Journal of Water and Soil Science

Integrated Water and Nitrogen Fertilizer Management Using Modeling of System Dynamics Approach in Sugarcane Farmlands with Subsurface Drainage Systems

J. Abedi Koupai

Using water and fertilizer management at the farm level can be increased water use efficiency and reduce the volume of drainage water, fertilizer losses, and other pollutants in farmland with deep underground drains such as Khuzestan agro-industrial Companies. In the present study, a comprehensive simulation model for the water cycle and the nitrogen dynamics modeling was used for water and fertilizer management modeling on farmland of sugarcane in Imam Agro-Industrial Company using a system dynamics approach. To reduce irrigation water consumption and nitrogen fertilizer losses, five different scenarios were considered including four scenarios of water management consist of 5, 10, 15, and 20 percent reduction in the amount of irrigation water (I₁, I₂, I₃, and I₄) compared to the current situation of irrigation in Imam agro-industrial Company (I₀), and one scenario of integrated water and fertilizer management (20% reduction in the amount of irrigation water and urea fertilizer 210 Kg/ha, I₄F). The results of modeling showed that the scenario of I₄F caused to reduce 31, 70, 71, 70, and 85 percent of the cumulative volume of drainage water, cumulative nitrate and ammonium losses, total losses of cumulative nitrate, and ammonium by tile-drain and cumulative losses of denitrification process, respectively. Thus, the implementation of this scenario, not only saves water and fertilizer consumption but also reduces environmental pollution effectively. So the scenario of I₄F (amount of irrigation water for six months 2656 mm and urea fertilizer 210 Kg/ha) is recommended for sugarcane in the Imam agro-industrial Company.

Temporal Variation Pattern of Runoff Generation and Rill Erosion in Different Soils and Slope Gradients

A.R. Vaezi

This study was conducted to investigate the temporal variations of runoff and rill erosion in various soil textures under different slope gradients. So, a laboratory experiment was set up in three soil textures (loam, clay loam, and sandy clay loam) and four slope gradients (5, 10, 15, and 20%) using the completely randomized design with three replications. Runoff production and rill erosion were measured at a flume with 4 m×0.32 m in dimensions using a simulated water flow with 0.5 lit min^-1 in discharge during 30 min. Results indicated that runoff and rill erosion and their interaction were significantly affected by soil texture and slope gradient (P < 0.001). Significant relations were found between rill erosion and runoff both in three soils and four slope gradients, and the strongest relations were in loam (R²= 0.86) and 15% slope gradient (R²= 0.94). Runoff and rill erosion varied considerably in the soil textures and slope gradients during the experiment. A 10-min pick time was found for runoff and rill erosion. In contrast to runoff, rill erosion appeared an irregular and gradual increasing pattern during the experiment which was associated with the frequency of transportable soil particles. Clay loam had more sensitive particles due to a higher percentage of fine particles and weaker structure, and most of them were washed in early times, and finally, rill erosion was reached to a constant pattern. This study revealed that temporal variation patterns of runoff and rill erosion are influenced by soil type (texture and structure) and slope gradient.

Interaction of Sugarcane Bagasse Biochars and Salinity on Availability and Fraction of Cadmium in a Clay Loam Calcareous Soil

