Re: [arts-users] Wind calculations (resent)

[Jana Mendrok]

Hi,

On Wed, Apr 3, 2019 at 4:30 PM Richard Larsson 
wrote:

> [...]
>
Indeed, the description seems coherent but difficult to follow.  Can you
> point out where in the docs you found this?  It should be updated to
> clearly read that a positive wind_v in 1D equates to a positive v in the
> standard (1 - v/c) Doppler shift expression.
>

I took my info from the built-in doc for wind_v_fields and sensor_los.
I did now update the wind field section in the AUG with an explicit
statement that positive/negative v-winds in 1D correspond to tail/head
winds, respectively (incl a reference to the AUG section Patrick pointed
out).


On Wed, Apr 3, 2019 at 6:40 PM Patrick Eriksson <
patrick.eriks...@chalmers.se> wrote:

> The definition of the winds do not change between 1D, 2D and 3D. That
> would at least lead to messy code.
>

I'm aware of this. And no critisism of that at all.


>
> This is in fact a question of the azimuth angle of the sensor. There is
> also a north direction in 1D and the sensor is assumed to be looking
> exactly North for 1D. This is explained in Sec 3.2 of AUG:
>
> 1D ... The sensor is assumed to by directed towards the North pole,
> corresponding to an azimuth angle of 0 ◦.
>
> Maybe few are looking in AUG so I just also added this information to
> the built-in doc of sensor_los.
>

I had indeed not checked the AUG, but used the built-in doc instead. Thanks
for adding the clarification there, Patrick!

However, the tricky part here (and on other occassions), particularly for
people not that familiar with ARTS, is to know what to look for and where
look for it specifically: for the 1D wind case, even with Patrick's
clarification, the crucial point is/was join the information available in
the wind_v_fields and the sensor_los/atm_dim documentation. I am not sure,
if I had taken a look into the AUG regarding that matter that this would
have included the atm dim (3.2) section.

But, that's why the mailinglist is so great! there are people out there
with better overview and better detail knowledge, who can answer questions
or at least point one in a (better) direction where to look for. :)
So thanks, Patrick & Richard (and everyone else contributing on other
occassions) for your valuable feedback!

Best wishes,
Jana


-- 
=
Jana Mendrok, Ph.D. (Geoscience)

Email: jana.mend...@gmail.com
Phone : +46 (0)708 860 729
=
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

Re: [arts-users] Wind calculations (resent)

[Patrick Eriksson]

Hi Jana, Hi all,

The definition of the winds do not change between 1D, 2D and 3D. That 
would at least lead to messy code.


This is in fact a question of the azimuth angle of the sensor. There is 
also a north direction in 1D and the sensor is assumed to be looking 
exactly North for 1D. This is explained in Sec 3.2 of AUG:


1D ... The sensor is assumed to by directed towards the North pole, 
corresponding to an azimuth angle of 0 ◦.


Maybe few are looking in AUG so I just also added this information to 
the built-in doc of sensor_los.


Bye,

Patrick



On 2019-04-03 17:08, Jana Mendrok wrote:

Hi,

thanks Richard for your feedback. I'm not sure, though, whether I really 
understand it (incl. wondering why you talk about jacobians here...). 
*scratches head*


we're trying to pin down where the wind is blowing in a 1D atm setup. 1D 
allows us to have vertical (w) and horizontal (v) winds that affect the 
signal. let's forget about the vertical, that one is clear. for the 
horizontal, there is two possible wind directions in a 1D case - a head 
wind (aka blowing in the observer's face) and a tail wind (aka blowing 
from the observer's back). which of these corresponds to a positive, 
which to a negative wind(_v) speed.


starting from your "For the 'absolute' wind speed option, a positive 
retrieved sign is equivalent to a red-shift" and equating a red-shift 
with observer moving away from source (or source from observer...), a 
positive wind should mean a tail wind.


which actually seems indeed in agreement with info I dig from the ARTS 
documentation (not that easy to find and combine for a beginner...): 
wind_v doc says positive v-winds are winds blowing south to north (which 
alone isn't very helpful for 1D setups - there is no north in 1D...); 
sensor_los doc says for the 2D case (allowing positive and negative 
zenith angles - in order to distinguish two possible viewing directions) 
that positive angles are equivalent to viewing towards higher latitudes, 
ie towards north. assuming this is valid for 1D, too (ie in 1D the 
observer always looks towards north) means for the 1D case positive 
winds (blowing to north) equate tail winds (for an observer looking 
north), negative winds are head winds.


