Roger concluded that his signal was in the noise most of the time.
He also concluded that he would need 2 W.
The analysis below shows that the conditions changed dramaticly after the 7 spots with 5 mW.
The header of the table shows the Calculated Lowest Possible Power in mW (Click for info),
which is calculated from the used power (Pwr Sent in mW) and the SNR.
The better the propagation, higher the SNR and
the lower the Calculated Lowest Possible Power will be.
9 UTC gives the spots that were made in the hour from 9:00 till 9:59 utc.
This first line shows, that the power that was needed for the spots was 1 W down to 2 mW
In this hour, half of the spots that were made with 2 W could be made with 100 mW.
The white circles indicate the power with which half of the spots could be made.
The lower the power needed, the better the propagation.
From 9 utc till 14 utc on the 18th there are 11 spots with a calculated lowest possible power of 5 mW or less. So it was possible to make spots with 5 mW.
WSPR with 5 mW
From 14 utc Roger reduces further to 5 mW and over 24 hours only 7 spots were made, shown in the yellow coloured part of the table. In the red part of the table there are no spots since the power that was used is 5 mW. In the red part I show the receiving stations and their SNR.
Back to 2 W
You would expect more spots, since the 24 hours before showed many more spots that could have been made with 5 mW. But the conditions changed dramaticly.
The last part of the table shows spots that were made with 2 W.
At 20 utc on the 20th at least 500 mW is needed.
And at 5 utc and 7 utc spots could have been made made 5mW