[ncl-talk] (advect_variable) is undefined, can't continue

[Dennis Shea]

The gradients are calculated via spherical harmonics. Hence, the
'global grid' requirement. NOTE: The spherical harmonic derived
gradients are "highly accurate".
You can see the code by looking at contributed.ncl.

Yes: typo (fixed):     U*(dT/dlon) + V*(dT/dlat)  =>  (m/s)*(K/m) => K/s

Also, you can see a new page that is actually associated more with
6.4.0 but could be useful:

https://www.ncl.ucar.edu/Applications/gradients.shtml

===
Updated doc:

https://www.ncl.ucar.edu/Document/Functions/Contributed/advect_variable.shtml


Cheers

On Mon, Jan 25, 2016 at 10:17 AM, Xi Chang <xi.chang01 at gmail.com> wrote:
> Thanks guys for a quick reply.
> I have one more concern, in the example of "temperature advection" why the
> units in m.K/s instead of K/s?? is it also a typo? how does this fubction
> compute the gradient? i assume the gradient must be in K/m in lon and lat
> direction.
>
> Chang
>
>
> On Monday, 25 January 2016, Dennis Shea <shea at ucar.edu> wrote:
>>
>> That was an error in the 6.3.0 release note. Apologies!
>>
>> Try using : advect_var reather than advect_variable
>>
>>
>> =====
>>
>>
>> The 6.3.0 had:
>>
>> New functions
>>
>> advect_var - Advect a variable horizontally: U*(dQ/dlon) +
>> V*(dQ/dlat). This function was inadvertently advertised as
>> advect_variable. The 6.3.1 release will support advect_variable name
>> as an alias.
>>
>> https://www.ncl.ucar.edu/future_release.shtml
>>
>>
>> If you look at the 6.4.0 release note
>>
>> Bugs fixed
>>
>> The advect_var function can also be accessed via an alias name:
>> advect_variable. This addresses a documentation error in the 6.3.0
>> release.
>>
>> On Mon, Jan 25, 2016 at 9:57 AM, Xi Chang <xi.chang01 at gmail.com> wrote:
>> > Hallo,
>> > Im using NCL v.6.3.0,  why I cannot use this function
>> >
>> > https://www.ncl.ucar.edu/Document/Functions/Contributed/advect_variable.shtml
>> >
>> > Thanks
>> > Chang.
>> >
>> > _______________________________________________
>> > ncl-talk mailing list
>> > ncl-talk at ucar.edu
>> > List instructions, subscriber options, unsubscribe:
>> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>> >