Digital twin implementation of the DSSAT cropping system model for crop simulation and optimization

Overview

This project explores implementing the DSSAT Cropping System Model (CSM) as a digital twin for agricultural applications. Digital twins are dynamic virtual representations of physical systems that integrate real-time data to simulate, predict, and optimize performance. The CSM is a widely used dynamic crop simulation platform that models the interactions between genotype, environment, and management to predict crop growth, development, and final yield. By converting CSM into a digital twin, the project aims to enable in-season optimization of crop management, support the design of climate-smart crop ideotypes, and extend simulation capabilities to crops outside the existing DSSAT ecosystem.

The initiative is led by a team responsible for advancing the DSSAT modeling ecosystem and its underlying source code, offering deep expertise in crop modeling, genetic integration, and agricultural data systems. Dry bean will serve as the model crop, building on prior work with the GeneGro model and recent advances linking quantitative trait loci to the CSM-CROPGRO-Dry bean model.

Technical specifications

Core platform:

• Built on the DSSAT Cropping System Model, a modular Fortran-based simulation engine known for computational efficiency
• Supports more than 40 crop modules with cultivar-level genetic responses
• Integrates gene-action linkages through the GeneGro framework and QTL-based modeling approaches

Digital twin capabilities:

• Ingestion and assimilation of near real-time environmental data from public databases and sensors
• Incorporation of observational data for continuous model calibration and validation
• Enhanced physiological-to-gene-action linkages using current scientific information
• Modular architecture designed for extensibility to crops not currently in the DSSAT ecosystem

Input requirements:

• Local weather and soil data
• Crop management records
• Genetic coefficients at species and cultivar levels

Open source availability:

• DSSAT and CSM source code released under the BSD-3-Clause license
• CSM source available via GitHub; DSSAT accessible through the DSSAT portal

Technology readiness level

The DSSAT Cropping System Model is a mature technology with extensive validation documented in hundreds of publications across crop management optimization, plant ideotype design, and gene-based simulation. The digital twin implementation represents an early-to-mid stage research effort, focusing on two parallel development tracks: real-time data assimilation and enhanced gene-action modeling. The project will produce a modular model framework ready for further validation and application expansion, with dry bean serving as the initial test case before generalization to additional crops.