[MAMS] Wind losses

[Dr. Gerald N. Johnson]

On 4/5/2011 9:41 AM, Lloyd Ellsworth wrote:
> I have set up and operated rover many times in very windy locations.
> It can be very frustrating, and destructive.
>
> Biggest suggestion. Advance planning. Testing. Better mechanical
> supports. Keeping to short paths, with short antennas, and horns
> rather than dishes and long antennas.
>
> As Jerry noted, the winds tend to blow away the nicely layered
> atmosphere, that promotes good DX conditions. I would support that
> observation.
>
> Conclusion I reached many years ago, along with many others, is that
> wind tends to scatter signals. I tend to use the term shred on
> microwaves. Winds seem normally to be gusty in nature. Changing speed,
> often, quickly, rather than being steady. Even direction. And on the
> Lake shore. Rarely, is it steady. This creates different atmospheric
> density zones, which affects signals, by shifting the signal direction
> slightly. As it passes from one density cell to a different density
> cell. The wind forming pressure waves. Ridges. The signal then
> slightly interfering with itself. Thus noise.

This is an area of interest, unfortunately historical weather 
observations and forecast models work on the contention that hourly 
averages and grids of 30 to 50 km are adequate detail. Which thoroughly 
messes up forecasting winds for wind farms. Those wind machines report 
wind speed (each has an anemometer on the nacelle unfortunately down 
wind of the hub) and AC power output at minute intervals. And since the 
shaft horsepower of the wind turbine varies according to the cube of the 
wind speed those power figures jump all over.

There are at least a dozen Planetary Boundary Layer models that try to 
couple near surface winds to winds aloft. I think none of them have been 
proven reliable comparing short term reports of winds aloft to winds on 
the ground. A few years back while visiting a NWS office (in the company 
of a private forecaster that I do software for) I asked about how the 
NWS forecast gusts. 1.4 times the average speed was their answer. Wind 
in the boundary layer is complex because its caused by pressure 
differences between points, but it has to go to zero at the surface 
(down between the blades of grass) and the surface shape and cover 
affects it down low. Or is it complex because motion of air causes the 
differences in pressure? The friction at the surface can make bodies of 
air travel in whorls or rolls, maybe in pressure waves. I'm sure that 
temperature, humidity, and dirt content varies all over the map too. All 
these affect density and so refraction in stable times making area 
inversions and ducts. Then you add in lift of storm clouds and the total 
motion gets really complex. The weather models run on a very fine scale 
with just the initialization of the area immediately around a potential 
storm will often model the storm details very accurately (at least after 
the fact) but the wider area (like a few states across) are very bad at 
predicting the locations and motion of fronts where storms tend to 
occur. Part of that is the coarse grid, part is bad modeling, part is 
too much averaging of wind details that obscures the wind shifts at a 
front and too much dependence on linear interpolation between grid 
points. When there's a wind shift between grid points its improper to 
interpolate wind speeds between those points because wind doesn't do 
that in the real world. It holds up speed and direction (somtimes 
increasing) until the front is very close, then it varies wildly from up 
drafts and down drafts of the storms (if any) along the front and a few 
miles past the front its steady from nearly the opposite direction, 
usually towards the front. If I was in a working mood, I'd probable be 
working on better ways to use the available model data to improve the 
forecasts of fronts.

While the observed changes in refraction causes apparent signal arrival 
direction changes, I don't think the signal is directly blown by the 
wind, just the refractive components in the atmosphere aren't sitting 
still and aren't of constant size. Which points to your comment on using 
lower gain antennas for windy days and planning on working shorter paths 
for at least a couple reasons, lower wind load and then greater 
acceptance of varying direction of arrival. And since the atmosphere 
varies so fast on a windy day its not practical to predict the return 
path is the same. Sometimes I'm sure that two narrow antennas get some 
common scattering volume and we make the contact, even though neither 
antenna is aimed along the great circle path between stations. I used to 
notice this when an Indiana station was working Minneapolis on 432. He 
peaked up 50 or 60 degrees north of the direct path from that common 
scattering volume. Where the antenna headings were nearly 90 degrees 
from being on the same great circle path.

Since the randomness of short term atmosphere isn't yet practical to 
predict, the other alternative we've used for a long time is to blat 
away on the presumed path and if the signal comes up out of the noise in 
a ten minute period, make the contact quick. Likely sometimes we 
illuminate aircraft, flocks of birds, and masses of insects for that 
short term path enhancement, or we luck out with refractions and density 
variation scatterers making the connection. But with the winds of the 
past weekend, I think the atmosphere was too well mixed to get any 
refraction enhancement and the smaller scatters, bird flocks and clouds 
on insects were walking.
>
> Also noted, the higher the band, the greater the effect.

Likely because the atmospheric anomalies are more wavelengths across and 
the antenna patterns narrower.
>
> Perhaps we need to study making use of the pressure waves, like a
> surfer uses waves near the beach shore. Gain advantage.

But then we need to be able to find and identify those pressure waves, 
which don't show on 1 hour 40 km grids. Otherwise we just keep trying a 
path until it works or we give up.
>
>
> 73, Lloyd NE8I/r
> EN74 etc

73, Jerry, K0CQ