H. Motaghian

Pollution of the environment by heavy metals, especially soil pollution with cadmium (Cd), is one of the most important environmental problems. Also, salinization of soils due to a decrease in irrigation water quality reduces plants growth. To investigate the effect of sugarcane bagasse biochars and salinity on Cd available and Cd fractions in a contaminated soil (15 mg kg^-1 Cd), an experimental factorial design in a completely randomized design including (1) biochar factor with control, 1% (w/w) bagasse, 1% (w/w) biochar 400 °C and 1% (w/w) biochar 600 °C, and (2) salinity factor with control, 20, and 40 mmol kg^-1 as sodium chloride were performed in 3 replications. The amount of available Cd was determined by the DTPA-TEA method and the Cd fractions were determined by Tessier et al. (1979). The results showed that biochar and salinity had no significant effect on soil pH (P > 0.05) but increased soil EC (P < 0.05). Interaction of biochar and salinity was not significant (P > 0.05) on available Cd and Cd fractions. Biochar application reduced Cd -DTPA-TEA (P < 0.05), whereas salinity increased Cd -DTPA-TEA (P < 0.05). Application of biochar prepared at 600 °C reduced (P < 0.05) exchangeable Cd (23.8%) and increased (P < 0.05) Cd associated with iron and manganese oxides (25.2%) and residual (15.6%) compared to the control. The results showed that salinity had no significant effect on the Cd fractions (P > 0.05) and soil treated with sugarcane bagasse biochar can reduce Cd available due to changing the Cd distribution from unstable forms to stable forms.

Assessment of Sources and Health Risks of Polycyclic Aromatic Hydrocarbons (PAHs) in the Street Dust of Dezful City

N. Ghanavati

Street dust is mainly affected by the pollution of polycyclic aromatic hydrocarbons (PAHs). PAHs are a group of organic pollutants consisting of two or more benzene rings and are mainly produced during incomplete combustion. PAHs have attracted widespread attention due to their high carcinogenic and mutagenic properties in humans. Therefore, the purpose of this study was to investigate the sources and extent of the impact of these compounds on human health and the environment. To achieve this goal, 30 dust samples were collected from the sidewalks of the main streets of Dezful and the concentration of PAHs was determined by gas chromatography-mass spectrometry (GC-MS). The total concentration of PAHs in street dust of Dezful varied from 562.85 to 51447.10 μg / kg. The ratio of carcinogenic compounds to total PAHs was in the range between 0.73 to 0.91. Low molecular weight and high molecular weight PAHs accounted for 12% and 88% of total PAHs, respectively. The most important sources of PAHs in Dezful are the combustion of fossil fuels and petroleum products and emissions from vehicles and traffic. Moreover, incremental lifetime cancer risk (ILCR) in pathways ingestion in children was higher than in adults, but the ILCR in pathways dermal contact and inhalation in adults was higher than in children. The total cancer risk (CR) for children (5.77×10^-3) was higher than adults (5.56×10^-3), which shows the high potential for CR in the study area.

Assessment of Comparative Advantage of Agricultural Crops in Semnan Province Based on Water Productivity

A.R. Tavakoli

Due to limited access to water, it is necessary to determine the comparative advantage of crops and horticultural products in different parts of the province and identify products that lack the desired productivity. Then, find ways to improve water productivity or replace products with higher comparative advantage with low-yield products. Based on this, the crop and economic productivity index of the province's agricultural and horticultural products under surface irrigation systems was determined. Based on the results of gross economic productivity of horticultural products, pistachio with 48690 Rials per cubic meter had the highest economic productivity, and grapes with 30220 Rials per cubic meter (62% of pistachios) was in second place. In addition, water quality for pistachios can never be generalized for grapes and other crops, and this is a tremendous advantage for pistachios that low-quality water resources can also be used. The lowest gross economic productivity of water for barley, alfalfa, and wheat is equal to 3790, 3990, and 4570 Rials per cubic meter, respectively. The study of fodder corn shows that the net profit from the cultivation of this crop in the surface irrigation system is equal to 51.78 million Rials per hectare and its net profit in the strip drip irrigation system (tape) is equal to 110.87 million Rials, which it has a 2.14-fold increase compared to the conventional irrigation method. The comparative advantage of horticultural products was higher than that of crops, and the replacement and development of orchards instead of some crops is recommended as a solution. In addition, solutions that can be recommended to improve the productivity index include the use of a drip irrigation system (tape) for crops (fodder corn, tomatoes, summer crops, and potatoes) and the use of drip irrigation (normal, subsurface, and subsurface modified drip) for horticultural products. Examination of practical experiences of using tape irrigation system for wheat and barley showed that this method has not improved the agricultural and economic productivity index, which indicates the lack of comparative advantage of wheat and barley cultivation in Semnan Province even with tape irrigation system.

