A Weak Temporal Association Between Multi-Week Geomagnetic Activity and Satellite-Derived Solar-Induced Chlorophyll...

Magnetic-field effects have been reported in controlled biological systems, but the relevance of weak geomagnetic variability to vegetation under natural conditions remains uncertain. We examined temporal associations between satellite-derived solar-induced chlorophyll fluorescence (SIF) and geomagnetic disturbance derived from the Disturbance Storm Time (Dst) index. We defined the sign-inverted Dst index, SII, as the sign-inverted daily mean Dst, so that stronger negative Dst excursions corresponded to larger positive SII values. The primary exposure was the trailing 28-day mean SII, excluding the day of the SIF observation. The primary analysis used Orbiting Carbon Observatory-2 (OCO-2) SIF at 771 nm with leave-one-year-out harmonic adjustment for the annual cycle and linear calendar-time trend. Sensitivity and robustness analyses examined a 21-day exposure, a more flexible cyclic-spline seasonal adjustment, spatial cluster bootstrap, and three temporal-surrogate null models. We also examined temperature and vegetation strata, land-cover and geographic controls, adjustment for an ECMWF Reanalysis version 5 (ERA5) surface solar radiation downwards (SSRD)-derived photosynthetically active radiation (PAR) energy proxy and vapour-pressure deficit, and comparisons with planetary Kp index, 10.7 cm solar radio flux index (F10.7), and supplementary SIF wavelengths. In the primary analysis, the 28-day trailing mean of SII was weakly negatively associated with SIF anomalies (Spearman ρ = -0.051; 95% cluster-bootstrap CI, -0.053 to -0.050), with an estimated linear change of -0.0381 residual-SIF units per 100 nT. Empirical p-values were 0.084 for year permutation, 0.011 for circular shift, and 0.001 for 30-day block permutation. The association remained negative at 21 days and in persistently vegetated cells, but its magnitude was substantially reduced with cyclic-spline adjustment (ρ = -0.013 in the pairwise-matched sample). Negative estimates were found in several vegetated land-cover classes, whereas estimates for the Barren land-cover class and Sahara geographic control were close to zero. The contribution of SII to explained variance remained below one percentage point across the examined temperature regimes. Overall, the results show a weak temporal association whose magnitude depends on analytical choices. Independent observational replication and controlled experiments are needed to determine whether it reflects a biological response to natural geomagnetic variability.