Abstract

Rainfall characteristics, differences in catchment storage, soil moisture dynamics, and geophysical properties influence spatial and temporal variability of runoff generation thresholds. This study addresses two questions: (1) Are there nonlinear thresholds of rainfall or antecedent moisture that trigger abrupt changes in agricultural catchments' hydrological response? (2) What physical characteristics determine runoff generation? We analyzed precipitation and streamflow variability during the 2025-2026 hydrological year at three agricultural catchments (36.4-70.8 km(2)) in central Chile. Relying on high-resolution observational data and statistical segmented regression, this study focuses on how rainfall, soil moisture, and physical characteristics trigger runoff activation after exceeding specific thresholds. Based on 63 rainfall events, runoff occurred in only 16 events, demonstrating a strong nonlinear response governed by antecedent moisture. Segmented regression revealed consistent activation thresholds across all catchments when combined precipitation and deep soil moisture (PTOT + ASM100) exceeded similar to 505 mm, evidencing strong subsurface control. However, runoff efficiency diverged sharply along the land-use gradient. The most intensively agricultural catchment exhibited the highest specific peak discharge and a unique rainfall intensity threshold (14.6 mm/h), indicating rapid infiltration-excess runoff driven by degraded soil permeability. Conversely, catchments with higher headwater or riparian forest cover buffered these rapid flows, sustaining baseflow and extending recession times.