Abstract

At a forested catchment in the Sierra Nevada Mountains of California, we address variability in snow-vegetation interactions using hemispherical canopy photography, a distributed instrument cluster design concept, and a physically-based snow model. As part of the Southern Sierra Critical Zone Observatory (CZO), four research sites spanning a range of elevation (2250 - 2660 m) and canopy density (48 to 72%) provide a unique opportunity to observe and model the energetic and hydrologic variability in a sub-canopy environment. Arrays of ultrasonic snow depth sensors and pyranometers showed site-scale (i.e. 2500-m2) snow accumulation and mean daily shortwave irradiance to vary by as much as 35 and 50%, respectively. The variability was found to be highly correlated with canopy sky view factor, a weighted measure of canopy openness above an effective horizon. Analysis of hemispherical photographs coincident with the six instrument arrays at each of the four sites yielded high resolution estimates of canopy parameters such as sky view factor and leaf area index. Sky view factor was positively correlated with seasonal maximum sub-canopy snow accumulation (R2=0.83, p<0.05) and mean daily shortwave irradiance (R2=0.78, p<0.05). The observed linear relationships were most significant at effective horizons specified with zenith angles of 15° and 55°, respectively. A hemispherical model was used to simulate the sub-canopy shortwave irradiance at a finer timescale (1-minute) and at more locations than pyranometer measurements allowed. Precipitation input at each array as estimated by a canopy interception model indicated that net seasonal precipitation one meter from a tree bole could be as much as 35% less than in a neighboring canopy gap. The physically-based SNOWPACK model forced with measured and modeled data at each of the twenty-four instrument arrays for water years 2007 and 2008 indicates that the canopy photography parameterizations accurately represent the variability in the observed snow states. Snow cover duration varied by one to three weeks at the plot scale, and as much as eight weeks between sites. The results indicate high variability in maximum snow accumulation and melt rates related to differences in identifiable canopy structure and orientation. Model results at the plot-scale demonstrate interannual similarity in SWE and melt patterns despite differences in precipitation and melt timing. The combined measurement and modeling methods elucidate important relationships between the conifer forest canopy architecture and snowpack dynamics such as snow covered duration, snow water accumulation, and the timing and magnitude of meltwater input to the local hydrologic system.

Citation

Musselman, K.N., Molotch, N.P., Margulis, S.A., Kirchner, P.B., Bales, R.C. (2009): A mechanistic approach for estimating snowpack dynamics in a conifer forest . Fall meeting, American Geophysical Union, December 2009. 90(52). Abstract C23D-07 .