Crop Simulation Modeling
Also known as: Crop Growth Simulation, Process-Based Crop Modeling, Crop Systems Modeling, Dynamic Crop Modeling
Crop simulation modeling uses process-based, dynamic computer models to predict how a crop grows and yields under specified weather, soil, and management, by numerically integrating mechanistic equations for development, photosynthesis, and water and nutrient balances on a daily time step. The two most widely used platforms are DSSAT, documented by James Jones and colleagues in 2003, and APSIM, whose modern architecture was described by Dean Holzworth and colleagues in 2014. Rather than fitting a statistical curve to yield data, these models encode the underlying biophysics — temperature-driven phenology, radiation-use efficiency, soil water and nitrogen dynamics — so they can extrapolate to weather, soils, and management combinations never directly observed. This makes crop models powerful tools for in silico experimentation, scenario analysis, and climate-change and management impact assessment where field trials alone would be impossibly slow or costly.
Key highlights
- Encodes biophysical processes, so it can extrapolate to weather, soils, and management never directly observed.
- Enables cheap, repeatable in silico experiments across vast input spaces impossible to cover with field trials.
- Couples crop, water, and nitrogen dynamics, supporting long-term rotation, resource-use, and risk analysis.
- Provides a rigorous framework for climate-impact, yield-gap, and ex ante management and technology assessment.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use crop simulation modeling when you need to predict crop and soil responses across combinations of weather, soil, and management that field trials cannot exhaustively cover — for climate-change impact and adaptation studies, designing fertilizer or irrigation strategies, yield-gap analysis, ex ante technology evaluation, or extrapolating trial results to new sites and years. It is appropriate when the minimum input data (daily weather, characterized soils, cultivar parameters, management) can be assembled and when at least some observed data exist for calibration and validation. It is less suitable when those inputs are unavailable or highly uncertain, when the question concerns processes the model does not represent (such as pests, diseases, or extreme events outside its mechanisms), or when a simple empirical relationship would answer the question more transparently. Models should complement, not replace, field experimentation.
Strengths & limitations
- Encodes biophysical processes, so it can extrapolate to weather, soils, and management never directly observed.
- Enables cheap, repeatable in silico experiments across vast input spaces impossible to cover with field trials.
- Couples crop, water, and nitrogen dynamics, supporting long-term rotation, resource-use, and risk analysis.
- Provides a rigorous framework for climate-impact, yield-gap, and ex ante management and technology assessment.
- Predictions are only as good as the input data and the cultivar calibration, which are often incomplete or uncertain.
- Models omit or simplify important processes (pests, diseases, weeds, extreme events), limiting validity outside their scope.
- Calibration can mask structural error, and good fit to one dataset does not guarantee transferability.
- Different models can disagree substantially under novel conditions such as elevated carbon dioxide, creating model-choice uncertainty.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
How does a process-based crop model differ from a statistical yield model?
A statistical model fits an empirical relationship — say, yield as a regression on rainfall and temperature — to observed data, and is reliable mainly within the range of conditions it was fitted to. A process-based crop model instead encodes the underlying biophysics: thermal-time phenology, radiation-driven growth limited by water and nitrogen, and soil water and nutrient balances integrated daily. Because the causal mechanisms are built in, it can extrapolate to management, soils, and climates never observed, which is exactly why it is used for climate-change and scenario analysis. The cost is greater data and calibration demands and the risk of structural error.
What inputs do DSSAT or APSIM need to run?
At minimum, daily weather (maximum and minimum temperature, solar radiation, and rainfall), a characterized soil profile with layered water-holding and nitrogen properties, cultivar-specific genetic coefficients that tune development and growth to a particular variety, and a management schedule covering sowing date and density, fertilizer, irrigation, and residue. Jones and colleagues call this the minimum data set. Output quality depends heavily on these inputs, and at least some observed phenology and yield are needed to calibrate the cultivar coefficients and validate the model before trusting its scenario predictions.
Can crop models be trusted under future climates they were never tested on?
With caution. Models extrapolate via their mechanisms, which is their strength, but novel conditions such as elevated carbon dioxide can expose structural differences, and different well-calibrated models sometimes diverge under these conditions. Best practice, formalized in efforts like AgMIP, is to validate models against the widest available data, run multiple models as an ensemble to expose this structural uncertainty, and report the spread rather than a single number. A model's projection is most credible when the change being studied stays within the range of processes the model represents and has been evaluated against.
Sources
- 1.Jones, J. W., Hoogenboom, G., Porter, C. H., Boote, K. J., Batchelor, W. D., Hunt, L. A., Wilkens, P. W., Singh, U., Gijsman, A. J., & Ritchie, J. T. (2003). The DSSAT cropping system model. European Journal of Agronomy, 18(3-4), 235-265.
- 2.Holzworth, D. P., Huth, N. I., deVoil, P. G., Zurcher, E. J., et al. (2014). APSIM - Evolution towards a new generation of agricultural systems simulation. Environmental Modelling & Software, 62, 327-350.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 23). Crop Simulation Modeling. ScholarGate. https://scholargate.app/food-agriculture-studies/crop-simulation-modeling