Cumulus Support

Now that I have my solar sensor installed, I've noticed a few issues in Cumulus regarding reading the solar data from the logger. When I first added solar data to Cumulus, it just displayed the values and didn't do anything else with them. So I didn't put in any code to read the values from the logger at start up - there would be no point as there would be nowhere for the data to go.

But now that the solar and UV values are going into the log files, it needs to read those values from the logger so that it can log them correctly. This is no problem for solar radiation and UV, I'll just add the code. But Evapotranspiration is a bit odd. Davis have decided that the value that is supplied 'live' (in the LOOP packet, for those of who know about these things) is the total ET for 'today' so far. This introduces the first problem: the Davis 'day' starts at midnight. Some of us have meteorological days that start at other times. To do it correctly, I am going to have to add code that keeps track of the totals and allows for a 'day' that doesn't start at midnight - assuming that it matters that ET total follows the meteorological day at all?

The second problem that Davis have created for me is that they have decided that in the archive ('logger') data, the ET value is not the daily total so far, instead it is the total for the previous hour. Furthermore, only logger entries on the hour have values; the others are zero. So I am going to have to try to maintain a running total with the archive data.

A possible solution to much of the above which may make things easier for me would be to define the Cumulus ET value to always be the ET for the previous hour. One tricky bit then is handling the period between midnight and 0100, when the total will have been reset. It's only tricky, though, not impossible. The other tricky bit is in the first hour after switching over from archive data to live data, and working out how much ET there has been in the previous hour.

I'm not really sure what people actually do with ET and how important any of this is.

Questions? Comments? Suggestions?

Steve...

I'm not really sure what people actually do with ET and how important any of this is.

As you indicated, most people know nothing about ET (evapotranspiration) - from what I have read thought, ET is widely used in the vineyards where it is important for the growers to know and compare the amount of evaporated water to the amount on rainfall (I make that information available one of my Tzouhalem-Maple Bay InstaWeather page under the heading 'Irrigation Index/ET' - there are about a dozen vineyards in my area).

The software I am using seems to provide, not only the daily, yesterday, weekly and monthly ET, but also current evapotranspiration rate calculated every minute - I do not know how it is done.

Davis have decided that the value that is supplied 'live' is the total ET for 'today' so far.

I think (not sure here) that the ET provided by a Davis station is a cumulative value (updated once an hour, on the hour), meaning that the ET value provided at 15:00 is the total amount of evaporation that has happened since midnight.

...assuming that it matters that ET total follows the meteorological day at all?

As indicated in the manual, the Vantage console uses the air temperature, the relative humidity, the average wind speed and the solar radiation to estimate the ET - if you use the data following the 09:00 start of the meteorological day then the information provided by Cumulus will be different than the information provided by the console as there will be up to 4 hours missing - the hours from sunrise to 09:00. If you decide to follow the 09:00 meteorological day (which appears to be used only in the UK) you will then have to increase the number of calculation required to produce a daily ET.

Re-reading the above does not seem to provide much useful information...

gemini06720 wrote:The software I am using seems to provide, not only the daily, yesterday, weekly and monthly ET, but also current evapotranspiration rate calculated every minute - I do not know how it is done.

Ah, yes, that's the solution. I can ignore the Davis ET value and get Cumulus to calculate it from the raw data (the formula they use is in their 'Derived Variables' application note). That way I can always be consistent, regardless of whether the data is live or from the logger, and I don't have to worry about the Davis midnight start of day.

steve wrote:I'm not really sure what people actually do with ET and how important any of this is.

The ET value is one of the most important functions that I use for watering my lawn in an effort to save water. I use a spreadsheet to track the total monthly ET for historical purposes and to calculate base watering needs in the absence of daily ET values and I use the daily ET value to calculate specific watering for that cycle. For example if my watering schedule is every 2 days, I take the two previous days ET totals and use that to determine how much water I need. The agriculture module that Davis has for the VP2 can run reports up to the 30 previous days of ET values.

I agree with gemini6720 that the software uses a cumulative value throughout the day for the Weatherlink daily ET display but only updates total monthly/yearly display/values at the end of the day and zeroes out the daily value at the beginning of the next day to start the new daily count.

