2026 Volume 49 Pages 83-88
In the present study, we utilised eight years of continuous monitoring data (2017–2024) to quantify the relationship between rainfall patterns and electrical conductivity (EC) dynamics in Sanbe Dam, Shimane Prefecture, Japan. A dimensionless EC index (EC*) was introduced to remove seasonal baseline variation. EC* reduction and time lag were calculated for 252 out of 276 rainfall events. Threshold analysis, conducted at 10 mm rainfall intervals, revealed that EC reduction and time lag characteristics underwent significant alterations at 30 mm and 60 mm thresholds. For moderate rainfall (30–60 mm) , persistent rainfall over 2–3 days produced EC reductions approximately three times larger than single-day events of equivalent total. These findings were subsequently corroborated by random forest analysis, which identified 5–6-day antecedent cumulative rainfall as the most significant predictor. The findings of this study indicate that 5–6-day cumulative rainfall can serve as a practical indicator for enhanced water quality monitoring in reservoir management.
