Laserfiche WebLink
<br />0.12 <br /> <br /> <br />......... <br />c <br />'-" <br />0.10 <br />en <br />..... <br />C <br />::J <br />o <br />E 0.08 <br />o <br /> <br />>- <br />1: <br />::J 0.06 <br />o <br />.s: <br /> <br />tot') <br />I 0.04 <br /> <br />...... <br /> <br />Ze = 155 51.. <br />N = 196 <br />R = 0.76 <br />Ranges: 24-49 km <br /> <br /> <br /> <br />~ <br />V <br />P' <br /> <br />en <br />CD <br />0'10.02 <br />o <br />C> <br /> <br />0.00 <br />0.00 0.02 0.04 0.06 0.08 0.10 0.12 <br />Radar-est. hourly amounts (in) <br /> <br />Figure 8. - Scatter plot of 196 pairs of hourly-observed snowfall for Denver gages No. 1 to 3 versus radar-estimated <br />snowfall directly overhead using the relation Ze = 155 51.6. The regression line is dashed, and the 1: 1 line is solid. <br /> <br /> <br />The R value for gage No.5 is the highest in table 10 and the a value is extremely low at 1.0. <br />Table 10 shows that the center of the 0.50 tilted beam was below the elevation of gage No. <br />5, so ground clutter and beam blockage would certainly be expected near this gage. The 9 <br />range bins in the vertical array over gage No. 5 were briefly inspected for many volume <br />scans. Range bins along about the 260 and 261 0 radials for ranges of90 to 92 km rarely had <br />any detectable returns. Such bins were assumed to represent zero 'snowfall in the processing <br />software: However, the 3 bins along the 2620 radial frequently had returns, albeit with <br />relatively low Ze values. The RDRHGT program discussed in section 3.3 shows almost total <br />blockage of the 0.50 beam near gage No.5 for the 260 and 2610 radials, and a large fraction <br />of the beam is unblocked for the 2620 radial. <br /> <br />Because the snowfall rate was calculated for each bin in an array and then averaged, the <br />usual inclusion of about 6 zero values over gage No.5 resulted in very low snowfall estimates. <br />The optimization scheme, in forcing radar estimates and gage observations to have the same <br />average values, produced the very Iowa. value and relatively high ~ value. A plot of the <br />resulting radar estimates and gage measurements (not shown) shows limited scatter, in <br />agreement with the R value of 0.89. But when the reasonable a. and ~ values of equation (9) <br />were used, most radar-estimated hourly snowfalls became trivial. <br /> <br />The 3 range bins along the 2620 radial over gage No.5 contained enough signal to give rise <br />to some optimism. These results suggest that radar returns that are not totally blocked or <br />suppressed may be used over mountainous terrain to provide some estimates of snowfall <br />intensity. Further work is warranted on the topic of radar estimation of snowfall on <br />mountainous terrain; including consideration of the 1.50 beam. <br /> <br />Upwind arrays were also used with gages No.1 to 5 in the optimization scheme. Some R <br />values increased slightly, but no more than 0.02 over those values in table 10. These results <br />suggest the current advection scheme offers no significant increase in predictive value in the <br />Denver area over simply using the range bins directly above the location of interest. <br /> <br />34 <br />