I have desperately been wanting to use Cumulus to replace the Weatherlink software but the lack of ET displays/reports has up to now prevented me from being able to switch.

steve wrote:Ah, yes, that's the solution. I can ignore the Davis ET value and get Cumulus to calculate it from the raw data (the formula they use is in their 'Derived Variables' application note). That way I can always be consistent, regardless of whether the data is live or from the logger, and I don't have to worry about the Davis midnight start of day.

I dont run my Weatherlink software 24/7 so I rely on my logger (using 10 minute interval) to catch up for the periods its not running. Would your method be able to mesh logger data and cumulus generated data correctly?

Assuming you can get all that to run correctly and there were some means of running a ET report or at least looking at the historical daily values within the Cumulus, I would be very happy as I could use that for my watering needs and finally be able to use something beside Weatherlink (which I really hate).

Paul

Slipstream wrote:I dont run my Weatherlink software 24/7 so I rely on my logger (using 10 minute interval) to catch up for the periods its not running. Would your method be able to mesh logger data and cumulus generated data correctly?

I believe so, yes. Cumulus would still be calculating the ET, whether from logger data or live data, and the parameters required are available from the logger data.

It seems odd to me that Davis have chosen not to make the latest hourly ET figure available in the 'LOOP' data. I guess this just reflects the way Weatherlink works; even when running 'live', it still downloads the archive data periodically. Cumulus doesn't do this; it only uses the archive data to catch up when it starts up. I suppose an alternative solution to this would be to change Cumulus to also download the archive data while it's running. It just seems like a big change for this one item.

I'll see if I can think of some more alternatives.

I agree that downloading archive data while Cumulus is running would be a big change for little value just to get the hourly ET values. Since we have the archived hourly readings from the logger data, wouldnt it just be easier to calculate the summary of the hourly readings pulled from the logger to come up with the cumulative daily ET so far and just add that to whatever Cumulus calculates as its running?

Paul

The main complication is a meteorological day that doesn't start at midnight. It's similar to the rainfall problem.

The easiest solution is probably for Cumulus to do the calculation. It could then also take humidity into account, which apparently the VP2 doesn't do.

Steve, if you were writing about the ET supplied by the Davis weather station, a little correction is required:

The easiest solution is probably for Cumulus to do the calculation. It could then also take humidity into account, which apparently the VP2 doesn't do.

According to the information provided in the manual, the Vantage Pro2 console uses the air temperature, the relative humidity, the average wind speed and the solar radiation to estimate the ET.

If you were writing about something else, excuse my correction.

gemini06720 wrote:According to the information provided in the manual, the Vantage Pro2 console uses the air temperature, the relative humidity, the average wind speed and the solar radiation to estimate the ET.

My info came from the Davis 'derived variables' application note. Perhaps when they say the Vantage Pro doesn't use the humidity in the calculation, they only mean the original VP rather than the VP2.

Ha, yes, the 'DERIVED VARIABLES IN DAVIS WEATHER PRODUCTS' publication - I have read that documentation a few times since I purchased the Vantage Pro2 weather station on order to find some answers to my many of my questions ... and the answers are almost always clear as mud!

I re-read the part on which method the designers at Davis Instruments had chosen to calculate the ETo for the Vantage Pro2 station:

The Vantage calculated ETo takes samples of Temperature, Wind Speed, and Solar Radiation over a one hour period and derives an average value in a manner similar to the GroWeather console1. Instead of sampling the humidity and deriving an “average humidity” for the hour, each time the temperature is sampled, the value of the saturation vapor pressure and actual water vapor pressure are calculated from the current values of temperature and humidity and sampled. These vapor pressure values (in kPa) are used to compute the average saturation vapor pressure and the average water vapor pressure for the hour. The Vantage has the capability to perform floating point arithmetic.

Note 1: The GroWeather console calculated ETo samples Temperature, Humidity, Wind Speed, Solar Radiation over a one hour period. This sampling is independent of sampling undertaken for the creation of archived data records. At the end of the hour, the arithmetic mean is calculated for each value by dividing the sum of the sampled data values by the number of samples taken. The number of samples is tracked for each sensor independently in case some sensors are not connected for some part of the period. In addition, the raw Barometer value (i.e. not corrected for altitude) at the end of the hour is read.

