Time Step and Concentration Calculation

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

Re: Time Step and Concentration Calculation

Post by mingwang »

No, I do not think so. I am sorry that I do not have much knowledge on that. Could you please give me an example? Thanks.
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Time Step and Concentration Calculation

Post by mingwang »

Do you mean something like 0.05Deg × π/180 = 0.0008727Rad? Thanks. :)
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Time Step and Concentration Calculation

Post by mingwang »

Dear alicec,

Thank you very much for your help. I really appreciate it.

I figured it out. The formula should be Area = (111198.323)^2 * Latspan * Lonspan * COS(Lat*π/180). In my case, the formula in Excel is Area = (111198.323)^2 * 0.05 * 0.05 * COS(RADIANS(E2)), where 0.05 (Latspan) * 0.05 (Lonspan) are the lengths (in degrees latitude, longitude) of the sides of the concentration grid, also called Spacing (in degrees) Lat, Lon in the CONTROL file, and E2 is the latitude for the concentration grid.

Now, I am able to get the correct number. :)

There is another thing that I am not quite clear about HYSPLIT, although I have looked at the Help file. For my simulation, I set deposition velocity [Vel(m/s) = 0.01].

Question 1: What is the definition of deposition velocity? Does it mean gravitational velocity? From what I known, in physics, the gravitational velocity for a particle should be changing over a course of dispersal. Why is there a constant for Vel(m/s) in HYSPLIT?

The reason I am asking is that unlike particles, the body shape of insects is not sphere. I do not think that I would be able to set values for Particle: Diameter (µm), Density (g/cc), and Shape Diameter (µm). Thus, I will need to set a reasonable value for Vel(m/s), because it can override Diameter (µm), Density (g/cc), and Shape Diameter (µm).

That would be great if you could give me a detailed description of Vel(m/s). How to estimate or measure Vel(m/s), if possible?

Your help would be much appreciated. Thanks. :)


Kind Regrads,

Ming
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Time Step and Concentration Calculation

Post by mingwang »

HYSPLIT takes a small shortcut and uses the location of the computational particle to calculate the grid area rather than the center of the concentration grid square. So the larger grid sizes may be off by a few particles if the particles are concentrated near an edge.

Please double check your numbers for the 0.01 case, because that one should be similar to the 0.05 grid.


Dear alicec,

I tried different spacing for the concentration grid, e.g. 0.01*0.01, 0.1*0.1, 0.25*0.25 and 0.5*0.5. Interestingly, the number of particles in the output file could not match the value shown in the MESSAGE file exactly, but they were still similar.

For example, for spacing 0.01*0.01, I got 4608 from the output file, but there should be 4620 shown in the MESSAGE file.

For spacing 0.1*0.1, 3987 (output file) vs. 3992 (MESSAGE)

For spacing 0.25*0.25, 3794 (output file) vs. 3793 (MESSAGE)

For spacing 0.5*0.5, 3698 (output file) vs. 3700 (MESSAGE)

The equation below only works for spacing 0.05*0.05.

Area = (111198.323)^2 * Latspacing * Lonspacing * COS(Lat*π/180)

Question: Did I use the equation in a wrong way? Why does it only work for spacing 0.05*0.05?

Your help would be much appreciated. Thanks. :)


Kind Regards,

Ming
alicec
Posts: 418
Joined: February 8th, 2016, 12:56 pm
Registered HYSPLIT User: Yes

Re: Time Step and Concentration Calculation

Post by alicec »

Please see this help page and some of the references quoted within.

https://ready.arl.noaa.gov/hysplitusersguide/S314.htm

The deposition velocity is simply defined such that the the flux of material to the surface
is the deposition velocity multiplied by the concentration above the surface.
e.g.
(kg/m^2/s) = (m/s) * (kg/m^3)

I don't know how you would set this value for insects - clearly their "deposition" mechanism is much different than for gasses or particles.
As stated before, HYSPLIT is not designed to model insects.
Maybe you have some empirical data relating the number density of insects on the ground to the number density in the air?
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Time Step and Concentration Calculation

Post by mingwang »

Thanks for your help, alicec.

