FLOOD01375

• County: Larimer
• Title: Mesoscale and Radar Observations of the Fort Collins Flash Flood of 28 July 1997
• Date: 2/1/1999
• Prepared By: American Meteorological Society
• Floodplain - Doc Type: Flood Documentation Report

nificantly larger for the R(KDI" ZDR) product compared
to the Z-R. As expected from Fig, 18a, the peak rain
rate from the blended polarimetric algorithm is located
in the downdraft several kilometers to the northwest
of the peak Z-R,
As discussed in Jameson (1985), KDP - W(1 - L)
where W is rain mass mixing ratio and L is the mass-
weighted rain drop axis ratio, Note that the value of
KDP at the peak R(KDP' ZOR) in Fig, 18b is only slightly
higher than at the peak Z-R, However, Z"R is at least
0,5 dB less at the peak R(KIJI" Z"R)' Since L -
[lOJZDRlIO]-()n) (Jameson 1991), the location of the peak
R(K"I" ZIJR) must contain a larger number of smaller
drops relative to the location of the maximum ZH' ZDR'
Of course, the Z-R relationship places the peak
rain rate at this point. Studies such as Blanchard and
Spencer (1970) suggest that in heavy rain the concen,
tration of smaller drops (D - 1-2 mm) may increase
with increasing rain intensity, This behavior of the
drop size distribution in heavy rain can result in an un-
derestimation of the peak rain rate estimated by the
NEXRAD Z-R (e,g" Smith et al. 1996) and cause
errors in the location of the peak Z-inferred rain rates
since reflectivity is biased by the presence of large
drops, Although we have insufficient high-resolution
rain gauge data to prove that the polarimetric rainfall
estimate was more accurate in rainfall amount and
location at this time, we will demonstrate in the
following section that the blended R(K,w' ZIJR) algo-
rithm did provide a better estimate of storm cumula,
tive precipitation,

d, Radar-estimated rainfall
The reflectivity (2) data collected by NEXRAD
(e,g" KCYS) and CSU-CHILL radars can be con-
verted to rain rate using ref1ectivity-rainfall (Z-R)
relationships (e,g" Battan 1973), Importantly, pola,
rimetric techniques (cL Doviak and Zmic 1993) can
also be utilized with the CSU-CHILL data, In opera,

FIG, 17. CSU-CHILL 2108 MDT 28 July 1997, Northwest,
southeast cross sections taken along a 2950 radial from CSU-
CHll..L. (a) Reflectivity (shaded), ZOR (black contour), and Knr
(light blue contour); contour interval 0.50 km-l (0.25 dB) for KDl'
(ZDR) starting at 10 km-' (I dB). (b) Rctlectivity and radial veloc-
ity; receding (approaching) radial velocities arc contoured using
solid (dashed) lines at an interval of 3 m S-I. (c) Reflectivity
(shaded), precipitation-sized ice fraction <1;, black contour), and
LDR (blue contour). Ice fraction contoured at 0.1,0.5, and 0.9.
LDR contours at-23.5 and -24.5 dB. The approximate height of
the OOC level is indicated by a bold hash mark at Z = 3.6 km. Grid
origin as in Fig. 9.

Bulletin of the American Meteorological Society

...-----=

pm, I::m,~=-I
10 20 30 35 40 45 50 54 58
a, CSU-CHILL 2108 MDT: d87JKdp/ZDR
9
NW SE
8
7
6
~
e 5
-=-
;C 4
""
]
- 3
2
0
-6 -5 -4 -3 -2 -I 0 2 3 4 5
h. CSU-CIULL 2108 MDT: dll7JVR
9
SE
8
7
6
~
e 5
-=-
<:: 4
"'5L
] 3
-
2
0
-6 -5 -4 -3 -2 -I 0 2 3 4 5
c. CSU-CIIILL 2lOI\ MDT: dl17JFl/LDR
9

NW SE
1\
7
6
~
-
-
.g 5
;C 4
""
'OJ
::: 3
2
0

-6 -5 -4 -3 -2 -I 0 2 3 4 5
Di,lance NW - SE of Taft and Drake (km)

209