to confirm this, Rita, you could have a look where the line peak moves 
(prefereably in a high-spectral resolution simulation) for positive and 
negative winds, respectively. for positive it should move to lower 
freqs, for negative to higher.


best wishes,
Jana

On Wed, Apr 3, 2019 at 12:03 PM Richard Larsson > wrote:


Hi Rita,

For the directional components, the sign of a retrieved component
with the ARTS Jacobian is positive along the axes themselves.  The
direction the sensor is looking only influences the scale of the
Jacobian, not the sign of the direction of the retrieval
calculation.  If you look parallel or anti-parallel to the field,
the values inside the Jacobian switch signs accordingly to work on
the external field.  As an example, you should be able to make
measurements at several different directions and, if you assume the
field is uniform, retrieve just one field without any transformation
on the components themselves inside the calculations.

For the 'absolute' wind speed option, a positive retrieved sign is
equivalent to a red-shift, or a Doppler shift below unity if you
wish.  This is much messier to work with during ARTS iterations, so
make sure your signal is easy to interpret if you use it.  This is
of course closer to the physics of what is going on with the signal
itself, since all we really can see of the wind is how much it is
blue- or red-shifting the signal.

With hope,
//Richard

Ps. Oliver, please sent another email when it is fixed.



Den ons 3 apr. 2019 kl 13:36 skrev Oliver Lemke
mailto:oliver.le...@uni-hamburg.de>>:

Hi all,

Sorry if you received the mail below already. We're currently
investigating a mail delivery problem to Gmail adresses. If
you've already received the mail from Rita yesterday please
disregard this mail.

Cheers,
Oliver


 > From: Rita Edit Kajtar mailto:rita.edit.kaj...@ltu.se>>
 > Subject: Wind calculations
 > Date: 2 April 2019 at 10:36:54 CEST
 > To: "arts_users.mi@lists.uni-hamburg.de
"
mailto:arts_users.mi@lists.uni-hamburg.de>>
 >
 >
 > Hello!
 >
 >
 > I have a short question regarding wind calculations.
 >
 > What is the general convention assumed for determining the
sign of the wind speeds having the measuring instrument as the
reference point? Are the winds blowing towards the instrument
considered positive and those blowing away from the instrument
negative or the other 

Re: [arts-users] Wind calculations (resent)

[Richard Larsson]

Hi Jana,

Den ons 3 apr. 2019 kl 17:08 skrev Jana Mendrok :

> Hi,
>
> thanks Richard for your feedback. I'm not sure, though, whether I really
> understand it (incl. wondering why you talk about jacobians here...).
> *scratches head*
>

I might have misunderstood the question slightly, but the easiest way to
understand this in the code is in relation to what a standard Jacobian
means, since it must relate to you the assumed internal signs and geometries


> we're trying to pin down where the wind is blowing in a 1D atm setup. 1D
> allows us to have vertical (w) and horizontal (v) winds that affect the
> signal. let's forget about the vertical, that one is clear. for the
> horizontal, there is two possible wind directions in a 1D case - a head
> wind (aka blowing in the observer's face) and a tail wind (aka blowing from
> the observer's back). which of these corresponds to a positive, which to a
> negative wind(_v) speed.
>

The documentation is too difficult for me to follow.  In the code, the
definition in the 1D case is the same as for the 3D fields.  Assuming
wind_u==0 is enforced somehow (I seldom use 1D geometry), a negative wind_v
value has a negative speed in the Doppler shift expression -- a blue shift
or a headwind -- and vice verse for a positive wind_v.


> starting from your "For the 'absolute' wind speed option, a positive
> retrieved sign is equivalent to a red-shift" and equating a red-shift with
> observer moving away from source (or source from observer...), a positive
> wind should mean a tail wind.
>

Yes.


> which actually seems indeed in agreement with info I dig from the ARTS
> documentation (not that easy to find and combine for a beginner...): wind_v
> doc says positive v-winds are winds blowing south to north (which alone
> isn't very helpful for 1D setups - there is no north in 1D...); sensor_los
> doc says for the 2D case (allowing positive and negative zenith angles - in
> order to distinguish two possible viewing directions) that positive angles
> are equivalent to viewing towards higher latitudes, ie towards north.
> assuming this is valid for 1D, too (ie in 1D the observer always looks
> towards north) means for the 1D case positive winds (blowing to north)
> equate tail winds (for an observer looking north), negative winds are head
> winds.
>

Indeed, the description seems coherent but difficult to follow.  Can you
point out where in the docs you found this?  It should be updated to
clearly read that a positive wind_v in 1D equates to a positive v in the
standard (1 - v/c) Doppler shift expression.

