Hidden Complexity in Reservoir Flow Regulation Revealed by Spectral Analysis

Abstract

Understanding reservoir regulation of streamflow is critical for hydrological modeling and ecohydrological assessments, yet our knowledge of how reservoirs preferentially modify flow variability remains limited. Here we apply spectral analysis to daily inflow-outflow data from 205 US reservoirs, decomposing regulation effects into operational mode (how reservoirs modify flow variability) and regulation intensity (magnitude of modification). We identify four distinct operational modes that transcend nominal design purposes, better distinguishing observed regulation behavior than purpose-based classifications. While the dominant mode strongly attenuates high-frequency flow variations, reflecting widespread flood control operations, three additional modes (i.e., seasonal amplification, seasonal smoothing, and high-frequency amplification) are also identified. We also find that larger reservoirs exhibit stronger regulation consistent with their characteristic operational modes. This work reframes our understanding of the complex reservoir regulation behaviors through a novel frequency-domain perspective. The insights hold potential to inform improved hydrological modeling.