[GRASS-user] Re: [GRASS-dev] Changes to r.sun made Feb 2006 [resolved]

José Antonio Ruiz Arias jararias at ujaen.es
Wed Dec 13 06:04:57 EST 2006


Hi Dylan,

I agree with you, probably is a better choice the correlation of Erbs
(equation [9] in Jacovides el al.).

It seems that you are calculating the Linke turbidity for all the days. As
you can read in the Cucumo paper, this parameter is used to estimate beam
and diffuse solar radiation in a determinated location for clear days. What
happens under overcast conditions (normal situation in winter) is that kt is
smaller than in a clear day, so kd is greater and you are underestimating
the beam component. For this reason, according to the expression [10] in
Cucumo et al., the Linke turbidity is overestimated.

That means that probably you should use some kind of filtering on your data
to rule out the cloudy days. For example, basing on the daily clearness
index, you could take only those days with daily kt greater then 0.6 more or
less.

Try this filtering. Probably you will improve the results

Cheers,

________________________________
 
José A. Ruiz Arias
Departamento de Física
Escuela Politécnica Superior
Edificio A-3, Campus Lagunillas
Universidad de Jaén
23071 Jaén Spain
Tlf. +34 953212474
Email:  jararias at ujaen.es
_____________________________________

> -----Mensaje original-----
> De: Dylan Beaudette [mailto:dylan.beaudette at gmail.com]
> Enviado el: martes, 12 de diciembre de 2006 23:27
> Para: grassuser at grass.itc.it
> CC: José Antonio Ruiz Arias
> Asunto: Re: [GRASS-user] Re: [GRASS-dev] Changes to r.sun made Feb 2006
> [resolved]
> 
> Hi Jose, some updates:
> 
> 
> On Tuesday 12 December 2006 03:59, José Antonio Ruiz Arias wrote:
> > Hi all,
> >
> > as I promised you (Dylan), here you have a useful reference to derive
> the
> > direct beam component from the global radiation.
> >
> > Jacovides, C.P., Tymvios, F.S., Assimakopoulos, V.A, and Kaltsounides
> N.A.
> > (2006): ' Comparative study of various correlations in estimating hourly
> > diffuse fraction of global solar radiation' Renewable Energy, 31, 2492-
> 2504
> >
> > This paper offers a comparative analysis of different correlations
> between
> > diffuse and global radiation. The references [2], [3], [4] and [9] on
> it, I
> > think are especially interesting. I mention only this paper because on
> it
> > you can find a summary of the most important correlations.
> >
> > Reference [3] is especially simple and easy to read although it only
> uses
> > measurements of Toronto. Also very interesting for you, perhaps, is the
> > reference [4] since it is based on five USA locations. For example, this
> > author proposes the correlation
> >
> > kd = 1.45-1.67kt for 0.3<kt<0.78
> > kd = 1.02-0.248kt for kt<=0.3
> > kd = 0.147 for kt>=0.78.
> 
> Ok. I found these papers and have read over them. I think that if I were
> to
> get this approach working* I will use equation 7 from Jacovides et al.
> 2006.... details below:
> 
> > where kt is the ratio of global to extraterrestrial hourly radiation and
> kd
> > is the ratio of diffuse to global hourly radiation. Therefore, to
> calculate
> > kt only need the measured global radiation and the extraterrestrial
> > radiation, theoretically calculated for a given location and time (see
> for
> > example, Iqbal, Muhammad (1983). An Introduction to Solar Radiation.
> > Academic Press, New York). Then you get kd from the correlation, and
> >
> > Direct Beam = Global*(1-kd).
> >
> > So, if you dispose of global measurements, you can get the direct beam
> > radiation directly applying the former expression.
> 
> Ok. this is the expression I used (after calculating kt and kd):
> 
> kt = global radiance at 12pm / extraterrestrial radiance
> kd = (from eqation 2 or 9)
> 
> unfortunately, when i compute a normal beam radiance from the kd values:
> 
> B = G(1 - kd)
> Bn = B/sin(alpha)
> 
> and then compute linke turbidity values (using the equation from Cucumo et
> al
> 2000) : I end up with very high Linke values in the winter months:
> 
> http://169.237.35.250/~dylan/temp/reindl_linke_bad-estimates.png
> 
> This may be a result of not having a *single* clear day for my winter
> months
> (I am using the max radiance at 12pm, over the course of 11 years to get
> as
> close as possible to clear sky conditions).
> 
> Note that linke turbidity values in the above figure track those that I
> had
> been estimating  in the summer months: with the diffuse = 10-20% of beam
> radiance assumption.
> 
> Quite possibly I am doing something incorrectly :)
> 
> Cheers,
> 
> Dylan
> 
> > By the way, I would like to know what you are trying to do, improving
> > r.sun? or just working in a personal project? I have used r.sun a little
> > bit and I would like to collaborate with r.sun. I could work with the
> model
> > and/or even developing some snippets of code.
