Cumulus Support

Time is useful. Normally the measurements are made at a specific time of day - this varies from nation to nation (in the UK it is 09:00), and enables comparison across different dates whilst minimising the difference in solar heating that would be present if the reading was performed at random times. It also enables the documentation of readings that are performed when the observer isn't able to measure at the standard time, consequently allowing for subjective assessment of what that reading could've been if performed at the correct time.

Allowance for DST will need to be done by the observer rather than the software. It is normal in the meteorological world to disregard DST. If DST is observed (in terms of the time recorded against a reading) then that adjustment is made by the observer.

@Mark: I agree about time but apparently it is feature in CMX.
If Paul agrees it can benleft out.

NB: I won't be available for a week so no hurry

Beautiful fall weather and colour here, but we know what is eventually coming and I have my snow blower ready to go.

Here is the link for the NOAA snow reporting provided by KFOKFL-13 the initial poster in this thread https://madis-data.ncep.noaa.gov/sample_snow_form.html
This is similar to the CoCoRaHS input other than it also has a "Last 6-hours" reporting.

Mark wrote:
I have converted the database and diary editor, but have a question about time.

I do not really see the value in adding the time to the entries, these are daily records, so unless you want the ability to add multiple records per day, I don't really see the value in adding the time?

Regarding the "time" question, again from my own practice and experience for CoCoRaHS reporting and the CU Diary entry I use "as at 8:00 am" as that is the most convenient and regular time to take and record the measurements, even though my CU automated records are at a midnight rollover. I do keep a manual record of CU precipitation since 8:00 am to midnight, plus midnight to 8:00 am, so I can compare to the manual reading for the same time period (CU1 console is great for giving the "Last hour" "Today" "Last 24 hours" and "Yesterday" rain values).

While I pride myself in trying to keep "good" records, the snow fall and accumulation is less precise compared to rain and the actual time of that is not of a great benefit. So I do not see a need to log the actual times just as long as the record has the "24-hour ending" date and somewhere an indication of the 24-hour period as set by the user.

I presume any charts that could be sourced from the records would be on a daily basis like Sunshine and Daily rain.

Enjoy,
Paul

Beautiful picture Paul.

Yes, the graphs in the dashboard - not done the default web site yet - are just daily plots like other daily values.

One quirk I have noticed (I'm using your diary db for testing) is that if you do not record a day with zero depth before or after days with snow depth the graph line ramps from the previous year's last value (usually a zero) up to the first value of the following year. I'll look at making the plots non-continuous if it's possible to do that. You can of course record a zero-depth day retrospectively.

Mark wrote:
One quirk I have noticed (I'm using your diary db for testing) is that if you do not record a day with zero depth before or after days with snow depth the graph line ramps from the previous year's last value (usually a zero) up to the first value of the following year. I'll look at making the plots non-continuous if it's possible to do that. You can of course record a zero-depth day retrospectively.

Yes, checking the Weather Diary I had been inconsistent, for some months only entering if there is actual snow data to record, and for other months an entry everyday even if no new snow or snow lying (entering 0 in Snow Depth box). Looks like I have a task to do that "record a zero-depth day retrospectively" for every day for the months of October through April. I have manual records since 2011 and have entered in CU1 Diary to current, but in CMX only since 2019 when Mark revived CMX. We've not had a snow day in the months of May through September since I started manually recording in 2011.

Enjoy,
Paul

I don't think you need to record every zero day, it only really affects the graphs and I have that sorted now. Is it safe to assume days without a record (as over the summer) have no snow to record?

Plus you may have a bit of a task coming with the new diary format to fill in all the 24h snow values

Hi Mark:

Is it safe to assume days without a record (as over the summer) have no snow to record?

Yes.

Plus you may have a bit of a task coming with the new diary format to fill in all the 24h snow values

What, no CreateMissing or Migrate tool?
So I just opened the Diary in DB Browser, highlight all the data and Copy, Paste 789 lines in Notepad++, now what?
Just kidding, I'll wait for the new diary update

Enjoy,
Paul

Another solution to the graph oddity would be to use a bar chart rather than a line. I'll see what it looks like, I was using a line for the depth and bars for the 24h values, using two bar series may look a little bit more confusing

Could you use a stacked bar with depth and fresh depth different colours?

That's an idea - but then you'd have to subtract the 24h snow figure from the depth figure for every day to retain the correct total.

I've played with overlaid bars like this...

mcrossley wrote: ↑Sat 26 Oct 2024 2:38 pm That's an idea - but then you'd have to subtract the 24h snow figure from the depth figure for every day to retain the correct total.

And if you just showed the depth (hid the 24h by clicking it then the totals would be wrong, hmm...)

I use two bars for my rainfall graphs (two rain gauges) and that looks okay.

What about doing SnowDepth as a line and SnowFall as bars and scale both identical

Using a line is the problem, if you do not have zero values logged in the days before and after snowfall is registered, then the line joins adjacent data points with a slop - that can be from the end of one snow season to the start of the next. If you use the same scale for both depth and 24h, then the small 24h values can be very hard to see against a scale that goes up to a metre or more for some people.