Refining the soil and water component to improve the MoSt grass growth model

Knowledge of previous and future grass growth is an important factor for grassland management decision making. It allows the farmer to predict the availability of grass for the herd on a short-term basis and adapt grassland management practise accordingly. The Moorepark St Gilles Grass Growth Model (MoSt GG) is used to predict grass growth weekly on 84 grassland farms across Ireland. The repeated use of the model on these farms has identified areas for improvement that have been addressed in this paper. Among these improvements, the soil sub-model component has been further developed to better represent different soil types and to account for different soil depths, improving the simulations of water and soil nitrogen fluxes (V2 V1 +soil ). A soil sub-layer of 10 cm was added to better simulate growth recovery after a drought period (V3 V2+water ). The radiation component was improved by including the day length in the grass growth estimation (V4 V3+rad ) instead of only accounting for daily cumulative solar radiation. These improvements were evaluated against several experiments conducted in Ireland and France. The developments improved model accuracy for every experiment evaluated. The RMSE in the original version of the model ranged from 322 to 1011 kg of DM/ha, whereas in the latest version of the MoSt GG model (V4 V3+rad ), the RMSE ranged from 312 to 671 kg of DM/ha. The further consideration of soil characteristics resulted in a higher variability in grass production and N leaching depending on soil type and weather conditions, leading to improved growth trend representation. The addition of the soil sub-layer (V3 V2+water ) improved the accuracy in drier years (French experiment) due to the more realistic grass growth recovery after a drought. The latest version of the model (V4 V3+rad ) simulates grass production more accurately than the previous versions and increases the reliability of grass growth prediction. • The soil and drought recovery aspects of the MoSt GG model were improved. • Better representation of water fluxes helped to distinguish leaching risk among soils. • Adding a topsoil sub-layer improved simulation of grass growth recovery post-drought. • Different versions of the model were assessed using empirical data from multiple sites. • The final version of the model showed improved accuracy and growth trend representation.