Robustness in the spring surface energy balance in a mountain basin
Robustness in the spring surface energy balance in a mountain basin
Abstract ID#: 35167
English Abstract:
Complex dynamical systems, such as hydrological systems, exhibit system-wide behaviours that are due to the interaction of numerous processes and may result in system robustness. Such robustness is characterized by persistent behaviours despite system perturbations. In process-based hydrological models, inter-process interactions are typically included to increase physical realism, often at the cost of increased complexity. Model robustness is characterized by a persistence of model behaviours to perturbations in initial/boundary conditions, system parameters, and underlying model assumptions. This model robustness may reproduce system robustness and allow up-scaled and possibly falsified model representations of the system to be successful. Model robustness is distinguished from model stability by considering not just the persistence of behaviours but by also considering the interplay between model dynamics, model structure, and model assumptions that leads to this persistence. Quantifying the details of why, and under what conditions, system and model robustness occurs is critically important to deciding how to design robust hydrological models of sufficient complexity that can perform well when faced with uncertainty in parameters and forcing data.
In this work, we explore model robustness through the use of point-scale, physically based, energy-budget snowmelt models of varying degrees of physical realism and model complexity. We consider snowmelt model robustness as a possible surrogate for snowmelt system robustness over complex mountain topography. These models are driven by data from a mountain research watershed to examine the effects of surface characteristics, perturbations to meteorological forcing, and level of process representation on robustness. The results have implications for the estimation of uncertainty in the representation of complex mountain terrain and snow processes in up-scaled hydrological models and land surface schemes.
In this work, we explore model robustness through the use of point-scale, physically based, energy-budget snowmelt models of varying degrees of physical realism and model complexity. We consider snowmelt model robustness as a possible surrogate for snowmelt system robustness over complex mountain topography. These models are driven by data from a mountain research watershed to examine the effects of surface characteristics, perturbations to meteorological forcing, and level of process representation on robustness. The results have implications for the estimation of uncertainty in the representation of complex mountain terrain and snow processes in up-scaled hydrological models and land surface schemes.