[Met_help] [rt.rap.ucar.edu #94201] History for MODE Time Domain outputs and attributes

John Halley Gotway via RT met_help at ucar.edu
Wed Mar 4 09:50:26 MST 2020


----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Hi,

I have some questions about the information of MODE Time Domain output. In
aranda_data I put the  MTD output (the ascii files), the MTD configure file
and some mode outputs as examples (postscript files).

1) Attribute units

The MET User's guide (V.7) says that in MTD the unit of length is the grid
resolution, and the unit of time is whatever the time separation between
the input files is. Speed and velocity are thus in grid units per time
unit. For other hand, The "grid_res" entry is the nominal spacing for each
grid square in kilometers (MET User's guide)

          a) So,  I understand that kilometers are  the units for
CENTROID_X and          CENTROID_Y in the *3d_single_cluster.txt and
*3d_single_simple.txt, but I'm not sure. And I have the same question for
SPACE_CENTROID_DIST in *3d_pair_simple.txt and  *3d_pair_cluster.txt

          b) I work with precipitation accumulation in 3 hours. Is the hour
the unit of time for the CENTROID_T? and for TIME_CENTROID_DELTA?

         c) What's about the unit for velocity (X_DOT, Y_DOT, SPEED_DELTA)?
is the grid unit/hour the velocity unit? ; and for the direction of
movement (DIRECTION_DIFF, AXIS_DIFF, AXIS_ANG), are the degrees units?

2) AXIS ANGLE
If I have a negative value for this attribute, it's means that the objects
moves backward with the time, is it correct?

3)  CDIST_TRAVELLED
I don't understand why I have CDIST_TRAVELLED = 0 in the
*3d_single_cluster.txt and *3d_single_simple.txt

4) INTEREST
In *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and for
CF002_COOO2 is 0.85. but the interest for these clusters  in the
*3d_pair_cluster.txt are -1. And I don't understand why.


I know there are a lot of questions but I will be very grateful for your
help!

Natalí


----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: MODE Time Domain outputs and attributes
From: Randy Bullock
Time: Tue Feb 18 11:37:30 2020

Hi Natalie -

Thanks for emailing us.  I'll try to answer your questions as best I
can.

1)  Attribute Units
   a)  CENTROID_X and CENTROID_y values are in grid units, not
kilometers.
Similarly, SPACE_CENTROID_DIST is also in grid units.  If you want to
calculate true distances in kilometers between centroids, then it's
best to
use the CENTROID_LAT and CENTROID_LON values for the two objects, and
then
calculate (for example) the great circle distance using those lat/lon
values.

   b)  The time unit is the difference in valid times between the 2D
data
files that you give as input to Mode Time Domain (MTD).  That's what
determines the time unit, not accumulation times.

   c)  Speeds and velocities are given in grid units per time unit.
Axis
angle values (amd difference values) are in degrees, and are measured
with
respect to the grid x and y axes.  Note that the (spatial) axis does
not
give direction of object movement,  For that, use X_DOT and Y_DOT.
Spatial
axis information relates the the orientation of the object with
respect to
the grid directions.

2)  Axis angle
   This relates to part c) above.  You're thinking that the axis gives
direction of movement,and that's not correct.  For that, use X_DOT and
Y_DOT.

3)  CDIST_TRAVELLED
    The reason this is always zero is because it's a bug in the MTD
code.
It will be fixed in the upcoming release.  Apologies for that.

4)  Interest
    To answer this I'd have to look at your data and run MTD myself to
look
at the output.  If you'd like for me to do that, let me know, and I'll
tell
you how to send your data and config files to us.  We're currently
working
to get the next release of MET out, though, and that's taking up all
our
time.  I'm perfectly willing to do this, if you want, just don't
expect a
super-quick reply because we're really busy right now.


Hope this helps!  If you have any more questions, or want further
explanation, please don't hesitate to email again!

