Modeling Nutrient Concentrations in a Greenhouse Cascade Soilless System

A dynamic simulation model was developed and calibrated to predict nutrient concentrations in a three-level cascade hydroponic system (CasH). The performance of the model was evaluated across donor/primary (tomato) and receiver secondary (mint, peppermint), and tertiary (sea fennel, lemon balm) crops. An independent dataset was available for both calibration and evaluation of the primary tomato crop, whereas data for the secondary and tertiary stages were used exclusively for model calibration. Following model evaluation, three distinct operational scenarios were simulated, ranging from a baseline monoculture to a comprehensive three-stage cascade system. The simulations forecasted the progressive nutrient accumulation across stages, identifying the operational point at which critical sodium and salinity thresholds are approached. The proposed model shows potential as a foundational component for digital twins of hydroponic systems, complementing rather than replacing physical sensors by providing dynamic hourly simulations of nutrient dynamics. The model serves as an effective tool for optimizing resource efficiency and mitigating salinity risks in sustainable, closed-loop horticultural systems. These findings confirm the feasibility of nutrient-specific tracking in closed-loop horticulture, though independent evaluation was restricted to the primary stage, and the empirical coefficients remain specific to the evaluated three-stage CasH configuration.