تحلیل منطقه‌ای منحنی تداوم جریان برای زیرحوضه‌های فاقد آمار (مطالعه موردی: حوضه آبخیز دریاچه نمک)

Flow Duration Curve (FDC) shows the percentage of time that the daily or monthly river flow in a basin in the course of time exceeds from a certain amount. The flow duration curve is widely used by hydrologists around the world in diverse applications associated with water, such as hydropower projects, irrigation system design and management of river pollution, erosion, reservoir sedimentation and erosion. The curve is widely used in water resources engineering, but the number of studies related to this issue is low compared to its importance. The flow duration curve in these studies has focused on water resources engineering, while some others cover regional FDCs. Because some water resources projects are in basins with no data but require flow duration curve, information can be used from adjacent basins to design hydrological projects in ungauged basins. Then regional flow duration curve is useful for analyzing the hydrologic behavior in order to consider the consequences of project designs.
The present study assesses regression relations as a parametric approach in order to draw flow duration curves for ungauged sub-basins. For this purpose, 18 hydrometric stations having adequate data in the Salt Lake basin with an area of 92670 square kilometers located central Iran were selected. The lowest and highest elevations of the basin are 795 m and 4315 m above sea level, respectively, while the annual average precipitation is 250 mm. The vegetation cover distribution in this part shows a diverse pattern.
The quality of data was checked and incomplete data were reconstructed based on the established regression equations. Then, index discharges of flow duration curve thresholds (Q10, Q20, Q95) were extracted. In the next step, the regression relations were established based on the stepwise method, between the index discharge values of flow duration curves and the selected variables including physiographic factors (e.g. basin area, length of main stream channel, basin average slope, minimum, average and maximum elevation of the basin, drainage density, elevation difference), precipitation as a climatic factor and land use including the area of rain-fed and irrigated lands, rocks and rangelands. Then the statistics of regression model including the prediction error sum of squares (PRESS) and mean square error (MSE) as well as Nash-Sutcliff coefficient was evaluated. The results showed that among the studied variables, main river length, watershed area, average annual rainfall and the basin slope had the largest role in river flow. Therefore, based on the significance of relations at the 95% threshold level and having the values of these variables, the obtained relations could be used as a regional index in drawing and establishing flow duration curves for ungauged basins.
The study classified FDC into five different conditions, including flooding, wet, average, dry and low flow status. The discharge with and exceedence level of 10% was considered for flooding threshold while 40% and 90% were classified as average and low flow levels. The values indicate the time in which flooding, wet, average, drought and low flow condition are prevailing status of the basin. The results of discharge occurrence percentage briefly show that flooding condition is mostly seen in the basin for discharge with the exceedence percentage of 0-10 and then to 10-40 percentage which shows the wet condition.
It was tried in this study to consider the most influential variables in the regression equations included in this respect, land use and geological maps were prepared. But the maps were without significant information that could determine the role of these factors in river discharge. Although the regression equation was significant, the role of the geologic formation permeability and Curve Number (CN) in different discharges cannot be ignored. According to the prevailing conditions in arid areas, permeability can be effective in flow characteristics. If the data are collected and used in future studies, it can be expected to achieve better results than current study. Also, evaluation of different approaches (parametric, non-parametric and statistical) can be useful to calculate the index discharges of ungauged basins to select the appropriate approach. Multiple linear regression approach was used to establish regression equations have been widely used in other research and not only the basin area but also other as independent variables used to model the flow duration curve. As a final conclusion, the area of the basin is highly correlated with the main stream, but it stated only the size of the catchment area. While the combined area of the watershed with the main stream length, shape and size of the river basin better shows the flow behavior and characteristics.