Crop Modeling for Yield Prediction Training Course

Duration 5 days 10 modules · 33 contact hours
Level Intermediate Certificate on completion
Who attends Agronomists, Agricultural researchers and others
Fee from USD 1,100 KSh 95,000 for Kenyan organisations
Where 52 cities + online See dates by city
Next start 9 Nov 2026 Course code SD-AA-051

Before you join: Participants should understand basic crop production concepts, including planting dates, crop stages, fertiliser, irrigation, soils and yield measurement.

Dates & prices

Two start dates a month in every city, and live online on weekdays or weekends. Fees are per participant; classroom fees include tuition, materials, lunch and refreshments, and the certificate.

Suggest a date or city

Showing the next six months. Later dates, other cities and private deliveries for your team are available on request.

Course overview

Crop yield estimates often guide seed and fertiliser procurement, irrigation allocation, food-security early warning, crop insurance, contract farming and market planning. Yet estimates can fail when weather records are incomplete, soil profiles are poorly defined, cultivar parameters are borrowed without testing, or model outputs are reported as certain predictions. Crop modelers need to connect field observations, weather data and management records to defensible projections that decision-makers can use.

This five-day course develops the technical and analytical skills required to build, calibrate, validate and apply crop simulation models for yield prediction and management decisions. Participants work with process-based modeling approaches used in DSSAT, APSIM and AquaCrop, covering weather and soil inputs, crop genotype and management files, calibration against observed data, validation metrics, sensitivity analysis and scenario design. The course also addresses seasonal climate scenarios, irrigation and nitrogen management, remote-sensing observations, district-level aggregation and communication of uncertainty.

Teaching combines focused technical briefings with guided software demonstrations, agricultural case work, group analysis and structured model interpretation. Cases reflect common decisions in commercial farming, agricultural extension, development programmes and food-security monitoring. Participants leave with a reusable crop-modeling workflow package, including input-data specifications, calibration and validation templates, scenario matrices, reporting formats and an application plan for a priority crop or production area.

The course is designed for agricultural professionals who already work with crop production data, agronomy, irrigation, climate information or agricultural programmes and need stronger methods for converting those data into documented yield forecasts and management advice.

What you will be able to do

  • Prepare weather, soil, crop and management datasets for process-based crop models
  • Configure DSSAT, APSIM and AquaCrop simulations for defined production systems
  • Calibrate cultivar and management parameters against observed phenology, biomass and yield
  • Validate model predictions using RMSE, bias, Willmott's d and residual analysis
  • Design weather, planting, irrigation and nitrogen scenarios for management decisions
  • Quantify prediction uncertainty through sensitivity analysis and Monte Carlo sampling
  • Interpret remote-sensing and field observations alongside simulated crop development
  • Produce a documented yield prediction brief with assumptions, ranges and action triggers

Benefits of attending

For you

  • Build credible yield forecasts from observed field, weather and soil data
  • Diagnose whether poor predictions arise from data, parameters or model structure
  • Compare planting, irrigation and fertiliser options before recommending interventions
  • Report forecast uncertainty without overstating model precision
  • Create an evidence trail for technical reports, donor submissions and management decisions

For your organisation

  • Improve traceability of seasonal production estimates and agricultural advisories
  • Test crop management options before committing scarce water, seed or fertiliser resources
  • Reduce reliance on undocumented expert judgement for yield forecasting
  • Strengthen food-security, procurement and insurance decisions with stated uncertainty ranges
  • Establish repeatable procedures for model inputs, validation, reporting and annual updates

Course modules

10 modules over 5 days

Participants can select an appropriate crop-modeling approach for a defined yield prediction or management decision.

  • Yield formation concepts: phenology, biomass partitioning, harvest index and sink limits
  • Model classes: empirical regressions, process-based models, remote-sensing and hybrids
  • Decision questions framed for seasonal forecasts, interventions, procurement and insurance
  • Spatial and temporal scales matched to farm, district and national use
  • Minimum data requirements distinguished from desirable measurements and assumptions
  • FAO crop modeling terminology: state variables, parameters, drivers and outputs

Participants can prepare and quality-check weather and soil inputs required for crop simulation.

