Houshmand Ata'ii; Sorayya Alijani Alijanvand
Volume 21, Issue 82 , September 2012, , Pages 57-63
Abstract
The present article shortly introduces Bazoft basin (one of Karun subbasins) and discusses different methods of predicting flood discharges, all of which requires specific local data. Due to lack of statistics in most basins and subbasins of the country, the suggested solutions require less ...
Read More
The present article shortly introduces Bazoft basin (one of Karun subbasins) and discusses different methods of predicting flood discharges, all of which requires specific local data. Due to lack of statistics in most basins and subbasins of the country, the suggested solutions require less hydrological statistics. Flood modelling is one of these solutions which requires fixed physiographic data and can be used for estimating and evaluating floods in basins which lack statistics. Bazoft subbasins have the potential to produce flood, while they face lack of statistics in many cases. With physiographic information and digital maps of the area and also with available statistical data in 30 studied basins (figure 3), the possibility of establishing logical relations between physiographic features and flood discharges with different return periods will be investigated. On the other hand, a GIS data bank seems necessary for easy access in later applications and capability of updating information and relations in any of these subbasins. Information on flood discharge (with different return periods), as the most important distinguishing parameter of floods, is especially important. In floods with shorter statistical period, establishing a regional relation between physiographic characteristics of the basin or subbasin seems logical. After completing and prolonging statistics, different statistical distributions will be fitted using SMADA software. Predicted value of statistical distributions (Log Pearson type3, Pearson type3, Gamble) and observational data were used to find the most appropriate distribution in least squares test and a distribution was selected for each station. With these distributions, flood discharges with different return periods of 2 to 1000 years were estimated. Then, physiographic features of the basin (like area, perimeter, and average slope of the basin, length of the main canal and shape coefficient) were linked with predicted discharges of different return periods using linear regression and multi-variable nonlinear regression in Minitab software. A larger number of parameters are involved in flood predicting models used for estimating discharges with short return periods.
