WMOD00399

• Title: WMO Training Workshop on Weather Modification for Meteorologists - Lecture Notes
• Date: 12/1/1979
• Weather Modification - Doc Type: Report

- 70 -

For heavy rains (> 25mm) the gaging requirements of Table 3 will
be red~ced (i:e. lesser density) somewhat and for light rains (< 2.5 mmO
they w1ll be 1ncreased (i.e. more gages per unit area). Although cautipn
should be exercised in extrapolating these estimates to other geograph-'
ical areas and weather regimes, they do provide a rough estimate of the'
measurement problem one might face in conducting a weather modification
experiment. .

3.2 Radar Estimation of Precipitation

Background

Weather radar is rece1v1ng increased attention in weather modi-
fication experiments, especially so, now that a computer processing
capability is a reality at many radar installations. Besides precipitaJ
tion detection and estimation, radar can be used to measure the vertical
and lateral extent of convective storms, their movemement and growth rate,
reflectivity profiles and, more recently, inf~rences of air motion can be
made with Doppler radars.

Radar is especially attractive for the estimation of precipita-
tion. In terms of rain detection, radar provides the equivalent of an :
infinitely dense raingage network which would make it a near perfect toql
for a convective precipitation measurement if the magnitude of the radar
precipitation estimates were without error. Unfortunately, this is fre-
quently not the case. Unless calibrated frequently, radar calibration is
always an uncertainty to some extent and the radar beam is usually not
uniformly filled with precipitation. Furthermore, the relationship of ;
radar reflectivity (2) to rainfall rate (R) is variable between storms ~nd
within storms even in the same geographical location and season (Stout and
Mueller, 1968, Joss et aI, 1968). One can also not count on "normal" re-
fractive conditions; with anomalous propagation, there is false echo and,
uncertainty as to what is being measured. All of these considerations
produce error in the radar estimation of rainfall.

Rather than attempt a quantitative correction for calibration
and beam filling uncertainties, anomalous propagations and Z-R variability,
it now appears more practical in many instances to calibrate the radar
against a few raingages. Radar defines the spatial variability and pro-I
vides a first estimate of the magnitude of the precipitation and the cali-
brating gages allow for its adjustment (Wilson, 1970). This still remaips
a rather controversial procedure. Several schemes (e.g. Brandes, 1975) :
have been proposed to accomplish the adjustment; our simplistic method ,
used in Florida will be described briefly here. For a comprehensive dis~
cuss ion of the use of radar in weather modification and other research, l
the radar is referred to a review paper by Atlas (1968) and to the lectures
by Melnichuk and Chernikov at the workshops.

Unadjusted Radar Performance in Florida

The radar systems and the methods of data processing in Florida
are described in some detail by Woodley et al (1975). Only the results
of the research will be described here.

~

Radar performance in estimating Florida convective rains was
determined by making comparisons of gage and radar-derived volumetric
rainfalls for gage clusters. In 1972, 40 recording raingages were in-