Exploring an Alternative Configuration of the Hydroclimatic Modeling Chain, Based on the Notion of Asynchronous Objective Functions
This study explores an alternative configuration of the hydroclimatic modeling chain around the notion of asynchronous objective-function (AOF). AOFs are calibration criteria purposely ignoring the correlation between observed and simulated variables. Within the suggested alternative configuration, the hydrologic model is being forced and calibrated with bias corrected climate variables over the reference period instead of historical meteorological observations. Consequently, the alternative configuration circumvent the redundant usage of climate observation operated within conventional configurations for statistical post-processing of simulated climate variables and calibration of the hydrologic model. AOFs optimize statistical properties of hydroclimatic projections, preserving the sequence of events imbedded within the forcing climate model. Both conventional and alternative configurations of the hydroclimatic modeling chain are implemented over a mid-size nivo-pluvial catchment located in the Saint-Lawrence Valley, Canada. The WaSiM-ETH hydrological model is forced with a bias-corrected member of the Canadian Regional Climate Model Large Ensemble (CRCM5-LE). Five AOFs are designed and compared to the common Kling-Gupta efficiency (KGE) metric. Forced with observations, AOFs tend to provide a hydrologic response comparable to KGE during the nival season and moderately degraded during the pluvial season. Using AOFs, the alternative configuration of the hydroclimatic modeling chain provides more coherent hydrologic projections relative to a conventional configuration.