Randy Bullock

On Mon, Feb 17, 2020 at 2:47 PM natali aranda via RT
<met_help at ucar.edu>
wrote:

>
> Mon Feb 17 14:47:05 2020: Request 94201 was acted upon.
> Transaction: Ticket created by natali.g.aranda at gmail.com
>        Queue: met_help
>      Subject: MODE Time Domain outputs and attributes
>        Owner: Nobody
>   Requestors: natali.g.aranda at gmail.com
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201 >
>
>
> Hi,
>
> I have some questions about the information of MODE Time Domain
output. In
> aranda_data I put the  MTD output (the ascii files), the MTD
configure file
> and some mode outputs as examples (postscript files).
>
> 1) Attribute units
>
> The MET User's guide (V.7) says that in MTD the unit of length is
the grid
> resolution, and the unit of time is whatever the time separation
between
> the input files is. Speed and velocity are thus in grid units per
time
> unit. For other hand, The "grid_res" entry is the nominal spacing
for each
> grid square in kilometers (MET User's guide)
>
>           a) So,  I understand that kilometers are  the units for
> CENTROID_X and          CENTROID_Y in the *3d_single_cluster.txt and
> *3d_single_simple.txt, but I'm not sure. And I have the same
question for
> SPACE_CENTROID_DIST in *3d_pair_simple.txt and  *3d_pair_cluster.txt
>
>           b) I work with precipitation accumulation in 3 hours. Is
the hour
> the unit of time for the CENTROID_T? and for TIME_CENTROID_DELTA?
>
>          c) What's about the unit for velocity (X_DOT, Y_DOT,
SPEED_DELTA)?
> is the grid unit/hour the velocity unit? ; and for the direction of
> movement (DIRECTION_DIFF, AXIS_DIFF, AXIS_ANG), are the degrees
units?
>
> 2) AXIS ANGLE
> If I have a negative value for this attribute, it's means that the
objects
> moves backward with the time, is it correct?
>
> 3)  CDIST_TRAVELLED
> I don't understand why I have CDIST_TRAVELLED = 0 in the
> *3d_single_cluster.txt and *3d_single_simple.txt
>
> 4) INTEREST
> In *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and
for
> CF002_COOO2 is 0.85. but the interest for these clusters  in the
> *3d_pair_cluster.txt are -1. And I don't understand why.
>
>
> I know there are a lot of questions but I will be very grateful for
your
> help!
>
> Natalí
>
>

------------------------------------------------
Subject: MODE Time Domain outputs and attributes
From: natali aranda
Time: Tue Feb 18 20:44:23 2020

Thanks Randy!

your answers helped a lot.

And now I have a new question. I use MODE and MTD to verify  the WRF
model
in South America, particularly, Argentina.  In MODE output, the
CENTROID_LAT and CENTROID_LON values are negative and it's ok because
the
objects are in Argentina ( in the south and in the west of the globe)
but
in MTD, the CENTROID_LON are positive. Is there any possibility that
MTD
forgot the minus sign for the longitude?

Also, I understand that CENTROID_LAT and CENTROID_LON in MTD are valid
during the life time of the object. For example, an object appears in
the
3rd step and finishes in the 9th step: does MTD do the average to  get
the
CENTROID_LAT and CENTROID_LON in the  *3d_single_simple.txt?

Thanks,
Natali

El mar., 18 feb. 2020 a las 15:37, Randy Bullock via RT
(<met_help at ucar.edu>)
escribió:

