Artificial neural network technique for rainfall temporal distribu-tion simulation (‍‍Case study: Kechik region)

در طي چند دهه اخير شبكه عصبي به عنوان ابزاري كارآمد در شبيه سازي بارش بكار گرفته شده است. محققين مطالعات متعددي براي تعيين شدت ميزان بارش بر اساس پارامترهاي هيدرو اقليمي در سطح حوزه هاي آبخيز انجام داده اند. در تحقيق حاضر، اطلاعات ايستگاه ثبات هواشناسي كچيك در بازه هاي زماني ده قيقه اي شامل دما، تبخير، سرعت باد، فشار هوا و رطوبت بعنوان ورودي مدل و بارش بعنوان خروجي مدل مورد استفاده قرار گرفت. براي شبيه سازي توزيع زماني بارش، شبكه پرسپترون تك لايه انتخاب گرديد. بر اساس نتايج، بهترين آرايش شبكه، الگوريتم LM بدست آمد و اين تاييدي بر كارايي بالاي شبكه عصبي با الگوريتم LM نسبت به الگوريتم هاي CG و GDXدر شبيه سازي توزيع زماني بارش مي باشد. همچنين، دما، تبخير و رطوبت مهمترين ورودي ها براي شبيه سازي بارش مي باشند.

Artificial neural networks (ANNs) have become one of the most promising tools for rainfall simulation since a few years ago. However, most of the researchers have focused on rainfall intensity records as well as on watersheds, which generally are utilized as input records of other hydro-meteorological variables. The present study was conducted in Kechik station, Golestan Province (northern Iran). The normal multi-layer perceptron form of ANN (MLP–ANN) was selected as the baseline ANN model. The efficiency of GDX, CG and L–M training algorithms were compared to improve computed performances. The inputs of ANN in-cluded temperature, evaporation, air pressure, humidity and wind velocity in a 10 minute increment The results revealed that the L–M algorithm was more efficient than the CG and GDX algorithm, so it was used for training six ANN models for rainfall intensity forecasting. The results showed that all of the parameters were proper inputs for simulating rainfall, but temperature, evaporation and moisture were the most im-portant factors in rainfall occurrence.