FLOOD00419

• County: Jefferson
• County: Park
• Community: Buffalo Creek
• Title: The Buffalo Creek Flash Flood of July 12, 1996 Draft - A Reconstruction of Rainfall and Meteorology
• Date: 7/12/1996
• Prepared For: CWCB
• Prepared By: Henz Meteorological Services
• Floodplain - Doc Type: Flood Mitigation/Flood Warning/Watershed Restoration

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

Additionally, it will present a detailed reconstitution of the rainfall on July
121h and contrast it to the rainfall associated with several additional storm events
which occurred over the Buffalo Creek basin during the post-fire period of 1996,
Finally, the report will present information on the meteorological causes of the
primary July 12'h flash flood event and contrast it to the causes of the other
lessor events which occurred,

2,0 Use of radar to describe storm rainfall

The ability to use radar to obtain a map of current rainfall has been
pursued for over 30 years by engineers, meteorologists and hydrologists, In
general, most current radar-rainfall techniques rely an assumed relationship
between the strength of the radar reflectivity and the intensity of the rainfall rate,
This relationship is described by the equation below:

(1 )

Z = A Rb

where, Z is the radar reflectivity, R is the rainfall rate, A is an empirically derived
co-efficient related to the cloud physics of the storm cloud water droplets and b is
another empirical co-efficient related to the type of storm cloud present. This
relationship has proven to produce highly variable results, Since the values of
both A and b must be assumed to allow the equation to be solved correctly,
many opportunities for errors in the calculation are possible.

The algorithms used to estimate the rainfall are standard for use around
the country and have not proven to be responsive to local cloud variations. The
r-squared or "goodness" correlation factor of the rain to radar reflectivity
statistical relationship has varied from 0.15 to 0,90 on a daily basis and for most
storm seasons has been about 0.60. The good "r's" (values >0.75) have been
achieved for the low volume and low intensity rain events, generally those of less
than 0.25"/hr accumulation rates. The high intensity, high volume, "front-end
dumper" thunderstorms have produced r-values of 0,15 to 0.45. Thus the
standard products appear to be unreliable at this point. The storm rainfall has
been both overestimated and underestimated for periods of less than three
hours for storms within 25 miles of each other.

Finally, hail "pollution" of the equation has proven to be a troublesome
problem, The strength of the radar return signal is related objectively to the
diameter of the rain droplet size, The strong radar return signal produced by
wet hail stones frequently causes an over-estimation of the rainfall rate.
Attempts have been made to reduce this over-estimation by adjusting the
coefficients A and b in Equation 1. A satisfactory solution to this problem
continues to prove quite elusive.

4