Last year, during the first phase of this project, various powders were tested to find which would closely replicate snow in a small scale drifting model. It was concluded that Cascade Dish Detergent was the best for modeling wind blown snow. In this year's continuation, Cascade powder was used to further evaluate the relationship of wind blown snow and snow fences.

Literature research and professional interviews provided information about water supply from drifted snow, wind flow, drifting patterns, wind tunnel modeling, and snow fence variables, but did not describe the effects of varying horizontal board thickness on the wind pattern or the resulting drifts. The intent of this project was to evaluate relationships of fence thickness and porosity to wind direction and speed, and drift geometry and volume.

The model testing system included six scaled model snow fences of varying thicknesses and porosities, a wind tunnel, and apparatus for wind direction and wind speed. Each fence was tested three ways: wind direction, wind speed, and drift geometry. Graphs for each fence were developed to show wind direction and speed, drift profile, and a comparison of wind speed to drift profile. Also, graphs were developed to compare total drift volumes and locations of drift apex for all fences.

Three of the four hypotheses were supported. As fence porosity decreased and board thickness increased, pattern of wind direction changed, wind speed decreased, and the drift apex occurred further upwind. The volume hypothesis for thickness was supported because increased thickness resulted in increased drift volume. However, the volume hypothesis for porosity was denied, because increased fence porosity resulted in increased drift volume.

Based on these results, snow fences can now be designed for specific intents. Other variables that could be tested next include those of terrain, buildings, and wind intensities. The most valuable extension is application of these fence model results to establish full-size fences to collect the snow in desired locations; this will collect more volume of winter wind blown snow to improve water supply in a drought period and keep critical areas free of drifting snow. This project proves potential to "Get the Drift!" into more helpful locations.

For the last four winters (01'- 04'), observations of the orographic effects of wind at Stevens Pass Ski Area have been recorded. This was accomplished by correlating data from the three ski area telemetry sites, simultaneous compass readings from various locations within the ski area, and visual observations of wind and its effects. This data was then compared with the forecasted free winds at 5,000 feet, and some selected telemetry sites around the state of Washington. Looking for consistency as well as anomaly between various observations and data, this data was then plotted on maps of the ski area to get an overall picture of what a free wind at 5,000 feet really means to our local area. When available, crown depths from shot records were looked at to confirm suspected loading of slopes, and "ski-pole" measurements were taken when shot records were not available.

The findings have shown that not only large terrain features such as major mountain passes and ridges can effect wind direction, but also rows of trees and more subtle terrain features such as gullies, small rises, and depressions. In other words, a west wind has the potential of loading a west slope in certain areas. The knowledge gained has helped to make operational decisions concerning which areas should be of priority focus when doing mid-day ski cutting or doing control work when the number of patrollers are limited.

The great interest of these new devices is the high frequency at which recordings are possible, so that investigations of flux response to wind gusts become possible. First, a one month wind speed recording at a rate of 10 Hz has shown the intricate structure of wind gusts on a mountainous terrain. Second, periodic recording during last two months have shown the possibility of drawing concentration profiles every four seconds for many periods of one or two minutes. These high-speed measurements clearly show that classical models are not efficient at such a time scale. Moreover, comparisons have been established between results from our automatic device (Flowcapt) and classical mechanical snow traps at different time rates and with different snow types.

However, transport rate non-stability even occurs in steady flows and can then be reproduced in a wind-tunnel. Some initial results including hot-wire anemometry and video at 200 images per second will lead to new developments so that the complex mechanisms of particle transport can be better understood.