Modeling stream discharge in forest catchments across Canada: hydraulic conductivity calibrations
Loading...
Files
Date
2012
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of New Brunswick
Abstract
This thesis informs about the process of (i) using soil texture, organic matter and soil depth to predict hydraulic conductivities at soil saturation (Ksat), and (ii) refining these predictions by way of catchment-level calibrations. The process also involves the application of the Forest Hydrology Model (ForHyM2). This model uses daily rain, snow, and mean air temperature, as well as basic soil (forest floor, rooted mineral soil, and subsoil) and catchment (slope, aspect, elevation) specifications as inputs. Model output refers to changes in soil moisture and temperature, snowpack ( depth, water equivalents, density) and flow rates (run-off, percolation, interflow, baseflow) at the daily level, summer through winter, with and without forest canopies. The catchment calibrations focused on matching the model output snowpack depth and stream discharge with actual data at six well-calibrated forest catchments across Canada (Nova Scotia: Moosepit Brook, Pockwock Lake; New Brunswick: Hayward Brook; Quebec: Lac Laflamme; Ontario: Turkey Lakes, Basin 31; and British Columbia: Rithet Creek). The resulting multi-year calibrations yielded good agreements with R² values of about 0.65, 0.77, 0.86, and 0.98, at the daily, weekly, monthly, and annual level, respectively. The multiplier adjustments concerning the predetermined hydraulic conductivity for downward flow (infiltration) varied from 0.5 to 2, but were still in general agreement with actual field-determined values. For lateral flow (interflow), these adjustments were more variable, because terrain and soil conditions would not uniform as set within the ForHyM2 formulation. The soil texture, organic matter and slope mediated impacts on modelled Ksat, interflow, infiltration, baseflow and stream discharge were analyzed for the Turkey Lakes and Moosepit Brook watersheds by way of a sensitivity analysis. As to be expected, increases in the sand content would increase infiltration and baseflow rates over interflow. Increasing the slope would favour interflow over baseflow. Changing the organic matter leads to non-linear responses, with optimal infiltration and interflow rates at an OM content of 15-25%, due to organic matter-induced soil granulation in mineral soil layers. For organic soils, Ksat as well as infiltration and interflow would be affected by the state of organic matter composition, being high within fibric layers, and orders of magnitude lower within sapric layers.