Dear Huang,
OK, now I understand a bit better what you have done, but could not
follow all details. So let me discuss the theory here in general terms.
If we forget about polarisation (stokes_dim=1), we can write the
radiance a satellite measures as:
I = I_d*r*exp(-tau) + (1-r)*B*exp(-tau) + I_u
I assume the nomenclature is clear here. This can be written as
I = I_s*exp(-tau) + I_u
with I_s = I_d*r + (1-r)*B
You seem to extract I_s correctly by storing the output of iySurfaceFastem.
But what about I_u? Please note that I_u is NOT equal to I_d. I get the
impression that you assume I_u = I_d, which is wrong.
(Don't get fooled by the fact that tau is independent of direction.)
In fact, it is not easy to calculate I_u by ARTS. To do this you need to
set r=0 and the surface temperature to zero! I don't think nobody has
tested this.
Another possible pitfall is that the "los" zenith angle" differs between
the surface and satellite altitude.
(Another possible error is that the equations above can only be written
in radiance, or RJ-Tb. Planck-Tb involves a non-linear mapping and that
would make the equations above invalid if I_u or I_d is small. But that
is likely not the problem here.)
Kind regards,
Patrick
On 08/11/15 13:23, 黄莹珠 wrote:
>
> Dear Patrick,
> The different colors line indicate the different frequency of the
> channels.It is 87,88.5,89.5,91GHz for channel-1(pic1 and pic2),and
> 118.66,118.68,118.82,118.84GHz for channel-2(pic3 and pic4),and
> 118.5,118.6,118.9,119GHz for channel-3(pic5 and pic6).The Ts is got from
> the WriteXML iy in the surface rtp agenda of fastem which is write as:
> AgendaSet( iy_surface_agenda ){
> #specular_losCalc
> #Extract(wind_direction,specular_los,0)
> NumericSet(wind_direction,0)
> Extract( surface_skin_t, sst_vec, 0 )
>
> iySurfaceFastem( surface_skin_t=surface_skin_t,salinity=salinity,
> wind_speed=wind_speed,
> wind_direction=wind_direction,fastem_version=3 )
> WriteXMLIndexed("ascii",forloop_index,iy)
> }
> and transfer to Planck BT.
> The trans is calculated from iy_aux_vars, ["Optical depth"]) with the
> equation of trans=-exp(Optical depth).
>
> and the Td is equals to Tup and calculated by setting the sensor_pos as
> Z_surface and sensor_los is the same as down look from the satellite.
>
> The direct simulation of arts with different incident angle is
> calculated by setting the sensor_pos as satellite's altitude and line
> scanned angle as sensor_los.
>
> I have not include any sensor_response and the stokes dim I set is 2,I
> calculate the quasi_v and quasi_h polarization with
> Tv=I+Q;Th=I-Q;Tv'=Tv*cos(theta)^2+Th*sin(theta)^2;The Unit is planck bt.
>
> Thank you again and best regards to you!
>
> Huang Yingzhu
> > -----原始邮件-----
> > 发件人: "Patrick Eriksson" <[email protected]
> <mailto:[email protected]>>
> > 发送时间: 2015年8月11日 星期二
> > 收件人: "黄莹珠" <[email protected]
> <mailto:[email protected]>>, [email protected]
> <mailto:[email protected]>
> > 抄送:
> > 主题: Re: [arts-
> users] Fw: some question about multiple sensor_los simulation using arts
> >
> > Dear Huang Yingzhu,
> >
> > Sorry, but I don't understand the figures at all. For example, what are
> > the different colours? You need to describe in detail how you have done
> > the calculations. How do you extract the terms Ts, trans and Td. What do
> > you compare Ts*trans+Td to? Do you include any sensor? What stokes_dim?
> > And so on ... Include as much information as possible.
> >
> > Kind regards,
> >
> > Patrick
> >
> > On 08/10/15 04:19, 黄莹珠 wrote:
> > > Dear Patrick,
> > > pic 1 and pic 2 is a couple,which pic 1 is the results of
> > > the equation of Ts*trans+Td I mentioned.and pic 2 is the results of
> > > direct simulation of arts with different incident angle.
> > > the pic 1 is the same results with the MPM93 model.
> > >
> > > pic 3 and pic 4 is a couple.which pic 3 is the results of
> > > the equation of Ts*trans+Td I mentioned.and pic 4 is the results of
> > > direct simulation of arts with different incident angle.
> > >
> > > pic 5 and pic 6 is a couple.which pic 5 is the results of
> > > the equation of Ts*trans+Td I mentioned.and pic 6 is the results of
> > > direct simulation of arts with different incident angle.
