I didn’t think this session was as good as the previous, which was worse than the one before that in spite of a few bright spots. Even with the weaker session, Larry, W7IUV / WH2XGP, was reported by Roger, VK4YB, as that path shifts from the central US to the Pacific Northwest.
Geomagnetic activity was elevated as a forecast G1 storm begins to manifest due to a geoeffective coronal hole. The Kp is elevated, the Bz is pointing to the South and solar wind velocities exceed 400 km/s. DST values have seen a dip as well:
Terrestrial noise levels were high this morning during my daily CW sked. Setting the bidirectional receive loop broadside to the storms in the cenral US provided significant relief and made the QSO possible. The main limiting factor at these frequencies is seasonal noise and that is very evident by looking at the S/N levels reported over this session. Even common, relatively short paths struggled to reach CW levels.
Ken, SWL/EN61, in Indiana, sent a note that he found a supply of large-value capacitors at Walmart of all places. These are popular with the car audio crowd and in some cases might be useful to operators interested in these frequencies below the AM broadcast band. You can see some Walmart’s offerings here.
Ken, K5DNL / WG2XXM, reported decreased activity and very noisy conditions. He decoded two WSPR stations and was decoded by fourteen unique stations during the session including a single report from KV8P, a new receive station on 630-meters. Ken notes that he had 34 early morning decodes from WH2XCR with three decodes after sunrise in Oklahoma and best at -10 dB S/N.
Neil, W0YSE/7 / WG2XSV, reports that he decoded WG2XIQ, WG2XXM, WH2XCR and WH2XGP on WSPR2 and was decoded by N6RY, VA7JX and WH2XGP.
Regional and continental WSPR breakdowns follow:
There were no reports from the trans-Atlantic, trans-African, or trans-Equitorial paths during this session.
In the Caribbean, Eden, ZF1EJ, reported WG2XXM and WG2XIQ:
Laurence, KL7L / WE2XPQ, decoded WH2XCR and WH2XGP during the session:
Merv, K9FD/KH6 / WH2XCR, had an almost carbon-copy session of the previous (minus the ZF1EJ reports) with reports around the US and Australia:
Jim, W5EST, presents “PART 2: HOW INTERPRET SNR PEAKS AND VALLEYS ON MF/LF?”:
“Yesterday, I offered some alternative ideas, with no assurance of any of them, for physically interpreting SNR peaks and valleys on one-hop MF/LF paths. Before I go further, note a few more preliminary considerations:
SNR is used as a proxy for signal strength in my thinking, while signal strength itself is not reported by the WSPR decoder (at least as it’s currently programmed). Because of this, much of spring and summer seasons in parts of North America are foreclosed from SNR-based propagation interpretation precisely because of prevalent storm noise on MF/LF.
SNR helps characterize path capability for communication, which after all, is what we amateurs and experimenters primarily care about. Signal strength information is more pertinent to analyze propagation. Notwithstanding the name WSPR (Weak Signal Propagation Reporter), noise is mixed into the information (SNR) that the decoder yields. That said, we’re nevertheless delighted to have a tool as powerful as WSPR indeed is.
Moreover it’s almost certain that SNR sequences, and indeed even signal strength sequences if we had them, aren’t enough to totally make sense of “what’s up there”—ionosondes and satellites give further information. The influences of GMF (geomagnetic field) and polarization rotation are not generally measured for a single specific MF/LF path. Notwithstanding, our oblique-incidence SNR sequences on LF/MF indeed do provide important information.
Today, let’s think about a way to use our one-hop SNR information according to the hypothetical idea that relative to a given path the reflecting nighttime ionosphere has a rough average underside “surface density” rho (ρ) of pertinent contours of electron concentration that you can picture as more-or-less reflective puckers or dapples on an imaginary surface. http://www.abc.net.au/news/2010-12-17/childs-play-at-goma/2379720 (scroll 1/3 to pool photo), http://reganforrest.com/2011/04/ (scroll down 2/3 to pool photos). The distribution of these features may be either varying or still, and as a whole they may be moving en masse in bulk or in acoustic waves as a whole, with an average speed s. For pictorial background, see HamCom 2016* slide 23:
https://www.dropbox.com/s/4hsk6ltvi65z8en/PDF_HamCom_PPT_Presentation_Final_061116.pdf?dl=0 and the accompanying audio for the presentation: https://www.dropbox.com/s/96es4o2p1etn4d8/HamCom_JimW5ESTAudio_Presentation_050316_061116.mp3?dl=0 .
Likewise, variable refractions through arc-shaped paths of reflection in the ionospheric volume get reduced onto the imaginary surface. Further, multiple reflections resulting in varying amounts of multipath self-interference become re-imagined onto the surface as ionospheric dapples having an average surface density included in surface density ρ.
