Any kind of atmospheric disturbance has a measurable effect on GPS
space and time precision, the question is more about how much the
impact is. Presuming that the GPS receiver is stationary, the antenna
has a clear visibility of the sky, then "ordinary weather" like
clouds, moderate rain or snow, should not have a large effect. The GPS
signal will be attenuated by rain and snow, but precision should not
be impacted provided that the GPS receiver is operating in stationary
mode (I'm referring to stationary mode as I presume the goal here is
to get precise GPS or UTC time, not position).

The signal quality will definitely be very poor if keeping the antenna
window mounted, especially if there are nearby metal surfaces. I see a
significant difference in quality between an antenna by the window in
my basement (up to and including complete periodic loss of GPS lock)
versus keeping the antenna in the attic (composition roof in my case).
In the basement there is only 1/4 sky direct view as opposed to the
attic which has essentially unobstructed view. The antenna in the
attic never gave me trouble.

I have not compared the difference in signal quality between an attic
mounted antenna versus an externally mounted antenna so I don't know
about that.

I haven't seen loss of GPS lock under light to moderate rain, let
alone simple clouds. Heavy rain doesn't seem to be a big problem
either. Your friend's experience of losing GPS lock on cloudy days
seems very, very strange to me. Barring malfunctioning receiver or a
defective antenna, I would first make sure to have the antenna
externally mounted, or at least in the attic if the roof is
composition or wood shingles.

Your friend could collect NMEA data on PDOP, HDOP, satellites view
with signal strengths and how many satellites are tracked and part of
the GPS solution. I definitely see better signal quality with the
BG7TBL GPSDO as opposed to a Garmin 18x with serial output and PPS.
The Oscilloquartz Star IV GPSDO seems to give even better results than
the BG7TBL, although having just recently bought them, I don't have
enough statistical data.

Another thing to try would be setting the minimum elevation threshold
for a satellite to be included in the position and time solution to
make sure to avoid surrounding hills or mountains, if any. Both the
Oscilloquartz Star IV GPSDO as well as the plain old UBLOX-7 GPS
receiver with PPS allow you to set the minimum elevation threshold.
The BG7TBL GPSDO does not allow you to do that. You should look at the
specifications of the two GPS receivers.

At the very high end of measurements, there are observable effects
during weather fronts -- which are of course much more severe than
simply ordinary clouds or light rain. This article is rather dated but
very interesting. With dual frequency receivers (L1 and L2 bands) they
can observe precision errors down to the order of about 3 - 10
centimeters  :

http://geodesy.unr.edu/publications/gpsworld.may98.pdf

-- Fio Cattaneo

Universal AC, can Entropy be reversed? -- "THERE IS AS YET
INSUFFICIENT DATA FOR A MEANINGFUL ANSWER."

On Mon, Oct 21, 2019 at 11:01 PM Dana Whitlow k8yumdoober@gmail.com wrote:

In message CADXevOaQ6a7eUoKciAttXhr5=w5Y6xxYbojFQMRB42x2QPLXfQ@mail.gmail.com, Fiorenzo Cattaneo writes:

Actually, it's even simpler than that:

Any electrical charge in the freshnell-zone between the two antennas
delays the signal.

In practice that means "any ion ..."

Rain clouds harbour significant ionization, long before they become
thunderstorms.

Apart from that, the lower atmosphere is pretty predictable with
respect to ionization.

--
Poul-Henning Kamp      | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG        | TCP/IP since RFC 956
FreeBSD committer      | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

I see no attenuation at 1376 MHz (close to GPS frequency) when observing
pulsars with a radio telescope. Even the brightest pulsar (Vela) is so much
fainter than a GPS signal which boom in when they happen to pass into the
telescope's beam.

It definitely happens at higher frequencies though. Up around 20 GHz it's a
different story.

On Tue, 22 Oct 2019 at 17:01, Dana Whitlow k8yumdoober@gmail.com wrote:

On 10/21/19 10:06 PM, Dana Whitlow wrote:

Clpuds and even light rain won't change the received SNR very much at L
band.