Assessment of Seasonal Variations of Pollutant Decay Coefficient of Talar River

K. Ebrahimi

Protection of rivers’ water quality as the most accessible source of the water supply has always been considered. In this paper, self-purification and the pollution decay coefficient values of Talar River, IRAN were studied based on field measurement of DO, BOD, pH, EC, Nitrate, Phosphate, and Temperature, in four seasons of the year 2018, in tandem with the river simulation and its calibration using QUAL2Kw model and the Streeter-Phelps method. In addition to the modeling and analysis results, the measured laboratory data values of the river water samples are also presented. Based on the results, the DO variations were ranged from 5.15 in summer to 7.47 mg/l in spring and BOD variations ranged from 1.88 in fall to 7.9 mg/l in summer. Also, according to the Streeter-Phelps method the decay coefficient values varied from 1.57 (1/day) in spring to 9.63 (1/day) in fall. The values of the Talar River decay coefficient also varied from 2 in fall to 7.7 (1/day) in summer involving the QUAL2Kw model.

Investigation of Surface Runoff of Plasjan River Using WetSpa Hydrological Model

S. Toghiani Khorasgani

In recent decades, water scarcity has become a global problem due to the growth of the world's population as well as the increase in per capita water consumption. Therefore, planning and managing water resources to prevent potential risks such as floods and drought in the future is one of the important measures of water resources management. One of the important measures to avoid potential risks and predict the future is rainfall-runoff modeling. The objective of this study was to investigate the efficiency of the WetSpa hydrological model in estimating surface runoff in the Eskandari watershed, which is one of the important sub-basins of the Zayandehrood watershed. In this study, Daran and Fereydunshahr synoptic stations have been used to collect meteorological information in the Eskandari watershed. Also, to study the flow of the Plasjan river, daily data of Eskandari hydrometric station, located at the outlet of the basin, have been used. Climatic data along with digital maps of altitude, soil texture, and land use were entered as input to the WetSpa model. Finally, the ability of the WetSpa model was evaluated in estimating river surface runoff. The observed flow at the basin outlet in the hydrometric station was used to evaluate and calibrate the model. The model was calibrated for the statistical period (1992-2000) and its validation was performed for the statistical period (2001-2004). In the calibration period, the trial and error method were used to calibrate the model parameters. The simulation results showed a good correlation between the simulated flow and the measured flow. In the present study, the Nash Sutcliffe coefficient in the calibration and validation stages was equal to 0.73 and 0.75, respectively which shows the good and acceptable ability of the model in estimating the surface runoff of the study basin.

Maximizing Economic Water Productivity Index in Agriculture in Different Levels of Irrigation

A. Mehrabi

Designing an optimal crop pattern and on-time water allocation of water resources along with deficit irrigation are among the optimal solutions to maximize the water economic efficiency index. In this paper, the simultaneous optimization of crop pattern and water allocation are discussed using the deficit irrigation method. The study area is located west of the Qazvin plain irrigation network. The six different levels of percentage reduction of irrigation rate (0, 0 to 10, 0 to 20, 0 to 30, 0 to 40, and 0 to 50%) in three climatic conditions consist of dry, normal, and wet years were compared. The best irrigation scenario was selected for each year, and the results were compared with the existing crop pattern of the same year. The new crop pattern included the main crops of the region and the addition of rapeseed. The objective was to reach the maximum net benefit per unit volume of water by considering the maximum extraction of monthly and annual surface and groundwater. The results showed that the best scenario in the dry year was maximum deficit irrigation up to 20%, in a normal year full irrigation, and a wet year maximum deficit irrigation up to 10%. The improvement of economic water productivity in a dry year was 52.2%, in a normal year 41.5%, and in a wet year is 19.6% compared to the existing crop pattern. The average percentage of annual irrigation supply increases from 64.3 to 91.7% in a dry year, from 70 to 100% in a normal year, and from 77.5 to 97.1% in a wet year. Also, the relative yield of all crops, especially wheat, alfalfa, and sugar beet significantly increases. Therefore, the gravitational search algorithm as an optimization model can be considered in selecting the suitable crop pattern and allocation of surface and groundwater resources concerning economic benefits in irrigation networks management.

