WMOD00417

• Title: Computer Processing of Digital Radar Data Gathered During HIPLEX
• Date: 4/4/1978
• Weather Modification - Doc Type: Report

..:

~.

499

APRIL 1978

MELVIN J. SCHROEDER AND GERARD E. KLAZURA

TABLE 1. Characteristics of HIPLEX SWR-75 radars.

does not interfere with the monitoring and recording
of the precipitation- data.

3. Calibration and transfer function

Radar and digitizer calibrations are crucial to attain-
ing useful data from the DVIP-produced numbers
recorded on the magnetic tape. Electronic and sphere
calibration procedures are used. Complete system
calibration is performed on each of the radars at the
beginning and end of each operational season. This
calibration consists of a measurement of the effective
antenna system gain and also incorporates a pre-, mid-
and post-season antenna boresighting, elevation and
azimuth error determinations and corrections using
solar evaluations.
Digitizer calibrations to relate the recorded binarv
eight-bit number in each bin to a received power at the
radar are made weekly. The DVIP is quite stable,
usually not deviating by more than i dB from one week
to the next. The DVIP calibration consists basically
of injecting RF test signals of specified power levels
through the directional cDupler into the receiver. For
each DVIP level the test signal and corresponding
, digitizer response are recorded (dBm vs DVIP units).
Later a polynominal regression analysis of these dBm/
DVIP data is carried out to compute the coefficients
of a third-order equation of the form

y= a+bx+cx2+dr,

where x is the DVIP value and y the corresponding
estimated dBm value.
Daily checks consist of measuring the pulse repetition
frequency, average power, transmitter frequency,
minimum discernable signal and checking three levels
of the DVIP. If the DVIP varies by more than! dBm
from the weekly calibration, a complete recalibration
is performed.
An independent evaluation of the Miles City radar
and calibration procedures was carried out by Smith
(1977).
Recorded data are not range-corrected, but subse-
quent data processing introduces the correction. The
2.5 dBm log-averaging correction (Lhermitte and
Kessler, 1965) and 0.2 dBm finite incremental trunca-
tion bias correction (Sirmans, 1972) are also made
during the computer processing stage.

4. Data acquisition and flow

a. Data acquisition

Recording of data is accomplished in two modes of
operation-surveillance and volume. While in the sur-
veillance mode, one 5 min volume scan (3600 in azimuth
and 120 in elevation starting at 10 and going in steps
of 10) is taken on the hour and 30 min past the hour.
The purpose of the surveillance mode is to' detect
echoes within 150 km, at which time recording of con-

Peak transmitter power
Pulse duration
Antenna type
Horizontal half-power beam width
Vertical half-power beamwidth
Effective system gain
Wavelength
Pulse repetition frequency
Receiver noise
Receiver
Receiver bandwidth
Receiver dynamic range
System noise figure

250 kW
2.0,..s
circular parabolic
0.90
1.00
43.7* dB
5.4cm
207/414 pps
2 dB
logarithmic
0.6 MHz
80 dB
<3 dB

* This figure incorporates all waveguide mismatch and similar
losses beyond the directional couplers as well as radome loss.

(1)

tinuous, sequential, 5 min volume scans begins. The
antenna azimuth sweep and elevation step sequence are
automatically regulated. The sweep rate is controlled
by the pulse repetition frequency (PRF) setting and
number of pulse samples per averaged return (SA). For
HIPLEX the PRF is set at 414 and the SA is set at 32
for a base elevation sweep and 16 for higher elevation
sweeps. This translates to a sweep duration (one
antenna rotation) of 34 s at 10 elevation and 17 s for
the other elevations. The returned signals are digitized
and linearly averaged over the range bin interval using
rectangular integration techniques. Pulse-to-pulse aver-
aging is initiated on the whole azimuth angle (e.g.,
273.00) and terminates after about 0.70 (e.g., 273.70)
in the 10 azimuth recording mode. Each !-km range bin
value is determined by adding four t km (in range)
samples, dividing by 4, and repeating and accumulating
until the number of samples specified by SA is fulfilled
at which time the sum is divided by SA to produce the
DVIP number.
A "blue-sky" elimination feature is wired into the
radar hardware so that only records containing at least
one radial with bin data above a preset threshold level
are recorded. The first and last records of a 'constant
elevation sweep are always recorded to maintain
antenna sweep history.

b. Description of magnetic tape record

Each magnetic tape record (Fig. 1) consists of 22
frames of housekeeping information (date, time, radar
settings, etc.) and four sets of 1) four frames of azimuth
and elevation data, 2) 250 frames of averaged-echo
range bin data, 3) three frames of project aircraft posi-
tions, and 4) three frames of automatic calibration data.
Housekeeping information includes a frame for
operator's notes which are loaded via a special key-
board. This is primarily used to record pertinent infor-
mation about project aircraft (i.e., locations, times,
seeding, cloud penetrations, etc.). The range delay sett-
ing (switch selectable 0-99) is used to specify the range
of the first data bin; scan mode inciit.ates whether the