If these dapples are stable (statistically anyhow) and randomly distributed over a surface so the dapples are moving as a whole with average speed s, what then? Well, if we can define and count a number N of SNR peaks and valleys for a given TX station as received at your RX for a period of time T, then a ratio N/T = R is the average rate R of these SNR peaks and valleys in the SNR sequence during the night. We can measure this rate R, even if only imperfectly.
Picture a square piece of this imaginary ionospheric horizontal underside surface having square dimensions X kilometers on a side. Let one side be parallel to a direction of dapples’ motion as a whole across the midpath reflection and with N number of dapples along a side. Average surface density ρ of the dapples as a number per square kilometer is ρ = N2/X2. That’s because the dapples are passing along the lengths of one side of the square across a place of RF signal reflection. So the number of dapples in the square itself is approximately N2. The area of the square is the product X2 of the lengths X of its sides. That makes the average surface density of the dapples, how densely (coarsely or finely) they are distributed over the surface, come out:
ρ = N2/X2 = (N/X)2.
Now, without upsetting anything, let’s divide by time T twice as shown:
ρ = (N/X)2 = [(N/T)/(X/T)]2.
N/T is the rate R in events per hour of the peaks and valleys we can measure overnight on an MF/LF band for a given pair of stations. X/T is the average speed s of the bulk motion. The average surface density ρ of the dapples then becomes the square of the ratio of the rate R divided by the speed s of the ionospheric dapples at the path midpoint relative to the Earth’s surface beneath it.
ρ = (R/s)2.
However, we hams and experimenters do not know what the speed s of the ionospheric dapples actually is at a path midpoint. (Nor do we know the direction of net motion, but that doesn’t matter for now). All we know is the rate R of peaks and valleys of SNR, which is the product of multiplying speed s times the square root of the spatial density of the dapples:
R = s √ρ.
Next, let’s consider the possibility that speed s is zero—the ionosphere is motionless. Then a bulk rate of change of reflectivity, call it r, of midpath reflectivity or absorption in events per hour comes into play. Further I write:
R = s √ρ + r.
Due to uncorrelated spatially random dapples and their overall motion versus bulk variation in reflectivity of the medium, the formula might instead be written something like
R = sqrt[s2ρ + r2] .
Nevertheless, the flow of discussion is the same from here on, so the simpler formula is used.
Now think about different MF/LF bands, like 630m and 2200m. Take a ratio of band-specific rates of peaks and valleys through the nighttime.
R630 / R2200 = (s √ρ630 + r) / (s √ρ2200 + r) = (√ρ630 + r/s ) / (√ρ2200 + r/s )
Units of r/s: (events/hr) / (km/hr) = events/km. √ρ has units: events/km.
Since the ionosphere is a physical entity, any bulk or acoustic wave motion s as a whole that it displays would be the same regardless of the band. Also, I presume any bulk values r of midpath reflectivity or absorption would vary in a correlated way for both bands even if their particular values may not be identical. Whatever difference in overall rates of peaks and valleys of SNR exists between the bands would probably result from the different average surface densities of the dapples of reflectivity, as RF signals on the two bands “see” them.
Because the wavelengths of 2200m and 630m are about a 3:1 ratio, they can at best resolve dapples in a 3:1 ratio of diameter (square root of ρ). That way, the rate R of SNR peak and valley events is greater on 630m because the shorter wavelength at 630m sees more dapples. If the quantity r/s is small compared to either square root, then this band-dependent effect should be plainly observable, and in my experience it is.
Summarizing: The dynamics of one-hop signal strength peaks and valleys that we observe with WSPR uses SNR as a proxy measure on storm-free nights. Even by reductively thinking in terms of an imaginary ionospheric surface, such a simplified semi-physical interpretation of the SNR peaks and valleys appears to resolve in a somewhat complex way into speed of ionospheric dapple motion, the square root of the surface density of ionospheric dapples of greater and less reflectivity, and the bulk rate of change of reflectivity of the ionospheric plasma. We can band-specifically compare the rate of SNR peaks and valleys on MF and LF through a night, thanks to both 630m and 2200m experimentation. With the tools we amateurs and experimenters are presently using on MF/LF, and possible improvements and additional tools we may acquire, our physical interpretation of SNR can get even further refined.
What could you tell us on this topic from the literature and from your own experience and good common sense? I look forward to any e-mails you’d like to send!
*See all the 630m Ham-Com 2016 presentations on TX, RX, and antennas using the links at:
Additions, corrections, clarifications, etc? Send me a message on the Contact page or directly to KB5NJD gmail dot (com)!