A data-driven assessment of agricultural sustainability in the EU using FAOSTAT data (1995-2024).
Sustainable biomass production is essential for the EU Green Deal and the European Bioeconomy Strategy. This project investigates how crop yields and land use have evolved across six EU agricultural producers (Germany, France, Italy, Spain, Portugal, and the Netherlands) over three decades. The analysis aims to identify which countries have achieved productivity gains without expanding their agricultural land, and whether the historical yield gap between Northern and Southern producers has narrowed over time.
- How have crop yields evolved in the EU over the last 30 years?
- Which EU countries achieve higher yields without expanding agricultural land?
- Are Southern EU countries converging or diverging from Northern EU yield levels?
- Convergence in grapes and maize: The Southern producers have closed the yield gap with the North. In maize, Spain has overtaken the Northern average since around 2015.
- Persistent divergence in sugar beet and wheat: The Northern yields remain considerably higher, and the gap is structural rather than narrowing.
Efficiency gains differ across crops:
- Grapes: Spain, Portugal, and Italy improved yields while simultaneously reducing harvested area.
- Maize: Spain and Portugal increased yields while reducing harvested area. Italy and France also reduced their harvested area, but their yield improvements were relatively modest.
- Sugar beet: All countries, except Portugal, achieved higher yields alongside reductions in harvested area. Portugal was excluded from this comparison due to missing values from 2017 onward.
- Wheat: Italy and Portugal increased yields while reducing harvested area. Spain also recorded a substantial yield increase; however, the reduction in harvested area was minimal.
- Data cleaning and interpolation of missing values within each country-crop-element group.
- Weighted regional yield computed as the sum of production divided by the sum of area harvested. This approach ensures that larger producers contribute proportionally to the regional figure.
- Three-year averages and rolling means to separate structural trends from year-to-year noise.
- Visualisation with Seaborn (
relplot, faceted scatter and line charts).
Python (Recommended Python version: 3.11+), Jupyter Notebook, pandas, NumPy, matplotlib, seaborn
git clone https://github.com/alirezahsni98/eu-crop-yield-analysis.git
cd eu-crop-yield-analysis
pip install -r requirements.txt
jupyter labThen run the notebooks in order (01 to 04).
eu-crop-yield-analysis/
├── data/
│ ├── raw/
│ │ └── FAOSTAT_data_en_1-8-2026.csv
│ └── processed/
│ ├── eu_stat_cleaned.csv
│ └── eu_stat_joined.csv
├── notebooks/
│ ├── 01_data_loading_cleaning.ipynb
│ ├── 02_yield_trends.ipynb
│ ├── 03_yield_efficiency.ipynb
│ └── 04_north_south_gap.ipynb
├── figures/
│ ├── grapes_yield_evolution.png
│ ├── maize_corn_yield_evolution.png
│ ├── sugar_beet_yield_evolution.png
│ ├── wheat_yield_evolution.png
│ ├── yield_vs_area_change_by_crop.png
│ └── north_south_yield_gap.png
├── README.md
├── LICENSE
└── requirements.txt
FAO. 2026. FAOSTAT Statistical Database. Accessed 8 January 2026. https://www.fao.org/faostat/. Licence: CC BY-4.0.
The findings indicate that the EU's productivity gains are unevenly distributed across crops. Two crops (grapes and maize) show genuine convergence, which suggests that productivity-enhancing technology and practices have transferred southward effectively. Conversely, two crops (wheat and sugar beet) show persistent gaps that are likely climatic in origin and unlikely to close through technology alone. Therefore, a sustainable EU biomass strategy should differentiate by crop. It should invest in Southern productivity for crops where convergence is achievable, and rely on regional specialisation where it is not.
The following peer-reviewed sources and institutional reports informed the discussion and interpretation of findings in this project.
On EU crop yield trends and climate fingerprints:
Moore, F. C., & Lobell, D. B. (2015). The fingerprint of climate trends on European crop yields. Proceedings of the National Academy of Sciences, 112(9), 2670-2675. https://doi.org/10.1073/pnas.1409606112
Hristov, J., Toreti, A., Pérez Domínguez, I., Dentener, F., Fellmann, T., Elleby, C., Ceglar, A., Fumagalli, D., Niemeyer, S., Cerrani, I., Panarello, L., & Bratu, M. (2020). Analysis of climate change impacts on EU agriculture by 2050 (EUR 30078 EN). Publications Office of the European Union. https://doi.org/10.2760/121115
On maize C4 physiology and temperature thresholds:
Crafts-Brandner, S. J., & Salvucci, M. E. (2002). Sensitivity of photosynthesis in a C4 plant, maize, to heat stress. Plant Physiology, 129(4), 1773-1780. https://doi.org/10.1104/pp.002170
Schlenker, W., & Roberts, M. J. (2009). Nonlinear temperature effects indicate severe damages to US crop yields under climate change. Proceedings of the National Academy of Sciences, 106(37), 15594-15598. https://doi.org/10.1073/pnas.0906865106
On climate change impacts on European viticulture:
Fraga, H., García de Cortázar Atauri, I., Malheiro, A. C., & Santos, J. A. (2016). Modelling climate change impacts on viticultural yield, phenology and stress conditions in Europe. Global Change Biology, 22(11), 3774-3788. https://doi.org/10.1111/gcb.13382
van Leeuwen, C., Sgubin, G., Bois, B., Ollat, N., Schultz, H. R., Zavaglia, V., & Quénol, H. (2024). Climate change impacts and adaptations of wine production. Nature Reviews Earth and Environment, 5, 258-275. https://doi.org/10.1038/s43017-024-00521-5
On EU grain production and technology-driven yield gains:
Pinke, Z., Decsi, B., Jámbor, A., Kardos, M. K., Kern, Z., Kozma, Z., & Ács, T. (2022). Climate change and modernization drive structural realignments in European grain production. Scientific Reports, 12, 7374. https://doi.org/10.1038/s41598-022-10670-6
While this project provides a data-driven assessment of crop yield trends and land-use efficiency across six EU countries, several limitations should be considered when interpreting the findings.
Initially, the analysis is restricted to six countries and four crops. While these were selected to represent the major EU agricultural producers and economically significant crops, they do not capture the full diversity of the EU-27. Consequently, the findings should not be generalised beyond the scope of this study without further investigation.
Secondly, the dataset does not include variables such as fertiliser application, irrigation infrastructure, soil quality, climate data, or cultivar information. While these factors are crucial for a comprehensive understanding of yield determinants, they were not available within the FAOSTAT extract used in this project. Therefore, the observed yield trends could be influenced by factors that this analysis does not account for.
Additionally, the annual yield was computed as production divided by harvested area, rather than imported directly as a pre-calculated variable. While this approach is methodologically sound and consistent with standard agronomic definitions, it implies that any measurement error in either the production or area figures could propagate into the yield estimates.
Furthermore, the North-South regional grouping is a simplification. Germany, France, and the Netherlands were grouped as Northern producers, while Italy, Spain, and Portugal were grouped as Southern producers. While this grouping reflects broad agro-climatic differences, it does not consider the considerable internal variability within each region.
Lastly, Portugal's sugar beet data required special handling due to missing values from 2017 onward. While the missing values were addressed through interpolation during the cleaning phase, Portugal was excluded from the efficiency comparison for sugar beet to avoid misleading conclusions.
AI assistance tools, including ChatGPT and Claude, were used for code review, visualization refinement, and documentation support; the research design, implementation decisions, and interpretation remain my own.
This project is licensed under the MIT License. See LICENSE for details.
Alireza Hosseini
LinkedIn: https://www.linkedin.com/in/alirezahs