The temperature is calculated in tenths of a degree F, the humidity is calculated in tenths of a percent, wind speed is calculated in miles per hour, solar radiation is calculated in watts per square meter, and atmospheric pressure is read in thousandths of an inch of mercury. All arithmetic is in integers. Values that use fractions are represented by multiplying by an appropriate value. The formulas given below that use functions more complicated than addition, subtraction, multiplication, and division are calculated with table lookups with linear interpolation where appropriate.

What I do not really comprehend is this statement: '...each time the temperature is sampled, the value of the saturation vapor pressure and actual water vapor pressure are calculated from the current values of temperature and humidity and sampled'. Although you wrote that 'It could then also take humidity into account, which apparently the VP2 doesn't do', it is my understanding that the current value of the humidity as well as the value of the current temperature are indeed taken into account when the values of the saturation vapor pressure and actual water vapor pressure are calculated. Or did I misunderstood both your statement and that of Davis?

gemini06720 wrote:If you decide to follow the 09:00 meteorological day (which appears to be used only in the UK)

We use it in Australia too, but it must be different to how you do it in the UK, since here its for the 24 hours to 9am, and the values are recorded on the day of that 9am, not the previous day, as Cumulus does. IE rain last night of 4.4mm is recorded by the BOM for today, Cumulus records it for yesterday.

Gordon-Loomberah wrote:We use it in Australia too, but it must be different to how you do it in the UK, since here its for the 24 hours to 9am, and the values are recorded on the day of that 9am, not the previous day, as Cumulus does. IE rain last night of 4.4mm is recorded by the BOM for today, Cumulus records it for yesterday.

Yes, I had a feeling that some countries worked that way; I'll try to incorporate that option at some point.

Gordon, what do you do with ET? You, or Australia in general - do you record the total for the day, 9am to 9am?

Steve, they generally use Evaporation, rather than a calculated ET here, from http://www.bom.gov.au/catalogue/observa ... ural.shtml

"The amount of water loss by evaporation from a pan of water measured in millimetres per day for the 24 hour period to 9am on the day of reporting. If the period since the last evaporation observation is more than 24 hours, the time in hours since the last observation is shown in brackets after the station name.
The pan is about 120cm in diameter and about 25cm deep. The water levels are measured and compared each day. The difference in the levels is the amount of evaporation."

I've seen the pans at met stations, but not many have them. I haven't really seen any reference to official use of ET in Australia, although I'd suspect it might be used in agriculture in some areas. I just keep an eye on it myself for determining if the vegie garden irrigation needs to run longer in dry weather. I'm stuck with the Weatherlink midnight to midnight timing though, but being able to show it 9 to 9 like everything else would be nice.

I'm adding a post to this long-dormant thread rather than starting a new one, but honestly my interest is well off the main line on which the thread developed. I own a Davis Vantage Pro 2, and have been running Cumulus since late May 2011, posting data from my station to WunderGround Albuquerque site 102 and to my own Cumulus-generated site.

Lately I've been toying with the idea of adding a UV and a Solar sensor. As this would run about US $350 plus shipping and a bit of setup effort, I am rather curious as to what I'd get before making a final purchase decision. To that end I'd appreciate links to both Wunderground Cumulus generated data and to direct Cumulus generated web sites for Davis installations with solar, UV, or both.

Before you ask, my station location is imperfect regarding solar/UV data collection in that in the early morning in summer months it will be in shade from our bathroom, and in late afternoon all year it will be partially shaded by Arizona Cypress, and there will even be a slight mid-morning shading from a small tree to the south. Still, I think many hours of zero shade data per day will be available, and as this is just for interest, and not agricultural business use, I'm not currently critically worried about the siting problems.

Here is a picture of the station viewed from approximately WSW. The very low anemometer position is the outcome of negotiating with neighbors who objected strenuously to the slightest appearance of such a thing on their horizon--one downhill, and one uphill, which blocked all rooftop installation possibilities.