PARENT SESSION
Contributed Oral Session 100: Modeling: Forest Systems
Wednesday, August 10, 1:30 PM - 5:00 PM, Meeting Room 519 B, Level 5, Palais des congrès de Montréal

Integration of hydrological modeling in small watersheds into industrial forest planning in Alberta: the Forest Watershed and Riparian Disturbance (FORWARD) project.

Prepas, Ellie^*,1, Smith, Daniel², Putz, Gordon³, Russell, Jonathan⁴, Burke, Janice¹, ¹ Lakehead University, Thunder Bay, Ontario, Canada² University of Alberta, Edmonton, Alberta, Canada³ University of Saskatchewan, Saskatoon, Saskatchewan, Canada⁴ Millar Western Forest Products Ltd., Edmonton, Alberta, Canada

ABSTRACT- The FORWARD project was designed to develop and calibrate a hybrid model (between deterministic (Soil and Water Assessment Tool or SWAT) and stochastic (Artificial Neural Networks or ANN) models) that can effectively link streamflow and water quality with spatially-explicit forest management planning on an industry-relevant scale. FORWARD is based largely in the Forest Management Area (FMA) of Millar Western Forest Products near Whitecourt, AB. Since 1998, field infrastructure has been installed and data have been collected on weather, soils, riparian vegetation, streamflow, and stream water quality in 12 small (5 reference, 5 harvested, 2 burned) and 4 large (2 reference, 2 burned) watersheds. Burning of 84 to 100% of some watersheds occurred during a wildfire in 1998. Harvesting of approximately 60% of some watersheds occurred in winter 1999-2000 (1 watershed) and 2003-2004 (4 watersheds). The harvest design also tests the effectiveness of fixed-width buffers in protecting surface water quality (buffer versus no buffer). To date, these data have demonstrated that precipitation patterns (relative importance of snow, timing and intensity of rain events), landscape features (peatland cover, slope, aspect), and disturbance (harvest, fire) influence runoff and water quality (phosphorus, inorganic nitrogen) in streams. For the Millar Western Detailed Forest Management Plan (DFMP) due in 2006, the two modeling approaches (SWAT and ANN) were adapted to first order watersheds in the study area. Model output will be used to predict how tree removal and soil disturbance affect streamflow, nutrient flux, forest growth, and biomass recovery. These predictions will be applied as constraints to Timber Supply Analysis in the DFMP process to determine the optimal harvest level within each watershed in the Millar Western FMA.

Key words: modeling, watershed, disturbance, management