Assessment of Empirical Methods of Runoff Estimation by Statistical test (Case study: Banadak Sadat Watershed, Yazd Province)

Estimation of runoff resulted from precipitation is the basis of studies in many plans of development
and utilization of water resource. Therefore, its measurement and calculation always has plenty problems
due to the environmental limitations. Consequently, considering the importance of output runoff and flood
volume estimation in watersheds to advance integrated watershed management of Iran, in this study, it has
been attempted to implement nine empirical methods of runoff estimation in Banadak Sadat watershed.
The achieved results are evaluated by the measured runoff at a hydrometric station located at the watershed
outlet, using paired t-test, MD, BIAS, RE and RMSE tests. Then, the best model was selected to estimate
runoff elevation having the maximum accuracy and efficiency in the studied watershed. After preparation
of the required maps using GIS software and statistical tests implementation by SPSS software, the results
showed that LACEY method was determined as the most appropriate empirical method for runoff estimation
in the studied area, with MD, BIAS, RE and RMSE values by 0.016, 0.007, 4.36 , 0.026 respectively, and
also no significant difference with the observed data at 95% confidence level, runoff elevation by 1.53 cm,
runoff volume by 2.29 million m3, and runoff coefficient by 18.79%. Then, the sensitivity analysis was
conducted using Excel software for LACEY model to determine the most effective reaches. According to
the achieved results, the parameter F/Z has the maximum impact on the model output at reaches 6-8, 8-10
and 18-20.