Sensitivity analysis of nitrogen river water loading model by random balance design variance based method, Case Study: Tweed river basin

The existence too much nitrogen compounds in water is one of the threatening factors of aquatic life and humans. Thus understanding and controlling the effecting variable on sources of nitrogen entering the water is crucial. Investigate of important variable symptom variety is impossible by sampling. Its can detect effective variable by using of mathematical technique and innovation computer approach simulation and neglected the variables that has little effect in practically that finance source consume for control in a count of optimize variable. Several models have been developed to estimate this phenomenon. In this article used of INCA-N to sensitivity analysis and variable reduce. INCA-N has different variable input thus detection of affectless variable of INCA-N is important. Variance based method doing well detecting of important variable and interaction between and sassing main sensitivity index of output model respected these variables. In this paper, further introduce variance based approach and estimation of sensitivity index with random balance design also defray to sensitivity analysis and reduce variables of INCA-N model in the tweed river. The results of sensitivity analysis in 500, 1000 and 2000 sample size and investigated uncertainty of result did by 50 repeating in samples. The result showed that four of seven variables (nitrate uptake rate by plants, denitrification rates, immobilization and mineralization) of INCA-N are important. Three variable nitrogen fixation, ammonium planet uptake and maximum nitrogen uptake is no important. These four necessary variables corresponding with main effect (0.456, 0.207, 0.204 and 0.09. Also Interaction between this variable are so weak (maximum= 0.036) that can surrender of them. Thus random balance design sensitivity analysis method has good efficiency in reduction of variable these phenomena.