مجله مهندسی آبیاری و آب ایران
پیاپی 1 (پاییز 1389)

Drought is one of the oldest natural calamities which has been known by human. Analysis of low flows is a method for estimating the hydrologic drought in the reigns which don’t have enough data. This study evaluates hybrid method for estimating low flow with index low flow method. The first, all the stations in case study reign was investigated and 24 suitable hydrometric stations were chosen. Then, low flow series with duration 7 days were estimated and was fitted distribution function suitable (log pyrson III) and calculated different return period. Case study was divided in index low flow based cluster analysis and in the hybrid method based area. Then low flow regional method was determined based both method and compared using train and error. Result show that hybrid method enjoys high degree as a compared to index low flow and is recommended in station that haven’t suffusion data.

The precision estimating of the carried sediments content by rivers and subsequently the entered sediment content into dams، are very important in many of water-sources majormanagement projects. For estimating of entered sedimentary load، will be applied، different methods among regressional relations as rating curves equations، which it is less considerable to procedure of use and also their precision is low، also، it will be performed to estimate content of suspensed sediments in rivers، normally، through the sediment rating curve due to high cost for continuous measurement of carried sediment and its purge. Therefor، present research intend to assess function of seven rating curve methods such as annual correlation، USBR، mean clusters، FAO-USBR، FAO-mean clusters، Fergosen-USBR and Fergosen- mean clusters، which it was performed to 34 statistical years in two stations namely Harirood-Pole Khatoon and Kashafrood-Pole Khatoon relevant to Doosti Dam. Finally، for all data، the research results indicate that better to FAO-mean clusters and Fergosen- mean clusters methods. The results of these methods were much similar. Also، by these methods were calculated useful content of rest reservoir after 34-years.

In this study، Doroodzan dam reservoir was sampled over 26-month period and twice per month at different depths and chloride concentration were measured. Based on the measured data، HEC5-Q model was calibrated and then used to simulate the reservoir. Mean Concentrations of observed and simulated chloride in the first year were 52. 9 and 55. 6 and in second year 79. 7 and 85. 8 mgL-1، respectively. The variations of measured and predicted Cl concentrations in the first year were between 33. 7-106. 4 and 33. 2 - 84. 9 mgL-1، respectively and in the second year 28. 4-177. 7 and 28. 6-169. 0 mgL-1، respectively. The maximum Cl concentrations in the first and second years were 106. 4 and 177. 7 mgL-1، respectively. Comparison of predicted concentrations and amounts of statistical parameters including d، RMSE، and MAE showed that predicted values in both years were in close agreement with observed data.

In this research، correlation between wind shelter index and snow depth (SD) was initially obtained by linear and polynomial non-linear regressions in one of Karoon sub-basins and appropriate relationships determined. Then، using obtained relationships and GIS techniques، spatial distribution of SD based on wind shelter index was estimated in the study area. Consequently، spatial distribution map provided by geostatistics models was used to evaluate the spatial distribution map of SD. The results of this study indicated that wind shelter index had direct and significant relationship (p-value < 0. 05) with SD. Correlation coefficient between these parameters in polynomial non-linear regression was more than linear regression. Spatial distribution map of SD based on wind shelter index was in good accordance with spatial distribution map obtained by ordinary kriging method. In general، it can be concluded that wind has significant effect on spatial distribution of SD.

Programminng and prioritizing are two important factors in basin management that needs different data as like as yield value and its fluctuations. The amount of average yield with different return periods is needed for different purposes as: Estimating river yield، planning for the operation of small and large dams، downstream farms management and other water resources and watershed management projects. Based on available references، there isn’t any comprehensive surveying on this matter. Results of this paper can be applied directly for estimating unguaged sites yields. In this study، hydrometric stations data were collected and analyzed and then the stations which had sufficient data (in regards of both quality and quantity) in the common period were selected. Annual average yields of these stations were calculated. Basins homogenizing was done based on yields and other effective parameters. Furthermore، a number of years in all stations were put away from regional analyzing for evaluating achieved equations. In addition to regional analyzing of annual average yields، probability analysis was also done; the best statistical distribution was fitted and yeild discharge with 2 to 100 years return periods were estimated. By applying basin’s area، probable yield regional equations with 2 years return period were extracted for homogeneous regions. Then based on regional growth curve (Dimensionless ratio of yields with different return periods devided by yield with 2 ears return period) yield values in the stations which didn’t participate in the calculation، were achieved and evaluated with observed yield values. So in each region appropriate equations were determined. As the results of this study shows، in this region homogenizing for yield estimation in unguaged sites was not an appropriate method using regional growth curve، but it can estimate yields with an acceptable accuracy. On the other hand، results shows that regression equations based on median were more accurate in comparison with mean.

Bridges are one of the most important river structures. One of the main causes of bridge failure is local scour at bridge piers. Yearly، Many bridges failed around the world because of disregarding hydraulic factors. Hence، an accurate prediction of scour depth at piers is essential for the safe design of the bridge foundation. In this study، local scour at Meshrageh Bridge is investigated. The Meshrageh Bridge crosses the Jarahi River near the city of Ramshir، Khusestan province. We use one dimensional equations Sain-Venant for simulation river dynamic in FASTER model. For evaluation of scour depth، 16 scour equations in subroutines were added to FASTER model. Comparison of computed scour depth showed Froehlich equation (1996) and Johnson equation (1996) appear to give a reasonable estimate of the local scour depth. Jain and Fisher equation (1979)، Melville equation (1997) and Melville and Sutherland equation (1988) overestimate scour depth.

Modeling water flow and solute transport in unsaturated zone requires knowledge of soil hydraulic properties such as water retention curve. Indirect determination of these functions from basic soil properties using pedotransfer functions (PTFs) has attracted the attention of researchers in a variety of fields. Nowadays using of ANNs models for PTFs deriving is very usual. In this research we introduced and evaluated new pseudo pedotransfer functions and compared with point and parametric pedotransfer function. Also impact of two different method of training، direct and indirect was calculated. All the calculation is done on 122 soil samples of north and north east of Iran that were selected randomly. Generally result of parametric functions was better than point functions. The average of r and RMSE indicators in parametric functions are 0. 93 and 0. 032 and in point functions are 0. 76 and 0. 034. On the other hand pseudo parametric functions improved the results of point functions with using the same information’s for training. The average of r and RMSE indicators in pseudo parametric functions are 0. 95 and 0. 026. It seems that increasing of samples in training phase and select the matric potentials as input are the most important factors in pseudo parametric functions success. Also results show that using local data for deriving pedotransfer functions (direct training) is better than using universal data base (indirect training).