A shot summary of what Cumulus does

[WestTx]

I've been trying to understand more about the Cumulus software works and would like some confirmation that what I think I've learned is correct. I think I've read the Wiki and most of the forum pretty thoroughly. I've also looked at the diagnostic files to see what Cumulus is doing.

Station information:
I'm using a Davis VUE and Cumulus ver. 1.9.2 build 1032. My meteorological day ends at midnight and I generally exit Cumulus every evening and restart it the next morning. (If I'm away from home, Cumulus might not be run for a week or so.) Both my console data interval and Cumulus log interval are set for 10 minutes.

Summary:
- When Cumulus runs, it reads the Cumulus.ini file and the current data files written previously.
- If there has been a date change since Cumulus was last running, the appropriate log files are updated based on the user's selected meteorological day setting.
- A backup is created.
- It sets the console date/time using the desktop computer date/time? (I have selected Synchronize the clock to the PC in the configuration)
- It reads all data in the logger from the date/time Cumulus was last stopped up to the current time and makes the necessary calculations for temp trend, temp change-in-past-hour, and pressure trend.
- It reads the YTD raintotal from the weather console?**
- As the data stream is read from the weather console, Cumulus continuously calculates highs/lows/records, etc. and updates displays and files immediately.

** Do any other values come only from the weather console?

Thanks,
Les

[steve]

- When Cumulus runs, it reads the Cumulus.ini file and the current data files written previously.

Yes

- If there has been a date change since Cumulus was last running, the appropriate log files are updated based on the user's selected meteorological day setting.

Yes; it does the end of day 'rollover' when it gets to processing the appropriate entry from the data downloaded from the logger.

- A backup is created.

Yes, but note that this isn't intended as a backup in the 'data security' sense; it exists solely to make it easier to 'rewind' to a previous point in time (in case of missing or incorrect data for whatever reason)

- It sets the console date/time using the desktop computer date/time? (I have selected Synchronize the clock to the PC in the configuration)

Yes, for Davis stations.

- It reads all data in the logger from the date/time Cumulus was last stopped up to the current time and makes the necessary calculations for temp trend, temp change-in-past-hour, and pressure trend.

Yes

- It reads the YTD raintotal from the weather console?

Yes, but only in the same way that it reads all of the other current readings. The figure is only used as an increasing counter from which to calculate all the other rainfall figures. It isn't actually used as the annual rainfall figure.

- As the data stream is read from the weather console, Cumulus continuously calculates highs/lows/records, etc. and updates displays and files immediately.

Yes.

[WestTx]

- It reads the YTD raintotal from the weather console?

Yes, but only in the same way that it reads all of the other current readings. The figure is only used as an increasing counter from which to calculate all the other rainfall figures. It isn't actually used as the annual rainfall figure.

I don't understand what you mean by this. Could you elaborate?

Thanks,
Les
(and if anyone finds my missing "r", please return it to the title of the original post please. )

[sfws]

Yes, but only in the same way that it reads all of the other current readings. The figure is only used as an increasing counter from which to calculate all the other rainfall figures. It isn't actually used as the annual rainfall figure.

I thought what Steve said was clear, but perhaps if I spell it out it will help. (It did take me a while to understand all this when I was actively using Cumulus a while ago and in my original positioning of my rain gauge the wind affected many of my rain measurements and I had to correct them, but you can search this forum for the word rainfall and the author steve to find lots of helpful comments).

• If you are reading a counter, you can calculate the difference between what it says now, and what it said before to get the rainfall in the period between readings in count units.
• Cumulus can then turn counts (which it logs) into your required rain (in your selected units) by subtracting the more recent count from an earlier count for a period like last hour to be displayed (or any other period displayed on the main Cumulus screen).
• Now that you can log the total rainfall (in your units) for a day, you can add those logged rainfalls to get monthly and annual totals. see http://wiki.sandaysoft.com/a/FAQ#Where_ ... ls_from.3F

This means if you have an error in today's rain, you can use the option on the edit menu to edit the rain today, and after the next rollover that edit is carried forward into monthly and annual totals. But if you have an error in the past, you will have to manually edit the counts in the logs (with Cumulus stopped) and really understand the counts. See http://wiki.sandaysoft.com/a/FAQ#My_sta ... _figure.29

[steve]

And if you find it strange that Cumulus does it this way rather than reading the rain figures from the station directly, there are two reasons for this:

1. Cumulus provides an 0900/1000 start of day option; no stations provide this option, so Cumulus has to work out the figures itself. It needs an incrementing counter to do this, and most stations just provide the raw 'tips since last reset' figure from the rain gauge. Davis don't do this, so I use the annual rainfall figure instead, and cope with it resetting on Jan 1st (or whenever).
2. Not all stations provide daily (etc) totals anyway, particularly in their logger data, and as much code as possible has to be common to all station types.

[WestTx]

Thanks to both of you for trying to help me understand.

It makes sense to me that Cumulus calculates rain totals from incremental rain rather than reading the total direct from the console. And that is why it seemed to make no sense to even read the YTD rain from the console. If the totals are calculated from incremental rain captured by the logger, then read individually by Cumulus, there seems to be no need at all for the YTD total from the console.

The wiki links provided by sfws explain very well but neither mentions reading the YTD rain from the console and using that data.

Thanks,
Les

[steve]

If the totals are calculated from incremental rain captured by the logger, then read individually by Cumulus, there seems to be no need at all for the YTD total from the console.

The YTD total from the console (this year's rain total so far) is the incremental counter which Cumulus uses to calculate all of the rainfall data. If it didn't use that, what else could it use?

[WestTx]

If the totals are calculated from incremental rain captured by the logger, then read individually by Cumulus, there seems to be no need at all for the YTD total from the console.

The YTD total from the console (this year's rain total so far) is the incremental counter which Cumulus uses to calculate all of the rainfall data. If it didn't use that, what else could it use?

It's my understanding that when Cumulus executes it reads data from the datalogger from the point it last read before Cumulus was stopped (the night before, in my case). That data is incremental snapshots every ten minutes showing temp, pressure, rain, etc. If it rained between 1:00 a.m. and 2:00 a.m., the 10-minute snapshots would show that rain as it occurred. Maybe .10 between 1:20 a.m. and 1:30 a.m. and another .20 between 1:30 a.m. and 1:40 a.m. If there was no more rain during the night then Cumulus would know there had been .30 inches of rain from that. Cumulus has already logged 15.14 inches of rain prior for the year and would just add the .30 inches for a YTD total of 15.44 inches.

Is that not right? I apologize if I seem obtuse... I'm really not trying to be.

Les

[steve]

Sorry, yes, if we're talking about catching up using logger data. The logger data doesn't hold the annual rainfall counter, so Cumulus has to update the total itself by adding the rainfall amount in each logger entry to the counter that it stored when it closed down, and then using the calculated total to determine the rain figures, in the usual way. Once it's finished processing the logger data and is reading live data again, it uses the current annual total, periodically read from the station, to determine how much rain has fallen.

[WestTx]

Thanks. I appreciate your help.

Les