> >
> > Cheers,
> >
> > Jose A.
> >
> > > -----Mensaje original-----
> > > De: grassuser-bounces at grass.itc.it
> > > [mailto:grassuser-bounces at grass.itc.it] En nombre de jararias at ujaen.es
> > > Enviado el: viernes, 08 de diciembre de 2006 13:09
> > > Para: grassuser at grass.itc.it
> > > Asunto: Re: [GRASS-user] Re: [GRASS-dev] Changes to r.sun made Feb
> 2006
> > > [resolved]
> > >
> > > Hi all,
> > >
> > > firstly, I really hope not to interfere in this conversation. I assist
> > > expectant to the thread cause is very interesting for me.
> > >
> > > It seems you want to get diffuse component from global radiation. If I
> > > have understood well, you estimate it as the 10%-20% of the global
> one.
> > > Just a little suggestion that I hope not to be very simplistic. Have
> you
> > > think about a correlation between the clearness-index and the
> proportion
> > > of diffuse? There exist several correlations around the world such as
> the
> > > studies of Orgill, Reindl, Erbs, Holland or Duffie, for example. They
> > > usually use hourly values. You can easily obtain a estimation for the
> > > hourly diffuse component (and, therefore, for the hourly direct beam
> one)
> > > from the hourly global irradiation measured with a pyranometer as the
> > > LICOR. If you are interesting in this approach just let me know and I
> > > will be delighted of sending you more references. Now I can't be more
> > > explicit because I don't have access to my personal data-base cause I
> am
> > > not at the office.
> > >
> > > Kind regards,
> > >
> > > José A.
> > >
> > > > On 12/7/06, Hamish <hamish_nospam at yahoo.com> wrote:
> > > >> Dylan Beaudette wrote:
> > > >> > > > > Get local Linke values from the SoDa Solar radiation
> Databases
> > > >> > > > >   http://www.soda-is.com/
> > > >> > > >
> > > >> > > > Tried this, and got unrealistic values for my site.
> > > >> > >
> > > >> > > how so? Have you calculated good Linke values from your light
> > > >> > > meters? (formula for that, please?)
> > > >>
> > > >> ..
> > > >>
> > > >> > Attached is a plot of estimated Linke turbidity values, compared
> to
> > > >> > the SoDa  derived data. Details on this calculation soon to
> follow
> > > >> > in a new thread.
> > > >
> > > > Hi Hamish,
> > > >
> > > > Looks like I sent an update this morning with an attachment that was
> > > > too large!
> > > >
> > > > I'll Include it at the bottom of my replies to you last message.
> > > >
> > > >> I assume the physical basis for the Linke turbidity factor is the
> > >
> > > amount
> > >
> > > >> of scattering dust, ash, water vapour, turbulence etc. in either
> the
> > > >> column of air directly above the sampling site or in the ray
> between
> > >
> > > the
> > >
> > > >> site and the Sun (which will pass through more atm in winter due to
> > > >> the lower angle). Maybe the paper sheds some light on that. So e.g.
> LA
> > > >> county would have very high turbidity numbers.
> > > >>
> > > >> I would figure in the winter the air to be crisper/clearer due to
> less
> > > >> water vapour, less turbulence, and less upwelling of aerosol laiden
> > >
> > > warm
> > >
> > > >> air from the Earth's surface. So lower turbidity numbers in the
> winter
> > > >> as the SoDa data shows. BUT if you live in a place which is prone
> to
> > > >> inversion layers in winter and many locals heat their homes with
> wood
> > >
> > > or
> > >
> > > >> coal, winter air quality may turn out to be much more turbid.
> > > >
> > > > Actually have a couple good papers on the defintion for the Linke
> > > > turbidity values:
> > > >
> > > > Louche, A.; Peri, G. & Iqbal, M. An analysis of linke turbidity
> factor
> > > > Solar Energy, 1986, 37, 393-396
> > > >
> > > > Rapti, A. S. Atmospheric transparency, atmospheric turbidity and
> > > > climatic parameters Solar Energy, 2000, 69, 99-111
> > > >
> > > > Kasten, F. The linke turbidity factor based on improved values of
> the
> > > > integral Rayleigh optical thickness Solar Energy, 1996, 56, 239-244
> > > >
> > > > My site is in a mediterranean, semi-coastal / semi-arid setting
> where
> > > > winter time moisture and summer time dryness are the usual weather
> > > > patterns. Summer through fall are usually cloud free, and quite
> warm.
> > > > Also- this site is rather remote, so urban ag-related dust
> influences
> > > > are minimal. These are some of the reasons that I was having second
> > > > thoughts about the SoDa values.
> > > >
> > > >> > I have calculated Linke turbidity for my site with the equation
> > > >>
> > > >> T_linke = ln( G / ( I_0 * sin(h) * 0.84) ) * ( sin(h) / -0.027 )
> > > >>
> > > >> > Unfortunately, there are some aspects to this equation which
> > > >> > I do not understand: namely the two constants present.
> > > >>
> > > >> perhaps they are empirically derived?
> > > >
> > > > Ok- big change in direction. I posted an update to the use of the
