| Title: | Assessment of Asfezari national database precipitation data in runoff evaluating and monitoring regional drought |
| Authors: | Eini, M.R., S. Javadi, M. Delavar and M. Darand |
| Year: | 2018 |
| Journal: | Iranian Journal of EcoHydrology |
| Volume (Issue): | 5(1) |
| Pages: | 99-120 |
| Article ID: | |
| DOI: | 10.22059/ije.2017.235625.643 |
| URL (non-DOI journals): | |
| Model: | SWAT |
| Broad Application Category: | hydrologic only |
| Primary Application Category: | drought assessment |
| Secondary Application Category: | input data effects |
| Watershed Description: | The Maharlu Lake basin is located in Fars province with an area of 4270 square kilometers. This study area is located in the basin of Iran's central plateau and in the watershed basin of Tashk-Bakhtegan and Maharlu lakes. The Maharlu Lake basin is located between the latitudes 29° 1’ 6” to 30° 6’ 0” north and 52° 12’ 0” to 53° 28’ 0” east (Fig. 1). Maharlu Lake is the outlet of Shiraz and Sarvestan catchments located in the northwest and southeast of the lake, respectively (Tajabadi et al., 2017a). In this study, we simulated northwest of Maharlu Lake basin. The elevation of the Maharlu Lake is 1460 m overhead sea level, and its water depth varies from zero to about 3 m during a year. The surface flow into the Maharlu Lake happens mostly during winter and early spring (wet seasons), that increases the lake water level and increases water body area on the lake ground (Tajabadi et al., 2017b). The ephemeral Khoshk River is made from the combination of two streams, namely, Chenar Sukhte and Nahre Azam, which begin from northwestern of basin. After passing through the city, the Khoshk River discharges into the Maharlu saline Lake. In dry months (May-October), apart from the upstream flows, Khoshk River receives discharges only from the municipal wastes (Salati and Moore, 2010). The weather of the study area is temperate dry (mean temperature 17.4 Celsius) with annual average rainfall of 430 mm and ET0 of about 2500 mm. In this research, four rainfall gauge stations are planned. In addition, three discharge stations of Nahre Azam, Roode Khoshk, and Baghe Safa were used for calibration and validation of the area. Three major crop products, namely spring wheat, olive, and grape, entered the model according to their crop area. In Fig. 2, the position of the climate stations and the position of the discharge stations within the area and the networks of streams are given. In this study, the period used to assess the accuracy of the climate datasets has been selected from 1980 to 2013. For simulation from 1980 to 1983, it was used during the warm-up period and from 1983 to 2010 for the calibration period and from 2010 to 2013 for validation purposes. Table 1 shows the data of the digital layers used. |
| Calibration Summary: | Calibration and validation for observation database showed that the parameters of base flow and surface runoff are highly sensitive. The model could show the coefficient of determination of for the three stations to be higher than 0.6 for the calibration period and above 0.65 for the validation period. The best results were related to the Rode Khoshk station with a value of 0.72 for calibration and 0.81 for the validation period, as well as for Nash Sutcliffe 0.72 for calibration period and 0.78 for the validation period. The evaluation and analysis of the uncertainty utility of the model results were performed using the P-factor and R-factor indices. The P-factor index represents the percentage of observation data in which the 95% confidence interval simulated by the model, and the R-factor index represents the mean uncertainty band 95% confidence interval divided by the standard deviation of observation data. The higher values of P-factor and fewer R-factor represent better model results in simulation. |
| Validation Summary: | Calibration and validation for observation database showed that the parameters of base flow and surface runoff are highly sensitive. The model could show the coefficient of determination of for the three stations to be higher than 0.6 for the calibration period and above 0.65 for the validation period. The best results were related to the Rode Khoshk station with a value of 0.72 for calibration and 0.81 for the validation period, as well as for Nash Sutcliffe 0.72 for calibration period and 0.78 for the validation period. The evaluation and analysis of the uncertainty utility of the model results were performed using the P-factor and R-factor indices. The P-factor index represents the percentage of observation data in which the 95% confidence interval simulated by the model, and the R-factor index represents the mean uncertainty band 95% confidence interval divided by the standard deviation of observation data. The higher values of P-factor and fewer R-factor represent better model results in simulation. |
| General Comments: | Database manager notes: There is no English title or abstract provided for this study; the lead author provided all of the information submitted for this database entry. The title listed in the table of contents for the issue on the journal webpage (see https://ije.ut.ac.ir/issue_8552_8553_Volume+5%2C+Issue+1%2C+Spring+2018%2C+Page+1-341.html) reads differently; i.e., "Assessment of precipitation data from Asfazari national database in runoff evaluating and regional drought monitoring". This appears to be a matter of how the original Farsi (Persian) text is translated.
Additional comments from author: Estimation of precipitation data of the Asfezari on precipitation estimation revealed that this database in the cold season, which accounts for 90% of precipitation in this area, could be used with high confidence, or in basins with a shortage of stations or in terms of time, short periods can be used from this database. Based on the findings from this research in recent years (from 2005 onwards), the accuracy of the Asfezari database for estimating runoff on study area compared with the pre-era period increased proportionally. |
| Language: | Farsi |
| Keywords: | Reanalysis database; Maharlu Lake; SPI and SDI Indexes; SWAT model |