> Hi Natalie -
>
> Thanks for emailing us.  I'll try to answer your questions as best I
can.
>
> 1)  Attribute Units
>    a)  CENTROID_X and CENTROID_y values are in grid units, not
kilometers.
> Similarly, SPACE_CENTROID_DIST is also in grid units.  If you want
to
> calculate true distances in kilometers between centroids, then it's
best to
> use the CENTROID_LAT and CENTROID_LON values for the two objects,
and then
> calculate (for example) the great circle distance using those
lat/lon
> values.
>
>    b)  The time unit is the difference in valid times between the 2D
data
> files that you give as input to Mode Time Domain (MTD).  That's what
> determines the time unit, not accumulation times.
>
>    c)  Speeds and velocities are given in grid units per time unit.
Axis
> angle values (amd difference values) are in degrees, and are
measured with
> respect to the grid x and y axes.  Note that the (spatial) axis does
not
> give direction of object movement,  For that, use X_DOT and Y_DOT.
Spatial
> axis information relates the the orientation of the object with
respect to
> the grid directions.
>
> 2)  Axis angle
>    This relates to part c) above.  You're thinking that the axis
gives
> direction of movement,and that's not correct.  For that, use X_DOT
and
> Y_DOT.
>
> 3)  CDIST_TRAVELLED
>     The reason this is always zero is because it's a bug in the MTD
code.
> It will be fixed in the upcoming release.  Apologies for that.
>
> 4)  Interest
>     To answer this I'd have to look at your data and run MTD myself
to look
> at the output.  If you'd like for me to do that, let me know, and
I'll tell
> you how to send your data and config files to us.  We're currently
working
> to get the next release of MET out, though, and that's taking up all
our
> time.  I'm perfectly willing to do this, if you want, just don't
expect a
> super-quick reply because we're really busy right now.
>
>
> Hope this helps!  If you have any more questions, or want further
> explanation, please don't hesitate to email again!
>
> Randy Bullock
>
> On Mon, Feb 17, 2020 at 2:47 PM natali aranda via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Mon Feb 17 14:47:05 2020: Request 94201 was acted upon.
> > Transaction: Ticket created by natali.g.aranda at gmail.com
> >        Queue: met_help
> >      Subject: MODE Time Domain outputs and attributes
> >        Owner: Nobody
> >   Requestors: natali.g.aranda at gmail.com
> >       Status: new
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201 >
> >
> >
> > Hi,
> >
> > I have some questions about the information of MODE Time Domain
output.
> In
> > aranda_data I put the  MTD output (the ascii files), the MTD
configure
> file
> > and some mode outputs as examples (postscript files).
> >
> > 1) Attribute units
> >
> > The MET User's guide (V.7) says that in MTD the unit of length is
the
> grid
> > resolution, and the unit of time is whatever the time separation
between
> > the input files is. Speed and velocity are thus in grid units per
time
> > unit. For other hand, The "grid_res" entry is the nominal spacing
for
> each
> > grid square in kilometers (MET User's guide)
> >
> >           a) So,  I understand that kilometers are  the units for
> > CENTROID_X and          CENTROID_Y in the *3d_single_cluster.txt
and
> > *3d_single_simple.txt, but I'm not sure. And I have the same
question for
> > SPACE_CENTROID_DIST in *3d_pair_simple.txt and
*3d_pair_cluster.txt
> >
> >           b) I work with precipitation accumulation in 3 hours. Is
the
> hour
> > the unit of time for the CENTROID_T? and for TIME_CENTROID_DELTA?
> >
> >          c) What's about the unit for velocity (X_DOT, Y_DOT,
> SPEED_DELTA)?
> > is the grid unit/hour the velocity unit? ; and for the direction
of
> > movement (DIRECTION_DIFF, AXIS_DIFF, AXIS_ANG), are the degrees
units?
> >
> > 2) AXIS ANGLE
> > If I have a negative value for this attribute, it's means that the
> objects
> > moves backward with the time, is it correct?
> >
> > 3)  CDIST_TRAVELLED
> > I don't understand why I have CDIST_TRAVELLED = 0 in the
> > *3d_single_cluster.txt and *3d_single_simple.txt
> >
> > 4) INTEREST
> > In *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and
for
> > CF002_COOO2 is 0.85. but the interest for these clusters  in the
> > *3d_pair_cluster.txt are -1. And I don't understand why.
> >
> >
> > I know there are a lot of questions but I will be very grateful
for your
> > help!
> >
> > Natalí
> >
> >
>
>

------------------------------------------------
Subject: MODE Time Domain outputs and attributes
From: Randy Bullock
Time: Wed Feb 19 15:24:32 2020

Hi again, Natalie.  Hope all's well with you.

There could very well be a bug in the MTD code for writing longitudes.
Internally, the map projection code considers west longitude positive,
so
we have to remember each time we write out a longitude to toggle the
sign.
Perhaps we forgot.  Wouldn't be the first time.  I'll look at the code
and
if it is wrong, I'll fix it for the upcoming release.

