Mohammad Nazeri Tahrudi; Keivan Khalili; Javad Behmanesh; Kamran Zeinalzadeh
Abstract
Introduction: Drought from the hydrological viewpoint is a continuation of the meteorological drought that cause of the lack of surface water such as rivers, lakes, reservoirs and groundwater resources. This analysis, which is generally on the surface streams, reservoirs, lakes and groundwater, takes ...
Read More
Introduction: Drought from the hydrological viewpoint is a continuation of the meteorological drought that cause of the lack of surface water such as rivers, lakes, reservoirs and groundwater resources. This analysis, which is generally on the surface streams, reservoirs, lakes and groundwater, takes place as hydrological drought considered and studied. So the data on the quantity of flow of the rivers in this study is of fundamental importance. This data are included, level, flow, river flow is no term (5). Overall the hydrological drought studies are focused on annual discharges, maximum annual discharge or minimum discharge period. The most importance of this analysis is periodically during the course of the analysis remains a certain threshold and subthresholdrunoff volume fraction has created. In situations where water for irrigation or water of a river without any reservoir, is not adequate, the minimum flow analysis, the most important factor to be considered (4). The aim of this study is evaluatingthe statistical distributions of drought volume rivers data from the Urmia Lake’s rivers and its return period.
Materials and Methods: Urmia Lake is a biggest and saltiest continued lake in Iran. The Lake Urmia basin is one of the most important basins in Iran region which is located in the North West of Iran. With an extent of 52700 square kilometers and an area equivalent to 3.21% of the total area of the country, This basin is located between the circuit of 35 degrees 40 minutes to 38 degrees 29 minutes north latitude and the meridian of 44 degrees 13 minutes to 47 degrees 53 minutes east longitude. In this study used the daily discharge data (m3s-1) of Urmia Lake Rivers.
Extraction of river drought volume The drought durations were extracted from the daily discharge of 13 studied stations. The first mean year was calculated for each 365 days using the Eq 1 (14).
(1) (For i=1,2,3,…,365)
That Ki is aith mean year, Yijis ith day discharge in jth year and n is number of period years. After the extraction the 1 to ndays drought duration, the years with no data were complete with Regression or interpolation methods. After the extraction, data initial evaluation (Trend, Independence and Stationarity) and completed the drought volume data, these data were fitted by the common distribution functions and select the best function based on Kolmogorov-Sminnov test. To read more information about the data initial evaluations see the NazeriTahroudi et al (15).
Log Pearson type 3 distribution Log Pearson type 3 distribution and its parameters is (7 & 12):
(2)
After selectingthe best distribution function based on Kolmogorov - Smirnov test, estimated the selected function parameter to evaluate the return period. For this purpose, there are many methods such as moments, Sundry Average method (SAM), Logarithm of applied moments observations and maximum likelihood that in this study were compared.
Results and Discussion: In this study, using daily flow data fromstations studied; the drought volume of days 1 to 60 was extracted, corrected, and completed. Before fitting the extraction drought volume river data with distribution function, the mentioned data were investigated with Wald-Wolfowitz (Independence and Stationary), Kendall (Trend) and Wilcoxon (Homogeneity) tests and the results of these tests were accepted in two significant levels of 1 and 5 percentages. Before estimatingthe Log Pearson type III parameters, first the drought volume river data were modeled by the Easy Fit software and common distribution functions and Log Pearson type III was selected by the Kolmogorov – Smirnov test as the best function. Results of two Anderson Darling and Chi Squared tests foraccurate evaluation were added. After initial evaluation of data and statistic tests, the time series of drought Volume River data of the studyarea were fitted by log Pearson type III. To estimatethe Log Pearson type III parameters used the sundry average method and to investigatethe accuracy of this method, 3 methods (moment, maximum likelihood and Logarithm of applied moment observations) were used and 4 mentioned methods for all of rivers were calculated. The most river drought relating to Gadar-Chai river with 1742 million cubic meters low volume and the lowest of it relating to Mardoq-Chai river with 68 million cubic meters low volume in 10000 year return period. After Gadar-Chai river the most low volume of discharge relating the Zarineh-rood river. Two Zarineh-rood and Gadar-Chai rivers among other rivers have a higher average discharge. Log Normal III, Gamma, Wikeby and GEV distributions have a good fitting on river flows data and no difference in investigation models that corresponded with Mosaedi et al (13) and NazeriTahroudi et al (15). The results of Grifits (7) also introduced the Wikeby distribution has a better than Beta distribution. Lee (12) also with evaluation the rainfall frequency in the study the rainfall concentration properties in Chia-Nan (Taiwan) introduced the Log Pearson type III as the best distribution function between the common distribution function. Results of Chi-Squared test in methods of parameter estimation showed that all methods are acceptable.
