[ncl-talk] Azimuthal average of vertical velocity

Thu May 16 02:52:59 MDT 2019

Thank you very much for your response.

I am trying to implement the same method for a variable chosen from the
model output which is in GRIB2 format.

I have many forecast files(126hr forecast) initialized at 1-Dec-2017 00:00
during the cyclone time. For time being I am attaching 5 forecast files
(f000, f006, f012, f018, f024) and I used "join" to concatenate all the
five files. (uploaded data files using ftp)

Each grib2 file has got the following variables. And my task is to compute
the radius-height cross-section of the variable "TMP_P0_L100_GLL0" for the
desired forecast hour and lat/lon location.

*Variable: vNamesType: stringTotal Size: 568 bytes            71
valuesNumber of Dimensions: 1Dimensions and sizes:    [71]Coordinates:
(0)    TMP_P0_L1_GLL0(1)    TMP_P0_L7_GLL0(2)    TMP_P0_L100_GLL0(3)
TMP_P0_L103_GLL0(4)    POT_P0_L7_GLL0(5)    DPT_P0_L100_GLL0(6)
DPT_P0_L103_GLL0(7)    LHTFL_P0_L1_GLL0(8)    SHTFL_P0_L1_GLL0(9)
SPFH_P0_L100_GLL0(10)    SPFH_P0_L103_GLL0(11)    RH_P0_L100_GLL0(12)
RH_P0_L103_GLL0(13)    PWAT_P0_L200_GLL0(14)    CLWMR_P0_L100_GLL0(15)
RWMR_P0_L100_GLL0(16)    SNMR_P0_L100_GLL0(17)    RIME_P0_L100_GLL0(18)
TCOLR_P0_L200_GLL0(19)    TCOLS_P0_L200_GLL0(20)    UGRD_P0_L7_GLL0(21)
UGRD_P0_L100_GLL0(22)    UGRD_P0_L103_GLL0(23)    VGRD_P0_L7_GLL0(24)
VGRD_P0_L100_GLL0(25)    VGRD_P0_L103_GLL0(26)    VVEL_P0_L100_GLL0(27)
ABSV_P0_L100_GLL0(28)    UFLX_P0_L1_GLL0(29)    VFLX_P0_L1_GLL0(30)
VWSH_P0_L7_GLL0(31)    CD_P0_L103_GLL0(32)    PRES_P0_L1_GLL0(33)
PRES_P0_L7_GLL0(34)    PRMSL_P0_L101_GLL0(35)    HGT_P0_L1_GLL0(36)
HGT_P0_L7_GLL0(37)    HGT_P0_L100_GLL0(38)    HPBL_P0_L1_GLL0(39)
DSWRF_P0_L1_GLL0(40)    USWRF_P0_L1_GLL0(41)    DLWRF_P0_L1_GLL0(42)
ULWRF_P0_L1_GLL0(43)    CICE_P0_L100_GLL0(44)    TCOND_P0_L100_GLL0(45)
TCOLW_P0_L200_GLL0(46)    TCOLI_P0_L200_GLL0(47)
TCOLC_P0_L200_GLL0(48)    CAPE_P0_L1_GLL0(49)    CIN_P0_L1_GLL0(50)
HLCY_P0_2L103_GLL0(51)    REFD_P0_L100_GLL0(52)    REFC_P0_L200_GLL0(53)
NLAT_P0_L1_GLL0(54)    ELON_P0_L1_GLL0(55)    LAND_P0_L1_GLL0(56)
SFCR_P0_L1_GLL0(57)    WTMP_P0_L1_GLL0(58)    LCH_P0_L103_GLL0(59)
APCP_P8_L1_GLL0_acc(60)    NCPCP_P8_L1_GLL0_acc(61)
ACPCP_P8_L1_GLL0_acc(62)    DSWRF_P8_L1_GLL0_avg(63)
USWRF_P8_L1_GLL0_avg(64)    DLWRF_P8_L1_GLL0_avg(65)
ULWRF_P8_L1_GLL0_avg(66)    lv_HTGL2(67)    lv_ISBL1(68)    lv_ISBL0(69)
lat_0(70)    lon_0*

But before proceeding to the averaging stuff, first I tried to understand
the input variable. I picked up "TMP_P0_L100_GLL0" (temperature) and this
is how it looks like after joining.

