New Formulation for Predicting Daily Reference Evapotranspiration (ET0) in the Mediterranean Region of Algeria Country:
This chapter aims to investigate the capabilities and usefulness of two new data-driven techniques: optimally pruned extreme learning machine (OPELM) and online sequential extreme learning machine (OSELM) newly applied and compared for predicting daily re
- PDF / 610,056 Bytes
- 19 Pages / 439.37 x 666.142 pts Page_size
- 85 Downloads / 175 Views
Contents 1 Introduction 2 Materials and Methods 2.1 Extreme Learning Machines 2.2 Online Sequential Extreme Learning Machine (OSELM)
S. Heddam (*) Laboratory of Research in Biodiversity Interaction Ecosystem and Biotechnology, Hydraulics Division, Agronomy Department, Faculty of Science, University 20 Août 1955, Skikda, Algeria e-mail: [email protected] O. Kisi School of Technology, Ilia State University, Tbilisi, Georgia e-mail: [email protected] A. Sebbar Soil and Hydraulics Laboratory, Hydraulics Department, Faculty of Engineering Sciences, University Badji-Mokhtar Annaba, Annaba, Algeria e-mail: [email protected] L. Houichi Department of Hydraulic, University of Batna 2, Batna, Algeria e-mail: [email protected] L. Djemili Research Laboratory of Natural Resources and Adjusting, Hydraulics Department, Faculty of Engineering Sciences, University Badji-Mokhtar Annaba, Annaba, Algeria e-mail: [email protected] Abdelazim Negm, Abdelkader Bouderbala, Haroun Chenchouni, and Damia Barcelo (eds.), Water Resources in Algeria - Part I: Assessment of Surface and Groundwater Resources, Hdb Env Chem, DOI 10.1007/698_2020_528, © Springer Nature Switzerland AG 2020
S. Heddam et al. 2.3 Optimally Pruned Extreme Learning Machine (OPELM) 2.4 Performances Criteria 3 Study Area and Data Description 4 Results 5 Discussions 6 Conclusion 7 Recommendations References
Abstract This chapter aims to investigate the capabilities and usefulness of two new data-driven techniques: optimally pruned extreme learning machine (OPELM) and online sequential extreme learning machine (OSELM) newly applied and compared for predicting daily reference evapotranspiration (ET0) in the Mediterranean region of Algeria. Using large data sets from east to west regions of Algeria, the models were developed using several well-known climatic variables as inputs: daily maximum and minimum air temperatures, wind speed, and relative humidity. The proposed models were compared using several well-known statistical indexes: root mean square error (RMSE), mean absolute error (MAE), and coefficient of correlation (R). The obtained results have shown that all the proposed models present high prediction accuracy and the OPELM models provide better overall performances compared to the OSELM models Keywords Algeria, Climatic variables, ET0, Extreme learning machine, Modelling, OPELM, OSELM
1 Introduction Nowadays, reference evapotranspiration (ET0) is one of the most important components of the hydrological cycle that have received great importance and has paid great attention by researcher’s worldwide [1]. ET0 is considered as an indicator of climate change and can be directly estimated from weather variables [2]. During the last decades, modelling ET0 using data-driven models has been broadly discussed and widely reported in the literature, and many models have been proposed. In-depth literature review demonstrates that generally the proposed models were based on the estimation of the ET0 at several time steps using several climatic variables as in
Data Loading...
