Trajectory "Hits" ground

Topics about the HYSPLIT trajectory model.
Post Reply
masisneros
Posts: 7
Joined: February 18th, 2014, 3:17 pm
Registered HYSPLIT User: No

Trajectory "Hits" ground

Post by masisneros » April 8th, 2014, 10:50 am

Hi I'm trying to understand what does it mean if I run a backward trajectory. I did 3 different heights (50, 500, 1500) at my location (31.79 N, 106.68 S) (EDAS 40K) date July 4, 2013 at 00 UTC for 72hr.

When I get the output file I see that after a certain time the lower trajectories appear to "hit the Ground" and after some more time the upper height trajectory does as well. After all 3 height trajectories are one the, the model keeps running but all 3 trajectories are following basically ground (that's what it looks like).

My first question: The last 24 hours are "along the ground" so, what does that mean? What does the particle hitting the ground mean?

in terms of my full 72 run, can I count on the last part of the model run to say the particle appears to be going thru this area.

Maria

ariel.stein
Posts: 639
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: Trajectory "Hits" ground

Post by ariel.stein » April 10th, 2014, 10:45 am

That is a great question that does not have an easy answer. Trajectories can hit the ground for a variety of reasons, for instance when influenced by a low pressure system. After a trajectory hits the ground it looses information about the vertical motion because by definition the vertical velocity is zero. However, the horizontal transport still follows the wind fields and the trajectories will continue moving. To solve this information loss, I would recommend you use a back dispersion calculation that includes vertical mixing processes instead of the trajectories. In this way the lagrangian particles that hit the ground can go back up because of the turbulent nature of the flow modeled by the dispersion component. For more information about backward dispersion please check the following link:
http://www.meteozone.com/home/tutorial/ ... _back.html

yuejun0026
Posts: 4
Joined: May 28th, 2014, 7:15 am
Registered HYSPLIT User: Yes

Re: Trajectory "Hits" ground

Post by yuejun0026 » November 10th, 2014, 4:36 am

I had met the same situation. After some hours the terrain height become <0 . And the pressure become >0. Is there some method to solve this bug?

ariel.stein
Posts: 639
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: Trajectory "Hits" ground

Post by ariel.stein » November 10th, 2014, 10:02 am

It is not a model bug. It is just one of the shortcomings of running single trajectories to determine source-receptor relationships. As stated in my comment above, in those situations I would recommend running a back dispersion calculation instead of a single trajectory.

yuejun0026
Posts: 4
Joined: May 28th, 2014, 7:15 am
Registered HYSPLIT User: Yes

Re: Trajectory "Hits" ground

Post by yuejun0026 » November 12th, 2014, 1:48 am

Thank you for your answer. I just running a back dispersion calculation instead of a single trajectory like you said , the height of some trajectories which i need their data of height still hit ground ,how can i solve this situation? Delete these trajectories or averaging these data of height?

ariel.stein
Posts: 639
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: Trajectory "Hits" ground

Post by ariel.stein » November 12th, 2014, 9:50 am

It is OK if some particles hit the ground. You will need to set a regular grid to calculate the average concentration.
See the following link for further details:
http://www.meteozone.com/home/tutorial/ ... _back.html

qudsia
Posts: 2
Joined: December 29th, 2015, 5:39 am
Registered HYSPLIT User: Yes

Re: Trajectory "Hits" ground

Post by qudsia » December 29th, 2015, 5:51 am

Hi, The above posts have been very helping as i was having the same problem of "trajectories hitting the ground". I have used the back dispersion calculation as you suggested and did it with one hour averaging on a square grid. The model output is being stored in a "cdump" file which does not open in the editors such as notepad etc. In this case how can we get the end points for these back dispersion trajectories? OR can we get the same kind of a notepad readable endpoints points (such as tdump for the trajectory model) also for the concentration model? Thanks.!!

ariel.stein
Posts: 639
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: Trajectory "Hits" ground

Post by ariel.stein » December 29th, 2015, 12:51 pm

It seems you are doing a back dispersion calculation, which is similar to a concentration calculation. The output file you will get from such calculation is a binary file. For more information about backward dispersion please see the following link:
http://ready.arl.noaa.gov/documents/Tut ... _back.html

qudsia
Posts: 2
Joined: December 29th, 2015, 5:39 am
Registered HYSPLIT User: Yes

Re: Trajectory "Hits" ground

Post by qudsia » December 30th, 2015, 1:43 am

Thank you for your reply, actually i am looking for the 'end points' back trajectories files obtained from the trajectory model but my trajectories hit the ground. i did the back dispersion calculation in order to get the data where the trajectories give zero heights. Now that the output is a binary file, my question is as how can i get the 'end points' file (with latitudes and longitudes etc) of this back dispersion calculation which is readable in notepad like editors, as i am using these end point files to plot trajectories in some other graphics software. Thanks..!!

ariel.stein
Posts: 639
Joined: November 7th, 2012, 3:14 pm
Registered HYSPLIT User: Yes
Contact:

Re: Trajectory "Hits" ground

Post by ariel.stein » December 30th, 2015, 1:55 pm

Unfortunatelly you will not be able to open the binary file with a notepad like software. However, you might want to take a look at the following link where the pardump ( also binary) file that contains the lagrangian particle description is described.
http://ready.arl.noaa.gov/documents/Tut ... _pard.html

Post Reply