Conclusion: Drought occurrence can be estimated bythe analysis of historical data for different regions and using the results of predicting problems can be reduced. In this research daily river flow of Lake Urmia basin applied to calculate drought volume of rivers. Log Pearson III distribution selected among current hydrological distribution functions for fitting drought volume of rivers. Using selected distribution function and Sundry Average Moment method for estimating parameters return period of drought from 2 to 10000 years extracted. Results showed that volume of drought for Shahar-chai , Barandoz-chai, Nazlu-Chai, Mahabad-Chai, Rozeh-Chai, Gadar-chai, Simineh-rood, Zola-chai, Aji-chai, Sofi-chai, Leilan-chai and Mardoq-chay rivers in the return period of 10000 years will be 92, 125, 228, 150, 110, 1742, 90, 77, 690, 280, 65, 68 Mm3 respectively.
ELNAZ Rezaei abajelu; KAMRAN Zeinalzadeh
Abstract
Introduction: Soil Hydraulic conductivity is considered as one of the most important hydraulic properties in water and solutionmovement in porous media. In recent years, variousmodels as pedo-transfer functions, fractal models and scaling technique are used to estimate the soil saturated hydraulic conductivity ...
Read More
Introduction: Soil Hydraulic conductivity is considered as one of the most important hydraulic properties in water and solutionmovement in porous media. In recent years, variousmodels as pedo-transfer functions, fractal models and scaling technique are used to estimate the soil saturated hydraulic conductivity (Ks). Fractal models with two subset of two (solid and pore) and three phases (solid, pore and soil fractal) (PSF) are used to estimate the fractal dimension of soil particles. The PSF represents a generalization of the solid and pore mass fractal models. The PSF characterizes both the solid and pore phases of the porous material. It also exhibits self-similarity to some degree, in the sense that where local structure seems to be similar to the whole structure.PSF models can estimate interface fractal dimension using soil pore size distribution data (PSD) and soil moisture retention curve (SWRC). The main objective of this study was to evaluate different fractal models to estimate the Ksparameter.
Materials and Methods: The Schaapetal data was used in this study. The complex consists of sixty soil samples. Soil texture, soil bulk density, soil saturated hydraulic conductivity and soil particle size distribution curve were measured by hydrometer method, undistributed soil sample, constant head method and wet sieve method, respectively for all soil samples.Soil water retention curve were determined by using pressure plates apparatus.The Ks parameter could be estimated by Ralws model as a function of fractal dimension by seven fractal models. Fractal models included Fuentes at al. (1996), Hunt and Gee (2002), Bird et al. (2000), Huang and Zhang (2005), Tyler and Wheatcraft (1990), Kutlu et al. (2008), Sepaskhah and Tafteh (2013).Therefore The Ks parameter can be estimated as a function of the DS (fractal dimension) by seven fractal models (Table 2).Sensitivity analysis of Rawls model was assessed by making changes)±10%, ±20% and±30%(in input parameters (porosity, fractal dimension and the intake air suction head).Some indices like RMSE, AIC and R2 were used to evaluate different fractal models.
Results and Discussion: The results of the sensitivity analysis of Rawls - Huang model, showed the least sensitivity to changes in porosity and suction entry air and the most sensitivity to changes in fractal dimension. The saturated hydraulic conductivity is underestimated by increasing the fractal dimension in Rawls - Huang model. The high sensitivity of the combined model to changes in fractal dimension, is considered as one of the model limitations.In other words, fractal dimension underestimation increased the error related to the hydraulic conductivity estimation. Sensitivity analysis of Ks regression model was done among parameters like bulk density, dry density, silt, sand, fractal dimension of particle size and porosity. Results showed less sensitivity to fractal dimension and porosity. The highest RMSE was 0.018 for fractal dimension and porosity (in the range of ±30% changes). The results showed that the amount of clay in the estimation of fractal dimension is of crucial importance. Statistical analyzes indicated the high accuracy of the PSF models based on soil texture data.Error indices showed the high accuracy of Rawls and three-phase fractal (pore- solid- fractal) models combination in estimating the Ks value. The results suggest that Huang and Zhang model, with the highest correlation, the least Root Mean Square Error and the least Akaike criteria among the studied fractal models for estimation of the Ks values. Fuentesand Hunt models, overestimated soil saturated hydraulic conductivity. Fuentes et al. (1996) as an experimental fractal model to estimate the saturated hydraulic conductivity indicatedvery poor results. Bird model had higher error values compared with the best model, (RMSE =0.73). This model fit well with the measured values compared to Sepaskhah and Taylor models particularly at low Ksvalues. Taylor's two-parameter model, which is similar to the Brooks - Corey and the Campbell model, was inserted in the fourth priority. The RMSE values of Sepaskhah and Taylor models were 0.62 (cm/h) and 0.55(cm/h) respectively. The fractal dimension is a function of soil texture. Heavy soils resulted in a larger fractal dimension and less hydraulic conductivity. Therefore, the Huang-Zhang model as a result of clay value using model (lower values for Ks), had a close fit with the measured data in probability distribution.
