WMOD00425

• Title: Demonstration of Improved Operational Water Resources Management Through Use of Better Snow Water Equivalent Information
• Date: 9/1/2002
• Weather Modification - Doc Type: Report

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

At altitudes just below the 0 oC level, melting ice (the bright band) exists, and nearly any
temperature profile is possible, including more than a l-km thickness of near-isothermal air and
multiple inversions. The altitude of the bottom of the bright band was noted and compared with
temperatures from rawinsondes. This altitude was temporally unstable, as were the altitudes of
other temperatures. It appeared, however, that the temperature of the bottom of the bright band
was usually about +4 oc. This is in agreement with typical experiences at the surface, where wet
snow is seldom seen at temperatures warmer than +4 oc. Though this threshold is more variable
than the melting level, a conservative transition from melting ice to rain is declared to be at the
lowest altitude of +4 oC, i.e., exclusively rain is assumed if the top of the radar beam is entirely
below the lowest altitude of +4 oc. In this instance, a Z-R instead of Z-S relationship is used to
produce rainfall rates. Super and Holroyd (1998) showed results near Albany, however,
indicating that these two relationships may not be substantially different. Our preliminary case
study results (section 6.6) confirm this similarity.

Any beam even partially sampling the layer between 0 oC and +4 oC is assumed by the new
algorithm to be contaminated by bright band effects. These data require a different relationship
between Ze and precipitation rate. Coding has recently been added to estimate the proportion of
the beam contaminated by the bright band, thereby decreasing the bright band correction with
range. This coding was being tested in early 2002. We adopted the earlier, more conservative
approach because of the large temporal and spatial variability of bright band structure observed
in the Northern Plains. Fabry and Zawadzki (1995) also reported great variability in bright band
intensities. A preliminary Z-R relationship viewing a mixture of snow, rain, and melting snow
("slush") was calculated from three case studies, as mentioned in section 6.6.

Once the precipitation type (snow, melting snow, or rain) is determined, the most appropriate
Z-R (or Z-S) equation (section 6.6) is applied to calculate a precipitation amount for that radar
bin. This amount is then extrapolated to the surface using the new climatological VPR and the
bin clearance C (section 6.2). Then, from the terrain file, the surface elevation under the bin is
determined, and its temperature is obtained from a sounding (see next paragraph). This
temperature is also compared to the 0 oC and +4 oC thresholds and, using the same logic as the
radar bin sample, a "final" precipitation type at the surface is prescribed. No SD is accumulated
in areas of rain; for areas of melting snow, a larger snow density p is used for SD than is used for
areas of dry snow. The entire precipitation typing is diagramed in figure 9.

We chose to extract the 0 oC altitude from vertical profiles of temperature and dew point from
numerical weather prediction (NWP) models ("model soundings") instead of rawinsonde
soundings. The latter may be unrepresentative if they are distant or if several hours have passed
since the rawinsonde was released (recall the 12-hour rawinsonde frequency). While model
soundings are subject to the same errors as NWP in general, they have additional advantages over .
rawinsonde soundings. First, models have recently offered soundings at hourly forecast intervals.
Second, these soundings are available for many more locations than those with rawinsonde
soundings. This is useful for obtaining soundings at WSR-88D sites that are not also upper-air
sites.

23