As for the 3D CENTROID_LAT and CENTROID_LON, they can be looked at as
averages of the related 2D attributes, sort of.  That's not quite what
they
are, however.  The 3D CENTROID_X is the average value, over the entire
3D
spacetime object, of the x coordinate.  Similarly for CENTROID_Y and
CENTROID_T.  Then for CENTROID_LAT and CENTROID_LON, we just use the
grid
projection information to translate CENTROID_X and CENTROID_Y as grid
(x,
y) coordinates and calculate the corresponding lat, lon.

Hope this helps,  And, as always, thanks for writing.

Randy

------------------------------------------------
Subject: MODE Time Domain outputs and attributes
From: natali aranda
Time: Mon Mar 02 18:59:34 2020

Hi Randy,

I'm writing again...
off course, your explanations helped me to understand the MTD tool.

Do you remember that I have a problem with the interest attribute (In
*3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and for
CF002_COOO2 is 0.85. but the interest for these clusters  in the
*3d_pair_cluster.txt are -1)? I'm trying to put the observation files
and
forecast files in the *incoming/irap/met_help/aranda_data. *I have 16
files
for the observation and 16 files for the forecast. Could you tell me
how
send the information?

In incoming/irap/met_help/aranda_data, I put the configuration file

Thanks,
Natalí



El mar., 18 feb. 2020 a las 15:37, Randy Bullock via RT
(<met_help at ucar.edu>)
escribió:

> Hi Natalie -
>
> Thanks for emailing us.  I'll try to answer your questions as best I
can.
>
> 1)  Attribute Units
>    a)  CENTROID_X and CENTROID_y values are in grid units, not
kilometers.
> Similarly, SPACE_CENTROID_DIST is also in grid units.  If you want
to
> calculate true distances in kilometers between centroids, then it's
best to
> use the CENTROID_LAT and CENTROID_LON values for the two objects,
and then
> calculate (for example) the great circle distance using those
lat/lon
> values.
>
>    b)  The time unit is the difference in valid times between the 2D
data
> files that you give as input to Mode Time Domain (MTD).  That's what
> determines the time unit, not accumulation times.
>
>    c)  Speeds and velocities are given in grid units per time unit.
Axis
> angle values (amd difference values) are in degrees, and are
measured with
> respect to the grid x and y axes.  Note that the (spatial) axis does
not
> give direction of object movement,  For that, use X_DOT and Y_DOT.
Spatial
> axis information relates the the orientation of the object with
respect to
> the grid directions.
>
> 2)  Axis angle
>    This relates to part c) above.  You're thinking that the axis
gives
> direction of movement,and that's not correct.  For that, use X_DOT
and
> Y_DOT.
>
> 3)  CDIST_TRAVELLED
>     The reason this is always zero is because it's a bug in the MTD
code.
> It will be fixed in the upcoming release.  Apologies for that.
>
> 4)  Interest
>     To answer this I'd have to look at your data and run MTD myself
to look
> at the output.  If you'd like for me to do that, let me know, and
I'll tell
> you how to send your data and config files to us.  We're currently
working
> to get the next release of MET out, though, and that's taking up all
our
> time.  I'm perfectly willing to do this, if you want, just don't
expect a
> super-quick reply because we're really busy right now.
>
>
> Hope this helps!  If you have any more questions, or want further
> explanation, please don't hesitate to email again!
>
> Randy Bullock
>
> On Mon, Feb 17, 2020 at 2:47 PM natali aranda via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Mon Feb 17 14:47:05 2020: Request 94201 was acted upon.
> > Transaction: Ticket created by natali.g.aranda at gmail.com
> >        Queue: met_help
> >      Subject: MODE Time Domain outputs and attributes
> >        Owner: Nobody
> >   Requestors: natali.g.aranda at gmail.com
> >       Status: new
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201 >
> >
> >
> > Hi,
> >
> > I have some questions about the information of MODE Time Domain
output.
> In
> > aranda_data I put the  MTD output (the ascii files), the MTD
configure
> file
> > and some mode outputs as examples (postscript files).
> >
> > 1) Attribute units
> >
> > The MET User's guide (V.7) says that in MTD the unit of length is
the
> grid
> > resolution, and the unit of time is whatever the time separation
between
> > the input files is. Speed and velocity are thus in grid units per
time
> > unit. For other hand, The "grid_res" entry is the nominal spacing
for
> each
> > grid square in kilometers (MET User's guide)
> >
> >           a) So,  I understand that kilometers are  the units for
> > CENTROID_X and          CENTROID_Y in the *3d_single_cluster.txt
and
> > *3d_single_simple.txt, but I'm not sure. And I have the same
question for
> > SPACE_CENTROID_DIST in *3d_pair_simple.txt and
*3d_pair_cluster.txt
> >
> >           b) I work with precipitation accumulation in 3 hours. Is
the
> hour
> > the unit of time for the CENTROID_T? and for TIME_CENTROID_DELTA?
> >
> >          c) What's about the unit for velocity (X_DOT, Y_DOT,
> SPEED_DELTA)?
> > is the grid unit/hour the velocity unit? ; and for the direction
of
> > movement (DIRECTION_DIFF, AXIS_DIFF, AXIS_ANG), are the degrees
units?
> >
> > 2) AXIS ANGLE
> > If I have a negative value for this attribute, it's means that the
> objects
> > moves backward with the time, is it correct?
> >
> > 3)  CDIST_TRAVELLED
> > I don't understand why I have CDIST_TRAVELLED = 0 in the
> > *3d_single_cluster.txt and *3d_single_simple.txt
> >
> > 4) INTEREST
> > In *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and
for
> > CF002_COOO2 is 0.85. but the interest for these clusters  in the
> > *3d_pair_cluster.txt are -1. And I don't understand why.
> >
> >
> > I know there are a lot of questions but I will be very grateful
for your
> > help!
> >
> > Natalí
> >
> >
>
>

