Just in case

[mcrossley]

Sorry, yes I found it in the calculation procedure page rather than the net long wave radiation page.

Source code changed for the next build.

Ah, but I was simply dividing the end result Rnl by 24 to get the same result - so as you were!

[Matt.j5b]

No worries Mark.

Thanks for that.

For my tests I based my code on your code and after that compared it the FAO document. And as you say that info is the calculation procedure pages, which is where I notice some things to check as provided.

[HansR]

Interesting, thanks.

I had coded the Stefan-Boltzman constant to be 4.903e-09, but as you point out that is a per day value. Since we are calculating by the hour it makes sense to divide that by 24.
However I cannot see a reference to doing that in the FAO document?

Yes there is in Chapter 4, search for Boltzman (roughly halfway you find devide by 24).

Interesting discussion. Though I don't see EVT be more than some kind of reference. Not something like a measurement, not even a derivative.
Interesting nonetheless!

[HansR]

Ah.... too late.
Nice read anyway

[mcrossley]

Interesting discussion. Though I don't see EVT be more than some kind of reference. Not something like a measurement, not even a derivative.
Interesting nonetheless!

Mainly for agriculturists, gives some idea of when watering may be required if you compare the cumulative figure against rainfall. The ET CMX and Davis calculates is just for the standard grass crop, for anything else it has to be corrected.

[Matt.j5b]

After the tweaks to the ET calculation in build 3152, I re did my comparisons for the year so far and is similar to the previous totals of when this was first included in CMX.

Code: Select all

	ET CMX	ET CMX v2  ET bom1  ET bom2
------------------------------------------------------------------
Jan	123.45	123.77	   164.5      156.8
Feb	101.62	101.9	   149.3      129.9
Mar	89.51	89.76	   122.6      106.9
Apr	84.34	84.5	   105.4      87.4
May	75.79	75.98	   86	      70
Jun	63	63.2	   72.4	      53.4
Jul	71.6	71.97	   81.2	      66.5
Aug	92.82	93.05	   106.2      68.9
Sep	112.58	112.94	   135.3      88.9
Oct	110.65	110.96	   143.9      115.9
------------------------------------------------------------------
Total  925.36   928.03    1166.8     944.6

ET_Bom1: As a comparison to a nearby BoM Station - Archerfield Airport that get fairly similar weather but is more exposed.

ET_Bom2: As a comparison to a nearby BoM Station - Brisbane that generally doesn't get as hot (closer to the coast) but not as exposed as Archerfield.

From what I can see, I pretty happy with the ET figures, and fits in roughly what I would expect.

I find the ET figures interesting to follow as a guide in relation to weather at the time and compared with rainfall. I will be looking to adding this data to my website.

[mcrossley]

Thanks for continuing to look at this. I'm getting low daily ET values as we head into winter - 1 mm or less - so it makes evaluation a little tricky.

From my quick scan of the results and comparing with my VP2 console...
The value derived from wind is better (I'm wondering if Davis have omitted the anemometer height compensation altogether - there is no input for it on the console).
The solar contribution appears to be lower. I may dump the current MX method, and just implement the FAO method - though I think their value is an approximation of what MX is using.

[Matt.j5b]

No worries. I would be happy to assist with this if you need any thing further - we get plenty of ET here. I already had some calculations worked out because I wanted to use the ET values because of the WLL not calculating the data from separate sensors.

Interesting comparisons there - can't compared with the console here (It's wasn't until I moved to the WLL that I could measure solar). I don't think Davis worried about the height of the anemometer as you say, unless that is undocumented but it would not know anyway. Davis however don't specifically explain how they calculate the net radiation, as in their specs document they say they use a modified calculation. I suspect that Davis approximated the calculations to some degree.

[mcrossley]

I suspect that Davis approximated the calculations to some degree.

With the limited CPU power in the console, they seem to have used lookup tables and interpolation for other values, I wouldn't be surprised if they did the same with some ET calculations.

[iandrews]

Is there a way to (re)calculate missing Evapotranspiration?

I used to get the figure but "lost" it when I moved my solar sensor to a separate transmitter (and got a WLL) a few years back. I noticed the CumulusMX "Calculate evapotranspiration" option at the beginning of the year, and turned it on, and I have the evapotranspiration since, but I have almost 2 years where it's missing.

Anything I can do?

[mcrossley]

Nope, sorry.

[iandrews]

Nope, sorry.

OK, thanks.