With hope,
//Richard

>
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

Re: [arts-users] Wind calculations (resent)

[Jana Mendrok]

Hi,

thanks Richard for your feedback. I'm not sure, though, whether I really
understand it (incl. wondering why you talk about jacobians here...).
*scratches head*

we're trying to pin down where the wind is blowing in a 1D atm setup. 1D
allows us to have vertical (w) and horizontal (v) winds that affect the
signal. let's forget about the vertical, that one is clear. for the
horizontal, there is two possible wind directions in a 1D case - a head
wind (aka blowing in the observer's face) and a tail wind (aka blowing from
the observer's back). which of these corresponds to a positive, which to a
negative wind(_v) speed.

starting from your "For the 'absolute' wind speed option, a positive
retrieved sign is equivalent to a red-shift" and equating a red-shift with
observer moving away from source (or source from observer...), a positive
wind should mean a tail wind.

which actually seems indeed in agreement with info I dig from the ARTS
documentation (not that easy to find and combine for a beginner...): wind_v
doc says positive v-winds are winds blowing south to north (which alone
isn't very helpful for 1D setups - there is no north in 1D...); sensor_los
doc says for the 2D case (allowing positive and negative zenith angles - in
order to distinguish two possible viewing directions) that positive angles
are equivalent to viewing towards higher latitudes, ie towards north.
assuming this is valid for 1D, too (ie in 1D the observer always looks
towards north) means for the 1D case positive winds (blowing to north)
equate tail winds (for an observer looking north), negative winds are head
winds.

to confirm this, Rita, you could have a look where the line peak moves
(prefereably in a high-spectral resolution simulation) for positive and
negative winds, respectively. for positive it should move to lower freqs,
for negative to higher.

best wishes,
Jana

On Wed, Apr 3, 2019 at 12:03 PM Richard Larsson 
wrote:

> Hi Rita,
>
> For the directional components, the sign of a retrieved component with the
> ARTS Jacobian is positive along the axes themselves.  The direction the
> sensor is looking only influences the scale of the Jacobian, not the sign
> of the direction of the retrieval calculation.  If you look parallel or
> anti-parallel to the field, the values inside the Jacobian switch signs
> accordingly to work on the external field.  As an example, you should be
> able to make measurements at several different directions and, if you
> assume the field is uniform, retrieve just one field without any
> transformation on the components themselves inside the calculations.
>
> For the 'absolute' wind speed option, a positive retrieved sign is
> equivalent to a red-shift, or a Doppler shift below unity if you wish.
> This is much messier to work with during ARTS iterations, so make sure your
> signal is easy to interpret if you use it.  This is of course closer to the
> physics of what is going on with the signal itself, since all we really can
> see of the wind is how much it is blue- or red-shifting the signal.
>
> With hope,
> //Richard
>
> Ps. Oliver, please sent another email when it is fixed.
>
>
>
> Den ons 3 apr. 2019 kl 13:36 skrev Oliver Lemke <
> oliver.le...@uni-hamburg.de>:
>
>> Hi all,
>>
>> Sorry if you received the mail below already. We're currently
>> investigating a mail delivery problem to Gmail adresses. If you've already
>> received the mail from Rita yesterday please disregard this mail.
>>
>> Cheers,
>> Oliver
>>
>>
>> > From: Rita Edit Kajtar 
>> > Subject: Wind calculations
>> > Date: 2 April 2019 at 10:36:54 CEST
>> > To: "arts_users.mi@lists.uni-hamburg.de" <
>> arts_users.mi@lists.uni-hamburg.de>
>> >
>> >
>> > Hello!
>> >
>> >
>> > I have a short question regarding wind calculations.
>> >
>> > What is the general convention assumed for determining the sign of the
>> wind speeds having the measuring instrument as the reference point? Are the
>> winds blowing towards the instrument considered positive and those blowing
>> away from the instrument negative or the other way around?
>> >
>> > The ARTS documentation provides information for how the sign should be
>> considered regarding the four cardinal directions, but not relative to a
>> measuring device.
>> >
>> >
>> > Best regards,
>> > Rita Kajtar
>>
>>
>> ___
>> arts_users.mi mailing list
>> arts_users.mi@lists.uni-hamburg.de
>> https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
>>
> ___
> arts_users.mi mailing list
> arts_users.mi@lists.uni-hamburg.de
> https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
>


-- 
=
Jana Mendrok, Ph.D. (Geoscience)

Email: jana.mend...@gmail.com
Phone : +46 (0)708 860 729
=
___
arts_users.mi mailing list

Re: [arts-users] Wind calculations (resent)

[Oliver Lemke]

Hi Richard,

> On 3 Apr 2019, at 14:03, Richard Larsson  wrote:
> 
> Ps. Oliver, please sent another email when it is fixed.

