North America: Dynamical downscaling experiments and hydrological modelling for Canada and Mexico
Weak (strong) thermal gradient and intense (weak) convective activity characterizes the tropical (extra tropical) atmosphere. For the tropics, the release of latent heat is the primary source of energy in the circulations of the region; while for the extra-tropics are the thermal gradients, Coriolis and the momentum fluxes associated to perturbations of all scales.
Earth System Models (ESMs) are a useful tool to contribute to the understanding of global circulation and generate climate change scenarios for the following decades. The resolution of these models is adequate to describe global circulation patterns, but not enough to describe the subcontinental processes important to North America, such as the effects of their abrupt orography or atmospheric processes at scales smaller than 10 km. That gap is filled up by performing dynamical downscaling, which requires regional models with the potential to increase to higher spatial resolutions.
Based on the contrasting nature of the tropics and the extra-tropics, this study proposes to carry out a dynamical downscaling experiment for North America (Mexico, the United States and Canada), using two of the CMIP6 ESMs with the “best performance” for the regions, in agreement with statistical metrics (Colorado et al., 2018), forcing the RegCM and WRF models, and evaluating their capabilities. The regional models will run under low resolution (45 km), medium resolution (15 km) and high-resolution (5 km) nesting grids, with the highest resolution covering two domains, one over a group of river basins located in the tropical zone of Mexico (namely Actopan, Antigua, Jamapa and Nautla) and a second one over a group of river basins located in the southern province of Quebec in Canada (namely Du Nord, Doncaster, Matawin, L´Assomption, Noire, Maskinonge, Du Loup and Taureau).
Some numerical experiments will be made with 2.5 km resolution over the group of river basins located in Mexico and Canada.
We propose to carry on a process-oriented verification in both subregions, considering its different regional processes, but applying the same methodologies to analyze the contributions of regional mechanisms under contrasting conditions that allow the understanding of high/low frequency atmospheric dynamics, acquiring a better-understanding of processes in contrasted environments.
Evaluating how well the models can reproduce the observations under different conditions is challenging since there are different modes of response to forcing with scales on the order of hours, days, months and years.
In the project, dynamical downscaling with the highest spatial resolution of 5 km using the “convection permitting” option will be applied, for the historical period: 1981-2010, near future: 2020-2049 and far future: 2070-2099, under two extreme climate change scenarios: SSP5-8.5 and SSP1-2.6, estimating the temporal and spatial variability over the two regions.
Some statistical downscaling techniques based on bias correction (such as Bias Correction/Constructed Analogues and Bias-Correction Spatial Disaggregation) and Perfect Prognosis will be applied to compare with the dynamical downscaling simulations. Comparisons with existing CORDEX simulations over the NAM and CAM domains will also be carried out.
And finally, the HYDROTEL distributed hydrological model will be run to better investigate the use of convection-permitting climate simulation outputs for the simulation of streamflows.