# Understanding the impacts of crop diversification in the context of climate change: a machine learning approach

**Authors:** Georgios Giannarakis, Ilias Tsoumas, Stelios Neophytides, Christiana Papoutsa, Charalampos Kontoes, Diofantos Hadjimitsis  
**arXiv:** [2307.08617](https://arxiv.org/abs/2307.08617)  
**Date:** 2023-07-17  
**Summary page:** https://www.summarizepaper.com/en/paper/2307.08617v1/

## Key points

- Sustainable intensification in agriculture prioritizes sustainability while maintaining productivity
- Effectiveness of sustainable intensification practices can vary depending on environmental conditions
- Crop diversification was investigated for its impact on productivity in the context of climate change
- Heterogeneous Earth Observation data and causal machine learning were used in the study
- Crop diversification had a significant positive effect on net primary productivity, increasing it by 2.8%
- The effect of crop diversification synergized well with higher maximum temperatures and lower soil moistures
- Crop diversification exhibits promising potential for adaptation in a warmer and more drought-prone climate
- Considering environmental conditions and climate change is important when implementing sustainable intensification practices
- Machine learning techniques and Earth Observation data provide valuable insights into the impact of crop diversification on productivity

## Summary

The concept of sustainable intensification in agriculture involves implementing management practices that prioritize sustainability while maintaining productivity. However, the effectiveness of these practices can vary depending on environmental conditions, which are expected to change due to climate change. In this study, conducted by Georgios Giannarakis, Ilias Tsoumas, Stelios Neophytides, Christiana Papoutsa, Charalampos Kontoes and Diofantos Hadjimitsis, the impact of crop diversification on productivity in the context of climate change is investigated using a machine learning approach. To understand how crop diversification may change in the future, the researchers utilized heterogeneous Earth Observation data and applied a data-driven approach based on causal machine learning. The study focused on the country of Cyprus over a four-year period. The findings revealed that crop diversification had a significant positive effect on net primary productivity of crops, increasing it by an average of 2.8%. Interestingly, this effect was found to synergize well with higher maximum temperatures and lower soil moistures. This suggests that in a warmer and more drought-prone climate, crop diversification exhibits promising potential for adaptation and can be considered as a sensible policy choice for enhancing agricultural productivity both presently and in the future. Overall, this research highlights the importance of considering environmental conditions and climate change when implementing sustainable intensification practices in agriculture. By leveraging machine learning techniques and Earth Observation data, valuable insights were gained regarding the impact of crop diversification on productivity. These findings contribute to our understanding of how agricultural systems can adapt to changing climatic conditions while maintaining sustainable production levels.