Location: Food Animal Environmental Systems Research
Title: Incorporating model uncertainty into predictions of soil test PAuthor
|
Submitted to: European Geosciences Union General Assembly Proceedings
Publication Type: Abstract Only Publication Acceptance Date: 1/25/2023 Publication Date: 4/28/2023 Citation: Bolster, C.H., Vadas, P.A. 2023. Incorporating model uncertainty into predictions of soil test P. European Geosciences Union General Assembly Proceedings. https://doi.org/10.5194/egusphere-egu23-7145. DOI: https://doi.org/10.5194/egusphere-egu23-7145 Interpretive Summary: Technical Abstract: In this study we conducted an uncertainty analysis using the Annual P Loss Estimator (APLE) model on predictions of STP. Using Monte Carlo simulations, we predicted P drawdown at multiple sites in Maryland, USA, under two cropping systems with different histories of P application over a 14-yr period. To incorporate model input error into prediction uncertainties, we assumed a triangular distribution of errors in all model inputs with a range of ± 20%. We then evaluated whether measured and assumed uncertainties in model factors resulted in model-prediction uncertainties that were similar in magnitude to measured variability in STP between replicates. While model prediction uncertainties were generally greater than the variability observed for measured STP, the mean measured STP for nearly all site years fell within the 95% confidence intervals of the STP predictions. Results from this study show the importance of accounting for model uncertainties when predicting long-term drawdown of STP. |
