Disentangling Climatic Volatility from Long-Term Productivity Trends in Semi-Arid Agriculture: An Empirical Mode Decomposition Approach

Objective: The variability of crop yields in semi-arid agriculture presents a significant threat to food and water security since crop yields are modulated by interacting climatic and systemic drivers with a variety of temporal interactions. Methods: This study employed Empirical Mode Decomposition (EMD) to a 22-year dataset (2001–2022) of pigeon pea yields from semi-arid India. From this data, eight distinct patterns called Intrinsic Mode Functions (IMFs) and a final residual trend were extracted. Findings: The study produced three main findings: i) The EMD method successfully distinguished between high-frequency volatility (short-term changes caused by annual weather shocks) and low-frequency patterns (long-term changes from decadal climate cycles and productivity trends). ii) The residual trend showed that after a long period of sustained growth, pigeon pea yields have been in a sharp decline since 2018. iii.) An analysis of variance revealed that long-term components (systemic issues and climate cycles) contributed the most to yield variability, having a greater impact than short-term "noise" from weather. This research offers a novel approach to analyzing agricultural systems that are non-stationary (i.e., their statistical properties change over time). Novelty: By separating random (stochastic) shocks from underlying (deterministic) trends, EMD serves as a powerful diagnostic tool for understanding agro-ecosystems. Applications: This method allows for the attribution of yield gaps to specific drivers across different time scales. This strengthens resilience assessments and supports more effective, climate-informed agricultural planning. The data from this study will allow for better planning and decision-making by farmers and other stakeholders in semi-arid areas by providing them with the knowledge to make better decisions on how to best manage their water resources. Keywords: Agricultural Resilience, Climate Variability, Crop Yield, Empirical Mode Decomposition, Semi-Arid