Development of a Conceptual Hydrological Model Based on Supply‐Demand Relationship and Its Applications

Monthly conceptual hydrological models can provide simple yet effective descriptions of hydrological processes. Most hydrological models were designed to understand physical processes of catchments, focusing on individual sub‐processes (Newtonian paradigm). However, few were developed to represent the overall behavior of hydrological systems (Darwinian paradigm). This study adapted the objective functions from water resources systems to simulate catchment water partitioning. Building on this, a supply‐demand‐based hydrological model (SDM) was developed. The proposed SDM was validated using data from 640 CAMELS‐US and 171 CAMELS‐AUS catchments, and compared with five parsimonious models: the Two‐parameter Water Balance Model, the WatBal Model, the Dynamic Water Balance Model, the Génie Rural 5‐parameter Model, and the Time Variant Gain Model. Results indicate that: (a) The model shows satisfactory results, with median NSE values of 0.65 and 0.74 for CAMELS‐US and CAMELS‐AUS catchments, respectively, during the validation period. (b) Compared to the other five models, the SDM achieves better performance in terms of the structural risk minimization metric, with median values of 0.61 and 0.42 during the validation period for the two data sets, respectively. (c) The SDM shows greater improvement compared to other models in catchments with low mean annual runoff or runoff ratio. This improvement, however, decreases when the fraction of precipitation falling as snow increases. This study offers a novel perspective on understanding water partitioning patterns in natural catchments by leveraging principles from human water resources management.