Investigation of Ground Waters Quality Using Multivariate Statistical Techniques (Case Study: Maragheh-Bonab Plain, East Azerbaijan Province)

M. Amini

Investigation and analysis of groundwater quality to monitor contamination and identify the most important pollutants and pollution points is one of the research fields. The objective of this research was to plan to improve groundwater quality on various spatial and temporal scales. Groundwater information of Maragheh-Bonab plain was collected from 26 wells in 10 years (2001-2011) with 454 sampling points from East Azerbaijan Regional Water Organization and was analyzed using multivariate statistical techniques such as DFA and PCA. Analyzed Variables are included Mg, Ca, Cation, K, Na, TDS, TH, SAR, EC, Anion, pH, Cl, SO4, CO3, and HCO3. Results of PCA showed that variables such as cation, HCO3، TDS، SAR، EC، Anion ،Cl, Ca, and TH were identified as important variables which they can great impacts on the groundwater quality of this region and in the other hand DFA showed which mentioned variables can discriminate land uses and geology formations in primary and normal distribution data with power discriminatory of 68.7 %, 92.2 %, and 66.5 %, 89.1 %, respectively. Investigation of the spatial position of elements using interpolation technique in Maragheh-Bonab plain showed that variables concentration in lowlands are high and 20 villages and their surrounding farms are exposed to high contamination risk of groundwater.

Evaluation of Heavy Metal Dust Pollution and their Origin in the Atmosphere of Kuhdasht, Lorestan Province

A. Jalalian

Dust is one of the environmental hazards in arid and semi-arid regions of the world. In some areas, under the influence of human activities, dust is contaminated by heavy metals. In this study, the dust of 10 stations in the Kuhdasht region of Lorestan province in four seasons of spring, summer, autumn, and winter, as well as adjacent surface soils (a total of 40 dust samples and 10 surface soil samples), were sampled and some heavy metals including Zn, Pb, Cd, Ni, Cu, and Mn were analyzed. The results revealed that the amount of Zn in the dust was much higher than the surface soils of the region (800 vs. 85 mg/kg). Contamination factor index calculation indicated that high contamination of Cd and Zn, significant contamination of Ni and Pb, and lack of contamination by Cu and Mn. The annual enrichment factor of Cd (33.9) and Zn (24.6) was very high, Ni (11.3) was significant, Pb (6.4) was moderate, Mn (1) and Cu (0.82) were low. Based on the enrichment factor values, Cd, Zn, and Ni seem to have a human origin, Pb has both human activities and natural origin, and Cu and Mn have an only natural origin.

Integrated Assessment of Ferizi Watershed Based on Water Accounting Framework Using SWAT Model

M. Hosseini

Integrated assessment of the watershed is critical in arid and semi-arid areas due to the severe water stress in these regions. Data and information are an essential part of decision making and water governance to obtain integrated water resources management at the watershed scale. Water accounting is a helpful tool to organize information and present them as the standard indicators to achieve this goal. Therefore, the objective of this study is to implement the Water Accounting Plus framework (WA+) in the Ferizi watershed located in the Khorasan-e Razavi Province. In this study, water accounting indicators of the Ferizi watershed for a period of 28 years (1990-2017) and wet (1990-1997) and dry (1998-2009) periods were calculated using the SWAT model. The calculated indicators showed that the amount of manageable water and usefulness of consumption (transpiration) is low in the watershed and a large part of the share of irrigation in the watershed is provided by groundwater resources. Generally, the results of this study showed that the use of the SWAT model, WA+ framework, and analysis of water accounting indicators play a significant role in assessing the agricultural and hydrological conditions of the watershed. The proposed approach in this study can help managers make enlightened decisions to keep the sustainability of the watershed.

