| Summary: | Given the significance of climate models for assessing climate change impacts, and recent increases in their resolution, there is a need to understand strengths and weaknesses of climate models in reproducing key atmospheric processes, and to assess their performance using accurate ground-based observations. This thesis first investigates the inconsistencies in ground-based observations for cold environments, second, the role of ground-based observations for empirical model validation over complex terrain, and third, uses both observations and model output, to describe a mesoscale process associated with precipitation and their changes in a simulated future climate. Regional climate modelling in a convection-permitting configuration improves simulation of mesoscale systems in which convection initiates and develops, adding value to estimates of convective precipitation compared to models that rely on deep convective parameterization schemes. On the leeside of the Canadian Rocky Mountains in extratropical regions, convective precipitation is influenced by a strong longitudinal gradient of low-level moisture across the foothills. Known as the dryline, this gradient is the result of the convergence of moist air from the interior of the continent and the dry air from the subsidence on the lee side of the Rocky Mountains. The dryline plays a key role in initiating convective precipitation. To find robust answers to questions about a future transient climate, a better understanding is needed of the dryline’s relationship to the location and timing of convective initiation. This research has three objectives: 1) to critically quantify the systematic bias of precipitation measurements on two sides of the northern Canada-U.S. border since the two countries use different standard instrumentation to observe liquid and solid precipitation; 2) to study if a convection-permitting model can reproduce the warm season’s diurnal cycle of precipitation at a continental scale, and 3) to describe a mesoscale mechanism related to the initiation of convective precipitation in the Rocky Mountains vulnerable to climate change at the end of the century. Results show that a correction due to wind-undercatch in monthly solid precipitation is up to 31% during January in the Yukon, whereas across the border in Alaskan stations, it is up to 136%. This correction leads to a smaller and inverted horizontal precipitation gradient in the northern part of the border. In July, the correction for monthly liquid precipitation is around 20% in Alaska and 4% in the Yukon. This inconsistency has to be considered in any regional study using precipitation in cold and windy environments. The research to validate the precipitation diurnal cycle characteristics using a convection-permitting model, uses ground-based observations and a gridded product. Results show that the convection-permitting model can represent the main continental patterns and also represent the precipitation peak transitions from the afternoon to night on the leeside of the Rocky Mountains. However, in the central and eastern region of the study domain, the convection-permitting model performance deteriorates during the diurnal cycle for observed morning peaks (in the central-east U.S.) and overestimates the magnitude of the observed diurnal cycle in the southeast region in the U.S. When a warmer climate scenario is simulated at the end of the century, persistent increase is shown both, in the amplitude of the precipitation diurnal cycle and in the precipitation intensity throughout the domain. The warmer climate simulation also presents an increase in precipitation frequency in the northern region in early summer. These increases may impact the agricultural sector and alter flood risk. Finally, it is found that the convection-permitting model can simulate the dryline, showing an average magnitude of 0.48 g kg-1 km-1 and its maximum intensity being in July. The dryline is present in 37% of the biggest precipitation events (storms with at least one day above the 85% quantile in 13 years period). Although the percentage of the dryline frequency associated with convective initiation in the future scenario is not substantially changed, the dryline is both more intense (0.55 g kg-1 km-1) and narrower. Furthermore, in the simulation of the future climate, an intensification in the north and a dissipation in the southern part of the region was found in a standardized number of occurrences of convective initiation east of the dryline. This finding is associated with a change in the thermodynamical forcing of the most intense precipitation on the selected events in the southern part of the region. By describing the dryline, this research provides a reference point to assess the convective initiation forecasts and offers information on precipitation changes in a warmer scenario at the end of the century.
|
|---|