> > >
> > > Best regards,
> > > Huang Yingzhu
> > >
> > > > -----原始邮件-----
> > > > 发件人: "Patrick Eriksson" <[email protected]
> <mailto:[email protected]>>
> > > > 发送时间: 2015年8月4日 星期二
> > > > 收件人: "黄莹珠" <[email protected]
> <mailto:[email protected]>>, [email protected]
> <mailto:[email protected]>
> > > > 抄送:
> > > > 主题: Re: [arts-
> > > users] Fw: some question about multiple sensor_los simulation using arts
> > > >
> > > > Dear Huang Yingzhu,
> > > >
> > > > Back from vacations today.
> > > >
> > > > I don't know what to comment. You need to give us more details. What
> > > > other models are you using? How big are the differences? Can you be
> more
> > > > precise about how you compare the models? Particularly, how do you
> > > > handle V and H polarisation? Can you send some figures?
> > > >
> > > > Kind regards,
> > > >
> > > > Patrick
> > > >
> > > >
> > > >
> > > > On 07/31/15 05:28, 黄莹珠 wrote:
> > > > > Dear Patrick,
> > > > >
> > > > > This email I send you before your vocation,and I still have no
> answer of
> > > > > that.So I send you again and waiting for your reply.
> > > > >
> > > > >
> > > > > Yes,it is very funny.If I set the surface as blackbody, the results
> is
> > > > > the same as other models.But if I set the surface as iyFastem, the
> > > > > results is changed.
> > > > > And I record the iyFastem's iy results, and calculated the download
> Td
> > > > > and transmition respectively.and I use the equation of
> Ts*trans+Td,then
> > > > > the results is the same as the total Tb the sensor received above
> the
> > > > > atmosphere that calculated by the other models.
> > > > >
> > > > > Why?
> > > > >
> > > > > Thank you again and best regards!
> > > > >
> > > > > Huang Yingzhu
> > > > >
> > > > >
> > > > > > -----原始邮件-----
> > > > > > 发件人: "Patrick Eriksson" <[email protected]_
> <mailto:[email protected]_>
> > > > > <mailto:[email protected]
> <mailto:[email protected]>>>
> > > > > > 发送时间: 2015年6月17日 星期三
> > > > > > 收件人: "黄莹珠" <[email protected]_
> <mailto:[email protected]_>
> > > > > <mailto:[email protected]
> <mailto:[email protected]>>>, [email protected]_
> <mailto:[email protected]_>
> > > > > <mailto:[email protected]
> <mailto:[email protected]>>
> > > > > > 抄送:
> > > > > > 主题: Re: [arts-
> > > > > users] Fw: some question about multiple sensor_los simulation using
> arts
> > > > > >
> > > > > > Dear Huang Yingzhu,
> > > > > >
> > > > > > The Ppath structure is described carefully in the ARTS user
> guide (AUG).
> > > > > > In the latest version it is section:
> > > > > >
> > > > > > 10.5 The propagation path data structure
> > > > > >
> > > > > > but if you use arts v2.2 the section number is different.
> > > > > >
> > > > > > Note that "start" here means the start of the propagation path,
> at the
> > > > > > surface, not at the sensor. That is, e.g. lat=10 means that the
> surface
> > > > > > intersection is 10 degrees away from the sensor (for 1D).
> > > > > >
> > > > > > For window channels, how the Tb varies with zenith angle depends
> > > > > > critically on the surface emission. Do you use the same surface
> > > > > > emissivity in the different models?
> > > > > >
> > > > > > Best regards,
> > > > > >
> > > > > > Patrick
> > > > > >
> > > > > >
> > > > > >
> > > > > > On 06/17/15 08:17, 黄莹珠 wrote:
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > -----原始邮件-----
> > > > > > > *发件人:* "黄莹珠" <[email protected]_
> <mailto:[email protected]_>
> > > > > <mailto:[email protected] <mailto:[email protected]>>>
> > > > > > > *发送时间:* 2015年6月16日 星期二
> > > > > > > *收件人:* [email protected]_
> <mailto:[email protected]_>
> > > > > <mailto:[email protected]
> <mailto:[email protected]>>
> > > > > > > *抄送:*
> > > > > > > *主
> > > 题:* some question about multiple sensor_los simulation using arts
> > > > > > >
> > > > > > > Hi, all
> > > > > > > I am confused that I use arts to simulate the multiple
> sensor_los
> > > > > > > of the measurement in microwave channels. The simulated
> Temperature
> > > > > > > brightness of windows channel such as 89GHz is bigger when the
> > > > > > > sensor_los is 180 then the one of the sensor_los is 126.5.see
> attached pic.