Experimental Study of Discharge Coefficient of Symmetric and Asymmetric Rectangular Labyrinth Weirs

O. Tayari

The discharge coefficient of labyrinth weirs increases with increasing the crest length in a certain width range. The present research was carried out in a laboratory flume with a length of 8 m, a width of 0.6 m, and a height of 0.6 m. The discharge coefficient of two-cycle symmetric and asymmetric rectangular labyrinth weirs was experimentally measured. The dimensional analysis by the Buckingham π theorem indicated that the discharge coefficient was dependent on S_e, B/W_avg, H_t/P, and W_L/W_R. According to the results, the discharge coefficient decreased with increasing the hydraulic head in the symmetric and asymmetric labyrinth weirs and the linear weir. Asymmetric labyrinth weirs with a W_L/W_R of 2.05 outperformed symmetric labyrinth weirs with a W_L/W_R of 1. Quantitatively, the discharge coefficient of the labyrinth weir with a B/W_avg of 3.1 was respectively 21% and 94% higher than that with a B/W_avg of 2.93 and 2.76. The discharge coefficient of the labyrinth weir with a W_L/W_R of 2.05 was 10-27% higher than that with a W_L/W_R of 1. The discharge coefficient of the linear weir was 60-250% higher than that of labyrinth weirs.

Estimation of Actual Evapotranspiration, Water productivity, and Irrigation Efficiency of Wheat Fields in Surface and Sprinkler Irrigation Systems Using Remote Sensing

S. Zand-Parsa

In arid and semi-arid regions, water resource management and optimization of applying irrigation water are particularly important. For optimization of applying irrigation water, the estimated values of actual evapotranspiration are necessary for avoiding excessive or inadequate applying water. The estimation of actual crop evapotranspiration is not possible in large areas using the traditional methods. Hence, it is recommended to use remote sensing algorithms for these areas. In this research, actual evapotranspiration of wheat fields was estimated using METRIC algorithm (Mapping EvapoTranspiration at high Resolution with Internalized Calibration), using ground-based meteorological data and satellite images of Landsat8 at the Faculty of Agriculture, Shiraz University, in 2016-2018. In the process of METRIC execution, cold pixels are located in well-irrigated wheat fields where there is no water stress and maximum crop evapotranspiration occurred. The estimated maximum values of evapotranspiration using the METRIC algorithm were validated favorably using the obtained values by the AquaCrop model with NRMSE (Normalized Root Mean Square Errors) equal to 0.12. Finally, the values of water productivity (grain yield per unit volume of evapotranspiration) and irrigation efficiency were estimated using the values of predicted actual evapotranspiration using remote sensing technique. The values of measured irrigation water and produced wheat grain yield in 179 ha were estimated at 0.86 kg m^-3and 75%, respectively.

Source Identification and Health Risk Assessment of Polychlorinated Biphenyls (PCBs) in Surface Soils from Vicinity of Arad-Kouh Processing and Disposal Complex, Tehran

B. Lorestani

Polychlorinated biphenyls (PCBs) can adversely affect human and environmental health according to long-term half-life and persistence in the environment. Therefore, this study was conducted to detect, identify, and health risk assessment of PCBs in surface soils collected from the vicinity of Arad-Kouh processing and disposal complex, Tehran, in 2020. A total of 30 surface soil samples was collected from 10 sampling sites near the Arad-kouh complex. After extraction of analytes, the gas chromatography/mass spectrometry (GC–MS) method was used to determine PCBs in soil samples. Based on the results, 15 congeners of PCBs were detected in the analyzed soil samples. Also, the minimum, maximum, and mean concentrations of total PCBs (µg/kg) were 269, 434, and 359, respectively. Moreover, the results of PCA and significant contribution values of low molecular weight homologs indicated that the presence of PCB compounds in the soil samples was connected with combustion processes in the soil. Besides, as among the detected PCBs, the TEF values only established for PCB105 showed that exposure to contaminated soil could be lead to a moderate level of carcinogenic risk through PCB105. Given that PCBs have adverse effects on the environment and human health, detecting, determining the concentration, source identification, and periodical monitoring of these compounds in different mediums to human health maintenance is strongly recommended.

