Climate Intelligence for Food

The project developed a prototype for regionalized life cycle assessments (LCA) that calculates and visualizes the environmental impacts of crops on a site-specific basis. In addition, a scalable data architecture and models for estimating regional crop yields were developed. The results provide a foundation for more precise environmental assessments that can be applied globally.

Niatsu developed a prototype that calculates and visualizes dynamic regional LCA results. This prototype serves as an example of how regionalized LCA models can be applied. It also highlights the relevance and added value of regional differentiation in environmental impact assessments. Specifically, Niatsu implemented regionalized LCAs for various crop species (such as wheat) in Switzerland, Austria, and Germany. In a second step, the model was expanded to include palm fruits and corn in Indonesia to demonstrate that the technology is also applicable on a global scale.

Niatsu developed a base model for regional LCA calculations, collected regional data on crop yields, fertilizer use, and emission factors, and processed this data into harmonized formats. The model was then implemented in an interactive visualization accessible via the cloud. Even at this early stage, the prototype showed that regional aggregations provide new insights: Significant fluctuations in environmental impacts can be observed within individual countries, confirming the need for dynamic LCA.
Niatsu further developed an approach to scale the method and integrate it into our system. To this end, we defined a software architecture that specifies how input data and results from the dynamic LCA are stored in a database. We identified the most important inputs for calculating regional environmental impacts—including crop yields, nitrogen fertilizers, and phosphorus fertilizers—and structured them into a clear schema. This work made the system scalable, transparent, and expandable.

To reduce our reliance on official statistics, we developed a yield engine that estimates crop yields at the regional level. By combining global datasets with satellite data, it provides consistent yield estimates with global coverage. This allows us to fill data gaps, enable comparisons between regions, and provide reliable inputs for our environmental assessments. These yield data are integrated into the database architecture described above, thereby strengthening the robustness of the overall system. Niatsu now has over 40 models for crops with global coverage. However, further validation of the results is still pending.

In addition, we have initiated initial research and the processing of global irrigation data at the regional level, with the goal of integrating this data into the dynamic LCA model. This will further increase the accuracy of regional impact assessments, as it takes into account differences in water consumption between crops and regions.