Modeling and forecasting river flow rate from the Melen Watershed, Turkey

The Melen Watershed is located in Western Black Sea region of Turkey. Buyuk Melen and Kucuk Melen Rivers are located in this watershed. By 2010 more than 50% of Istanbul's water demand is supplied from the Buyuk Melen River. This paper presents a new approach using an artificial neural network (ANN) technique to improve precipitation forecast performance. Missing value predictions and the future precipitation value estimations were researched throughout this study. A case study was performed in Bolu and Duzce provinces of Turkey that are located in Black Sea Region. The most crucial objective of this study was to estimate missing values and to generate quantitative forecasts for future precipitation data of Duzce. Monthly average daily precipitation data of Bolu and limited number of Duzce precipitation data were used for this purpose. Ultimately, monthly river flow rate from the Melen Watershed of Turkey was modeled and forecasted through the SWAT Model using the generated precipitation data and other required spatial and temporal data. Results show that there is a considerable relation between the simulated model and observed results. This study also shows that water supply for Istanbul can be managed based on precipitation forecast. (C) 2012 Elsevier B.V. All rights reserved.