Effect of slurry spreading time-of-day on ammonia emissions from cattle slurry under Irish weather conditions

This dataset contains the full output of a Monte Carlo simulation study evaluating the effect of slurry spreading time-of-day on ammonia emissions from cattle slurry under Irish weather conditions. Ammonia emission fractions were predicted using the ALFAM2 process-based volatilisation model (Hafner et al., 2024; R package version as documented in the companion paper). Three spreading start times were compared: morning (06:00), afternoon (14:00), and evening (20:00). Spreading scenarios were constructed from 22 Met Éireann synoptic weather stations for the period 2013–2025. Agronomically realistic spreading dates were identified using a soil moisture deficit (SMD) model (Schulte et al., 2005) and filtered to exclude days with rainfall exceeding 0.5 mm h⁻¹, in line with the Irish Nitrates Action Programme. For each scenario, a 168-hour (7-day) weather sequence was extracted beginning at the spreading start time, and ALFAM2 was run to predict cumulative ammonia emission over this period. Uncertainty was characterised using two sources: (1) ALFAM2 parameter uncertainty, represented by 101 parameter sets (the central set alfam2pars03 plus 100 bootstrap sets from alfam2pars03var); and (2) variability in slurry composition and application rate, represented by 50 Monte Carlo draws per parameter set from the distributions documented in slurry_distributions_corrected_1.csv. This yields 5,050 individual ALFAM2 runs per scenario × time-of-day combination, and approximately 20.6 million rows in total in mc_results_master.csv. Slurry was characterised as unacidified cattle slurry applied by trailing shoe, consistent with the dominant spreading practice in Ireland. The raw Met Éireann hourly weather data used as model inputs are not included in this deposit (subject to Met Éireann data licensing); they may be obtained directly from Met Éireann (https://www.met.ie/climate/available-data/historical-data). The R scripts used to generate this dataset are archived separately.