This session was much improved over the previous two which is a little surprising given the level of geomagnetic activity that has occurred over the last 36 hours. WSPR decode numbers were two to three times better at my station during this session which may strictly be a function of increased activity. A review of my S/N numbers, however, suggests that propagation, noise level, or both were improved.
Geomagnetic activity continues to be at unsettled to storm levels. The Bz has been variable and solar wind currently exceeds 600 km/s.
Edgar Twining of Moonah, Tasmania submitted the following CW capture of Phil, VK3ELV, on the VK 600m reflector:
Regional and continental WSPR breakdowns follow:
There were no reports from the trans-Atlantic, trans-African, or trans-Equitorial paths. UA0SNV was present during the session but no reports have been filed at this time.
In the Caribbean, Eden, ZF1EJ, reported my station during this session:
Laurence, KL7L / WE2XPQ, decoded WH2XCR and WH2XGP during this session:
Merv, K9FD/KH6 / WH2XCR, had another pretty good night in spite of antenna damage:
Jim, W5EST, presents, “PART 4: HOW TO DATA-INTERPRET SNR FOR PEAKS AND VALLEYS ON MF/LF?”:
“The last three blog-days, as you recall, have been considering a few main things:
1) Different possible features of the reflective ionosphere relevant to 630m/2200m one-hop paths in or near the lower-48 USA.
2) Whether interpretation of SNR slot-by-slot dynamics can meaningfully go beyond merely computing an all-night statistical spread of SNRs.
3) What governs the spatial resolution of SNR measurements, in view of the relatively long MF/LF wavelengths and the 2-minute slots and transmit percentages TxPct used for WSPR2.
So far, my thinking suggests that the effects of ionospheric waves and/or drift en masse probably make a considerable contribution to 630m/2200m SNR dynamics from slot to slot across each nighttime hour. Regarding resolution of our 630m/2200m SNR measurements, they appear to be limited more by the WSPR2 two-minute slots and TxPct setting than by the wavelengths of these bands themselves.
Today, let’s look at some actual data at 630m and 2200m for essentially one same AZ-WA path, XXP-xgp and XND-w7iuv, taken within two nights of each other early last December, 2015. One can rightly object to the use of single band-night data on two different nights. Also, unascertained weather might have featured some light rain in the WA region. It would be far preferable to have same-night comparisons on many different nights. But that’s life as we know it.
The example SNR data that I graphically present here was copied from the WSPR database at the time. At least you can consider some steps to analyze and critique SNR information using the example data for these two bands and hope for a better trove of data someday.
For this blog post, the band-specific data streams are transcribed to a spreadsheet. The illustration scatterplots the spreadsheet SNRs versus progress of time Zulu from right to left. The 630m and 2200m SNR scatterplots have the same SNR scale and are aligned to the same times Zulu. Peak WSPR SNRs were favorable– around +0dB– and the SNR variations were generally more rapid than I would expect from a slow moving rainy weather pattern across WA.
The overall overnightstatistical variabilities represented by 630m & 2200m SIQs (SNR interquartile differences) were 12dB on 630m, and 6dB on 2200m. SIQ was obtained by sorting the SNRs on the spreadsheet and subtracting the 25th percentile SNRs (-18 &-8 dB respectively) from the 75th percentile SNRs (-6 & -2 dB respectively). The 12dB 630m SIQ is considerably more than the 9dB 630m SIQ I recorded for XXP-xgp a month earlier on Nov. 6, 2015. So the weather on Dec. 6 may indeed have been problematic for this work.
On an hourly basis, the 191 WSPR 630m SNRs spread out up and down over each hour much more than the 167 SNRs for 2200m did. That’s remarkable considering that XXP operated 630m 50% TxPct, while XND operated 2200m at about 25% TxPct. As the scatterplots show, the 630m SNRs were far more scattered and did vary far more over intervals of even a few minutes than the 2200m SNRs.
I suggest a definition of SNR valley or peak as that where the SNRs appeared to have steadily decreased and then steadily increased (or vice-versa) 6dB or more in an hour or less.
2200m numerical data on the spreadsheet and the 2200m scatterplot suggest about five (5) 2200m SNR nighttime dips (i.e., valleys, see green arrows). Further, these dips punctuated a general all-night increasing trend from negative single digit SNRs in the evening on 2200m to low positive single digit SNRs a couple of hours before sunrise (SR). If not the weather, this SNR trend may indicate that the reflectivity of the ionosphere was gradually increasing.
On 630m, the corresponding nighttime regime for 630m displays as much as 15dB scatter in an hour or less, but with very little indication of steady decline followed by steady increase to define a valley, except perhaps at 1030z. It’s as if the 630m SNRs randomly sampled a rapidly varying ionospheric terrain or randomly sampled a rapidly varying mixture of signal strength and weather noise.
Over the entire night, the 630m data trend is roughly hump-shaped with SNR nearing its -1 dB high sometimes in the 0430z-08z interval. This trend may have been weather-noise related since it’s not clear to me why the 630m reflectivity of the ionosphere would decrease after increasing over hours.
In the pre-sunset (pre-SS) regime, the 2200m SNRs remarkably executed a serpentine rise-fall-rise at far right, while the 630m decodes only began at SS.
At far left, the pre-sunrise (pre-SR) regime on both bands displayed a 15 to 20dB SNR decline in the hour prior to SR. The post-sunrise (post-SR) regimes showed ~6dB scatter for both bands, depressed 10dB below the pre-SR levels.
Summarizing, I’d say that if the 630m nighttime SNRs were not corrupted by weather noise, their dynamics fails to define more than one SNR valley (1030z, Dec. 6) at the resolution of 50% TxPct WSPR2. By contrast, 2200m SNRs (Dec. 4)were more nearly deliberate in their variation and displayed five dips or valleys. If ionospheric terrain was their cause, then its terrain variations were within the limit of resolution by 25% TxPct WSPR2.
Do you have better interpretations or better data you’ve found to compare the 630m and 2200m bands? We look forward to your contributions now and in the future! GL.”
Additions, corrections, clarifications, etc? Send me a message on the Contact page or directly to KB5NJD gmail dot (com)!