Simulating scenarios under climate change is essential to understanding vegetation dynamics, ensuring the survival of forage species, and minimizing uncertainties in project costs and timelines. This study aimed to simulate historical probabilities and develop a biomass production model using PHYGROW software (Texas A&M University, College Station, TX, USA), combined with Monte Carlo Simulation (MCS) in the @RISK program (Ithaca, NY, USA), to evaluate long-term biomass production in a native pasture area of the Caatinga biome. The results show strong agreement between software estimates and field data. For 2016, PHYGROW estimated 883 kg/ha, while field measurements reached 836.8 kg/ha; for 2017, 1117 kg/ha was estimated, while 992.15 kg/ha was observed. For 2018, the model estimated 1200 kg/ha compared with 1763.5 kg/ha in the field, and for 2019, 1230 kg/ha was estimated versus the 1294.3 kg/ha observed. The Monte Carlo simulations indicated that the Weibull distribution best fitted the synthetic series, with 90% adherence. Biomass production values ranged from 618 to 1427 kg/ha with a 90% probability. Only 5% of the simulations projected values below 600 kg/ha or above 1400 kg/ha. Moreover, there was a 95% risk of production issues if planning was based on biomass values above 1000 kg/ha. These findings highlight PHYGROW's potential for pasture management under semi-arid conditions for predicting and avoiding degradation scenarios that could even lead to areas of desertification.