Laserfiche WebLink
<br />I <br /> <br />I <br /> <br />Goodison (1978) showed that both the Universal (also called Belfort) and Fisher-Porter gauges, used in <br />the United States for decades, had about a 50% undercatch with a wind speed of only 5 mph if not <br />shielded. By 10 mph the gauge catch had decreased to only about 35% of ground truth for the Universal <br />and 25% for the Fisher-Porter. These are obviously huge errors with even light to moderate wind speeds, <br />and undercatch increases with stronger winds. Use of Alter wind shields, the general type used in the <br />U.S. and for the 2003/04 Utah program, increased the 10 mph catch to about 55% of ground truth for the <br />Universal and 45% for the Fisher-Porter. While use of Alter wind shields improves gauge catch, it is <br />clearly not the total solution. Nipher wind shields used in Canada have a superior catch efficiency <br />(Goodison 1978), but must frequently be cleared of snow, making them impractical for remote mountain <br />locations serviced at approximately weekly intervals. <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />It is well documented that accurate snowfall measurements cannot be obtained in windy locations with <br />the possible exception of large and elaborate double fence structures for shielding gauges known as the <br />Double Fence Intercomparison Reference or DFIR with a 12 m (39 ft) diameter. DFlR equipped gauges <br />were used at several locations world wide in a recent and comprehensive snow measurement study for the <br />World Meteorological Organization (WMO) reported by Goodison et al. (1998). Construction of such <br />structures was beyond the resources of the Utah project. Moreover, it is still uncertain how well gauge <br />catch within a DFlR shield compares with "ground truth" which is difficult to determine. The WMO <br />study recommends use ofa shield of bushes encircling the gauge, and cut off to gauge orifice level, as the <br />primary standard. Bushes of that type do not exist at high altitudes on the Wasatch Plateau. The WMO <br />study recommends the DFIR as a secondary standard, and also recommends clearings in forest with the <br />distance of trees from the gauge roughly equal to the height of the trees. It was not specified whether the <br />forest should consist of deciduous or coniferous trees but the former can only provide limited protection <br />from the wind once they shed their leaves in the fall. <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />The WMO-recommended clearing size falls into the over-protected class of Brown and Peck (1962). <br />They published a subjective classification system to evaluate snowfall measurements in Utah. Seasonal <br />snowcourse data, generally obtained in sheltered forest locations, was compared with nearby gauge <br />seasonal accumulations. Gauge site classifications ranged from over-protected to very windy based on <br />nearby objects and general terrain. The experience of the authors of this report caused them to favor an <br />over-protected natural clearing in a conifer forest as the best mountain gauge location. Such clearings <br />provided very close agreement between snowcourse and gauge seasonal totals according to Brown and <br />Peck (1962). However, they recommended larger clearings in their "well-protected" class, sheltered in all <br />directions by objects subtending 20 to 30 degrees vertical from the gauge orifice. But such clearings <br />often have significant wind and gustiness during snowfalls. Only clearings with the highest conifer tree <br />tops about 45 deg above the gauge orifice for all common wind directions during storms, and generally <br />encircling the gauge, reduce winds to generally light. Use of such clearings in the Bridger Range <br />Experiment in Montana showed high correlation coefficients of about 0.95 for daily snowfall totals <br />between sheltered gauge sites located about 4 kIn apart (Super and Heimbach 1983). Such high values <br />could not be achieved unless the clearings were consistent in limiting gauge level winds. Gauge <br />comparisons were made with co-located snowboards in some of the clearings, the latter to provide <br />"ground truth." Resulting correlation coefficients were an excellent 0.99. One of the authors has used <br />similar gauge sites in mountains of Colorado and Arizona and during several previous projects in Utah, <br />always with good results. Consequently, only "over-protected" clearings in conifer forest were <br />considered suitable for gauge operation in the 2003/04 Utah project as specified in Utah's proposal to <br />Reclamation for the project. <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />I <br /> <br />Funding was sufficient for purchase of six digitized load cell gauges, recorded by data loggers, and <br />equipped with Alter type wind shields. The latter were made of clear Lexan plastic, which is durable and <br />resistant to deterioration in sunlight. These gauge systems were purchased from ETl Instrument Systems, <br />Inc. of Fort Collins, CO. Four gauges were proposed to be located along the likely seeding plume <br />trajectory, with two of the four between the HAS seeding site and the plateau top instrumented target, <br />TAR (See Fig. 1). One was located 320 m east-southeast of the TAR in a clearing used in previous <br /> <br />I <br /> <br />I <br /> <br />17 <br /> <br />I <br />I <br />