Laserfiche WebLink
<br /> <br />-38- <br /> <br />missed samples can be identified and adjusted values estimated from such <br />graphs. Experience has shown that at any given time the densities of <br />individual samples along a SnoW course generally vary less than lO percent; <br />thus, comparison of densities may yi~ld a clue to an inconsistent sample. <br /> <br />Where re-surveys are impossible, some questionable snow-course sur- <br />veys can be adjusted by direct correlation with records from other courses <br />surveyed at the same time. Very high correlation sometimes exists bE1tween <br />adjacent courses (coefficients of determination as high as 0.98). <br /> <br />Consistency in values of snow density can be tested against repcrted <br />precipitation and snow depths. Such, tests should be accepted with caution, <br />in part because field observers, lacking a snow board and convenient means <br />for measuring density, may have reported precipitation according to the <br />empirical assumption of "lO-percent'" density for new fallen snow. <br /> <br />Reference has been made to isochion maps and profiles fcr isolating <br />specific errors. These may afford a,test for consistency in snow records <br />of diverse kinds, althOUgh they are commonly unreliable for very rugged <br />terrane. <br /> <br />Publication <br /> <br />, Data on snowfall (precipitation, as snow) and on depth of accumulated <br />s)low ("snow on the ground") have been and are published by the U. S. <br />Weather Bureau in its climatologicseriesl "Climatological Data" and <br />"storage-gage Precipitation Data". 1lecords from snow-course surveys are <br />published by the U. S. Soil Conservation Service, ,currently for about <br />1,300 courses in the western United States. These are issued separately <br />as a "Water Supply Outlcok" for each State, except California, as of the <br />1st day of the month, beginning with January or February and ending with <br />M;ay or June of each year. Thus, publication corresponds to the schedule <br />of surveys, which usually are made within a three-day or four-day <br />interval at the beginning of the month. At times the interval is longer <br />when weekends or operational difficulties intervene. <br /> <br />The snow-survey data published by the Soil Conservation Service are <br />expressed in terms ofa IS-year average, the base period for which has <br />been updated every five years. Presently, the base period is 1948-1962. <br />For courses whose record is incomplete but spans 10 years or more in the <br />base period, a l,-yearaverage is extrapolated. Specifically, a short- <br />term average is extended in simple ratio to the standard l5-year average <br />from an adjacent course provided that, for the period of common record, <br />that ratio has been reasonably stable from month to month and from year <br />to year. Alternatively, single months that are missing can be interpo-. <br />lated or extrapolated, and not correlated with an adjacent record, <br />prOvided month-to-month changes at the course in question hlilve been <br />reascnably stable for ten years or more. For courses less thliln ten years <br />old, averages are computed for the period of record only, and are so <br />identified in the publication. If a course has less than fiveyesrs of <br />record, usually no average is computeo or published. <br />