Author/Authors :
Milad Jajarmizadeh، Department of Hydraulic and Hydrology in Universiti Teknologi Malaysia , Sobri Bin Harun، Department of Hydraulic and Hydrology in Universiti Teknologi Malaysia , Marjan Nikudel، Isfahan University of Technology
كليدواژه :
SWAT , SUFI-2 , Land-Use update , Arid to Semi-arid region
چكيده لاتين :
Prediction of surface runoff by semi-distributed models is a challenging topic for most hydrologists
because of water resource management and developing of models in accuracy of performing. These
days, arid and semi-arid areas are being one of the critical regions for developing in contrary scarcity
water; meanwhile, most of the arid and semi-arid areas are potential for preserving surface waters.
Therefore, modeling of these regions can be useful for both, evaluation the capability the models in arid
to semi-arid regions and developing them due to practical application for water resource management
and future planning of watersheds management. In this study, SWAT model is applied for modeling of
daily flow in Roodan watershed, Iran. The climate of Roodan is arid to semi-arid and agriculture and
gardening are the main activities during the past 30 years. The area of Roodan is 10570km2 and
Esteghlal (Minab) Dam is located after the outlet of the Roodan watershed to collection of the stream
flow from Roodan basin. For modeling of Roodan is prepared required data digital elevation map, Land
use map from pasture map, satellite image and developing statistic agriculture field. Then soil layer is
prepared by FAO codes, and then it is evaluated by considering the geology map and available sample
soil in Roodan. Finally, meteorological data are collected near 21 years from 1988. In this research three
scenarios is defined for land use map due to continuously update of this layer during the 21 years (1988-
1993,1994,2002,2003-2008) in modeling stream flow. Then, SUFI-2 Algorithm is performed for
calibration and uncertainty analysis. Results showed a promising modeling in climate arid to semi-arid
regions of southern part of Iran (Hormozgan Province). Strength of calibration by P and R factors were
for calibration and validation periods 50%, 0.16 (calibration) and 72%, 0.14 (validation) respectively.
Generally model was performed in a acceptable values of R2 (coefficient correlation) and NS( Nash-
Sutcliff). In this study has been tried to achieve a logical range with best solution due to modeling of
Roodan by SWAT for R2 and NS, which they have been obtained in consistent values in last iteration.
Values of R2 and NS are resulted similarly 0.72 for calibration and 0.60 for validation period
