Agrometeorological Yield Model — Weather-Based Crop Yield Prediction
Agrometeorological Crop Yield Prediction Model · Also known as: crop yield model, agroclimatic yield model, weather-based yield model, meteorological crop model
An agrometeorological yield model is a quantitative framework that relates observed or forecasted weather variables — temperature, precipitation, solar radiation, humidity — to the final grain or biomass yield of a crop. Grounded in plant physiology and agricultural climatology, the approach is used worldwide in food security monitoring, insurance underwriting, irrigation planning, and climate-change impact assessment.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Agrometeorological yield models are appropriate when the research or operational goal is to explain, monitor, or forecast crop yields from weather data. They are well-suited to national and regional food security monitoring, early-warning systems for drought or heat-related crop failure, agricultural insurance loss estimation, and climate-change impact studies. The method requires at least a decade of co-located weather and yield records for calibration. It is less suitable when plot-scale precision is needed (favour process-based crop simulation models such as DSSAT or APSIM) or when weather data coverage is sparse and unreliable.
Strengths & limitations
- Operationally efficient — runs on routinely collected meteorological station data or gridded reanalysis products without requiring detailed soil or management inputs.
- Scalable from field to national level; widely used in FAO and WFP food security monitoring systems.
- Pre-harvest yield forecasting capability enables timely policy responses weeks before harvest.
- Transparent stress-response framework makes results interpretable to agronomists, policymakers, and insurers.
- Adaptable to both mechanistic (FAO water-balance) and statistical (regression, machine-learning) implementations.
- Calibration requires at least 10–15 years of reliable, co-located weather and yield data, which may be unavailable in data-sparse regions.
- Mechanistic variants assume crop parameters (Ky, rooting depth) are known and stable across environments; mis-specification degrades accuracy.
- Does not explicitly represent soil fertility, pest pressure, or agronomic management decisions, which can be large yield determinants.
- Statistical variants risk spurious correlations and poor out-of-sample performance when trained on short records or non-stationary climate series.
Frequently asked
What is the difference between an agrometeorological yield model and a crop simulation model like DSSAT?
Agrometeorological yield models relate weather variables to yield through relatively simple stress-response functions or statistical regressions; they are fast, data-light, and suited to regional monitoring. Crop simulation models such as DSSAT or APSIM simulate soil-plant-atmosphere processes at a daily time step and require detailed soil, cultivar, and management inputs. Simulation models are more mechanistically complete and accurate at field scale but are far more demanding in data and parameterisation effort.
How many years of data are needed to calibrate a reliable model?
A commonly cited minimum is 10–15 years of concurrent weather and yield observations. Shorter records increase the risk of overfitting and reduce the model's ability to capture inter-annual variability driven by rare extreme events such as severe droughts or late frosts. Where records are short, cross-validation and regularised regression are essential safeguards.
Can satellite data replace ground-based weather stations in these models?
Increasingly, yes. Satellite-derived products — MODIS NDVI, CHIRPS rainfall estimates, ERA5 reanalysis temperature — extend spatial coverage in regions with sparse station networks. However, satellite estimates carry their own uncertainties, particularly for precipitation, and should be validated against available station data before being used as the sole input.
Is the yield-response factor Ky fixed or does it vary by region?
Ky values published in FAO Irrigation Paper No. 33 are based on pooled experimental data and serve as starting-point estimates. In practice, Ky varies with cultivar, soil type, and management practices, so local calibration against observed yield data is recommended whenever sufficient data are available.
Sources
- Doorenbos, J., & Kassam, A. H. (1979). Yield Response to Water. FAO Irrigation and Drainage Paper No. 33. Food and Agriculture Organization of the United Nations, Rome. link ↗
- Lobell, D. B., & Burke, M. B. (2010). On the use of statistical models to predict crop yield responses to climate change. Agricultural and Forest Meteorology, 150(11), 1443-1452. DOI: 10.1016/j.agrformet.2010.07.008 ↗
How to cite this page
ScholarGate. (2026, June 3). Agrometeorological Crop Yield Prediction Model. ScholarGate. https://scholargate.app/en/agronomy/agrometeorological-yield-model