Hadi Memarian
Firozeh Amirafzali - M.Sc. watershed management, University of Birjand, Birjand, Iran
Seyed Mohammad Tajbakhsh - Watershed Management Dept, Natural Resource Faculty, University of Birjand, Birjand, Iran
Abolfazl Akbarpour - Department of Civil Engineering, University of Birjand, Birjand, Iran

Due to the fact that ecosystems are more fragile and sensitive in arid and semi-arid regions of the world, the phenomenon of floods and the resulting damage and loss will be more severe in these areas. The subject of river canals and their morphology is one of the key topics in engineering and river management, which can be used to obtain a useful set of information about the geometric shape, bed shape, longitudinal profile, cross sections, over time. Default geometric relationships in hydrological models such as Kineros۲ are based on field measurements in US watersheds and cause uncertainty in model results. Therefore, the purpose of this study is to determine the regional statistical relationships between the width and depth of the canal with the area of the upstream watersheds for employing in hydrological models such as Kineros۲, and the width and depth of the canal with other basin characteristics. The data in this research are mainly topographic maps of Neishabour Bar watershed. Thus, according to topographic maps, preliminary studies were carried out to identify sub-basins. In this study, the basin is divided into ۳۴ sub-basins that involve Lar Formation (in ۱۶ sub-basins), and marl (in ۱۸ sub-basins), where, ۲۷ sub-basins in upstream are non-orchard lands. The linear and nonlinear regressions and equations were studied and evaluated using the software such as Excel, SPSS, Curve Expert, and XLSTAT. The results of correlation of physical parameters with canal width and depth, nonlinear regression analysis and analysis of variance in the relationships between canal depth/width with upstream area in the whole basin (R۲=۰.۵۸ for canal depth and R۲=۰.۱۴ for canal width), in upstream Lar formation, in downstream and non-orchard lands, showed a greater impact of the upstream acreage on the canal depth relative to the width (due to the higher coefficient of determination and less error). Furthermore, the separation of sub-basins in terms of geological formation and the presence of orchards had a significant effect on improving equations and reducing errors. The greater impact of the canal depth than the width from the upstream area is mostly related to successive droughts and the absence of flash floods in the area to change the canal depth, while the width of the canals has been mainly a function of human manipulations on the river bed. Stepwise linear regression analysis also showed a higher correlation between canal depth than the canal width with the physical parameters of the basin (R۲=۰.۸۵ and R۲=۰.۷۷, respectively).

Canal width and depth, Regression analysis, River morphology, Semi-arid region, Statistical equations, Regional equations, watershed