Dear Huang Yingzhu,

The across-track scanning operational instruments are problematic when 
it comes to polarisation, as you have noticed. For this reason there is 
a special treatment in ARTS of such sensors. Here you should use arts 
2.3, and the "metmm" system. It is demonstrated in

arts/controlfiles/instruments/metmm/sensor_descriptions

For questions regarding this part, I refer to Alex and Stefan in Hamburg 
(see Cc).

I have never combined this with qpack. Has someone else done it?

Anyhow, I should here recommend to set the sensor following option 2:

 >> qinfo(@qarts,'SENSOR_RESPONSE')

/--------------------------------------------------------------------\
| SENSOR_RESPONSE:                                                   |
| As the arts WSV with the same name. Three main options exist:      |
| 1: The first option is to specify each variable associated with    |
| the sensor individually. This field can then be the name of a XML  |
| file or a Matlab sparse matrix. Other fields that must be          |
| specified for this option include: ANTENNA_DIM, MBLOCK_DLOS_GRID   |
| and SENSOR_RESPONSE_F/DLOS/POL.                                    |
| 2: The name of control files to include (ARTS_INCLUDES             |
| recognised). Given as an array of strings.                         |
| 3: Calculate the response from basic data. This field shall then   |
| be a structure, following the definitions in *qartsSensor*. Type   |
| 'qinfo(@qartsSensor);' for definition of required data             |
| fields.Other fields that must/may be specified here include        |
| ANTENNA_DIM and MBLOCK_DLOS_GRID.                                  |
\--------------------------------------------------------------------/


That is, you define the sensor by creating control file part, as other 
"include cfiles".

Bye,

Patrick Eriksson



On 07/31/15 05:44, 黄莹珠 wrote:
> Hi,all
>
> I am trying to use qpack to retrieve the AMSUB measurements which Ta
> polarization is quasi-Vertical pol with the form of
> Tv*cos(sensor_los)^2+Th*sin(sensor_los)^2.
>
> How to set the parameters of Q structure when I retrieve the atmosphere
> state of the quasi-Vertical polarization measurement provide?
>
> I  try to hack the art_oem.m with yf and J as follows:
>
> %--- Run ARTS --------------------------------------------------------
>
> if nargout == 3
>    do_j  = 1;
>    cfile = R.cfile_yj;
> else
>    do_j  = 0;
>    cfile = R.cfile_y;
> end
> %
> arts( cfile );
> %
> y2 = xmlLoad( fullfile( R.workfolder, 'y.xml' ) );
> I=y2(1:2:end);
> Q=y2(2:2:end);
> tv=I+Q;
> th=I-Q;
> y=tv*cos((180-theta)*pi/180)*cos((180-theta)*pi/180)+th*sin((180-theta)*pi/180)*sin((180-theta)*pi/180);
> %
> if do_j
>
>    % Load Jacobian
>    J2 = xmlLoad( fullfile( R.workfolder, 'jaco bian.xml' ) );
>    IJ=J2(1:2:end,:);
>    QJ=J2(2:2:end,:);
>    VJ=IJ+QJ;
>    HJ=IJ-QJ;
>    J=VJ;
>
> But I failed.
>
> If somebody can give me some advise?
>
>
> Thanks very much and best regards
>
> Huang Yingzhu
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
> qpack mailing list
> [email protected]
> https://www.sat.ltu.se/mailman/listinfo/qpack
>
_______________________________________________
qpack mailing list
[email protected]
https://www.sat.ltu.se/mailman/listinfo/qpack

Reply via email to