Investigating the effect of climate factors on fig production efficiency with machine learning approach

This study employs a machine learning approach to investigate the impact of climate change on fig production in Turkey. The eXtreme Gradient Boosting (XGBoost) algorithm is used to analyze production performance and climate variable data from 1988 to 2023. Fig production is a significant component of Turkey's agricultural economy. Therefore, understanding how climate change affects fig production is essential for the development of sustainable agricultural practices.

Despite an observed increase in fig production between 2005 and 2020, potential yield may be negatively impacted by climate variables. Identifying the specific climatic factors affecting fig production efficiency remains a challenge. In the study, two different machine learning models are created: one for fig production yield per decare and another for fig production yield per bearing fig sapling. Eight climate variables (16 variables considering day and night values) serve as independent variables in the models. The models reveal that temperature change has the highest impact, with a percentage contribution of 41.30% in the first model and 43.90% in the second model. Thermal radiation (day and night) and 2 m temperature also significantly affect individually fig production. Wind speed, precipitation and humidity contribute to a lesser extent.

This study illuminates the intricate interrelationship between climate change and fig production in Turkey. The utilization of machine learning as a predictive tool for future production trends and an instrument for informing agricultural practices is a valuable contribution to the field. © 2024 The Author(s). Journal of The Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.