Simulation of Soil Nitrate Transport Process under Center Pivot Irrigation Using the HYDRUS-PSO Optimizer Model

Climate change phenomenon and warming earth as well as drought, have been reducing water sources in the world wide at the last decades, especially in arid, semi-arid regions. On the other hand, soil qualities factors including organic matter and fertility, in addition to soil salinity are the most critical issues in agricultural management and environmental conservation. Applying chemical fertilizers is one of the management methods in arid and semi-arid regions to amends decreasing water resources and supply required plant nutrients in recent decades. Although using fertilizers cause to increase production and crop yields, but meantime this management way leads to destroying the soil structures. Soil nitrate pollution is the most important challenge in agriculture and environment. The numerical models were used in predicting element concentrations under various climates, management methods, and study conditions (laboratory or field,). Estimated soil hydraulic and solute transport parameters are the most important part of applying numerical models to evaluate trends in soil element concentration changes under management methods. Hence, in this research soil hydraulic and solute transport parameters was predicated under center pivot irrigation using particle swarm optimization algorithm and the obtained results of optimization in soil water and nitrate transport along soil profile was used to survey soil pollution on the farm under four-year- alfalfa cultivation. In this study a farm with 3 hectares under center-pivot irrigation and four- year- alfalfa cultivation was chosen in an arid and semi-arid region. In the first step, a profile on the farm was drilled and sampling were done at the two depths (0-40 cm and 40-100 cm), finally soil physical-chemical properties were measured. In the second step, the particle swarm optimization algorithm linked to the HYDRUS model and parameters of soil water and nitrate transport process was predicated under management conditions. In the final step, the trend of soil water content, ammonium, and nitrate concentration changes were evaluated using the HYDRUS-PSO with the statistical indices (NSE and RMSE). The obtained results of the study soil profile showed that heterogeneous condition dominates the soil profile. Along the soil profile gravity of soil and sand content was changed in the specific. The results of the HYDRUS-PSO model were indicated different values of soil hydraulic and solute transport parameters in both studied two depths. The two indices of NSE and RMSE, were used to evaluate the calibration and validation of the HYDRUS-PSO model. The obtained results show that the HYDEUS-PSO model simulated soil water flow with NSE equal to 0.89 and RMSE equal to 0.001, ammonium transport with NSE equal to 0.98 and RMSE equal to 0.001 and nitrate transport with NSE equal to 0.94 and RMSE equal to 0.035 in the surface soil. Also, the evaluated model in the subsurface soil indicated that the HYDRUS-PSO model reasonably predicated soil water, ammonium and nitrate variation with NSE equal to 0.94, 0.96 and 0.97 respectively. Decreasing accuracy and the effectiveness of the HYDRUS-PSO model along the profile in predicting soil water content, soil ammonium, and nitrate concentrations were indicating the effects of heterogeneous condition on soil water in solute transport processes. Of course, irrigation and fertilizer methods are the other factors that influence the HYDRUS-PSO model performance. In addition, the field studies and uncontrolled conditions on the farm level are other factors to make a heterogeneous environment that causes non-uniform behavior of the model. Decreasing model performance in predicting soil water content could be related to sand content variation and the difference between the saturated hydraulic conductivity of the surface and subsurface soil. The trend changes in soil nitrate concentration in both two studied depths were indicating the amount of soil nitrate did not reach the level of toxicity and did not accumulate nitrates in the soil. Precipitation intensity, irrigation volume, and soil porosity are the most effective factors in the leaching soil nitrate issue. Based on the obtained results it could be mentioned that applied irrigating and fertilizer methods do not lead to accumulating soil nitrate. In addition, the accumulation of root systems and their extension are the other factors in the no accumulation of soil nitrate. Although, these factors had an effective role in the model performance and intensified the soil heterogenic. According to the obtained results it could be mentioned that the HYDRUS-PSO could simulate soil water and solute variations along soil profile under center pivot irrigation. Therefore, the metheuristic algorithms could use in predicting parameters as a powerful and practical tool to increase the efficiency of the numerical models in water and soil studies.