Agricultural Modeling Services

Predictive Models for Crops, Carbon & Climate

From process-based crop models to neural networks to hybrid approaches—we build the modeling solution that fits your prediction challenge.

Yield Prediction Carbon Quantification Geospatial Analysis Climate Scenarios
What We Build

The Full Modeling Spectrum

Every project is different. We meet you where you are and build what you need.

Process-Based Models

Mechanistic crop and ecosystem models grounded in plant physiology, soil science, and biogeochemistry.

  • APSIM, DSSAT, ecosys implementation
  • Crop growth & development simulation
  • Calibration & validation
  • Scenario analysis & projections

Data-Driven ML

Neural networks and statistical models trained on observational data for pattern recognition at scale.

  • Deep learning & neural networks
  • Geospatial & remote sensing ML
  • Time series forecasting
  • Satellite imagery analysis
Our Specialty

Hybrid KGML

The best of both: ML that learns from data while respecting physical and biological constraints.

  • Physics-constrained neural networks
  • Process model surrogates (7000x faster)
  • Knowledge-guided pretraining
  • Interpretable, scientifically consistent

We recommend the right approach for your data, timeline, and accuracy requirements—or combine them for optimal results.

Research Foundation

Peer-Reviewed Research Portfolio

Our methods are grounded in published, peer-reviewed research—ensuring scientific rigor and reproducibility in every project.

Uncertainty quantification maps for yield and nitrogen
Uncertainty Quantification

Constraining Large-Scale Yield from Crop Models with a Field-Scale Census

Novel approach combining field-scale census data with crop models to constrain regional yield predictions and quantify spatial uncertainty in agricultural forecasting.

Environmental Research Letters
Carbon reanalysis time series showing above ground woody biomass, LAI, and soil carbon
Data Assimilation

A Model-Data Fusion Approach for Improved Ecosystem Carbon Estimates

Data assimilation methods that fuse process-based ecosystem models with observations to produce improved carbon flux and stock estimates with uncertainty quantification.

Geoscientific Model Development
US maps showing biochar application response ratios and effectiveness probability
Carbon Sequestration

Spatial Targeting of Agricultural Soil Amendments

Machine learning framework for optimizing biochar application across the US, mapping response heterogeneity and effectiveness probability to guide targeted soil amendment strategies.

Environmental Research Letters
LAI time series comparing site-level, global, and HPDA optimization schemes
Hybrid Modeling

Remote Sensing-Integrated Crop Model Calibration

Integration of satellite-derived LAI data with APSIM crop model through hierarchical parameter data assimilation for improved regional yield prediction.

Remote Sensing (MDPI)

These peer-reviewed publications form the scientific foundation for our consulting work—bringing academic rigor to real-world agricultural challenges.

Prediction Outcomes

What You Can Predict

Turn your agricultural data into actionable predictions.

Yield Prediction

Field-level and regional yield forecasts accounting for weather, soil, and management.

Carbon Quantification

Soil carbon changes, GPP, and ecosystem carbon flux for sustainability reporting.

Environmental Outcomes

N₂O emissions, nutrient leaching, water use efficiency predictions.

Climate Scenarios

Projections under future climate including drought, heat stress, and extreme events.

Geospatial Analysis

Satellite imagery fusion, spatial variability mapping, remote sensing integration.

In-Season Forecasts

Real-time predictions that update as the season progresses with new data.

Let's Build Your Prediction Solution

Tell us about your data and prediction goals. We'll recommend the right modeling approach for your needs.

Start the Conversation