Laserfiche WebLink
<br />Finally, equation (9) was applied to all five gages with the result that the R values were <br />basically unchanged from table 10, similar to the Cleveland findings, with the exception of <br />gage No.5 which had its R value reduced from 0.89 to 0.82. Moreover, the radar-estimated <br />snowfall accumulations were reasonable except at gage No.5. The average radar-estimated <br />hourly snowfall accumulation over gage No.5 was 0.002 inch, and the gage-observed value <br />was 0.025 inch h-1. So, with the exception of the blocked and cluttered mountainous regions, <br />use of equation (9) appears justified in the Denver area, at least until testing with additional <br />data is possible. <br /> <br />9.3 Significance of Different Ze-S Relations for Cleveland and Denver <br /> <br />The recommended initial a and ~ values are 318 and 1.5 for Cleveland and 155 and 1.6 for <br />Denver. These values are very similar to the top two curves of figure 2. <br /> <br />It is reasonable to ask whether these curves produce sufficiently different snowfall <br />accumulations to make a practical difference. One way to address that question is to apply <br />the Cleveland relation of equation (8) to Denver data and the Denver relation of equation (9) <br />to Cleveland data. This application produced the results shown on figures 9 and 10, <br />respectively; <br /> <br />Examination of figure 9 reveals Cleveland equation (8) significantly underpredicts Denver <br />hourly snowfall accumulations. The dashed regression line is well above the 1:1 line, and the <br />average radar-estimated snowfall accumulation is 0.0126 inch, 61 percent of the gage-, <br />observed 0.0205 inch. Conversely, figure 10 shows use of Denver equation (9) with Cleveland <br />radar measurements seriously overpredicts snowfall accumulations as shown by the <br />regression line well below the 1:1 line. The average radar-estimated snowfall accumulation <br />in this case is 0.0289 inch, 144 percent of the gage-observed average value of 0.0201 inch. <br />Use of a Ze-S relationship appropriate to the geographical region clearly is important for <br />quantitative snowfall estimation. <br /> <br />It is not known why the different Ze-S relations resulted from the Cleveland and Denver area <br />data. The difference may be related to microphysical differences between the lake effect <br />storms that predominated over the Cleveland gages and the upslope storms that affected the. <br />Denver area. Ice crystal observations were routinely made near Denver but not near <br />Cleveland. However, microphysical observations of lake effect storms have been published <br />which may be compared with the past winter's Denver observations in a future report. <br /> <br />35 <br />