L7. Current state of hydrological modelling: Need for a new approach?

Most practicing hydrologists and researchers still use hydrological models like Sacramento, MIKE-SHE, SWAT, SIMHYD, SWMM and many others that were devised several decades ago. In many cases, other than vastly improved user-friendliness of the modelling softwares, not a great deal has changed in terms of concepts on which models are based and how they are calibrated and used in the last 20 years. Most progress in hydrological modelling has been in the areas of creating modelling frameworks and front/back end graphics interfaces rather than in the understanding of the underlying hydrological processes and improving relationships built into the model structure. Despite the advent of multiple parameters models ‘black-box’ processes and inappropriate simplification still decide the model results. This is compounded by the ‘equifinality’ due to over-parameterisation and redundancy. The reasons for this are well articulated and documented by a few authors which include lack of field studies that would have helped understand the hydrological processes.

This session will invite researchers to present their

(i) ideas to how do we make hydrological modelling results more robust by using internal processes based calibration rather based on catchment streamflow only
(ii) case studies to demonstrate attempts to establish credibility of their modelling outcome and to overcome uncertainty in their model results
(iii) thoughts on ways to improve hydrological modelling using a “tailor-made” approach to address specific problems at hand.