[600MRG] TA Propagation Metrics

[James Hollander]

       Improving propagation increases the  number  of SNRs above the  threshold which translates to a  dramaticincrease  in the  number o f WSPR spots--timeslots that get filled  in  and decoded.  [Even so, most transatlantic (TA) SNRs are  within 15dB of the WSPR decoder threshold.] The  number  of spots dramatically increases because  it  is also  strongly affected by the  number of  TX and RX stations active  during the  TA opportunity time window and  by the WSPR transmit percentage TxPct that  each TX operator chooses. But that's not all--number of WSPR spots can actually make  a  TA night like Oct. 2 look  average  or  even poor when propagation was actually quite good.
     I'm studying two propagation metrics for 630m TA.  I'm leaning toward a  metric based on the  median number of spots from the  various TX/RX pairs of stations on a given TA night. 
     Any comments or suggestions are welcome--contact me direct at mrsocion at aol.com.
73, Jim H  W5EST
 
2-WEEKCOMPARISON OF TA 10LOG#SPOTS AND 3xMEDIAN#SPOTS METRICS
Metric        Dates (2015)
               9-24  9-25 9-26  9-27  9-28 9-29   9-30  10-1 10-2  10-3  10-4 10-5  10-6  10-7
10Log      3        14    0     10      11      0       10      10      13     7       11      20    11      17
3xMedn  6          9    0      11       8      0         9        6      26     6       12      18      9      15
 
    Tripling the  median#spots makes the  10Log method and 3xMedian method easier to  compare. 
Notice the 10Log method makes 10-5 look much better  than 10-2.  But  I  think that 10-2 was  a  better night for propagation than  the  10Log method shows. On 10-2 VE1HF & VE1VDM RX off,  G8HUH TXwas only  qrv  0402-0440z. DJ0ABR was only  qrv in 4 timeslots  all  evening 10-2, DF6NM 3x/hr 10-2 but 5x/hr 10-5. 
    Conclusion: The 3xMedian method  is  much more resilient to variations  in number  of  stations and TxPct.  With an average of  eight spots per  receiving station and given two  more  receivingstations and  one more  transmitting station than  the  2 TX & 2 RX on 10-2, the  10Logmethod would have  scored much higher 20 =10 log10 (8x3x4 =96). Both3xMedian method and 10Log method are noisy with fewer stationsactive, but  I  think 3xMedian method does  a  better job  of  reflecting what was  actually goingon 10-2 compared to 10-5.  On  all the  other nights, the  two methods give remarkably similar results.  Both methods look back  over  each night, and results for the  different nights vary widely. We can hope  that  strong TA propagation nights  actually become more frequent as they appear to sporadically be doing. --end--