Influence of Terrain and Land Cover on the Isotopic Composition of Seasonal Snowpack in Rocky Mountain Headwater Catchments Affected by Bark Beetle Induced Tree Mortality

Abstract:

Session H103: The Hydrology–Vegetation–Climate Nexus: Identifying Process Interactions and Environmental Shifts in Mountain Catchments

Influence of Terrain and Land Cover on the Isotopic Composition of Seasonal Snowpack in Rocky Mountain Headwater Catchments Affected by Bark Beetle Induced Tree Mortality

Evan L Kipnis, Melanie A Murphey, Alan Klatt, Scott N Miller, David G Williams

Snowpack accumulation and ablation remain difficult to estimate in forested headwater catchments. How physical terrain and forest cover separately and interactively influence spatial patterns of snow accumulation and ablation largely shapes the hydrologic response to land cover disturbances. Analysis of water isotopes in snowpack provides a powerful tool for examining integrated effects of water vapor exchange, selective redistribution, and melt. Snow water equivalence (SWE), δ²H, δ¹⁸O and deuterium excess (D-excess) of snowpack were examined throughout winter 2013-2014 across two headwater catchments impacted by bark beetle induced tree mortality. A USGS 10m DEM and a derived land cover product from 1m NAIP imagery were used to examine the effects of terrain features (e.g., elevation, slope, aspect) and canopy disturbance (e.g., live, bark-beetle killed) as predictors of D-excess, an expression of kinetic isotope effects, in snowpack. A weighting of Akaike’s Information Criterion (AIC) values from multiple spatially lagged regression models describing D-excess variation for peak snowpack revealed strong effects of elevation and canopy mortality, and weaker, but significant effects of aspect and slope. Snowpack D-excess was lower in beetle-killed canopy patches compared to live green canopy patches, and at lower compared to high elevation locations, suggesting that integrated isotopic effects of vapor exchange, vertical advection of melted snow, and selective accumulation and redistribution varied systematically across the two catchments. The observed patterns illustrate the potential for using D-excess to identify origins and timing of snowmelt runoff in streams and assessing the relative magnitude of different accumulation and ablation processes in snowpack evolution.