Abstract Numerous dams and reservoirs have fundamentally altered streamflow regimes and terrestrial hydrology. However, large‐scale hydrological models still depend on simplified and often unrealistic reservoir operation schemes due to limited understanding of real‐world operation complexities and inaccessibility of long‐term in situ operation data. Building on the Generic Data‐Driven Reservoir Operation Model (GDROM), this study introduces C+GDROM, a hybrid empirical and conceptual reservoir operation model designed to reduce data requirements while improving compatibility with hydrological models. Retaining GDROM's module‐based structure, where each module represents a distinct operational action under specific conditions (e.g., floods or droughts), C+GDROM explicitly predefines module formulations and their transition conditions using synthesized real‐world operation experiences and conceptual storage regulation curves derived from either historical observations or remote sensing storage data, which helps maintain physically consistent storage dynamics and guide release decisions. C+GDROM presents a General Model and two operation purpose‐specific variants: Flood Control Model and Irrigation Model , which is evaluated on 256 U.S. reservoirs. Results show that C+GDROM General Model significantly outperforms existing conceptual reservoir models in daily release and storage simulations under data‐scarce condition (no in situ data, using default rule parameters), while calibration with limited in situ data (2‐year) further enhances the performance of the two purpose‐specific models. The improved performance and reduced data requirements lie in the generality Inherited from both GDROM general module types and the conceptual storage regulation curve. These results demonstrate the potential of C+GDROM, with low reliance on in situ data, as an improved reservoir operation representation in hydrological models.
Chen et al. (Fri,) studied this question.