REPLY to Discussion of “Precipitation Forecasting Using Wavelet-Genetic Programming and Wavelet-Neuro-Fuzzy Conjunction Models”

Abstract

The authors would like to express their gratitude for the interest shown by the discussers in the paper and for their comments on the subject. We have tried to clarify all the points raised by them in this closure. The discussers say in their discussion that “... the computational mechanisms used in this study are all relatively new and therefore require more detailed reporting throughout all stages of the modelling process, accompanied by better justification of decisions influencing model development.” The authors do not agree with this because this method was used in lots of study by various researchers before (Wang and Ding 2003; Anctil and Tape 2004; Kisi 2008; Adamowski 2008; Nourani et al. 2009; Partal and Cigizoglu 2009; Adamowski and Sun 2010; Adamowski and Chan 2011) and detailed methodology can be found in many related textbooks (e.g. Burrus et al. 1998; Welnut 2004). Partal and Kisi (2007) arbitrarily used ten resolution levels in their study. We employed three resolution levels following the suggestion of Wang and Ding (2003). According to the Wang and Ding (2003), in general there is log (n) resolution level, where n is the length of the time series. In our study, 2550 daily data were used in training period. This (log2550) approximately gives three resolution levels. We agree with the discussers that Partal and Kisi (2007) used same data and found different results in their study. They divided data into two sets, training (12 years daily data) and testing (3 years daily data). Shiri and Kisi (2011) divided data into three parts, training, test and validation. They used 7 years daily data for training, 4 years for testing and remaining 4 years for validation. The authors were used db5 wavelet in their study because this is one of the commonly used wavelets (Wei et al. 2011). The time series A+D2+D3 was obtained by removing D1 Water Resour Manage (2012) 26:3663–3665 DOI 10.1007/s11269-012-0060-y