For stations at lower latitudes, it would be tough to really complain about the band last night in spite of a few quirks. QSB was active so stations reporting positive S/N numbers one minute might be reporting near detection limit levels only minutes later. Of course this behavior was exacerbated at higher latitudes. Fortunately a large number of reports suggest stable enough band conditions for QSO’s in a variety of modes during the evening and overnight. The variability keeps things interesting.
Geomagnetic activity was quiet and the Bz was stable through the session with solar wind velocities in the low category, less than 325 km/s.
Hideo, JH3XCU, provided the following information about a new MF and LF award sponsored by the JARL:
Neil, W0YSE/7 / WG2XSV, reports that he decoded Doug, K4LY / WH2XZO, on WSPR2 for only the second time in 2016 and provided the following details:
Doug, K4LY / WH2XZO, points out that the last few sessions may have been the seasons best in spite of recent antenna improvements:
John, WA3ETD / WG2XKA, noted some of the differences from the previous session:
Larry, WD0AKX, in Minnesota, created the following video on his You Tube channel about WSPR and specifically about WSPR on 630-meters. Larry was also kind enough to plug this blog near the end of the video, which I am eternally grateful. Please support Larry’s channel by subscribing. He is introducing a completely different set of operators to what we do on 630-meters.
Regional and continental WSPR breakdowns follow:
There were no trans-Atlantic or trans-African reports during this session. UA0SNV was present but no reports were found in the WSPRnet database.
The Canary Islands are busy with two vacationing stations from continental Europe providing reports during this session.
In the Caribbean, Eden, ZF1EJ, and Roger, ZF1RC, provided reports for most of the stations QRV across North America. Eden also reported WH2XCR on KH6.
In Alaska, Laurence, KL7L / WE2XPQ, continues his receive-only operations for the moment, reporting as far south as WH2XXP and WH2XCR in addition to the “regulars” in the Pacific Northwest.
In the Pacific, Merv, K9FD/KH6 / WH2XCR, commented on this season as well as the coming seasonal transitions in North America. Merv continues to report and receive reports from stations in Australia on a nightly basis. The Japanese path remains closed off at the moment.
In Australia, Phil, VK3ELV, and Roger, VK4YB, receive reports from Japan and Hawaii:
Additional statistics, anecdotes, comments and information:
Larry, W7IUV / WH2XGP, reports that he decoded ten WSPR stations and was decoded by 37. He also notes two-way reports with WH2XCR before sunset in KH6 with reports for a while after sunrise in Washington. WH2XXP continued to be decoded 20 minutes after sunrise.
Ken, K5DNL / WG2XXM, reports that he decoded eleven WSPR stations and was decoded by 36.
Jim, W5EST, presents Part 2 in his discussion on daytime propagation on 630-meters:
“PART 2: 630M DAYTIME PROPAGATION & SOLAR FLARES
Short Summary: Forget flares. During the period Nov. 1-Jan. 15, predict 630m daytime reception every day.
Recall from the Feb. 25 and 27 blogs that 44 daytime 630m sky wave receptions significantly clustered by dates and accounted for 36% of the 123 days Oct. 25-Feb. 25 (four months). TABLE I shows these clusters.
TABLE I: DATE INTERVALS OF 630M DAYTIME RECEPTION, & DESCRIPTION
Oct. 25. Singleton. October had 1 day of daytime prop we know of.
Nov. 2-6, 9-15, 18-20, 22-25 4 Clusters occupy 19 days of Nov. November had 20 days.
Nov. 30, Dec. 2, 8, 10-12, 17. Scattered, 1 Cluster. December had 9 days of daytime prop.
Dec. 29–Jan.3, Jan 5-12. 2 Clusters January had 13 days of daytime prop.
Jan. 20 & 30. Feb. 6. Scattered. February has had 1 day of daytime prop thru 2/25.
Today, I summarize the solar flare data, which involves solar X-ray intensity ranges or classes that are designated C, M, X. Solar X-rays are the very high energy part of the sun’s electromagnetic radiation that has a far shorter wavelength than ultraviolet light. Less than C-class X-radiation is not regarded as a “flare” and yet does have some intensity. I evaluate possible predictors of 630m daytime propagation based on solar X-ray levels.
Out of the 4 month period with its 123 days, solar flares of at least C-class occurred 85 days. Flares at least C5 level occurred 31 days, and stronger flares at least M1 or higher happened 13 days. X-flares are the strongest flares, and none occurred in this period.
