Large-scale climate oscillations shape interannual variability in Indian cereal yields, yet most assessments consider single indices, assume stationarity, or omit explicit tests of incremental value when combining multiple climate drivers with local meteorology. We analyze rice, wheat, and maize yields (1966–2017) for representative districts across representative districts of fourteen agro-climatic zones using a structured wavelet-coherence framework. First, Bivariate Wavelet Coherence (BWC) quantifies time–scale-localized associations between yields and four teleconnections—the El Niño–Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), North Atlantic Oscillation (NAO), and Pacific Decadal Oscillation (PDO)—summarized by Average Wavelet Coherence (AWC) and the Percentage of Significant Coherence (PoSC). Second, we have implemented Multiple Wavelet Coherence (MWC) through a systematic coherence elimination approach (CEA) supplemented by a rigorous Gain-Loss (GL) analysis of PoSC measure. In the GL analysis of two-, three-, and four-factor combinations we retain combinations only for 5% PoSC change relative to the best combination to screen out incidental gains. Finally, we append annual rainfall and temperature to dominant teleconnection sets to test the added value of proximal meteorology. Across zones and crops, ENSO is the most frequent dominant single predictor, while PDO rarely dominates alone but materially strengthens pairs—especially ENSO–PDO—at 4–8-year scales. Adding rainfall and temperature yields zone-specific PoSC gains, most notably where irrigation is limited or late-season heat risk is high. Phase diagnostics indicate predictor-leading relationships at interannual scales, implying seasonal-to-annual predictability windows. The framework provides a transparent, reproducible path from teleconnection diagnostics to parsimonious, operational predictors for crop-yield early warning and climate-smart agronomy.
Sankaran et al. (Tue,) studied this question.