In the western United States, anywhere from 60% to 90% of the precipitation comes in the form of snow.  Snow accumulates in high altitude locations that are often characterized by complex topography and heterogeneous canopy cover.  The complexity of the topography and land cover leads to spatiotemporal variability in snow accumulation, energetics, and melt.  Improved understanding of the snow cover and canopy interaction at the process-scale will improve regional-scale modeling and allow for the assessment of the potential effect of canopy change on snow accumulation, energetics, and melt. 

The most accurate method used to model snow accumulation, energetics, and melt utilizes the snow cover energy balance.  Topography, vegetation, and meteorological information are needed to run these types of models.  The primary energy input to the snow cover comes from net radiation and has been widely studied.  The secondary input to snowmelt is turbulent fluxes (defined as the sum of sensible and latent heat fluxes).  Though secondary to radiation, turbulent fluxes dominate snowmelt during rain-on-snow events, foehn wind events, and early season melt.  Also during low snow years or years with early melt out dates, turbulent fluxes are more and more important.  For these reasons, the general purpose of this research is to improve our knowledge of measurements and modeling of turbulent fluxes of the snow cover energy balance.  Specifically, my research will contribute to the body of knowledge regarding canopy effects on the measurements and modeling of the turbulent components of the snow cover energy balance at the process-scale with implications for improved large-scale modeling in complex, heterogeneous systems. 

The interaction between canopy cover and snow energetics has been studied at both the individual tree scale and stand scale.  Most of these studies have focused on canopy effects on net radiation.  The turbulent exchange of sensible and latent heat between the snow cover and the atmosphere can represent 60% to 90% of the snowmelt energy during a rain-on-snow event and early season melt can be dominated by sensible heat exchange.  Vegetation absorbs, emits, and transmits radiation, moderates winds, warms and differentially heats the air immediately surrounding trees, and modifies vertical variation of temperature and humidity.  Improved understanding of the manner in which canopies affect air temperature, humidity, and wind would allow one to more accurately simulate below canopy turbulent fluxes over snow, which would improve simulation of snow cover energetics and melt.  Though canopy has a clear effect on the variables that control turbulent fluxes over snow, few published studies and very little data exist to evaluate how a canopy modulates meteorological forcings to affect turbulent fluxes over snow.  This limits one’s ability to model the development and melt of the seasonal snow cover in heterogeneous mountain landscapes which contain a wide diversity of canopy conditions.  Furthermore, an improved understanding of the processes involved in turbulent exchange in heterogeneous, vegetated terrain could assist in parameterizing snow cover energetics in regional and continental scale models, specifically for models that address changes in canopy cover and/or structure. 

Almost all of the components of the snow cover energy balance can be validated with direct measurements except for turbulent transfer.  Turbulent transfer is not commonly measured and therefore, not often validated for snow cover energy balance models.  Currently, the most direct way to measure turbulent transfer over snow is with eddy covariance (EC).  Eddy covariance utilizes fast response instrumentation to measure instantaneous fluctuations from the mean to arrive at a measured flux.  Fluxes of heat, moisture, and carbon dioxide are most commonly measured but the methodology is also used to measure fluxes of other gases.  EC systems typically take fast response measurements at 10 hertz (ten times per second) to generate fluxes over an appropriate averaging timestep.  Though turbulence is complex and variable, improvements on the instrumentation have made EC measurements more reliable and improved computer power and storage have made post-processing the data more efficient.  However, a high degree of uncertainty is associated with these measurements.  Energy balance closure discrepancies are often attributed to differences in measurement scale of the energy balance components, systematic bias in instrumentation, and the loss of low and/or high frequency contributions.  The advantages to EB over snow are that snow cover integrates scale mismatches and snow cover change (e.g. melt and accumulation) is readily measured.  Over snow, even in complex, heterogeneously vegetated terrain, the snow cover EB is simulated with closure typically better than 5%, and frequently better than 1%.  I hypothesize that EC measurements can be used to validate the general trends of simulated turbulent fluxes and that the EB closure in the snow cover simulations will allow us to validate the effectiveness of EC instrumentation over snow in mountain regions. 

Experiments using EC are expensive and time consuming.  EC systems generate massive amounts of data (more than 1 gigabyte per month), require much more power than a standard micro-meteorological station, and require exensive post-processing and correction.  The result of these costs is that in remote areas EC measurement sites are necessarily limited to a few select locations.  One way to reduce the cost of maintaining an EC system is to use data loggers that only store summary EC data instead of all of the 10 hertz data.  This reduces the data volume, extends the possible time for unattended operation, and largely eliminates post-processing of the data.  Unfortunately, without post-processing, it is difficult to know if the measured fluxes need to be corrected or adjusted for site effects.  In mountain locations these corrections can be substantial and may result in bias in the summary EC fluxes.  At locations where both summary and fast response data are available, it may be possible to understand the biases of the summary fluxes in the context of the canopy and site condition.  Therefore, allowing for the development of a first-order correction of the summary data.  If the inherent biases in the summary data could be corrected, it would be possible to compare measured fluxes over snow across a wide range of hydro-climatic conditions and mountain environments.

The overall objective of this study is to improve understanding of spatiotemporal variability in turbulent fluxes in heterogeneous canopy and complex mountainous terrain.  The proposed study will facilitate understanding of how turbulent fluxes are effected by canopy cover over snow and offer an opportunity to validate snow cover energy balance modeling.  The study will focus on the processes at the small catchment scale but implications for larger scale models are anticipated.  The specific objectives of this study are: 

•Determine the effectiveness of EC measurements of turbulent fluxes over snow cover at complex sites.

•Quantify the differences between open site and below canopy measured turbulent fluxes over snow. 

•Determine modifications to the turbulent flux algorithm that is currently used in an existing snow cover mass and energy balance model.  Quantify measured to modeled turbulent fluxes both before and after turbulent flux algorithm modification.