Saskatoon (52N,107W) MF Radar Winds, monthly quick look plots
A prototype coherent receiver was first used for a dedicated Doppler
velocity measurement during ~1985-1987. Subsequently Rob Strother-Stewart
designed very good, fast gain switching, coherent receivers to be used
in the Canadian Network for Space Research collaboration (CNSR, 1991-1995, which comprised Saskatoon, Robsart,
Sylvan Lake, London, Ont.)
The Saskatoon (often called "PARK", because it is located in what was
then Park Municipality, but
now amalgamated into Corman Park) first coherent system ran in a
system consisting of a receiver inteface with 2MHz 6502 CPU, A/Ds etc.),
an Apple II+, and a Commodore C128, for full correlation analysis (as did
Robsart and Sylvan Lake). As in the RTW (non-coherent) system, data
were converted to bit amplitudes relative to long-term running means for
each height gate (32, 40-133 virtual Km) and component
(I1,Q1,I2,Q2,I3,A3,I4,Q4), and lagged complex auto and cross correlations
consisted of arrays of numbers of 1-matches.
In this system, all receivers are sampled simultaneously.
During this time at Park we used a locally designed and built 50KW
(nominal) Transmitter (20μsec pulse) - but high power tubes were
becoming very expensive.The other CNSR sites used 25KW TOMCO/ATRAD
(Australia) solid state units. When CNSR ended (August 1995), we
turned off the tube Tx and installed one of the solid state units
at Parksite (the other went to Platteville).
In August 1996 we upgraded the analysis to use 8 bit amplitudes in a
386 PC.
There was a considerable increase in the number of wind values produced.
In fall 2013 we suffered 3 break-ins at the site. In the last our
Tx antenna open wire feeders were cut (around 3AM!). I assume
thinking copper (they weren't copper - just copper clad steel.)
So we made some Tx baluns and fed three E-W receiving dipoles in parallel
( feeders are underground!) ,
but we receive on N-S dipoles, so there was significant loss
of data at lower heights. In August 2014 we ran an
(over ground) new coax feeder to 1/2 the Tx antenna to return to N-S
transmission, and it survived(!) for more than a month. The subsequent
change back to N-S polarization on Sept. 30 resulted in at least one
extra height at the bottom. A diagram of the antenna arrays is
here.
So far (April 2015) it hasn't been chewed and the
cows haven't trampled it (much). In summer 2016 we had a disconnect
in the Tx antenna feed (cowsdragged the on-the-ground feeder?) and
lost 2 weeks.
Grant applications to run the site have not been funded for the last few
years but we are scrounging and continuing on less than a shoestring.
The end game(cm, my view):
There are two of us left - both "retired".
It is not excessively expensive to run the radar (~$7K Can/ yr), as
long as there is no radar array damage that we can't work around
by ourselves (e.g. by changing to good antennas, not by fixing them).
For that we get to contribute to a couple of global studies a year,
since single site research topics are rather limited, and MF radar
has lost in popularity to meteor radar.
Come September 2019 we will have had 41 yrs of continuous operation
(and years before that of episodic or limited operation, e.g. 1 hr
per day).
There are data features to study which don't require antennna work,
e.g. the statistical nature of MF scatter, single scatterer vs.
diffraction pattern, but there are sites
with much bigger arrays which can do, and have done, more.
So, barring unforseen circumstances (the final lease runs out in July 2020,
and time is needed to decommission the site), we will cease the Parksite
radar operation in Sept. 2019. (In fact we have shut down in early
November 2019.)
Such a long data set needs a long term study. And one has been done -
but due to reasons including internal differences of opinion,
it may never be submitted or, if submitted, published. However we still
regard the results as a reliable statement of change we see in the
40 years of operation. A skeleton version of the zeroeth draft appears
here.
The figures and captions are intended to be final; the text is
obviously not (cm).