Dust Storm Frequency in Connection with Climatic Change in the Arid Region of Iran

R. modarres

Dust storms in arid and semi-arid regions have harmful impacts on the environment, the economy, and the health of local and global communities. In this study, the frequency of annual dust events in twenty-five stations and five climatic variables including rainfall, maximum annual wind speed, average annual wind speed, maximum annual temperature, and average annual temperature in arid regions of Iran up to 2014 were used to show the effects of climatic change on dust storms. Annual correlation coefficient time series between climatic variables and dust storms were first calculated based on monthly observations. Then, the trend in climatic variables, dust storm frequency, and their correlation were assessed using the Mann-Kendal method. Results indicated that the correlation coefficients had fluctuations in time and are both significant and insignificant in different years that reach from 0.6 to 0.9 for wind speed and temperature and -0.2 to -0.6 for precipitation. This trend in correlation has the same direction with climatic variables and shows co-movement between climatic change and dust storm fluctuations in central Iran. Results also showed that wind speed and temperature have a high impact on dust storm fluctuations and rainfall reduction has an increasing effect on dust storms.

Evaluation of ANFIS and ANFIS-PSO Models for Estimating Hydraulic Jump Characteristics

M. Akbari

In this study accuracy of the ANFIS and ANFIS-PSO models to estimate hydraulic jump characteristics including sequence depth ratio, the jump length, the roller length ratio, and relative energy loss was evaluated in stilling basin versus laboratory results. The mentioned characteristics were measured in the stilling basin with a rectangular cross-section with four different adverse slopes, four diameters of bed roughness, four heights of positive step, three Froude numbers, and four discharges. The average statistical parameters of NRMSE, CRM, and R² for estimating hydraulic jump characteristics with the ANFIS model were 0.059, -0.001, and 0.989, respectively. While, the mean values of these parameters for the ANFIS-PSO model were 0.185, 0.002, and 0.957, respectively. The results indicated that these models were capable of estimating hydraulic jump parameters with high accuracy. However, the ANFIS model was moderately more accurate than the ANFIS-PSO model to estimate the sequence depth ratio, the jump length, the roller length ratio, and relative energy loss.

Investigating the Effect of Microbial Induced Carbonate Precipitation on Reducing Wind Erodibility of Soils in Segzi Desert Area, Isfahan

M. Mirmohamad Sadeghi

Desertification has become one of the main problems of human societies living in the vicinity of desert areas in recent years. One of the methods that have been considered in recent years and are rapidly expanding in the field of soil mechanics is the Microbial Induced Carbonate Precipitation (MICP). In this method, urea-positive organisms that are naturally present in the soil can stabilize the soil and improve its engineering parameters by using urea and calcium chloride. Recently, attempts have been made to use this method to create a crustal layer on the soil to prevent wind erosion. In the present study, the effect of environmental conditions in deserts such as temperature and sand bombardment on microbial soil treatment has been investigated using this new method. The soil of the Segzi region as one of the main centers of dust in the Isfahan region was studied in this research. Therefore, the improved samples are subjected to regional temperatures which increased the surface layer resistance with increasing temperature. Also, the sandstorm conditions of the region were simulated using three different grain sizes of sand inside the wind tunnel. The results of these experiments showed that stabilized soil could withstand the conditions at wind speeds of 7 and 11 m/s. However, by increasing the wind speed to 14 m/s and the grain size, the crustal layer destroys and increases the wind erosion of the soil. Also, the resistance of the surface layer increased by increasing temperature in the tested samples. This increase in resistance continued up to 24 degrees with a high slope, but from 24 degrees onwards, this slope decreases. Based on the results of this research, it can be said that the microbial improvement method can be used as an alternative method in the future to stabilize desert soils.

