Compute Robustness Assessment Test (RAT) Indicator
compute_RAT_X.RdEvaluates the robustness of a hydrological model performance metric (e.g., Bias) with respect to an explanatory variable (e.g., discharge characteristics), using a Spearman correlation test. The function returns a Boolean indicating whether a significant correlation exists, implying potential lack of robustness.
Based on the RAT (Robustness Assessment Test) methodology introduced by Nicolle et al. (2020), this function tests whether the variability of a model performance criterion can be explained by an explanatory variable, suggesting a lack of robustness if so.
Arguments
- Bias
Numeric vector of model performance scores (e.g., Bias) across sub-basins, time periods, or experiments.
- X
Numeric vector of an explanatory variable (e.g., flow regime characteristic) of the same length as
Bias.- thresh
Numeric threshold for the p-value of the Spearman correlation test. Default is 0.05 (5% significance level).