  • Weather quality control using range checks, consistency tests and gap flags
  • Reference evapotranspiration calculated with FAO-56 Penman-Monteith procedures
  • Solar radiation estimated from sunshine hours using Ångström-Prescott relationships
  • Soil profile inputs: texture, bulk density, organic carbon, field capacity
  • Plant available water derived between field capacity and wilting point
  • Missing rainfall and temperature handled through documented infilling protocols

Participants can assemble crop, cultivar and field-management records for model configuration and calibration.

  • Cultivar coefficients linked to thermal time, photoperiod and phenological stages
  • Management files specify sowing, density, fertiliser, irrigation and residue practices
  • Rooting depth, emergence, flowering and maturity observations prepared for calibration
  • Crop cuts convert quadrat samples into dry biomass and yield records
  • Treatment histories reconciled with plot maps, dates and input quantities
  • FAO-56 crop coefficients interpreted alongside model-specific water stress functions

Participants can configure baseline simulations using DSSAT, APSIM and AquaCrop model structures.

  • DSSAT experiment files structured around weather, soil, genotype and management
  • APSIM simulations configured through soil, crop, clock and manager modules
  • AquaCrop projects define canopy cover, water productivity and irrigation schedules
  • Model selection matched to crop, data availability, decision horizon and users
  • Baseline scenarios documented with versioned inputs, assumptions and run identifiers
  • Batch runs organised using parameter sets and scenario naming conventions

Participants can calibrate crop-model parameters using observed crop development, biomass and yield data.

  • Calibration targets prioritised using phenology, biomass, yield and soil moisture
  • Sequential calibration adjusts phenology before growth, yield and water parameters
  • Objective functions compare observed and simulated values across measured variables
  • RMSE, nRMSE, MAE and bias quantify model error consistently
  • Parameter bounds prevent biologically implausible cultivar and soil values
  • Calibration records preserve data sources, changes, rationale and model versions

Participants can evaluate prediction quality and communicate the uncertainty associated with model outputs.

  • Independent datasets separated from calibration records before validation begins
  • Willmott's d, R-squared and Nash-Sutcliffe efficiency interpreted carefully
  • Residual plots reveal systematic errors by season, soil or treatment
  • Cross-validation assesses transferability where multiple sites or seasons exist
  • Sensitivity analysis ranks influential weather, soil, cultivar and management inputs
  • Monte Carlo sampling expresses yield ranges rather than single-point predictions
  • Validation reports distinguish predictive skill, limitations and intended uses

Participants can design model scenarios for seasonal planning and crop-management decisions.

  • Seasonal forecast products translated into rainfall, temperature and onset scenarios
  • Ensemble weather sequences generated from historical analogues and forecast probabilities
  • Planting-window simulations compare establishment risk, maturity dates and expected yield
  • Irrigation scheduling scenarios test timing, depth and water allocation constraints
  • Nitrogen response curves identify economically relevant application rates and timings
  • Scenario matrices separate controllable management choices from external climate uncertainty

Participants can combine field, satellite and model information for spatial yield assessment.

  • Satellite vegetation indices compared with simulated canopy development and biomass
  • NDVI and LAI observations checked for clouds, timing and spatial mismatch
  • Yield gaps partitioned into water, nutrient, management and genetic constraints
  • District aggregation weights field simulations by crop area, soils and practices
  • Ground-truth protocols align harvest samples, coordinates, dates and sensor observations
  • Data assimilation concepts distinguish state updating from parameter recalibration

Participants can present yield predictions with clear assumptions, confidence ranges and decision implications.

  • Prediction briefs state forecast date, area, crop stage and confidence range
  • Maps and fan charts communicate spatial variation and scenario uncertainty
  • Trigger thresholds connect predicted outcomes to procurement, irrigation or advisory actions
  • Assumption registers expose data gaps, model limits and interpretation risks
  • Ethical communication avoids false precision in food-security and insurance decisions
  • Stakeholder review meetings test whether outputs answer operational decision questions

Participants can establish a repeatable workflow for routine crop-model updates and reporting.

