APSIM Maize Growth Simulation
From preparing site data to running the APSIM NG maize model, analysing how climate and management practices affect yield.
Steps
Prepare weather data
Python / ExcelFormat daily site weather data into APSIM .met files with variables such as maximum temperature, minimum temperature, rainfall and radiation.
Check missing values, year range and site names first, so APSIM doesn't fail to find the matching .met file.
Configure soil parameters
APSIM NGSelect or build the soil profile for the site and check that parameters such as LL15, DUL, SAT and BD are physically plausible.
SAT should exceed DUL, and DUL should exceed LL15; the number and depth of layers should match the crop root settings.
Set management practices
APSIM NGConfigure management rules for sowing date, plant density, cultivar, fertilisation, irrigation and harvest conditions.
Get one representative site running first, then scale out to multiple sites.
Calibrate and validate the model
APSIM NG + RCompare simulated and observed yields and adjust cultivar parameters, RUE, grain filling parameters and management parameters.
Don't look only at R²; also check bias, RMSE and year-by-year errors.
Run batch scenarios
APSIM ToolkitBatch-run different sites, years, cultivars, irrigation and fertilisation schemes to produce a unified results table.
Keep the configuration files and logs for each run so results can be traced back.
Analyse and visualise results
R / PythonPlot observed-vs-simulated scatter plots, interannual variation, yield potential and yield gap charts.
For paper figures, keep fonts, font sizes, axis units and legend formats consistent.