Cumulus Support

Yes, delete them. They look to have come from corrupt logger entries. See the FAQ for what items you may need to correct:

http://wiki.sandaysoft.com/a/FAQ#My_sta ... _figure.29
http://wiki.sandaysoft.com/a/FAQ#My_mon ... _are_wrong

I'm assuming you restarted Cumulus at around that time, so those corrupt entries came from the station logger rather than from Cumulus running 'live'? It's hard to see how Cumulus could log entries with the same timestamp like that, unless your PC clock went haywire. Do you have the correct station type selected in Cumulus - i.e. with or without solar sensors as appropriate?

Doubt that I was even up at that time of the morning. It is possible that there could have been a power cut that would have caused the system to reboot when the power came back on. However, my UPS has been working well in that regard and so I doubt that. My memory is poor, I will check some other files to see if there was anything happening around that time.

The diags files will show what was going on at the time; if you zip up the diags folder and attach it, I'll take a look.

Well, looks as if it did reboot sometime between 07:53 and 07:57 as Cumulus splits diag files then. I have attached the second one. There appear to be some comments re. abnormally high rainfall being ignored but, presumably, it didn't. It was doing a lot of WU updating at that time, presumably catchup from the system being down.

There was some kind of problem with the station or the USB interface when it restarted - possibly related to whatever caused the restart - such that the data was offset by one byte, and was thus interpreted incorrectly. It did manage to ignore a lot of that garbage data, but it's hard to deal perfectly with what was effectively random numbers. As it had restarted almost immediately, it didn't actually need most of the data from the logger anyway. It's possible that the latest version would have coped slightly better than the old version you're using; I'm constantly doing tweaks to avoid problems caused by the hardware's bad behaviour.

Restarting from the backup files at the time would probably have automatically fixed the data, but three days later it's probably not worth doing that now.