I'm going to guess that there's something that is cloud correlated -
open and closing ventilation shutters? Soil or ground cover moisture
changes, which changes the strength of a multipath signal. Some sort of
thermal expansion effects changing the spacing of siding panels when the
sun shines on them vs no.

On 10/22/19 1:13 AM, Poul-Henning Kamp wrote:

That's still a pretty small effect for RF propagation at L-band

and here's a paper discussing just such effects
https://iopscience.iop.org/article/10.1088/1742-6596/1112/1/012021/pdf

what happens here is that charge in a large thunderstorm affects the ion
density in the ionosphere.

However, I don't think that's "cloudy day vs sunny day"

On 10/22/19 2:23 AM, Jim Palfreyman wrote:

On a sensitive receiver I operated at 13.402 GHz in New Mexico, I could
easily tell when it was overcast or not by the noise temperature. There
was some discussion about whether we were seeing the physical
temperature of the clouds, or we were seeing a reflection of the earth
off the water particles in the clouds.  We didn't go any further, since
our project was to measure a radar signal, but it was interesting.

Hi jimlux -- unfortunately my domain expertise is in the digital
domain (computer science) and I'm having a hard time in following this
paper. I do understand that "space weather", i.e. CMEs (coronal mass
ejections) will significantly impact GPS/GLONASS operations when the
charged particles reach earth.

One question I would have for you is on the ionospheric delays due to
weather conditions like moderate or heavy rain, thunderstorms and more
severe weather fronts. My understanding of it is that when it comes to
GPS usage for positioning purposes, the use of DGPS or some kind of
augmentation system like WAAS provides "good enough" corrections to
compensate for ionospheric perturbations. Whereas for timing
applications, using stationary mode (i.e. either manually programming
station coordinates into the GPS receiver or having the GPS receiver
compute them over several hours by averaging out positioning data) is
enough for "good" enough timing measurements.

How much is "good enough" is not quite clear to me. What would be the
magnitude of timing errors with GPS in stationary mode, assuming my
understanding of stationary mode is correct?

My current Oscilloquartz DGPS, BG7TBL DGPS and GPS + Symmetricom BCP
635 timing board setups gives me an estimated best time accuracy in
the order of 5 to 50 microseconds (my own estimate -- unfortunately I
don't have access to a lab with calibrated reference cesium
oscillators), which is enough for my "time-nuts" hobby.

I wonder if I would be able to measure ionospheric delays -- perhaps I
could measure them by comparing the difference of my "well known" GPS
coordinates once Oscilloquartz DGPS locks in stationary mode versus
the GPS coordinates I receive with a UBLOX-7 vanilla receiver?

Apologies for the many questions.

-- Fio Cattaneo

Universal AC, can Entropy be reversed? -- "THERE IS AS YET
INSUFFICIENT DATA FOR A MEANINGFUL ANSWER."

On Tue, Oct 22, 2019 at 6:06 AM jimlux jimlux@earthlink.net wrote:

I don't think clouds is the direct cause, but of course clouds in the sky
can be correlated with wet foliage.

Especially if the GPS Field Of View has a lot of angle taken up by tree
canopy, wet foliage can substantially degrade not just GPS reception but
other VHF and UHF signals.

I notice this the most in the fall when the leaves are large and mature and
are start losing their waxy coat, but have not yet fallen.

Even after all the leaves have fallen, wet branches have attenuation
ability as well.

Some research papers have connected windy conditions with UHF attenuation
especially deep fading as well (https://ieeexplore.ieee.org/document/1337081
).

Tim N3QE

On Tue, Oct 22, 2019 at 2:01 AM Dana Whitlow k8yumdoober@gmail.com wrote:

As far as I know, my friend's GPS dropout problems have been associated with
simple clouds with no rain (at least no rain reaching the surface).

Dana

On Wed, Oct 23, 2019 at 2:00 PM Tim Shoppa tshoppa@gmail.com wrote: