Diameter (µm) Density (g/cc) and Shape

Topics about the HYSPLIT dispersion model.
Post Reply
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Diameter (µm) Density (g/cc) and Shape

Post by mingwang »

Dear all,

I am using HYSPLIT to model insect dispersal including dry deposition. Insects are not a perfect sphere. I am not sure how to set ‘Diameter (µm)’, and ‘Shape’.

I guess that ‘Diameter (µm)’ refers to the diameter of the cross section (a circle) of a particle. If so, would I be able to use the area of an insect’s body to calculate ‘Diameter (µm)’?

For example, I could use Area = π*(Diameter/2) to calculate Diameter inversely, since the cross section of a particle in HYSPLIT is a circle.

How about ‘Shape’? Should I set it 2? According to the manual, 1 represent a perfect sphere and a tiny insect is not a perfect sphere.

It would be great if you could explain ‘Diameter (µm)’, ‘Density (g/cc)’, and ‘Shape’ in detail. Your help would be much appreciated. Thanks. :)


Kind Regards,

Ming
alicec
Posts: 416
Joined: February 8th, 2016, 12:56 pm
Registered HYSPLIT User: Yes

Re: Diameter (µm) Density (g/cc) and Shape

Post by alicec »

You may find this technical report by Richard Dare useful.
It explains the calculation of the terminal fall velocity of the particles in detail.

https://cawcr.gov.au/technical-reports/CTR_079.pdf

Here is a relevant section of the HYSPLIT user manual. You may also wish to consult the reference here.

https://ready.arl.noaa.gov/hysplitusersguide/S314.htm
If gravitational settling is on and the Shape is set to a negative value then the Ganser (1993) calculation is used to replace Stokes equation for estimating particle fallspeeds. The absolute value of the Shape factor is used for the calculation. The Stokes equation overestimates particle fallspeeds for particles larger than about 20 micron diameter. As this diameter often lies within size distributions of volcanic ash particles, it is desirable to use the Ganser formulation so that particle fallspeeds can be computed accurately. Ganser, G.H., 1993: A rational approach to drag prediction of spherical and nonspherical particles. Powder Technology, 77, 143-152. .
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Diameter (µm) Density (g/cc) and Shape

Post by mingwang »

Dear alicec,

Thank you very much for your help. I will look at these papers and get back to you later.

I have another two questions now. I use NCEP/NCAR Global Reanalysis Data to run HYSPLIT.

Question 1: Do I have to set ‘starting time (year, month, day, hour)’ as the UTC time in HYSPLT, regardless of meteorological data? I guess that I should not set as the local time?

Question 2: For each grid cell on the ground in the output file, the coordinate (longitude and latitude) represents the centre point of a grid cell? If so, does it work the same for the grids in the meteorological data (NCEP/NCAR Global Reanalysis Data)?

Your help would be much appreciated. Thanks. :)


Kind Regards,

Ming
Fantine
Posts: 150
Joined: November 8th, 2012, 3:41 pm

Re: Diameter (µm) Density (g/cc) and Shape

Post by Fantine »

HYSPLIT uses the UTC time. More info about the CONTROL file can be found in the user guide (starting p24): https://www.arl.noaa.gov/documents/repo ... _guide.pdf

For the conc grid (output file from HYSPLIT runs), it is in lat/long coordinate and is independent of the input met data. You can set up the conc grid (horizontal/vertical resolution and size of the domain) in the CONTROL file. More info is in the user guide p96.
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Diameter (µm) Density (g/cc) and Shape

Post by mingwang »

Dear Fantine,

Thank you very much for your reply. I am sorry that I did not describe Question 2 well.

I was asking whether the lat/long coordinate represents the centre of a grid cell or the corner of a grid cell (e.g. the top left). For example, each conc grid cell (output file from HYSPLIT runs) is 25km (0.25°) by 25km (0.25°). A grid cell (-25°, 135°) is roughly a square (25km × 25km).

Is the coordinate (-25°, 135°) located in the centre of the square, or the corner of the square?

Another question is for the input meteorological data (NCEP/NCAR Global Reanalysis Data). The resolution of NCEP/NCAR global reanalysis data is 2.5° × 2.5°. Within this domain (2.5° × 2.5°), the meteorological data would be the same, right?

For example, there are two cells: (-25°, 135°) and (-26°, 136°) in an output file from HYSPLIT runs. If they are located in a same meteorological domain (2.5° × 2.5°), do they have a same concentration value after dispersion calculation.

Your help would be much appreciated. Thanks. :)


Kind Regards,

Ming
alicec
Posts: 416
Joined: February 8th, 2016, 12:56 pm
Registered HYSPLIT User: Yes

Re: Diameter (µm) Density (g/cc) and Shape

Post by alicec »

The lat lon coordinate for the HYSPLIT concentration grid is generally the center of the cell (for computational particles on a square concentration grid).

Met data output is valid for the point at which it is output.
For computational particles which are between met grid points, the met data is interpolated (in space and time).
See the tutorial for more information on how trajectories are calculated.
https://www.ready.noaa.gov/documents/Tu ... _eqns.html

Lagrangian models are able to provide concentration at higher resolution than the input met data.
Two concentration grid cells which are located in the same meteorological domain will NOT necessarily have the same concentration.
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Diameter (µm) Density (g/cc) and Shape

Post by mingwang »

Thank you very much for your help, alicec.

I am not clear about the sentence below. Could you please rephrase it? Your help would be much appreciated. Thanks.

'Met data output is valid for the point at which it is output.'
Post Reply

Return to “Dispersion Model”