|Wednesday, June 02|
Improving Quality of Winter Discharge Records Using Acoustic Doppler Velocity Meters
* Zachary Bishop, Environment and Climate Change Canada, Canada
Emily Anderson, Canada
Curtis Bertrand, Canada
Derek Elliott, Canada
Jean-Pascal Faubert, Canada
Jeff Lipton, Canada
Craig Paul, Canada
Sandrina Rodrigues, Canada
"Computing discharge under ice conditions can be a very complex and time intensive procedure. The fluctuating backwater effect caused by slush and ice render the stage-discharge rating curve invalid as flow does not always increase with stage. The traditional method to compute discharge under ice conditions is to conduct discrete measurements and use the difference between measured discharge and rated discharge to shift the rating to account for the backwater. This shifted rating is only valid as long as conditions are stable. Between these stable periods, technologists use the interpolated discharge method to make corrections based on historical trends, meteorological data and comparison stations. Ice affected discharge data is published at Water Survey of Canada with associated grades and qualifiers to indicate reduced certainty due to these computational limitations. A project to investigate improving the quality of winter discharge records was created under the National Hydrological Services Renewal. A goal within the project is to investigate the use of Acoustic Doppler Velocity Meters (ADVM), which measure water velocity using the Doppler shift allow for the calculation of discharge without the traditional stage-discharge relationship. By using an ADVM for under ice discharge computations, we are able assess not only velocity data but also the additional parameters such as the vertical stage measurement and diagnostic parameters. In 2020, four test stations across Canada were established to determine the feasibility of this computational method. This presentation will describe the process of using ADVM data to compute under ice discharge, the limits and conditions of this workflow, and will compare the new method to traditional and discrete measurement methods. "
Deriving synthetic rating curves to link simulated stream flows to inundated areas: A global sensitivity analysis
* Camila Gordon, INRS, Canada
Etienne Foulon, Canada
Alain N Rousseau, Canada
Global sensitivity analysis (GSA) provides salient information about the interaction between input factors and model response. To carry out this type of analysis, a large number of simulations is required to achieve an adequate level of accuracy, which can be costly computationally intensive. The VARS framework is based on the variogram analysis of response surfaces to assess the global sensitivity of a model; that is with respect to the full spectrum of parameter variations. This framework was applied to the conceptual model HAND (Height Above Nearest Drainage) to assess the sensitivity of the computed synthetic rating curves of its forcing parameters. HAND was implemented within PHYSITEL, a specialized GIS for distributed hydrological models. The St-Charles River Basin in Quebec, Canada was used as a case study. Using a digital elevation model, HAND computes at any point on the landscape the height required for this location to be flooded. The synthetic rating curves are obtained from geometric properties of a river segment and the Manning equation, providing a mean of linking simulated stream flows to potential inundated areas. The GSA was conducted for five parameters (water and forest manning coefficients, river segment length and shape). This application of VARS allowed for comparing three types of GSA (Morris, Sobol, and STAR VARS), while defining the accuracy of the derived synthetic rating curves at various locations of the basin. This should prove to be a very useful tool for areas with low density of hydrometric data.
BasinMaker - An automated GIS toolbox for watershed delineation with lakes
* Ming Han, uwaterloo, Canada
"Lakes and reservoirs have critical impacts on hydrological, biogeochemical, and ecological processes, and they should be an essential component of regional-scale hydrological and eco-hydrological models. This is particularly important in Canada with its tens of thousands of lakes. Past large-scale hydrologic modelling efforts tend to either ignore the impacts of all lakes or explicitly simulate the behaviour of only the largest lakes in a watershed. This is practically due to difficulties inherent in representing thousands of lakes in various hydrological models. This research derives an open source GIS toolbox (BasinMaker) that can automatically and efficiently build vector-based hydrologic routing structures including any number of lakes. The toolbox is hydrologic model independent in that by default it outputs the routing structure and corresponding compete hydrologic routing characterization in shapefile format. BasinMaker supplies all the necessary hydrologic routing model inputs, including network topology, subwatershed geometry, channel characteristics (slope, length, roughness, and geometry), and lake characteristics (area, volume, and outlet description). Two main functions are included in BasinMaker: 1) It can use any DEM dataset and any lake polygon dataset to delineate a lake-river routing network. 2) It includes a set of useful functions to post-process any existing lake-river routing product generated by BasinMaker. BasinMaker has been tested to process multiple DEMs at various resolutions successfully including the Multi-Error-Removed Improved-Terrain DEM (MERIT) and HydroSHEDs DEM. BasinMaker is also configured to utilize the flow direction grid (Enhanced Flow Direction) from the Ontario Integrated Hydrology (OIH) Dataset. Example post-processing operations supported by BasinMaker include: a) Extract the lake-river routing product for the region of interest from the existing lake-river routing product; b) Simplify the existing lake-river routing product by i) increasing the size of catchment and reducing the density of river network; and ii) by removing the lakes from it; c) Define HRUs and generate model setup files for the Raven hydrologic modeling framework based on user provided landuse polygon and soil polygon, etc. BasinMaker is utilized to build a complete North American vector-based lake and river routing product including all North American lakes over 10 ha and incorporating all Water Survey of Canada and United States Geological Survey streamflow gauges as subbasin outlets. The average size of the routing subbasins is just over 10 km2. This routing product can be downloaded for any North American watershed using a web-based download tool. BasinMaker is unique compared with other routing network definition tools in terms of lake representation and is especially helpful for the hydrological modelling community in Canada interested in explicitly representing lakes in their models. "
