عنوان مقاله [English]
Background and objectives: Water resources development, requires frequency water recognition, temporal and spatial water distribution, also detailed assessment of its performance. Yusop et al (2007) used HEC-HMS model for rainfall-runoff modeling in a small watershed in Malaysia and reported the Satisfactory results of both calibration and verification periods (20). Planning in ungauged areas is required providing appropriate data. Thus using satellite data, is one of the methods solving this problem. For this purpose, large-scale models of the earth's surface such as GLDAS, have been updated based on satellite observations, are important tools for providing hydrological parameters. The global coverage of GLDAS, using data of the model have been considered in rainfall-run off studies. GLDAS model to evaluate the components of the water balance and energy changes in the Earth's surface, attempted to produce meteorological and hydrological parameters of high quality in the period 1948-2015. The purpose of this study, presented integrated model to estimated discharge in ungauged data.
Materials and Methods: In this study, the GLDAS model integrated with hydrological model, WMS/HEC-HMS. Integrated model was used in rainfall-runoff studies in Polroud area in Guilan province. In this case, the Tol-lat station observation data was used during 3 years (2003-2005) for calibration and 1 year 2006 for verification through two SMA and SCS methods. In this regard, for the SCS method, model calibrated parameters such as curve number, lag time and initial abstraction. For the SMA method, model calibrated soil storage, soil percolation rate, groundwater 1 storage coefficient, groundwater 2 storage coefficient and other parameters. After evaluating GLDAS results, analyzed the precipitation, surface runoff, subsurface runoff and temperature data in the pixels corresponding to the Tol-lat gauging station along the 10 (2004-2013) years. Then, using WMS/HEC-HMS, simulated runoff in the watershed, using two losses method SCS and SMA, 2004-2009 for coefficient calibration and 2010-2013 for verification. The results indicated that SMA and SCS method offer better results through integrated model. Evaluation based on criteria coefficient of determination (R2), Nash coefficient (E), the standard error of Bias, a root mean square error (RMSE) and Error showed that integrated HEC-HMS and GLDAS models is a useful tool for estimating run off in ungauged watershed.
Results: The results show that integrated HEC-HMS and GLDAS models in SMA losses in 2013 (verification period), with 0.8 coefficient of determination, 0.77 and Nash coefficient, 1.5 RMSE and 6.1 Bias error. The integrated model with SCS method in 2005 (calibration period), with 0.9 coefficient of determination, 0.86 and Nash coefficient, 0.78 RMSE and 2.5 Bias error have highest efficient. Also the model in estimating the flood peak moment is better than the non-flood values. Although in both cases, the model results were acceptable.
Conclusion: The integrated model HEC-HMS and GLDAS presented in this study is an acceptable tool for predicting runoff in inaccessible and ungauged watershed.