This session invites presentations that explore the use of ‘hydrological consistency’ as a basis for hydrologic model development and evaluation using large-sample data sets. Hydrological consistency (Martinez and Gupta, 2011) aims at increasing the robustness of model evaluation. Classical robust statistical model evaluation metrics are needed but not sufficient: multi-criteria assessments based on multiple hydrological signatures, can help better characterize hydrological functioning. Large-sample data sets greatly facilitate: (i) improved understanding through rigorous testing and comparison of competing model hypothesis and structures, (ii) improved robustness of generalizations through statistical analyses that minimize the influence of outliers and case-specific studies, (iii) classification, regionalization and model transfer across a broad diversity of hydrometeorological contexts, and (iv) estimation of predictive uncertainties at a location and across locations (Mathevet et al., 2006; Andréassian et al., 2009; Gupta et al., 2014)
The challenge is for us to combine the principle of hydrological consistency with the use of large-sample data sets: it should become possible to draw much more general and robust conclusions, distinguish between real trends and meteorological, modeling or other artifacts, and improve hypothesis testing, model selection and intercomparison. We therefore invite presentations that investigate and make contributions to these and related issues, in the pursuit of improved hydrological understanding.
This session is organized by the new IAHS Panta Rhei Working Group on Large Sample Hydrology, which seeks to (a) collaboratively develop an extensive and representative worldwide sample of watershed data sets that cover all of the continents (and hence hydrometeorological regimes) of the globe, and (b) develop improved statistical methods that support rigorous hydrological model development and evaluation. Beginning with an initial compiled sample of ~2000 catchments, we seek members to help achieve our ambitious goal of incorporating catchments from every (nearly) country on the planet. During this session, we will also provide a short overview on the possible opportunities and issues related with establishing such global database. To become a member of the Working Group, please contact the session conveners.
Contributions addressing the following (and related) topics are solicited:
a)Methods and metrics to diagnose hydrological consistency
b)Comprehensive and systematic evaluation and identification of hydrologic/land-surface models based on hydrological consistency and/or large-sample data sets
c)Regional watershed datasets