*Variable: tempType: floatTotal Size: 584384920 bytes            146096230
valuesNumber of Dimensions: 4Dimensions and sizes:    [ncl_join | 5] x
[lv_ISBL0 | 46] x [lat_0 | 721] x [lon_0 | 881]Coordinates:
lv_ISBL0: [200..100000]            lat_0: [57..-33]            lon_0:
[14.2..124.2]Number Of Attributes: 12  center :    US National Weather
Service - NCEP (WMC)  production_status :    Operational products
long_name :    Temperature  units :    K  _FillValue :    1e+20  grid_type
:    Latitude/longitude  parameter_discipline_and_category :
Meteorological products, Temperature
parameter_template_discipline_category_number :    ( 0, 0, 0, 0 )
level_type :    Isobaric surface (Pa)  forecast_time :    0
forecast_time_units :    hours  initial_time :    12/01/2017 (00:00)*

Since this is forecast, there is only*  forecast_time :    0* information
available from individual grib files. Now, how must I proceed to assign the
time information for the temp variable when I use "join"? Which option
would fit better to concatenate the forecast grib files ("join" or "cat")?

ncl 0>dnam = getvardimnames(temp)
ncl 1>print(dnam(0))

*Variable: temp (subsection)Type: floatTotal Size: 20 bytes            5
valuesNumber of Dimensions: 1Dimensions and sizes:    [ncl_join |
5]Coordinates: Number Of Attributes: 15  lon_0 :    14.2  lat_0 :    57
lv_ISBL0 :    200  initial_time :    12/01/2017 (00:00)
forecast_time_units :    hours  forecast_time :    0  level_type :
Isobaric surface (Pa)  parameter_template_discipline_category_number :    (
0, 0, 0, 0 )  parameter_discipline_and_category :    Meteorological
products, Temperature  grid_type :    Latitude/longitude  _FillValue :
1e+20  units :    K  long_name :    Temperature  production_status :
Operational products  center :    US National Weather Service - NCEP
(WMC)(0)    1e+20(1)    1e+20(2)    1e+20(3)    1e+20(4)    1e+20*

 Shouldn't the output of the above command be 0,1,2,3,4 instead of 1e+20? I
am new to NCL and trying my best to figure out the things. Kindly help me.

Thanks in advance for your time and support.

On Sat, May 11, 2019 at 4:11 AM Dennis Shea <shea at ucar.edu> wrote:

> For sure, the [nice] plot was generated by NCL [panel plot with common
> labelbar].
> ---
> There was a 2017 thread on this topic. A rough approach [tst_polar.ncl]
> was attached to that thread.
> I cleaned it up and made it a function. A pic of the methodology used by
> the function (or tries to do) is here:
>
> ; A conceptual sketch: A pic:
> ;
> *https://cs.stackexchange.com/questions/43744/fast-algorithm-for-interpolating-data-from-polar-coordinates-to-cartesian-coordi*
> <https://cs.stackexchange.com/questions/43744/fast-algorithm-for-interpolating-data-from-polar-coordinates-to-cartesian-coordi>
>
> This works in degrees not km.
>
> However, if your convert the 'radius' [degrees] to those that coincide
> with km distance, this should work!  I hope!!!
>
> The attached uses *random data *so the plot [azimuthal_avg.png] does not
> look too good.
> =====
>
> A better function would use "*geolocation_circle*
> <https://www.ncl.ucar.edu/Document/Functions/Contributed/geolocation_circle.shtml>
> "
> This optionally allows 'degrees' or 'km' as input and would allow for
> *much* higher spatial sample at each radii.
> Some example of its use are at:
>
> *https://www.ncl.ucar.edu/Applications/polyg.shtml*
> <https://www.ncl.ucar.edu/Applications/polyg.shtml>
> Examples*:*
>
> *polyg_24, polyg_25*
> *polyg_29, polyg_30 and polyg_31*
>
> Good luck
>
> On Fri, May 10, 2019 at 7:19 AM David Warner <
> davidwarnerdavid1992 at gmail.com> wrote:
>
>> [image: image.png]
>>
>> On Fri, May 10, 2019 at 6:46 PM David Warner <
>> davidwarnerdavid1992 at gmail.com> wrote:
>>
>>> Dear NCL community,
>>>
>>> Is there any function available in NCL to compute the azimuthal average of a 3D variable.
>>>
>>> For example, I need to plot something similar to the image displayed which is the height-radius crosssection of azimuthal average of vertical velocity.
>>>
>>> Given the lat and lon locations by the user, the function must compute the azimuthal mean of the variable.
>>>
>>> Any suggestions would be highly appreciated. Kindly reply at your earliest convenience.
>>>
>>>
>>> [image: image.png]
>>> (Image courtesy: Hui Wang et al., 2014)
>>>
>> _______________________________________________
>> ncl-talk mailing list
>> ncl-talk at ucar.edu
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20190516/ceb514e0/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 333221 bytes
Desc: not available
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20190516/ceb514e0/attachment-0001.png>