  • Operational calendars assign data collection, run dates, review points and owners
  • Standard operating procedures define input checks, approvals, archives and releases
  • Model performance dashboards track error metrics across crops, locations and seasons
  • Recalibration triggers set thresholds for new varieties, soils or management shifts
  • Application plans specify one priority decision, dataset and implementation timetable
  • Peer review checklists support reproducible analyses before management reporting

Exercises and examples are adapted to each cohort's sectors and roles.

What you take back to work

  1. Crop-model input data specification template
  2. Cultivar calibration and validation worksheet
  3. Yield forecast accuracy metrics dashboard
  4. Seasonal management scenario matrix
  5. Prediction uncertainty reporting brief
  6. Operational crop-modeling application plan

Who should attend

  • Agronomists — need defensible yield forecasts and management scenario analysis
  • Agricultural researchers — need repeatable methods for crop simulation studies
  • Food-security analysts — need documented production estimates for early-warning decisions
  • Irrigation specialists — need to compare water allocation and scheduling options
  • Agricultural extension managers — need evidence for seasonal farmer advisory services
  • Agribusiness planning managers — need crop outlooks for procurement and supply planning

Skills and tools covered

Crop simulation modelingModel calibrationYield forecast validationClimate scenario analysisIrrigation scenario planningUncertainty communication

Tools and standards

DSSAT, APSIM, AquaCrop, FAO-56 Penman-Monteith

How the course runs

Training methodology

The course uses short technical briefings followed by guided interpretation of crop-model inputs, outputs and performance diagnostics. Participants examine agricultural cases involving rainfall variability, irrigation limits, cultivar choice and nutrient management. Group work focuses on checking data quality, setting calibration priorities, interpreting validation results and comparing scenarios. Trainer-led simulations show how DSSAT, APSIM and AquaCrop represent crop growth and water stress. The final sessions convert technical results into forecast briefs, decision triggers and individual application plans.

Requirements and prerequisites

Participants should understand basic crop production concepts, including planting dates, crop stages, fertiliser, irrigation, soils and yield measurement. Experience handling agricultural datasets in spreadsheets is expected, along with confidence reading tables, charts and simple descriptive statistics. Familiarity with weather data, GIS, R or crop simulation software is useful but not required. The course explains model configuration and performance metrics from first principles. Advanced programming, prior use of DSSAT, APSIM or AquaCrop, and formal training in statistics are not required.

A typical training day

08:30 – 10:30First session
10:30 – 10:45Refreshment break
10:45 – 12:30Second session
12:30 – 13:30Lunch and networking
13:30 – 15:00Third session
15:00 – 15:15Refreshment break
15:15 – 16:30Workshop and daily review

Live online deliveries follow the same structure in East Africa Time, with shorter screen blocks and longer breaks.

What the fee includes

  • Instruction by a practitioner facilitator
  • Full course workbook and materials
  • Exercise files, templates and case studies
  • Refreshments and lunch on classroom deliveries
  • Certificate of completion
  • Post-course application plan

Frequently asked questions

Participants need basic knowledge of crop production, soils, weather and yield measurement, plus confidence using spreadsheets. Previous experience with crop-modeling software, GIS, programming or advanced statistics is not required.

Bring a laptop with spreadsheet software and, where possible, non-confidential examples of weather, soil, crop or yield data from your work. Training datasets and guided model files are supplied for the classroom work.

It suits agronomists, researchers, food-security analysts, irrigation specialists, extension managers and agribusiness planners. It is intended for professionals who need to interpret or produce crop yield predictions, rather than for participants seeking basic agronomy training.

General agronomy courses focus on crop husbandry, while precision agriculture courses commonly focus on sensors, machinery and spatial field operations. This course focuses on using weather, soil, crop and management data within simulation models to estimate yield and test future scenarios.

Participants can use the workflow to prepare model inputs, assess prediction accuracy and compare management scenarios for a priority crop or location. The reporting formats also support communication with programme managers, farm managers, procurement teams and technical partners.

The course runs from 08:30 to 16:30 each day, providing about 6.5 contact hours daily. Time is divided across model concepts, data preparation, software-guided analysis, calibration and validation, scenario work, communication and workflow planning.

Questions about this course?

Ask about dates, fees, group bookings or running it for your team. Our training team replies by email.

Ask on WhatsApp info@skillsetdevelopment.com

Related courses in Agriculture & Agribusiness