Whidden, Ian W. 2023. Forest canopy effects on snow depth and density of seasonal and transient mountain snowpack in a maritime snow climate. Corvallis, Oregon: Oregon State University. 144 p. M.S. Thesis.
Relatively little is known about how various factors influence snow water storage in forested mountain landscapes in maritime (warm winter) climates. This study took advantage of multiple snow data sources including long-term data, synoptic sampling, remote sensing, and modeling to examine factors influencing snow dynamics in the H.J. Andrews Experimental Forest, western Cascade Range, Oregon. The study combined two field campaigns, long-term snow survey data, long-term data from meteorological stations, lidar-derived snow depth maps, and snow modeling to quantify effects of forest canopy structure, landscape position, and season on snowpack depth and density in a ~52 km2 portion of forested watershed ranging from 800 to 1600 m in elevation. Methods included (1) collecting field measurements of snow in March of 2022 and 2023, (2) comparing long-term (1978-2022) snow measurements at the paired forested vs. open sites, (3) examining SWE and snow depth data at four meteorological stations (1997-2022), (4) using a lidar-derived snow height model of the upper elevations of the Andrews Forest acquired March 2022 to assess spatial patterns of snow depth, and (5) simulating snow water equivalent at forest and open sites using a snow model forced with meteorological data for the period 2014 to 2018.
Field sampling indicated that snow depth in March 2022 was as much as 130 cm greater in the clearing at the Upper Lookout meteorological station (1298 m) than in the adjacent forest plantation. Field sampling was used to validate lidar-derived snow depth.
Long-term snow survey data from paired sites in forest gaps created by roads and beneath adjacent forests (1978-2022) indicated that average annual snow depth was two to three times greater in openings than in adjacent forest sites. Snow density was similar (about 365-366 kg/m3) in forest and open snow survey sites. Snow disappeared earlier at snow survey sites under forest than in openings, but snow melt was more rapid in openings than under forest at snow survey sites.
Analysis of daily snow depth and SWE records from three meteorological stations (1997-2014) revealed a concave upward curve of average daily snow density over the snow season. Snow density ranged from 200 to 350 kg/m3 in November to January, 380 to 400 kg/m3 in February to early April, to higher values in May. Analysis of the 1995-2022 precipitation and SWE data from the Upper Lookout meteorological station also revealed that maximum snow water equivalent on April 1 was on average 38% of cumulative precipitation in the water year (Oct 1 to April 1).
Analysis of the lidar image of 52 km2 of the upper elevation portion of the Andrews Forest (600 to 1600 m) indicated that lidar-derived snow depth in March 2022 varied with landscape position and forest canopy structure: deeper snow occurred on north- and east-facing slopes, in valley bottoms, and openings such as avalanche tracks, montane meadows, access roads, and forest clearings created for meteorological stations. Lidar-derived snow depth was greater in plantations than in adjacent mature-old forest in valley bottoms affected by cold air drainage but was not related to forest canopy height in other landscape positions, based on a small sample of paired 60x60m polygons. Lidar-derived snow depth tended to overestimate field measurements, indicating that the post-survey elevation correction (+295 mm) may not be appropriate. The total volume of water stored in the snowpack in March 2022 based on the lidar-derived snow depth model (average depth 54 cm, average density 380 to 400 kg/m3) was approximately equivalent to 30 days of mean daily flow at Lookout Creek, which drains the H.J. Andrews Experimental Forest.
Modeling of snow using the SUMMA model reproduced the measured snowpack in an opening at the Upper Lookout meteorological station when the model was forced with data from that station. However, modeling of snowpack under forest using parameterizations of forest interception developed at a nearby site in the Umpqua National Forest did not reproduce snow measured under forest in the field or at the long-term snow stakes. Snow modeling results demonstrated the importance of early winter air temperature effects on snowpack formation and persistence.
These results demonstrate that conifer forests substantially reduce snow accumulation in a maritime climate, but these effects vary among years, with elevation, and with canopy structure. Further study is needed to better understand how landforms, climate, and conifer forest canopy influence snow accumulation and melt in the transient to seasonal snow zones of maritime climates.