------------------------------------------------
Subject: MODE Time Domain outputs and attributes
From: John Halley Gotway
Time: Mon Mar 02 19:55:29 2020

Natali,

Randy is out of the office for the next 2 weeks. But I can answer your
question.

Matching/merging is done in MODE on the simple objects from the fcst
and
obs fields.  An interest value is computed for each fcst/obs pair. The
result of that process is cluster fcst and obs objects.  But it is the
interest values of the simple objects which determine the
matching/merging.  As an extra step, MODE also recomputes interest
values
for the clusters... even though they don’t impact the logic in any
way.

The same logic is used in MTD as well.  However MTD does NOT recompute
the
interest values for the matched clusters.  Randy wrote -1 as the
interest
value for those lines as a flag value.  I believe it’s not being done
in
MTD to get it to runs as fast as possible.

Am I correct in thinking that you’d like MTD to be modified to
recompute
the interest values for matched clusters?

Thanks,
John Halley Gotway

On Mon, Mar 2, 2020 at 7:00 PM natali aranda via RT
<met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201 >
>
> Hi Randy,
>
> I'm writing again...
> off course, your explanations helped me to understand the MTD tool.
>
> Do you remember that I have a problem with the interest attribute
(In
> *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and for
> CF002_COOO2 is 0.85. but the interest for these clusters  in the
> *3d_pair_cluster.txt are -1)? I'm trying to put the observation
files and
> forecast files in the *incoming/irap/met_help/aranda_data. *I have
16 files
> for the observation and 16 files for the forecast. Could you tell me
how
> send the information?
>
> In incoming/irap/met_help/aranda_data, I put the configuration file
>
> Thanks,
> Natalí
>
>
>
> El mar., 18 feb. 2020 a las 15:37, Randy Bullock via RT (<
> met_help at ucar.edu>)
> escribió:
>
> > Hi Natalie -
> >
> > Thanks for emailing us.  I'll try to answer your questions as best
I can.
> >
> > 1)  Attribute Units
> >    a)  CENTROID_X and CENTROID_y values are in grid units, not
> kilometers.
> > Similarly, SPACE_CENTROID_DIST is also in grid units.  If you want
to
> > calculate true distances in kilometers between centroids, then
it's best
> to
> > use the CENTROID_LAT and CENTROID_LON values for the two objects,
and
> then
> > calculate (for example) the great circle distance using those
lat/lon
> > values.
> >
> >    b)  The time unit is the difference in valid times between the
2D data
> > files that you give as input to Mode Time Domain (MTD).  That's
what
> > determines the time unit, not accumulation times.
> >
> >    c)  Speeds and velocities are given in grid units per time
unit.  Axis
> > angle values (amd difference values) are in degrees, and are
measured
> with
> > respect to the grid x and y axes.  Note that the (spatial) axis
does not
> > give direction of object movement,  For that, use X_DOT and Y_DOT.
> Spatial
> > axis information relates the the orientation of the object with
respect
> to
> > the grid directions.
> >
> > 2)  Axis angle
> >    This relates to part c) above.  You're thinking that the axis
gives
> > direction of movement,and that's not correct.  For that, use X_DOT
and
> > Y_DOT.
> >
> > 3)  CDIST_TRAVELLED