> > > > above equation earlier today but it was rejected due to a large
> > > > attachment. Details below....
> > > >
> > > >> I take it I_0 is pure-clear air intentensity?; h is latitude?; and
> is
> > > >> G measured intensity, or ..?
> > > >
> > > > G = global radiance measured on the ground (beam+diffuse)
> > > > I_o = extraterrestrial radiance
> > > > h = solar elevation (angle)
> > > >
> > > >> > 1. Becker, S. Caclulation of Direct Solar And Diffuse Radiation
> in
> > > >> > Israel International Journal of Climatology, 2001, 21, 1561 -
> 1576
> > > >>
> > > >> I will have to track down a copy of that. Holy cow, for once our
> > >
> > > library
> > >
> > > >> is actually subscribed to the e-journal. Downloading the PDF now..
> > > >
> > > > ... i know how that can be sometimes!
> > > >
> > > >> What software did you use to make that plot? Looks nice.
> > > >
> > > > Thanks. It was made in R, a fun command-line based environment for
> > > > plotting, analysis, etc. I can post the code used after a small
> > > > cleanup.
> > > >
> > > > Ok, here is the text from the message rejected this morning:
> > > >
> > > >
> > > > Some updates on my findings with respect to obtaining optimal Linke
> > > > turbidity values via local pyranometer measurements.
> > > >
> > > > Chatting with a local atmospheric scientist, I learned that it is
> > > > possible (with a loss of some precision of course) to dissaggregate
> > > > beam radiance from global radiance (as measured by pyranometer), for
> > > > use in an equation [1] for the linke turbidity factor. Compared to a
> > > > previous paper I mentioned on grass-dev [2], the equation presented
> in
> > > > [1] is a much more appropriate approach to calculating Linke
> turbidity
> > > > values.
> > > >
> > > > The key points to using the equation from [1] are:
> > > >
> > > > - historical pyranometer data from a weather station, with
> coordinates!
> > > > - r.sun mode 1 to calculate solar elevation at this station, for
> each
> > > > day at some time. extraterrestrial radiance is also calculated in
> this
> > > > step
> > > > - dissagregation of the beam component from global radiance with the
> > > > assumption that diffuse radiance is approx 10%-20% that of the beam
> > > > radiance
> > > > - conversion of dissaggregated beam radiance to beam radiance on the
> > > > normal (B/sin alpha)
> > > >
> > > > here is a link to a bit of my thesis on this topic: (rather
> > > > incomplete!) http://169.237.35.250/~dylan/temp/linke-cucumo2000.png
> > > >
> > > > Here is a link of linke turbidity values, as calculated by the
> method
> > > > described above, along with a comparison to the SoDa-derived values.
> > > > There is considerable difference between the two sources!
> > > >
> > > > http://169.237.35.250/~dylan/temp/cucumo-
> > >
> > > estimated_daily_linke_values.png
> > >
> > > > With the equation from [1] it is possible to estimate the linke
> > > > turbidity value for each day (using the 12pm calculation as
> > > > representative for an entire day), and use these estimates in r.sun
> > > > mode 2. here is a summary of the incorporation of estimated linke
> > > > values into r.sun - as compared to the weather station data:
> > > >
> > > > http://169.237.35.250/~dylan/temp/11-yr_variation_vs_modeled.png
> > > >
> > > >
> > > > I'll post some updates as they come. If I have anything wrong please
> > > > don't hesitate to point it out!
> > > >
> > > > Cheers,
> > > >
> > > > Dylan
> > > >
> > > >
> > > > 1. Cucumo, M.; Kaliakatsos, D. & Marinelli, V. A calculation method
> > > > for the estimation of the Linke turbidity factor Renewable Energy,
> > > > 2000, 19, 249-258
> > > >
> > > > 2. Becker, S. Caclulation of Direct Solar And Diffuse Radiation in
> > > > Israel International Journal of Climatology, 2001, 21, 1561 - 1576
> > > >
> > > > _______________________________________________
> > > > grassuser mailing list
> > > > grassuser at grass.itc.it
> > > > http://grass.itc.it/mailman/listinfo/grassuser
> > >
> > > _______________________________________________
> > > grassuser mailing list
> > > grassuser at grass.itc.it
> > > http://grass.itc.it/mailman/listinfo/grassuser
> > >
> > > __________ Información de NOD32 1914 (20061211) __________
> > >
> > > Este mensaje ha sido analizado con  NOD32 antivirus system
> > > http://www.nod32.com
> >
> > _______________________________________________
> > grassuser mailing list
> > grassuser at grass.itc.it
> > http://grass.itc.it/mailman/listinfo/grassuser
> 
> --
> Dylan Beaudette
> Soils and Biogeochemistry Graduate Group
> University of California at Davis
> 530.754.7341
> 
> __________ Información de NOD32 1918 (20061212) __________
> 
> Este mensaje ha sido analizado con  NOD32 antivirus system
> http://www.nod32.com






More information about the grass-user mailing list