Page 1 of 1

Instantaneous wind fields as input?

Posted: July 7th, 2020, 2:26 pm
by MarcelMeyer
Hi,
so far I have always used gridded meteorological data with mean wind values, e.g. averaged over a 3 hourly time-period. Now we have gridded meteorological simulation data with 3-hourly temporal resolution and instantaneous wind velocities. I am wondering how to approach this, e.g. how to estimate the particle advection in this case? Is it common practise to simply use the same turbulence parameterization assuming that the instantaneous wind velocities can be used as mean values (which is clearly not the case but I suppose would be one naive option of which I dont know how bad it is)? Are there other schemes in Hysplit for estimating turbulence based on the wind profile or other inputs?
Many thanks!

Re: Instantaneous wind fields as input?

Posted: July 8th, 2020, 3:57 pm
by alicec
Yes - it is common practice to use the instantaneous winds instead of averaged. This is adequate in many situations, but not all.

HYSPLIT has several different options for estimating the turbulent velocity variance.
see KBLS, KBLT, KDEF
https://ready.arl.noaa.gov/hysplitusersguide/S410.htm
as well as the technical paper or BAMS paper
https://www.arl.noaa.gov/hysplit/hyspli ... formation/
as well as this reference
The evaluation of mixing methods in HYSPLIT using measurements from controlled tracer experiments
By: Ngan, Fong; Loughner, Christopher P.; Stein, Ariel
ATMOSPHERIC ENVIRONMENT Volume: ‏ 219 Article Number: 117043 Published: ‏ DEC 15 2019

The turbulent velocity variance is estimated by one of the methods described.
Then a turbulent velocity is computed for each computational particle at each time step as descibed here (or in tech memo)
https://www.ready.noaa.gov/documents/Tu ... _eqns.html

So the parameterization of the turbulence is not really affected by whether you have averaged or instantaneous winds.

Re: Instantaneous wind fields as input?

Posted: July 10th, 2020, 6:32 pm
by MarcelMeyer
Thank you very much for this very useful response! We will be looking at horizontal scales that resolve or partly resolve convection 2.5 - 5km grid box in the tropics. How would you recommend dealing with convection in that case / does this have affects on which turbulence parameterization is most appropriate?

Re: Instantaneous wind fields as input?

Posted: July 15th, 2020, 4:16 pm
by alicec
3 hourly and 2-5km is a fairly big mismatch in resolutions.

The mean vertical velocity tends to be a problem (values are often too large especially in convective regions) when you have low temporal resolution and high spatial resolution.
There are two vertical motion methods that were designed to help - if you use them, please let us know how they perform as they are somewhat experimental.
https://ready.arl.noaa.gov/hysplitusersguide/S212.htm

7- works under the assumption that a spatial average can be substituted for a time average.
8 - simply damps the magnitude of the vertical velocity.

As far as the turbulence parametezations go you may wish to utilize one of the CAPEMIN options to increase mixing in convective regions.
https://ready.arl.noaa.gov/hysplitusersguide/S640.htm

Re: Instantaneous wind fields as input?

Posted: July 16th, 2020, 6:03 am
by MarcelMeyer
Thank you very much. The reply is very helpful for guiding our setup. Cheers