Can Hydrologic Complexity Simplify Watershed Modeling.

Many water quality protection strategies have been developed based on results from watershed models (e.g., SWAT, GWLF, HSPF, ANGPS), which link runoff and pollutant concentrations almost solely to land use. As a result, we have sometimes dogmatically developed nonpoint source pollution control practices based too heavily on specific land uses and largely ignored the interaction between land management and physical, landscape scale processes. These models were never intended to be applied at field scales, yet they represent some of the most common model applications in the literature. Another potential drawback is that the manner in which they are initialized and calibrated creates largely stochastic, neural network type models that impart relatively little field scale information. In order to correctly apply these models at the field scale we need to explicitly incorporate pertinent physical controls. We contend that incorporating hydrologic complexity into these watershed models can reduce calibration needs, increase model parsimony, and better represent landscape features without adding undue data collection needs. As a proof-of-concept we demonstrate incorporating hydrologic complexity by considering linkages between field/hillslope complexity and watershed simplicity that are scale invariant or scalable: i) storage vs. discharge relationships, ii) storage vs. saturated area relationships, and iii) terrain attributes. With this framework we are able to better represent landscape and field level processes such as variable source area hydrology and critical zone physics. These linkages allow us to define where pollutant generation and hydrologic activity co-exist and underscore how consideration of field-scale processes can result in substantial impacts to watershed management and water quality and may help to inform spatially targeted water resource management decisions and future modeling efforts.