Sonnenzeitberechnung bei Vantage

[weneu]

Helo JackT,
thanks for your ueseful graphics, but I think this is not the reason. Because the concept of WSWIN is like this:
Below 0,83 degrees of sun-position there will be no calculation of suntime.
That means the sun must have a certain position above the horizon before there can calculation of suntime.
If this is not correct, I think, Werner will correct me.

[Werner]

Hallo,

wenn der Faktor zu "stark" reduziert werden muss, habt Ihr vermutlich das Kabel verlängert oder der Sensor liefert einen zu kleinen Wert (verschmutzt ?!).

In diesem Fall hilft es, beim Solarsensor einen Korrekturfaktor >1 zu verwenden.

Das Programm geht von Normwerten bei einem bestimmten Standort und Sonnenstand aus, wenn aber Eure "Normwerte" zu niedrig sind, kann das nicht mehr funktionieren.

Werner

[jackT]

Hallo,

wenn der Faktor zu "stark" reduziert werden muss, habt Ihr vermutlich das Kabel verlängert oder der Sensor liefert einen zu kleinen Wert (verschmutzt ?!).

In diesem Fall hilft es, beim Solarsensor einen Korrekturfaktor >1 zu verwenden.

Das Programm geht von Normwerten bei einem bestimmten Standort und Sonnenstand aus, wenn aber Eure "Normwerte" zu niedrig sind, kann das nicht mehr funktionieren.

Werner

Hello Werner,

In english again...sorry...
I agree that a "dirty" sensor will measure low value of solar radiation. But...

The theoritical solar radiation curve is a "semi-sinusoid" curve, it is normal.

But it seems that the solar radiation measured by a Vantage Pro pyranometer is not "semi-sinusoid" but "sigmoid" or "gaussian".

Better than theory here are data from a Vantage Pro in Brittany - France measured on February 14th, during a day with a full blue sky and not clouds :


The red curve shows the measured solar radiation, and the yellow curve shows the %sunstart% value. Standard parameters for the Vantage Pro are used for sunshine time calculations, and the location of the weather station is correctly entered . The sensor is not dirty, and data are transmitted wirelessly.
Due to the sigmoid or gaussian shape of the red curve , radiation measurements were below the %sunstart% value form 8h43 to 9h15, and no sunshine time were calculated by wswin during this period, but the sun was fully visible and no clouds at all. Same thing before sunset.

Here is the full graphic :

[FokkerPlanck]

Hello JackT,

the times for sun rise and sun set are reached if the position of the sun is 50'=0.8333° below the horizon. If the elevation is -0.8333° in the WsWin program this should be correct and in fits with the conventions.

See here in Davis' app-note, page 20 ff for more details:
"These formulas indicate the true geometric position of the sun. When the sun is on the horizon
(as in the case of sunrise and sunset), refraction by the atmosphere will alter the apparent
position of the sun. Under average conditions, the sun will appear at the horizon when it is
actually 34' (0.567°) below the horizon. Since sunrise and sunset is defined as when the upper
half of the sun is visible on the horizon, and the radius of the sun when on the horizon is 16'
(0.267°), sunrise and sunset are defined when the geometric position of the sun is 50' (0.833°)
below the horizon."

www.davisnet.com/product_documents/weat ... ote_28.pdf

[jackT]

Hello JackT,

the times for sun rise and sun set are reached if the position of the sun is 50'=0.8333° below the horizon. If the elevation is -0.8333° in the WsWin program this should be correct and in fits with the conventions.

Hello Fokker,
Thanks for this explanation. I understand now .
Yes, the elevation is -0.8333° in Wswin at the time of sunrise.

Regards

[FokkerPlanck]

I agree that a "dirty" sensor will measure low value of solar radiation. But...

The theoretical solar radiation curve is a "semi-sinusoid" curve, it is normal.

But it seems that the solar radiation measured by a Vantage Pro pyranometer is not "semi-sinusoid" but "sigmoid" or "gaussian".

The red curve shows the measured solar radiation, and the yellow curve shows the %sunstart% value. Standard parameters for the Vantage Pro are used for sunshine time calculations, and the location of the weather station is correctly entered . The sensor is not dirty, and data are transmitted wirelessly.
Due to the sigmoid or gaussian shape of the red curve , radiation measurements were below the %sunstart% value form 8h43 to 9h15, and no sunshine time were calculated by wswin during this period, but the sun was fully visible and no clouds at all. Same thing before sunset.

Hi JackT,

the correlation between the measured radiation of the VP+ solar-sensor and the criteria if the sun is shining or not is more complicated. The solar sensor is measuring the sum of two radiation parts (the sum is called global radiation Eg and only this is measured): short wave radiation from the sun that directly pass the atmosphere and is attenuated by clouds and air (=direct radiation Edir). This part varies significantly with the suns angle over the horizon h: Edir = Edir_p*sin (h), where Edir_p is the radiation if the sensor measures perpendicular to the sun.
The second part is the diffuse part and is generated by the scattering and is falling from the whole sphere onto the sensor (=diffuse radiation Ediff). The diffuse radiation is the blue light of the sky and white light of clouds.

The global radiation is then: Eg = Edir_p*sin (h) + Ediff.

The criteria for sunshine or not is linked to the direct radiation only: Sunshine is, when Edir_p = Edir/sin(h) > 120W/m2. Thus, strictly spoken there is no straight forward correlation between what the solar sensor is measuring and sunshine. But in pratice correlations can be constructed (I am working on them).

But I think the way WsWin is relating sunshine to global radiation could be better - strictly spoken: it is wrong! Nevertheless it may cause such missalignments like your observations. I guess (unfortunately I get no answer on my request on confirmation if the following is right or not one year ago) WsWin is ignoring the diffuse part and uses the criteria is this form (which has no physical meaning):

Sunshine is, if Eg/sin (h) > 120 W/m2, or some modified value. Thus, this is more often fulfilled than the correct definition and WsWin should cause overestimated sums of sunshine times.

Regarding your question about the curves: at sun rise (before and shortly after) the diffuse part is dominating the global radiation. Thus, this explains the deviation from the sinoid-curve.