Measurement of differences in snow accumulation, melt, and micrometeorology due to forest harvesting

Previous studies in the Pacific Northwest have suggested a link between for est logging and increased streamflows. We present a brief overview of previ ous experimental efforts designed to identify mechanisms for these increase s, especially during rain-on-snow (ROS) and spring snowmelt events. To exte nd the applicability of these data, we describe our measurement of snow acc umulation, melt, and micrometeorology in existing uncut forest and shelterw ood units as part of the Demonstration of Ecosystem Management Options (DEM O) study. Uncut forest units are dominated by over 100-year-old Douglas-fir (Pseudotsuga menziesii) with stand basal areas of ca. 40 m(2)/ha. Micromet eorology is measured 2 m above the ground surface in each 13-ha unit. Snow accumulation and melt are measured via weekly snow courses. Snowpack outflo w is obtained from two 2.6 m(2) non-weighing snow lysimeters in each unit. Large (25 m(2)) weighing lysimeters provide a continuous record of snowpack evolution at scales unaffected by variability in canopy throughfall. Snow interception is measured by continuous weighing of cut trees. Harvest treat ments will be imposed during summer 1998 and additional post-harvest data w ill be collected. Pre-harvest data indicate up to 60% greater (33 mm) 3-day production of runoff and 150% greater (22 mm) 3-day snowmelt (outflow minu s throughfall) in the shelterwood than in the uncut forest unit during ROS events. Snowmelt during radiation-dominated spring events is ca. 50% greate r (15 mm) in the shelterwood. Results from the weighing lysimeters show hig h correlation with snow course data from the shelterwood but poor correlati on with those from the uncut forest due to small scale spatial variability in tree canopy cover.