashhad Synoptic Station). Geographical Studies of Arid Regions.2013; 3 (11): 53-73.[In Persian].
    27. Golkar F, Farahmand A, Farahmand F. Evaluation of ANN application in prediction of Shiraz rainfall. National conference of Water Crisis Management. Islamic Azad University of Marvdasht. 2009.[In Persian].


Volume 4, Issue 1
March 2017
Pages 29-37
  • Receive Date: 10 November 2016
  • Revise Date: 26 December 2016
  • Accept Date: 09 January 2017
  • First Publish Date: 21 March 2017