> >     The reason this is always zero is because it's a bug in the
MTD code.
> > It will be fixed in the upcoming release.  Apologies for that.
> >
> > 4)  Interest
> >     To answer this I'd have to look at your data and run MTD
myself to
> look
> > at the output.  If you'd like for me to do that, let me know, and
I'll
> tell
> > you how to send your data and config files to us.  We're currently
> working
> > to get the next release of MET out, though, and that's taking up
all our
> > time.  I'm perfectly willing to do this, if you want, just don't
expect a
> > super-quick reply because we're really busy right now.
> >
> >
> > Hope this helps!  If you have any more questions, or want further
> > explanation, please don't hesitate to email again!
> >
> > Randy Bullock
> >
> > On Mon, Feb 17, 2020 at 2:47 PM natali aranda via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > Mon Feb 17 14:47:05 2020: Request 94201 was acted upon.
> > > Transaction: Ticket created by natali.g.aranda at gmail.com
> > >        Queue: met_help
> > >      Subject: MODE Time Domain outputs and attributes
> > >        Owner: Nobody
> > >   Requestors: natali.g.aranda at gmail.com
> > >       Status: new
> > >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201
> >
> > >
> > >
> > > Hi,
> > >
> > > I have some questions about the information of MODE Time Domain
output.
> > In
> > > aranda_data I put the  MTD output (the ascii files), the MTD
configure
> > file
> > > and some mode outputs as examples (postscript files).
> > >
> > > 1) Attribute units
> > >
> > > The MET User's guide (V.7) says that in MTD the unit of length
is the
> > grid
> > > resolution, and the unit of time is whatever the time separation
> between
> > > the input files is. Speed and velocity are thus in grid units
per time
> > > unit. For other hand, The "grid_res" entry is the nominal
spacing for
> > each
> > > grid square in kilometers (MET User's guide)
> > >
> > >           a) So,  I understand that kilometers are  the units
for
> > > CENTROID_X and          CENTROID_Y in the *3d_single_cluster.txt
and
> > > *3d_single_simple.txt, but I'm not sure. And I have the same
question
> for
> > > SPACE_CENTROID_DIST in *3d_pair_simple.txt and
*3d_pair_cluster.txt
> > >
> > >           b) I work with precipitation accumulation in 3 hours.
Is the
> > hour
> > > the unit of time for the CENTROID_T? and for
TIME_CENTROID_DELTA?
> > >
> > >          c) What's about the unit for velocity (X_DOT, Y_DOT,
> > SPEED_DELTA)?
> > > is the grid unit/hour the velocity unit? ; and for the direction
of
> > > movement (DIRECTION_DIFF, AXIS_DIFF, AXIS_ANG), are the degrees
units?
> > >
> > > 2) AXIS ANGLE
> > > If I have a negative value for this attribute, it's means that
the
> > objects
> > > moves backward with the time, is it correct?
> > >
> > > 3)  CDIST_TRAVELLED
> > > I don't understand why I have CDIST_TRAVELLED = 0 in the
> > > *3d_single_cluster.txt and *3d_single_simple.txt
> > >
> > > 4) INTEREST
> > > In *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89
and for
> > > CF002_COOO2 is 0.85. but the interest for these clusters  in the
> > > *3d_pair_cluster.txt are -1. And I don't understand why.
> > >
> > >
> > > I know there are a lot of questions but I will be very grateful
for
> your
> > > help!
> > >
> > > Natalí
> > >
> > >
> >
> >
>
>