Classification of Scouring Form of Homogeneous and Sticky Earth Dams in Overtopping

S. salehi

The objective of this study was to investigate how the earth dam is destroyed due to the effect of upstream and downstream slope of the body in overflow conditions. Therefore, eight models were provided that each model is constructed from the embankment dam with different upstream and downstream slopes (1:1, 2:1) and the soil properties (Sc) on breach formation. The time and method of dam break for flood discharges were investigated. The results showed that the upstream side slope of the embankment dam has less effect than the downstream side slope on the scour process resulting from the phenomenon and by increasing the downstream side slope of the embankment dam, the amount of erosion in the scour hole increases 28 %. Then, using nonlinear regression, relationships were presented to estimate the output flow rate and the location of the waterfall. A to the erosion and formation of the waterfall inside the body of sticky earth dams, two main outlines of the great waterfall and a series of waterfalls were presented. Finally, the formation of these waterfalls due to the effect of shear stress created during sediment erosion relative to the critical shear stress of the dam constituents was investigated and evaluated. Considering the limitations based on shear stress, the formation status of the type and the leaching pattern of the body of the cohesive earth dams during the overpass were estimated. Then, a general plan was presented to predict the behavior of the overflow stream in homogeneous and sticky soil.

Simulation of Scour Pattern Around Cross-Vane Structures Using Outlier Robust Extreme Learning Machine

S shabanlou

In this research, the scour hole depth at the downstream of cross-vane structures with different shapes (i.e., J, I, U, and W) was simulated utilizing a modern artificial intelligence method entitled "Outlier Robust Extreme Learning Machine (ORELM)". The observational data were divided into two groups: training (70%) and test (30%). Then, using the input parameters including the ratio of the structure length to the channel width (b/B), the densimetric Froude number (F_d), the ratio of the difference between the downstream and upstream depths to the structure height (Δy/h_st), and the structure shape factor (φ), eleven different ORELM models were developed for estimating the scour depth. Subsequently, the superior model and also the most effective input parameters were identified through the conduction of uncertainty analysis. The superior model simulated the scour values by the dimensionless parameters b/B, F_d, Δy/h_st. For this model, the values of the correlation coefficient (R), the variance accounted for (VAF), and the Nash-Sutcliffe efficiency (NSC) for the superior model in the test mode were obtained 0.956, 91.378, and 0.908, respectively. Also, the dimensionless parameters b/B and Δy/h_st were detected as the most effective input parameters. Furthermore, the results of the superior model were compared with the extreme learning machine model and it was concluded that the ORELM model was more accurate. Moreover, an uncertainty analysis exhibited that the ORELM model had an overestimated performance. Besides, a partial derivative sensitivity analysis (PDSA) model was performed for the superior model.

Comparison of Artificial Neural Network, Decision Tree and Bayesian Network Models in Regional Flood Frequency Analysis using L-moments and Maximum Likelihood Methods in Karkheh and Karun Watersheds

M. Vafakhah

Proper flood discharge forecasting is significant for the design of hydraulic structures, reducing the risk of failure, and minimizing downstream environmental damage. The objective of this study was to investigate the application of machine learning methods in Regional Flood Frequency Analysis (RFFA). To achieve this goal, 18 physiographic, climatic, lithological, and land use parameters were considered for the upstream basins of the hydrometric stations located in Karkheh and Karun watersheds (46 stations with a statistical length of 21 years). The best Probability Distribution Function (pdf) was then determined using the Kolmogorov-Smirnov test at each station to estimate the flood discharge with a return period of 50-year using maximum likelihood methods and L-moments. Finally, RFFA was performed using a decision tree, Bayesian network, and artificial neural network. The results showed that the log Pearson type 3 distribution in the maximum likelihood method and the generalized normal distribution in the L moment method are the best possible regional pdfs. Based on the gamma test, the parameters of the perimeter, basin length, shape factor, and mainstream length were selected as the best input structure. The results of regional flood frequency analysis showed that the Bayesian model with the L moment method (R² = 0.7) has the best estimate compared to other methods. Decision tree and artificial neural network were in the following ranks.