NIT-DRAIN model to simulate nitrate concentrations and leaching in a tile-drained agricultural field
Résumé
Agricultural pollution, due to excessive land fertilization, affects the quality of both surface and ground water. This paper presents a conceptual and reservoir-based model that simulates the time variation of nitrate concentrations [NO3-] at the outlet of a subsurface drainage network, referred to herein as "NIT-DRAIN." The model is based on a simplified scheme of water and nitrate transfer processes by decomposing the soil profile located between the drain and the mid-drain into three interconnected and conceptual compartments. Hence, the fast transfer throughout the macroporosity located above the pipe is described by the first compartment while the low and residual nitrate transfers are managed by the two other compartments. The model assumes that all seasonal nitrogen transformations at the field scale lead to the remaining pool of nitrate available for leaching at the beginning of the winter season (RNBW). This variable is updated by the model at the beginning of each hydrological year. The NIT-DRAIN model requires two input data: the measured or simulated drainage discharge and the RNBW. Seven parameters were introduced to manage nitrate leaching and water flow through the three soil compartments using two different transfer functions. The calibration and validation step (C/V) was performed to assess the performance of the model in reproducing the nitrate concentration observations acquired at Rampillon (355 ha, data for 6 years), located east of Paris, France. The results showed performance criteria of KGE greater than 0.5 and of RMSE less than 5 mgNO3-/l, which confirm the ability of the model to simulate the observed nitrate concentrations with a good agreement. The annual model calibration framework, implemented to assess the parameter variability, confirms the model stability and consistency. However, the use of simulated drainage discharges as input underlines the need to reproduce with accuracy the earlier low discharge values by the drainage model. The innovative management of nitrate transfer through the conceptual compartments based on the annual update of the RNBW variable enables the simulation of the nitrate concentrations with a high performance using a limited number of input data. In addition, the periods of flushing and dilution are clearly identified by the NIT-DRAIN model thanks to the accurate estimation of the nitrate storage variation inside each model compartment.