------------------------------------------------
Subject: MODE Time Domain outputs and attributes
From: natali aranda
Time: Tue Mar 03 09:51:19 2020

Hi John!

thanks for the quick response and for the explanation. I thought there
was
a  problem with my configure mtd file or my input files  and for that
I
have different interest values for the same pair object.
I can take the interest information in the 3d_pair_simple.txt instead
the
3d_pair_cluster.txt.

Thanks!
Natalí

El lun., 2 mar. 2020 a las 23:55, John Halley Gotway via RT (<
met_help at ucar.edu>) escribió:

> Natali,
>
> Randy is out of the office for the next 2 weeks. But I can answer
your
> question.
>
> Matching/merging is done in MODE on the simple objects from the fcst
and
> obs fields.  An interest value is computed for each fcst/obs pair.
The
> result of that process is cluster fcst and obs objects.  But it is
the
> interest values of the simple objects which determine the
> matching/merging.  As an extra step, MODE also recomputes interest
values
> for the clusters... even though they don’t impact the logic in any
way.
>
> The same logic is used in MTD as well.  However MTD does NOT
recompute the
> interest values for the matched clusters.  Randy wrote -1 as the
interest
> value for those lines as a flag value.  I believe it’s not being
done in
> MTD to get it to runs as fast as possible.
>
> Am I correct in thinking that you’d like MTD to be modified to
recompute
> the interest values for matched clusters?
>
> Thanks,
> John Halley Gotway
>
> On Mon, Mar 2, 2020 at 7:00 PM natali aranda via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201 >
> >
> > Hi Randy,
> >
> > I'm writing again...
> > off course, your explanations helped me to understand the MTD
tool.
> >
> > Do you remember that I have a problem with the interest attribute
(In
> > *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89  and for
> > CF002_COOO2 is 0.85. but the interest for these clusters  in the
> > *3d_pair_cluster.txt are -1)? I'm trying to put the observation
files and
> > forecast files in the *incoming/irap/met_help/aranda_data. *I have
16
> files
> > for the observation and 16 files for the forecast. Could you tell
me how
> > send the information?
> >
> > In incoming/irap/met_help/aranda_data, I put the configuration
file
> >
> > Thanks,
> > Natalí
> >
> >
> >
> > El mar., 18 feb. 2020 a las 15:37, Randy Bullock via RT (<
> > met_help at ucar.edu>)
> > escribió:
> >
> > > Hi Natalie -
> > >
> > > Thanks for emailing us.  I'll try to answer your questions as
best I
> can.
> > >
> > > 1)  Attribute Units
> > >    a)  CENTROID_X and CENTROID_y values are in grid units, not
> > kilometers.
> > > Similarly, SPACE_CENTROID_DIST is also in grid units.  If you
want to
> > > calculate true distances in kilometers between centroids, then
it's
> best
> > to
> > > use the CENTROID_LAT and CENTROID_LON values for the two
objects, and
> > then
> > > calculate (for example) the great circle distance using those
lat/lon
> > > values.
> > >
> > >    b)  The time unit is the difference in valid times between
the 2D
> data
> > > files that you give as input to Mode Time Domain (MTD).  That's
what
> > > determines the time unit, not accumulation times.
> > >
> > >    c)  Speeds and velocities are given in grid units per time
unit.
> Axis
> > > angle values (amd difference values) are in degrees, and are
measured
> > with
> > > respect to the grid x and y axes.  Note that the (spatial) axis
does
> not
> > > give direction of object movement,  For that, use X_DOT and
Y_DOT.
> > Spatial
> > > axis information relates the the orientation of the object with
respect
> > to
> > > the grid directions.
> > >
> > > 2)  Axis angle
> > >    This relates to part c) above.  You're thinking that the axis
