Identifying links between monsoon variability and rice production in India through machine learning

Area of Science:

• Agricultural Science
• Climate Science
• Data Science

Background:

• Climate change significantly threatens global food security, particularly for agriculture dependent on monsoon rainfall.
• Monsoon variability presents a critical challenge to agricultural systems, necessitating advanced analytical methods for understanding impacts.
• Traditional models often fail to capture the complex, non-linear relationships between weather patterns and crop productivity.

Purpose of the Study:

• To apply machine learning for a deeper understanding of monsoon variability's impact on agricultural productivity.
• To assess the effectiveness of random forest modeling in representing rice production variability.
• To identify key climate and agricultural factors influencing rice yield and harvested area.

Main Methods:

• Utilized random forest modeling to analyze the relationship between monsoon weather variability and rice production.
• Quantified the percentage of variation in detrended anomalies for rice yield and harvested area explained by monsoon weather predictors.
• Identified the most influential weather variables and agricultural management practices (e.g., irrigation) impacting rice production.

Main Results:

• Random forest models explained 33% of rice yield anomalies and 35% of harvested area anomalies.
• Downwelling shortwave radiation flux was the key weather variable for yield anomalies, while irrigation proportion was the most significant overall predictor.
• Weather extremes influence agricultural productivity through changes in both per-area yields and harvested area, highlighting a critical pathway for production losses.

Conclusions:

• Random forest modeling accurately represents crop-climate variability in monsoonal agriculture, offering insights beyond traditional parametric models.
• The study identified complex, non-linear responses of rice yield and area to factors like irrigation, monsoon onset, and season length.
• Machine learning provides a powerful tool for understanding and potentially mitigating the impacts of climate variability on rice production and food security.