Calibration of the APEX Model to Simulate Management Practice Effects on Runoff, Sediment, and Phosphorus Loss

Area of Science:

• Environmental Science
• Agricultural Engineering
• Computational Modeling

Background:

• Process-based computer models are increasingly used for Phosphorus (P) Index assessment.
• The reliability of models simulating P loss under novel agricultural managements remains largely untested.

Purpose of the Study:

• To evaluate the Agricultural Policy Environmental eXtender (APEX) model's accuracy in simulating P loss with varying managements.
• To assess model performance when applied to data outside its calibration set.

Main Methods:

• APEX model calibration using field-scale data from eight managements across two locations.
• Validation of management-specific and location-wide models against observed runoff, sediment, total P, and dissolved P loss.
• Performance evaluation using Nash-Sutcliffe efficiency and percent bias metrics.

Main Results:

• Management-specific APEX models achieved satisfactory performance (>0.50 NSE, ±35% bias for runoff) within their calibration managements.
• Model accuracy decreased substantially when applied to different managements, meeting criteria in only one-third of tests.
• Location models, calibrated with diverse management data, demonstrated superior performance across all tested managements compared to management-specific models.

Conclusions:

• APEX model predictions for P loss are reliable only within the scope of the calibration management.
• Broadening calibration datasets to include multiple management systems improves model applicability for assessing P loss and P Index evaluation.
• Caution is advised when extrapolating model results beyond their calibrated management conditions.