Assessing future changes in hydroclimatic processes in the Metropolitan Region of Belo Horizonte, Brazil, with the expanded Bluecat framework

General circulation models (GCM) have comprised ubiquitous tools for supporting water resources planning and decision-making under changing climate conditions. However, GCMs are often highly biased, which may limit their utilization for representing future trajectories of the hydroclimatic processes of interest. In addition, assessing the predictive uncertainty of climate models, which is paramount for simulation purposes, is not straightforward. For tackling these problems, in this paper we resort to the expanded Bluecat framework, which utilizes empirical conditional distributions for providing a complete stochastic representation of GCM outputs simultaneously to bias correction. The stochastic model was employed for assessing future trajectories of monthly rainfall and temperatures, under three Shared Socioeconomic Pathways, namely, SSP1-2.6, SSP2-4.5, and SSP5-8.5, in the Metropolitan Region of Belo Horizonte, Brazil. Our results indicated that e-Bluecat properly corrected bias for both variables and provided coverage probabilities close to the theoretical ones. Nonetheless, the resulting uncertainty, as materialized by confidence intervals, was deemed too large, which implicitly reflects the inability of the GCMs in describing the observed processes. In addition, in median terms, the bias-corrected estimates suggest considerably smaller increases in temperatures (~1°C), as compared to the climate models (up to 5°C), in all future scenarios. These findings suggest that deterministic outputs of GCMs may present limitations in effectively informing adaptation strategies, necessitating complementary approaches. Moreover, in view of the large uncertainty levels for the projected climate dynamics, simulating critical trajectories from the stochastic model is paramount for optimizing the allocation of financial resources over time in the study area.