In TABLE 2, a plus sign as in C5+ means that if a solar flare were C5 or stronger than C5 a given day, then the C5+ predictor predicts 630m daytime propagation would happen that day. A minus sign, as in C5-, means if a flare were weaker than C5 then it predicts 630m daytime propagation would happen that day. I studied predictors for C1+, C5+, M1+, C1-, C5-, M1-.
The solar flare predictors C1+, C5+ and M1+ surprised me. All were poor performers. For instance, the C5+ predictor showed only 35% Precision and sky-high 75% False Negatives. Regarding M-flares, only Nov. 4 & 9 and Jan.1 yielded 630m daytime prop with an M-Flare while 10 other M-Flare days had no 630m daytime prop.
If anything, 630m daytime prop mostly avoided strong solar flares. Even so, the solar flare predictor results for both reverse predictors C1- and C5- were mediocre, and the M1- reverse predictor gave bad precision.
I also combined low solar flare level with calendar-based prediction to make a mixed predictor named “Nov1Jan15&C5-”. This mixed predictor meets the quality criteria of False Negatives <25% and Precision >50%, but still does not outperform the simpler calendar-based predictor “Nov-Jan15”. (The jargon of False Negatives and Precision was explained in the Feb. 25 blog.)
Due to the clustering of days for 630m daytime prop, these poor results did not arise as if perhaps from a simple offset error of daytime prop days from flares that might have caused them. Because so many daytime prop days occurred on low-flare or non-flare days, I did not make a distinction between nighttime and daytime flares, nor did I look at advance timing of flares ahead of times of daytime prop.
I studied thunderstorm (Tstorm) data to understand the results better. Few Tstorm masses affected many scattered receiving sites at once. Also, daytime reception is resilient to merely regional Tstorms–daytime 630m SNRs not infrequently rose 10dB or more above WSPR’s decode threshold.
That said, Dec. 13 and Dec. 23-28 were dense with Tstorms in Texas and elsewhere that help explain the gap in 630m daytime receptions then. Jan. 21, 27 and Feb. 2 and 24 are four other days that had Texas Tstorms when no 630m daytime receptions occurred. If anything, 630m daytime prop is probably more frequent than daytime receptions tell us in any one year. Summing it all up:
The “Nov-Jan15” pure calendar-based predictor outperformed all the C1, C5, and M1 predictors regardless of the above/below sense of their threshold (e.g. C5+/C5-).
More daytime TX and RX station activity by more stations can reveal more days of 630m daytime prop especially in the winter season. Indeed, the number of Tstorms and the days they will affect will differ considerably from year to year. A year ago in late winter 2015, for instance, 630m daytime prop arose Jan 31, Feb. 24, 26, 27, and March 17. Also, I have virtually no data about spring and summer seasons of 630m daytime sky wave receptions nor about such propagation in full sun.
Is 630m daytime prop really only due to low angle sun elevation in temperate latitudes up to the aurora zone and within plus/minus several weeks of winter solstice? If you know something to add on this topic, e-mail us your experiences and educate all of us! Thanks & GL.
TABLE 2: PREDICTORS OF 630m DAYTIME PROP DURING 4 MONTHS 10/25/15 – 2/25/16
Predictor False Negatives Rate Precision Remarks
C1+ 43% = 19/44 29%=25/(25+60) Bad False Negs & Precision
C5+ 75% = 33/44 35%=11/(11+20) Bad False Negs & Precision
M1+ 93% = 41/44 23% = 3/( 3+10) Bad False Negs & Precision
Reverse Flare-based Predictors
C1- 57% = 25/44 50%=19/(19+19) Bad False Negs.
C5- 27% = 12/44 35%=32/(32+60) Bad Precision.
M1- 7% = 3/44 37%=41/(41+69) Bad Precision.
Calendar-based & Mixed Predictors
“Solstice+/-6wks” 16% = 7/44 44% = 37/84 Date symmetry. Under 50% Precision.
“Nov-Jan15” 9% = 4/44 53% = 40/76 Best calendar predictor w/ this data.
“Nov1Jan15&C5-” 34%= 15/44 52% = 29/(76-22)*Bad False Negs w/Mixed Predct
“Nov1Jan15&M1-” 16%= 7/44 54% = 37/(76-8)*Good Mixed Predictor
*NOTE: 22 days C5+, 8 days M1+, in Nov1–Jan15.”
Additions, corrections, clarifications, etc? Send me a message on the Contact page or directly to KB5NJD <at> gmail dot (com)!