> > > > > > > It different from the results simulated by the other models.
> > > > > > > when I check the ppath of them I find the start pos have a
> latitude of
> > > > > > > 10 when the sensor_los is 126.65 but the latitude is 0 when the
> > > > > > > sensor_los is 180 , like follows:
> > > > > > >
> > > > > > > <?xml version="1.0"?>
> > > > > > > <arts format="ascii" version="1">
> > > > > > > <Ppath><Index name="AtmosphericDimensionality">1</Index>
> > > > > > > <Index name="NumberOfPositionInPropagationPath">669</Index>
> > > > > > > <Numeric
> name="PropagationPathConstant">69249.3251344851</Numeric>
> > > > > > > <String name="RadiativeBackground">"surface"</String>
> > > > > > > <Vector name="StartPositionOfPropagationPath" nelem="2">
> > > > > > > 0
> > > > > > > 0.0720929145706464
> > > > > > > </Vector>
> > > > > > > <Vector name="StartLOSOfPropagationPath" nelem="1">
> > > > > > > 179.378097497658
> > > > > > > </Vector>
> > > > > > > <Numeric name="StartLstepOfPropagationPath">0</Numeric>
> > > > > > > <Matrix name="PropagationPathPointPositions" nrows="669"
> ncols="2">
> > > > > > > 63965.423084626 0.0659158935548021
> > > > > > > 63865.4288625289 0.0659254544247858
> > > > > > > 63765.4346406078 0.0659350155916627
> > > > > > > 63665.4404188674 0.0659445770553191
> > > > > > > 63565.4461973086 0.0659541388158118
> > > > > > > 63465.4519759286 0.0659637008731693
> > > > > > > 63365.4577547302 0.0659732632273631
> > > > > > > 63265.4635337079 0.0659828258784501
> > > > > > > 63165.4693128662 0.0659923888264871
> > > > > > > 63065.4750922015 0.0660019520713888
> > > > > > > 62965.4808717147 0.066011515613269
> > > > > > > 62865.4866514094 0.0660210794520424
> > > > > > > 62765.4924312793 0.0660306435875668
> > > > > > >
> > > > > > > <?xml version="1.0"?>
> > > > > > > <arts format="ascii" version="1">
> > > > > > > <Ppath><Index name="AtmosphericDimensionality">1</Index>
> > > > > > > <Index name="NumberOfPositionInPropagationPath">821</Index>
> > > > > > > <Numeric
> name="PropagationPathConstant">5788170.73405349</Numeric>
> > > > > > > <String name="RadiativeBackground">"surface"</String>
> > > > > > > <Vector name="StartPositionOfPropagationPath" nelem="2">
> > > > > > > -488.411979784258
> > > > > > > 11.822568283672
> > > > > > > </Vector>
> > > > > > > <Vector name="StartLOSOfPropagationPath" nelem="1">
> > > > > > > 114.827431716328
> > > > > > > </Vector>
> > > > > > > <Numeric name="StartLstepOfPropagationPath">0</Numeric>
> > > > > > > <Matrix name="PropagationPathPointPositions" nrows="821"
> ncols="2">
> > > > > > > 80520.0870012585 10.3120928628466
> > > > > > > 80412.9346699752 10.3140130944758
> > > > > > > 80305.789593217 10.3159333898205
> > > > > > > 80198.6517713424 10.3178537488795
> > > > > > > 80091.5212047147 10.3197741716519
> > > > > > > 79984.3978936942 10.3216946581362
> > > > > > > 79877.2818386406 10.3236152083316
> > > > > > > 79780.5901730424 10.3253490196361
> > > > > > > 79683.904421174 10.3270828828595
> > > > > > >
> > > > > > > what does them means?who can help me.Thank you!
> > > > > > >
> > > > > > > Huang Yingzhu
> > > > > > >
> > > > > > >
> > > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > _______________________________________________
> > > > > arts-users mailing list
> > > > > [email protected] <mailto:[email protected]>
> > > > > https://www.sat.ltu.se/mailman/listinfo/arts-users
> > > > >
> > >
> > >
> > >
>
>
>
>
>
>
>
>
>
> _______________________________________________
> arts-users mailing list
> [email protected]
> https://www.sat.ltu.se/mailman/listinfo/arts-users
>
_______________________________________________
arts-users mailing list
[email protected]
https://www.sat.ltu.se/mailman/listinfo/arts-users
