WMOD00283 - Laserfiche WebLink

Home Browse Search

I I I I I I I I I I I I I I I I I I I radar at OOZ and 12Z on January 5, 1997, revealed no expectation of bright band effects or any other reason for this storm's unusual VPR to the north and northeast. This case suggests that the approach of averaging snowfall estimates all around the radar may be too simplistic in some cases. The VPS slope varies among storms. Table 1 shows the ratios between the calculated storm average SWE at the highest and lowest sampling levels, based on each storm's regression equation. It might be expected that the regression equations would be extrapolated to ground level rather than the height of the lowest tilt with a beam center 380 m above the radar. However, Ze-SWE relationships so far reported by Super and Holroyd (1997b) have used optimization of lowest-tilt Ze measurements over gages with the underlying gage observations. Comparisons were made only with gages from about 25 to 60 km from the radar because ground clutter, or ground clutter suppression algorithms, often contaminated closer Ze measurements while known range underestimation made comparisons suspect with farther gages. Hence, the VPS slopes were not extrapolated below the lowest radar-observed level. Six of the 9 ratios of table 1 are in the 0.49 to 0.69 range. This result indicates that the appropriate correction near 150 km range, where the calculated beam center for the lowest tilt of 0.5 degree is at the same elevation as the beam center for the 5th tilt (4.3 degrees) at 35 km range, is a multiplication factor of 1.5 to 2.0 for most storms. The median value of 0.56 corresponds to a correction factor of 1.8 for 150 km range. Obviously, lesser (greater) corrections are appropriate for closer (farther) ranges. However, discussion will be limited to the correction needed at 150 km range for illustration purposes. The eventual correction scheme will be applied to ranges from perhaps 50 to 150 km and possibly beyond. If the median adjustment of 1.8 at 150 km range was used for each of the 9 storms of table 1, corrected values would range from about 50 to 150 percent of (presumed) true values, with 6 of the 9 storm total accumulations from 88 to 124 percent of "true" values. Extreme corrected values within a factor or 2 of "true" at 150 km range, with most within about 20 percent of "true," certainly can be considered reasonable agreement. Failure to apply this average correction would result in the underestimates given by the ratios of table 1. That is, 150-km SWE accumulation estimates would range from 29 to 82 percent of "true," with most values between 50 to 70 percent of "true." Clearly, use of even an average correction factor for the nine storms is superior.to using no correction. This finding is in agreement with the results of Joss and Waldvogel (1989) based on data from Switzerland. 7