gives
> > > direction of movement,and that's not correct.  For that, use
X_DOT and
> > > Y_DOT.
> > >
> > > 3)  CDIST_TRAVELLED
> > >     The reason this is always zero is because it's a bug in the
MTD
> code.
> > > It will be fixed in the upcoming release.  Apologies for that.
> > >
> > > 4)  Interest
> > >     To answer this I'd have to look at your data and run MTD
myself to
> > look
> > > at the output.  If you'd like for me to do that, let me know,
and I'll
> > tell
> > > you how to send your data and config files to us.  We're
currently
> > working
> > > to get the next release of MET out, though, and that's taking up
all
> our
> > > time.  I'm perfectly willing to do this, if you want, just don't
> expect a
> > > super-quick reply because we're really busy right now.
> > >
> > >
> > > Hope this helps!  If you have any more questions, or want
further
> > > explanation, please don't hesitate to email again!
> > >
> > > Randy Bullock
> > >
> > > On Mon, Feb 17, 2020 at 2:47 PM natali aranda via RT <
> met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > Mon Feb 17 14:47:05 2020: Request 94201 was acted upon.
> > > > Transaction: Ticket created by natali.g.aranda at gmail.com
> > > >        Queue: met_help
> > > >      Subject: MODE Time Domain outputs and attributes
> > > >        Owner: Nobody
> > > >   Requestors: natali.g.aranda at gmail.com
> > > >       Status: new
> > > >  Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=94201
> > >
> > > >
> > > >
> > > > Hi,
> > > >
> > > > I have some questions about the information of MODE Time
Domain
> output.
> > > In
> > > > aranda_data I put the  MTD output (the ascii files), the MTD
> configure
> > > file
> > > > and some mode outputs as examples (postscript files).
> > > >
> > > > 1) Attribute units
> > > >
> > > > The MET User's guide (V.7) says that in MTD the unit of length
is the
> > > grid
> > > > resolution, and the unit of time is whatever the time
separation
> > between
> > > > the input files is. Speed and velocity are thus in grid units
per
> time
> > > > unit. For other hand, The "grid_res" entry is the nominal
spacing for
> > > each
> > > > grid square in kilometers (MET User's guide)
> > > >
> > > >           a) So,  I understand that kilometers are  the units
for
> > > > CENTROID_X and          CENTROID_Y in the
*3d_single_cluster.txt and
> > > > *3d_single_simple.txt, but I'm not sure. And I have the same
question
> > for
> > > > SPACE_CENTROID_DIST in *3d_pair_simple.txt and
*3d_pair_cluster.txt
> > > >
> > > >           b) I work with precipitation accumulation in 3
hours. Is
> the
> > > hour
> > > > the unit of time for the CENTROID_T? and for
TIME_CENTROID_DELTA?
> > > >
> > > >          c) What's about the unit for velocity (X_DOT, Y_DOT,
> > > SPEED_DELTA)?
> > > > is the grid unit/hour the velocity unit? ; and for the
direction of
> > > > movement (DIRECTION_DIFF, AXIS_DIFF, AXIS_ANG), are the
degrees
> units?
> > > >
> > > > 2) AXIS ANGLE
> > > > If I have a negative value for this attribute, it's means that
the
> > > objects
> > > > moves backward with the time, is it correct?
> > > >
> > > > 3)  CDIST_TRAVELLED
> > > > I don't understand why I have CDIST_TRAVELLED = 0 in the
> > > > *3d_single_cluster.txt and *3d_single_simple.txt
> > > >
> > > > 4) INTEREST
> > > > In *3d_pair_simple.txt, the interest for CF001_COOO1 is 0.89
and for
> > > > CF002_COOO2 is 0.85. but the interest for these clusters  in
the
> > > > *3d_pair_cluster.txt are -1. And I don't understand why.
> > > >
> > > >
> > > > I know there are a lot of questions but I will be very
grateful for
> > your
> > > > help!
> > > >
> > > > Natalí
> > > >
> > > >
> > >
> > >
> >
> >
>
>

------------------------------------------------


More information about the Met_help mailing list