Since you (and Jana, Lukas and I) received this mail via gmail, I have to 
assume it is fixed. Although it is possible that Google blocks mails depending 
on the sender address. We had that issue some time ago when @yahoo mails that 
went through a mailing list where blocked by Google. We'll find out when Rita 
replies to your mail. I'll keep you posted.

Cheers,
Oliver



smime.p7s
Description: S/MIME cryptographic signature
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

Re: [arts-users] Wind calculations (resent)

[Richard Larsson]

Hi Rita,

For the directional components, the sign of a retrieved component with the
ARTS Jacobian is positive along the axes themselves.  The direction the
sensor is looking only influences the scale of the Jacobian, not the sign
of the direction of the retrieval calculation.  If you look parallel or
anti-parallel to the field, the values inside the Jacobian switch signs
accordingly to work on the external field.  As an example, you should be
able to make measurements at several different directions and, if you
assume the field is uniform, retrieve just one field without any
transformation on the components themselves inside the calculations.

For the 'absolute' wind speed option, a positive retrieved sign is
equivalent to a red-shift, or a Doppler shift below unity if you wish.
This is much messier to work with during ARTS iterations, so make sure your
signal is easy to interpret if you use it.  This is of course closer to the
physics of what is going on with the signal itself, since all we really can
see of the wind is how much it is blue- or red-shifting the signal.

With hope,
//Richard

Ps. Oliver, please sent another email when it is fixed.



Den ons 3 apr. 2019 kl 13:36 skrev Oliver Lemke :

> Hi all,
>
> Sorry if you received the mail below already. We're currently
> investigating a mail delivery problem to Gmail adresses. If you've already
> received the mail from Rita yesterday please disregard this mail.
>
> Cheers,
> Oliver
>
>
> > From: Rita Edit Kajtar 
> > Subject: Wind calculations
> > Date: 2 April 2019 at 10:36:54 CEST
> > To: "arts_users.mi@lists.uni-hamburg.de" <
> arts_users.mi@lists.uni-hamburg.de>
> >
> >
> > Hello!
> >
> >
> > I have a short question regarding wind calculations.
> >
> > What is the general convention assumed for determining the sign of the
> wind speeds having the measuring instrument as the reference point? Are the
> winds blowing towards the instrument considered positive and those blowing
> away from the instrument negative or the other way around?
> >
> > The ARTS documentation provides information for how the sign should be
> considered regarding the four cardinal directions, but not relative to a
> measuring device.
> >
> >
> > Best regards,
> > Rita Kajtar
>
>
> ___
> arts_users.mi mailing list
> arts_users.mi@lists.uni-hamburg.de
> https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi
>
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

[arts-users] Wind calculations (resent)

[Oliver Lemke]

Hi all,

Sorry if you received the mail below already. We're currently investigating a 
mail delivery problem to Gmail adresses. If you've already received the mail 
from Rita yesterday please disregard this mail.

Cheers,
Oliver


> From: Rita Edit Kajtar 
> Subject: Wind calculations
> Date: 2 April 2019 at 10:36:54 CEST
> To: "arts_users.mi@lists.uni-hamburg.de" 
> 
> 
> Hello! 
> 
> 
> I have a short question regarding wind calculations. 
> 
> What is the general convention assumed for determining the sign of the wind 
> speeds having the measuring instrument as the reference point? Are the winds 
> blowing towards the instrument considered positive and those blowing away 
> from the instrument negative or the other way around?  
> 
> The ARTS documentation provides information for how the sign should be 
> considered regarding the four cardinal directions, but not relative to a 
> measuring device. 
> 
> 
> Best regards, 
> Rita Kajtar




smime.p7s
Description: S/MIME cryptographic signature
___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi

[arts-users] Wind calculations

[Rita Edit Kajtar]

Hello!



I have a short question regarding wind calculations.

What is the general convention assumed for determining the sign of the wind 
speeds having the measuring instrument as the reference point? Are the winds 
blowing towards the instrument considered positive and those blowing away from 
the instrument negative or the other way around?

The ARTS documentation provides information for how the sign should be 
considered regarding the four cardinal directions, but not relative to a 
measuring device.


Best regards,
Rita Kajtar

___
arts_users.mi mailing list
arts_users.mi@lists.uni-hamburg.de
https://mailman.rrz.uni-hamburg.de/mailman/listinfo/arts_users.mi