So deposition velocity is actually not gravitational velocity?

From what I understand, the deposition velocity is a kind of velocity that is used to calculate the deposition flux. The value of deposition velocity depends on the types of particles or gases. That is why it is a constant value. Would I be correct? Thanks.
alicec
Posts: 418
Joined: February 8th, 2016, 12:56 pm
Registered HYSPLIT User: Yes

Re: Time Step and Concentration Calculation

Post by alicec »

1. Small particles reach a terminal fall velocity very quickly. So the gravitational settling velocity is a constant. There is extensive
literature on how to estimate the terminal fall velocity of a particle given shape, size, and density. This is probably not relevant for insects though.

2. Yes - deposition velocity is defined in order to calculate the deposition flux. Thus a gas which has no settling velocity does have a deposition velocity.
https://ready.arl.noaa.gov/hysplitusersguide/S314.htm
Dry deposition calculations are performed in the lowest model layer based upon the relation that the deposition flux equals the velocity times the ground-level air concentration.

So it might be possible to define a deposition velocity for insects if you have information relating the number of insects on the ground to the number in the air above the ground. I can see a lot of potential problems though. But I know very little about insect behavior.
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Time Step and Concentration Calculation

Post by mingwang »

Dear alicec,

Thank you very much for your help.

'HYSPLIT takes a small shortcut and uses the location of the computational particle to calculate the grid area rather than the center of the concentration grid square. So the larger grid sizes may be off by a few particles if the particles are concentrated near an edge.

Please double check your numbers for the 0.01 case, because that one should be similar to the 0.05 grid.
'

I have looked at the output file and MESSAGE again. It shows 4608 (output file) vs. 4620(MESSAGE).

I have run some simulations today with 100,000 particles and 0.25 spacing and simulation duration is 168 hours. The difference between output file (853) and MESSAGE (759) can be approximately 100 particles. Is this normal?

Your help would be much appreciated. Thanks. :)


Kind Regards,

Ming
alicec
Posts: 418
Joined: February 8th, 2016, 12:56 pm
Registered HYSPLIT User: Yes

Re: Time Step and Concentration Calculation

Post by alicec »

Please check that your concentration grid is large enough so that the particles are not being deposited outside of your grid.
That would cause the number of particles that you are calculating from the output file to be lower than you expect.
mingwang
Posts: 47
Joined: May 8th, 2019, 9:19 pm
Registered HYSPLIT User: No

Re: Time Step and Concentration Calculation

Post by mingwang »

Dear alicec,

Thank you very much for your reply.

I have run two simulations today again, and I set the sampling domain: Span (80.0, 100.00) and Spacing (0.25*0.25). It should be large enough to sample the deposited particles.

Simulation 1: one source location.

Simulation 2: six source locations (they were close to each other, because I did not want any particles deposited outside of the sampling domain).

For both simulations, I made each particle equivalent to 1 mass unit. The total number of particles released were 10,000.

Interestingly, the first simulation (only one source) could generate the number of particles (85) in the output binary file that matches the number (85 = 10,000 - 9915) shown in MESSAGE exactly.

However, for Simulation 2 (six sources), there was a difference (32 particles) between the number of particles (100) in the output binary file and the number (132 = 10,000 - 9868) shown in MESSAGE.

As you mentioned, 'HYSPLIT takes a small shortcut and uses the location of the computational particle to calculate the grid area rather than the centre of the concentration grid square. So the larger grid sizes may be off by a few particles if the particles are concentrated near an edge.’.

Does this difference caused by that as well? I have run lots of different simulations. I found that I often could not get the exactly same number, but similar.

I am sorry that I could not get my head around. I am afraid that I might make some mistakes on CONTROL and SETUP.

That would be great if you could look at the CONTROL, SETUP and MESSAGE files, and reproduce the simulations.

Your help would be much appreciated. Thanks. :)


Kind Regards,

Ming
Attachments
source_setup.txt
(547 Bytes) Downloaded 416 times
six_sources_control.txt
(394 Bytes) Downloaded 401 times
one_source_control.txt
(299 Bytes) Downloaded 427 times
Post Reply

Return to “Dispersion Model”