|Thursday, June 03|
Future flood frequency analysis using large climate simulations ensemble
* David Huard, Ouranos, Canada
Jacinthe Clavet-Gaumont, Canada
Jean-Luc Martel, Canada
Phil Slota, Canada
Kristina Koenig, Canada
Anne Frigon, Canada
"For the construction and maintenance of major dams and dikes, the standard engineering practice essentially consists in fitting a statistical extreme value distribution to observed river flows and compute the 1,000 and 10,000-year flows. Both recent historical observations and climate-change simulations show increasing trends in the frequency and intensity of extreme precipitation events over Canada. Climate change adds a layer of complexity as engineers are increasingly asked to demonstrate that infrastructures are climate-proof. Manitoba Hydro, Ontario Power Generation, Hydro-Québec, Rio Tinto, and Ministère de l’Environnement et de la Lutte aux changements climatiques du Québec partnered with the Ouranos Consortium to incorporate information on future projected changes into the design, operation, or maintenance of hydropower facilities. The goal of this collaboration was to propose a methodology to include future climate projections into the estimation of 1:1,000 and 1:10,000 flood events and to identify appropriate adaptation options. We will describe the methodology developed through a collaborative effort with working groups comprising of dam owners, regulators, engineering firms, climate scientists and professional associations, and academic researchers. After a review of current state-of-the-art design flood estimation, the methodology developed was tested over five selected watersheds, to be ultimately applied to 533 Canadian watersheds. The approach uses two large climate simulations ensembles driving a parsimonious hydrological model. The results are provided through publicly available geospatial layers describing relative change in frequential floods across Canada for the 2080-2100 horizon. "
Climate evolution in ensembles of GCM simulations for the Hudson Bay System
* Alida Thiombiano, University of Calgary, Canada
Marco Braun, Canada
Michael J. F. Viera, Canada
Tricia A. Stadnyk, Canada
Climate impact studies often require a reduction of the ensembles of opportunity from the Coupled Model Intercomparison Project (CMIP) when the simulations are used to drive impact models. In our study, we use simulated data from the Global Climate Models (GCMs) participating in phase five of the Coupled Model Intercomparison Project (CMIP5) to calculate 10 of the climate extremes indices recommended by the Expert Team on Climate Change Detection and Indices (ETCCDI). Based on these climate indicators, we then present and compare how annual and seasonal climatologies across the terrestrial domain of the Hudson Bay Drainage Basin (HBDB) are projected to change ‘globally’ (across the domain) and regionally (in seven major drainage basins) from 1981 to 2070 for two multimodel ensembles: a terrestrial ensemble of 19 simulations and an oceanic ensemble of 5 simulations, the second being a subset of the first. Our results show that the smaller ensemble captures generally similar climate change signals as the larger ensemble, but with more consistency for temperature-based indices than precipitation ones. Among the projected changes in climate across the HBDB during the study period, we have: (1) cold nights are projected to warm more than hot days, and mainly in the northern area (Foxe Basin); (2) increasing precipitation follows a West-East gradient of increase; and (3) Arctic amplification results in significantly more warming in the northern regions of the HBDB. Our findings provide an overview of moderate and more extreme indices projected future changes throughout space and time in the HBDB terrestrial domain that drive changes in its hydrology.
Down-Stream impact assessment of eutrophication and acidification potential in regional/local watersheds: Case study QuAppelle watershed, Saskatchewan
* Shayan Jamil, University of Regina, Canada
Water for domestic, industrial or agricultural use is typically sourced from the fresh water resources including streams, rivers, and lakes. These natural resources are exposed to contamination from effluent discharges, fertilizers and atmospheric depositions. When scaled over an entire watershed, even minor contaminant influxes per unit area can produce significant environmental impact. Water quality is regulated and monitored through water sampling programs designed to assess the concentration of contaminants through onsite testing and lab analysis results. Eutrophication and acidification are the impact categories used in this study for classification of nutrients, sulphur and related compounds resulting in algal blooms, depleted oxygen levels and overall deteriorated water quality. The characterization model used for assessing the eutrophication and acidification impact categories is based on Life cycle assessment framework to quantify the contaminant concentrations in the watershed and track the contamination to the source of origin. A mathematical and geo-statistical analysis of the inventory data is presented in the conclusion of the study with a process model to utilize the workflow for assessing similar watersheds.
Assessing the impacts of reservoir regulation in the Arctic
* Jenna Merth, University of Calgary, Canada
Andrew Tefs, Canada
Tricia Stadnyk, Canada
"The pan-Arctic region is home to large rivers that contribute freshwater discharge to the Arctic Basin. Reservoir regulation is prominent in some of these rivers including the Mackenzie River in northern Canada and the Yenisey River in Eurasia and one its main tributaries, the Angara River, both of which are the subjects of this study. With reservoir regulation altering the timing and magnitude of river discharge, we examine the regulation impacts on freshwater discharge to the Arctic basin over a long time period. Stations on each respective river were chosen based on location and completeness of daily timeseries of river discharge. Analyses ranged from 1950-2018 and were carried out for 30-year climatological time periods with 10-year offsets. The Mann-Kendall Trend test was applied to detect increasing and decreasing trends in discharge across both time and space. Spectral analysis was used to analyze the frequency domain of discharge timeseries over different time periods at all stations. Understanding the long-term impacts of regulation and hydropeaking signals can be helpful to better understand impacts to future freshwater discharge under the influence of both regulation and climate change. "
Multisite calibration of the CEQUEAU hydrologic and thermal Model in a large Canadian watershed
* Mostafa Khorsandi, INRS, Canada
André St-Hilaire, Canada
Richard Arsenault, Canada
"The Nechako River is an important habitat to a number of key fish species, including salmonids and white sturgeon. Ectotherm fish have specific water temperature requirements that can vary during their life history. In impounded systems such as the Nechako, water resources managers face the challenge of adapting their water release schedules and volumes to account for these requirements in the context of climate change. Hydro-thermal models are therefore required to be calibrated, validated and implemented on such systems to investigate different scenarios. In this study, single-site and multisite calibration approaches were used to estimate hydrological/thermal model parameters. For this, model calibration and simulations are carried out using a hydrological-thermal model (CEQUEAU) in two steps. First, the hydrological module calibration is performed using 25 parameters as tuning variables. The thermal module is then calibrated by estimating its ten parameters. Two calibration techniques were compared: (1) Averaging of single-site calibrated parameters, and (2) averaging the objective-function of the three sites during the calibration to return a “best compromise” solution. The Pattern Search (PS) single-objective optimization algorithm is applied to calibrate the model parameters using observed streamflow data at three monitoring sites within the Nechako Watershed. The KGE metric is used as an objective function for hydrological calibration, and the RMSE metric is used for the thermal model calibration. Single site calibration provides the best performance for each of the sites when taken individually, but these parameters perform poorly when applied to the other two sites. Both the multisite calibration and the averaging of the single-site calibrated parameters perform adequately for all sites. Multisite calibration performed better for the overall hydrograph and during high-flow seasons, while the averaging of parameters produces more accurate hydrographs during the low flow seasons. Overall, the application of different optimization schemes in the Nechako Watershed showed that both multisite calibration parameters and the average of single-site parameters could provide promising and less biased results for CEQUEAU model thermal calibration/validation. These results are expected to help the users of CEQUEAU and other distributed hydrologic models understand the sensitivity of semi-distributed hydrologic simulation to different calibration methods and demonstrate the advantages and disadvantages of single-site and multisite hydrologic parameter estimation methods on thermal parameter calibration and simulation."