Flood Modeling in Arid and Semi Arid Areas Using HEC-HMS Model (Case Study: Esteghlal Minab Basin)

Introduction :

The Flash Floods occurrence are the main features of dry areas. Iran is not exception, because of locating in an arid and semi-arid region of the world, so that there have been registered 3700 damaging flood in the past 50 years (1330-1380). This situation is somewhat more critical in areas like Hormozgan Province. For example, two thousand billion rials have been damaged by only a 110 mm shower in Minab city on March 5, 2013. Arid areas are introduced as one of the most vulnerable parts of the world in terms of shower characteristics and unpredictable floods. Unprotectability character are going to be intensified in the arid and semi-arid regions like Iran where rainfall and runoff data are incomplete. The arid regions, especially the ones have large catchments, require to installation, setting up and maintenance of equipment and systems of flood recording and it isn’t easy to preparing such facilities. So, it is necessary to calibrate some methods that can be estimated the amount of runoff from rainfall in the lack or incomplete statistics basins for which flood forecasting are essential in hydrological studies and designs, water resources management. So far, many methods have been considered by many researchers to simulate runoff rainfall to predict flood. Their development has led to the production of hydrological simulation models using computer models. One of the successful models in this field is the software HEC-HMS, by which many researchers have used to predict the flood. In this research, the Minab dam basin has been used as an example of arid and semi-arid regions due to the presence of data to simulate rainfall-runoff using HEC-HMS software. It is located in the southwest of Kerman province and in the northeast of Hormozgan province. The basin is one of the largest and most important catchments in the southern coast of Iran with an area of 9845 square kilometers. Climate of the basin is hot and dry and it has a Mediterranean rain regime with a mean annual rainfall of 230 mm.

Material and Methods:

The three methods studied in this study were implemented using a HEC-HMS software with the entry of the required values in the losses section and in the runoff and base water section. Three methods, SCS, Clarke and Schneider were compared and analyzed in the runoff section. In the loses section in all three methods, the SCS method was used and a fixed monthly model was used to calculate the base current. The parameters and their values required for each of these methods to using for simulation are: CN amount and the initial losses and delay time for the SCS method, the storage coefficient and the concentration time for the Clark method, and the standard lag time and the peak coefficient for the Schneider method. After initial simulation to determine the optimal value of these parameters, calibration of the model was performed for five rainfall-runoff events. The peak discharge error percentage and Nelder and Mead searching method have been calculated during the calibration process of the model with the objective function, and then repeated correction of the parameters and the calculation of the best fit between calculated and observed hydrograph were performed. It was detected the value of the parameter obtained from the acceptance model calibration . by the suiting the fitness and production of the nearest hydrograph to the observed hydrograph. Validation of the model was carried out based on two selected rainfall-runoff events that was not involved in calibration. In the validation section, flow simulation is performed with a new parameter, independent of the parameters used in the calibration step. The new parameters result from the numerical average of the calibrated parameters. The approximation rate of computational flow to the observational flow indicates that the model is valid. Otherwise, the accuracy and validity of the model will not be accepted. Finally, sensitivity analysis was performed to determine the effect of each of the parameters on simulation accuracy.

Results and Discussion:

 The results of this study showed that SCS unit hydrograph method was determined as the appropriate method for the studied basin, given the lower difference percentage between observational and computational peak discharge, so that the values of RMSE for each of the three SCS, Clark, and Schneider models are 353.0, 117.75, and 79.2920, respectively. Also, sensitivity analysis of the model to different parameters showed that the most influential factor on the model is CN with 591.191, initial losses with a sensitivity of 1/335 and delay time with the value of 0.813, respectively. 

Conclusion:

 Since this area is located in a arid and semi-arid area, and many of the predictions of flood failure for arid areas based on SCS, it can be concluded that this model is considered a suitable model for drylands. This method can be more reliable for where have geomorphologic conditions similar to the upstream of the minab dam . However, the study of geomorphological characteristics and the emphasis on modeling based on these features can lead to accurate models and predictions.