Conclusions: The results showed that the soil clay percent had a significant role in fractal dimension calculation.
sarvin zamanzad ghavidel; K. Zeinalzadeh
Abstract
Introduction: A total dissolved solid (TDS) is an important indicator for water quality assessment. Since the composition of mineral salts and discharge affects the TDS of water, it is important to understand the relationship of mineral salts composition with TDS.
Materials and Methods: In this study, ...
Read More
Introduction: A total dissolved solid (TDS) is an important indicator for water quality assessment. Since the composition of mineral salts and discharge affects the TDS of water, it is important to understand the relationship of mineral salts composition with TDS.
Materials and Methods: In this study, methods of artificial neural networks with Levenberg-Marquardt training algorithm, adaptive neuro fuzzy inference system based on Subtractive Clustering and Gene expression programming were used to model water quality properties of Zarrineh River Basin at upstream of Boukan dam, to be developed in total dissolved solids prediction. ANN and ANFIS programs code were written using MATLAB programming language. Here, the ANN with one hidden layer was used and the hidden nodes’ number was determined using trial and error. Different activation functions (logarithm sigmoid, tangent sigmoid and linear) were tried for the hidden and output nodes and the GeneXpro Tools 4.0 were used to obtain the equation of the best models. Therefore, water quality data from two hydrometer stations, namely Anyan and Safakhaneh hydrometer stations were used during the statistical period of 18 years (1389-1372). In this research, for selecting input variables to the data driven models the stepwise regression method was used. In the application, 75% of data set were used for training and the remaining, 25% of data set were used for testing, randomly. In this paper, three statistical evaluation criteria, correlation coefficient (R), the root mean square error (RMSE) and mean absolute error (MAE), were used to assess model’s performances.
Results and Discussion: By applying stepwise method, the first significant (at 95% level) variable entered to the model was the HCO3. The second variable that entered to the model was Ca. The third and fourth ones were Na and Q respectively. Mg was finally entered to the model. The optimal ANN architecture used in this study consists of an input layer with five inputs, one hidden and output layer with three and two neurons for Anyan and Safakhaneh hydrometer stations, respectively. Similar ANN, ANFIS-SC5 model had the best performance. It is clear that the ANFIS with 0/4 and 0/7 radii value has the highest R and the lowest RMSE for Anyan and Safakhaneh hydrometer stations, respectively. Various GEP models have been developed using the input combinations similar ANN and ANFIS models. Comparing the GEP5 estimations with the measured data for the test stage demonstrates a high generalization capacity of the model, with relatively low error and high correlation. From the scatter plots it is obviously seen that the GEP5 predictions are closer to the corresponding measured TDS than other models. As seen from the best straight line equations (assume the equation as y=ax) in the scatter plots that the a coefficient for GEP5 is closer to 1 than other models. In addition to previous operation, Gene expression programming offered mathematical relationships in the stations of Anyan and Safakhane with the correlation coefficients, respectively 0.962 , 0.971 and with Root-mean-square errors, respectively 12.82 , 29.08 in order to predict dissolved solids (TDS) in the rivers located at upstream of the dam. The obtained results showed the efficiency of the applied models in simulating the nonlinear behavior of TDS variations in terms of performance indices. Overall, the GEP model outperformed the other models. For all of applied models, the best result was obtained by application of input combination (5) including HCO3, Ca, Na, Q and Mg. The results are also tested by using t test for verifying the robustness of the models at 95% significance level. Comparison results indicated that the poorest model in TDS simulation was ANN especially in test period. The observed relationship between residuals and model computed TDS values shows complete independence and random distribution. It is further supported by the respective correlations for GEP5 models (R2 = 0.0011 for Anyan station and R2 = 0.0123 for safakhaneh station) which are negligible small. Plots of the residuals versus model computed values can be more informative regarding model fitting to a data set. If the residuals appear to behave randomly it suggests that the model fits the data well. On the other hand, if non- random distribution is evident in the residuals, the model does not fit the data adequately. On the base of these results, we propose GEP, ANFIS-SC and ANN methods as effective tools for the computation of total dissolved solids in river water, respectively.
Conclusion: It can be concluded that the ANN, ANFIS-GP, ANFIS-SC and GEP models can be considered as promising tools for forecasting TDS values, based on water quality parameters. It is notable from the results that the prediction accuracy of all applied models increases by increasing the number of input combinations. With attention to the aim of current research that is presenting the feasibility of artificial intelligence techniques for modeling TDS values, it is notable that the results presented in this paper are for research purpose and applying the abstained results for real-world needs some complicated steps and building artificial intelligences methods, based on complete data and parameters maybe affected the TDS values.