Analyzing the Effects of Spatio-Temporal Discretization on Hydrological Response of Natural Catchments
* Siavash Pouryousefi Markhali, ÉTS, Canada
Annie Poulin, Canada
Marie-Amélie Boucher, Canada
Annie Poulin, Canada
Understanding the spatio-temporal scale of governing equations representing hydrological processes, and confronting the issue of scale mismatch within inter-connected hydrological units are two major challenges in hydrological modelling. To better understand complexity (heterogeneity) of the hydrology system, which takes place under continuous internal change (e.g. land use change) and boundary conditions (e.g. changing climate) distributed hydrology models have been used across different spatio-temporal scales. However, the models themselves suffer of adequately simulating the hydrological processes due to lack of scale-relevant theories. Therefore, part of the modelling uncertainty is due to the extent with which the physiographic characteristics of the catchment are described, more or less finely, by the model. Such uncertainty is normally ignored in modelling practices, and is the focus of the presented research. More precisely, we aimed at quantifying the relative roles of the spatial resolution of the physiographic characteristics, and that of the model's parameters obtained by calibrating the model using different spatio-temporal representation of catchments. To do so, two different distributed hydrological models (WaSiM and Hydrotel) have been used, as well as six catchments with different sizes across Quebec, all regrouped in an ensemble-based approach. The impacts of the choice of a particular level of spatio-temporal discretization on streamflow simulation in natural catchments have been studies by analyzing the respective roles of catchment area and characteristics, the time-step of the simulation, the model structure and parameters as potentially important determinants of a hydrological model's response. The ensemble approach showed a significant uncertainty (more than 100% of error for estimation of extreme streamflow) linked to spatial discretization of the modelling. Regarding the role of catchment descriptors the results showed that first, there is no meaningful link between the uncertainty of spatial discretization and catchment size as such uncertainty can be seen across different catchment sizes. Second, the temporal scale has only minor impact on determining uncertainty of spatial discretization. Third, the more finely distributed and physically realistic a model is, the more sensitive to changes in spatial resolution it is; and fourth, the uncertainty related to model parameters is dominant (larger) than that of catchments descriptors (DEM resolution, land use, soil texture).
Evaluation of net changes in evaporation resulting from the impoundment of a boreal hydropower reservoir - A case study of the La Romaine complex in eastern Quebec
* Marco Alvers, Université Laval, Canada
Daniel F. Nadeau, Canada
Antoine Thiboult, Canada
Habiba Kallel, Canada
Adrien Pierre, Canada
Alain Rousseau, Canada
François Anctil, Canada
To support climate change adaptation and mitigation plans, the use of indicators quantifying the human environmental footprint is booming. Recent studies have suggested that hydroelectricity has a substantial water footprint, meaning that a large amount of water is lost through evaporation per unit of energy produced. The approach used in most of these studies may be misleading because it does not consider evaporation from the pre-flooded ecosystems that would occur without the presence of a human-made reservoir. A more robust approach is to evaluate net changes in evaporation by calculating the water footprint, which is the difference between the evaporation after and before the reservoir impoundment. This study aims to assess the net changes in evaporation following the creation of Romaine-2 reservoir (~50°N, ~64°W), a boreal hydropower reservoir located in the Côte-Nord region of eastern Québec, Canada. The Romaine-2 reservoir, created in 2014, covers 85.8 km2 and is part of the Romaine river's hydroelectric complex that comprises four power plants and their reservoirs. Direct measurements of evaporation and other surface variables are being carried out over three different ecosystems (reservoir, coniferous forest, and wetland) since 2018 to evaluate changes in evaporation caused by the landscape modification. The in-situ observations are used to validate the evaporation simulated by a 1D lake model (the Canadian Small Lake Model, CSLM) and a land surface model (the Canadian Land Surface Scheme Including biogeochemical Cycles, CLASSIC). These models are used to estimate evaporation over the entire complex and over a historical period beyond that of the observations. Both surface models are driven by ERA5-Land reanalysis, which provides consistent climate data at high spatiotemporal resolution. CSLM models the evaporative water losses from the reservoir, while CLASSIC simulates evaporation from the different ecosystems that existed before the impoundment. Results from this work will provide a clearer picture of the ecological footprint of hydropower in the boreal environment so that decision-makers can better choose their energy supply.