When the Ecowheataly model starts up, thousands of Italian farms are placed inside a large digital map: from the rolling hills of the Marche region to the plains of Puglia, all the way to mountain farms, every corner of Italy enters the simulation with its own distinct identity.
This is the starting point of the model presented at the 4th CREA Meeting: a simulation that blends real data, artificial farms and global market dynamics to understand how agricultural policies can shape the future of Italian durum wheat.
Where we begin: a complete picture of the territory
To build a credible agricultural representation of Italy, the model relies on the same datasets used in national research:
• 1,846 real RICA farms
• 121 province–altimetry combinations from ISTAT
• 242 artificial farms created to fill the territorial gaps where RICA has no observations
This ensures uniform coverage across all provinces and altimetric zones.
The process is distributed across multiple “ranks,” each receiving a balanced number of farms (for example, 522 farms managed by rank 0 and 522 by rank 1).
The result: no part of Italy is left empty in the simulation.
How artificial farms are created
When real data are missing for a particular province, the model reconstructs coherent, territory-specific farms using:
- provincial agricultural parameters (yield, farmed area, altitude, statistical distributions)
- estimates derived from RICA and the Agricultural Census
- missing values filled through extraction from estimated distributions
It’s a way of generating “digital twins” of real farms: not random inventions, but structures shaped by the mathematical patterns of the Italian agricultural sector.
A multi-level structure: the model behaves like a real system
Italian agriculture does not operate in isolation. Global wheat prices, international policies and external shocks influence farmers’ decisions.
For this reason, the model is organized across two levels:
International level (Rank 0)
• simulates 12 global production areas
• simulates 24 demand areas
• manages prices, shocks and global policies
National level (Rank 1–N)
• includes all Italian farms
• receives international prices every month
• recalculates production strategies, also considering LCA results
It is a continuous dialogue: prices and policies flow from the top; production and environmental impacts rise from the bottom.
The monthly cycle: how farms make decisions
Every month, the model goes through three phases:
1. International market
Global supply and demand interact, generating new prices.
2. National decisions
Each Italian farm decides how much to sow, how to manage land and which techniques to adopt, balancing economic outcomes and environmental impact.
3. Feedback
Italian production updates the global supply, which in turn influences the prices of the following month.
A system that “breathes,” cycle after cycle.
Why this model is robust
The slides highlight several checks already completed:
- correct distribution of farms across ranks
- full geographic consistency
- reduced initialization times
- optimal computational load balancing
Next steps will include sensitivity analyses, climate shock simulations and comparison with historical RICA data to further validate the model’s performance.
What we will use it for
The framework is designed to answer real-world questions:
- What happens if an ecoscheme changes?
- How do farms react to new incentives or restrictions?
- How resilient is Italy’s wheat system to global or climatic shocks?
- Which techniques truly reduce environmental impacts?
It is a digital laboratory where agricultural policies can be tested without putting real production at risk.
In summary
The Ecowheataly model is a unique tool in Italy: it integrates real farms, artificial farms, global prices and LCA analyses into a dynamic system that simulates, month by month, how our wheat responds to a changing world.
A living, complex map built to understand how to truly support Italian cereal production.
Sources:
– Multi-Rank Model for the Simulation of the Agricultural Market – by Gianfranco Giulioni and Alessandro Ceccarelli